How Important is “Downtown”?

In North America, the word downtown invites us to imagine the densest and most walkable part of any city, the place where transit and other non-car modes naturally thrive more than anywhere else.  And where this is actually true, it's logical for all kinds of intercity and local transit services to focus there.

But when we project this model of downtown onto every city, we encounter fatal confusions.  Downtown implies a single place; there's just one per city or metro area.  But some cities aren't like that.  Los Angeles and Houston, two take two famous examples, have a place called downtown, but it's really just a slightly larger cluster of towers among many clusters of towers dotted across the region.  Downtown in this model is not like a center of energy around which the whole city revolves.  It's like the brightest of a bunch of stars in a constellation, and not even the brightest by much.  

So if you cling to the notion that downtown means "focal point of travel demand and especially transit", then you have to embrace the concept that an urban area may be a constellation of many downtowns.  This is not just an American sunbelt idea.  Most of the population of the Netherlands is in a single metro area called the Randstad; it's made of many small cities each with its own downtown, with a shared airport near its centroid.  Many Asian cities are so uniformly and extremely dense that almost any point in the city could be called downtown, so long as there is some mixture of uses.  All of Paris is about equally dense, with government, business, and retail districts all over the city and the tallest towers around the edge of it, so good luck finding "downtown" there.  

In Citylab, Eric Jaffe gives us the supposedly bad news that the proposed Dallas-Houston High Speed Rail (HSR) line won't go to "downtown" Houston.  Instead it will end at Northwest Mall, just outside the I-610 loop in the northwest of the city.  

But most of the Houston transit-advocates I've talked with aren't sounding nearly as upset.  That's because:

  • the proposed terminal is close to the centroid of Houston as a whole.  It's also very close to Uptown-Galleria, the region's second downtown, and to Northwest Transit Center, the busiest transit hub in the western 2/3 of the city.
  • the terminal station area is massively redevelopable.  You could easily build yet another downtown there, and if HSR is built, they probably will.
  • the project will provide great impetus for light rail or Bus Rapid Transit linking the station to the original downtown.  These projects have been sketched many times and could include either I-10 nonstop links or a refurbishment of Washington Street, a promising old streetcar street linking the two nodes.
  • in high speed rail, the cost of the last miles into an historic downtown can be a huge part of the cost and grief of the whole project.  So if you want high-speed rail to happen at all, provoking this battle is not always a sensible part of Phase 1.

The bigger challenge, for folks from strongly single-centered cities, is to notice the limits of the term downtown.  As cities grow, there is no correlation between the sustainability of a city and its single-centeredness.  On the contrary, single-centered cities present huge problems for transportation, because they use capacity so inefficiently.  New York, for example, is spending over $10 billion on a project to fit more Long Island commuter trains into Manhattan, and to put them closer to jobs there.  The demand is mostly one-way, so this requires either storing trains all day on expensive Manhattan real estate, or running them all empty in the reverse-peak direction.   It's very inefficient compared to the transit problem in a multi-centered place like Paris or Los Angeles, where demand is flowing two-way most of the time.

So growing a single downtown isn't the key to becoming a great transit city.  Quite the opposite, it's best to have a pattern of many centers, all generating high demand, and supporting balanced two-way flows between them that let us move more people on less infrastructure.  This is the great advantage of Paris or Los Angeles or the Dutch Randstad over Chicago or Manhattan.

So remember: when it comes to the efficiency and abundance of transit — or roads for that matter — "downtown" isn't all it's cracked up to be.  For transit, big clumps of density and walkability are great, but several are better than one.

37 Responses to How Important is “Downtown”?

  1. Rob J November 20, 2015 at 11:12 am #

    So for a metro region like Hampton Roads, VA (Virginia Beach-Norfolk-Newport News), where we have 9 cities, each with a downtown (or 2) spread over the whole region, transit becomes a difficult proposition because it has to serve all (or most) of these CBDs at all times of the day. I work at the MPO, so I’m not intimately familiar with how are two agencies manage this, but I can’t imagine its easy, especially when you add in all the water crossings (bridges, tunnels).

  2. simval November 20, 2015 at 11:28 am #

    I wouldn’t use Paris as a good example. Sure, transit use is high, but the dispersion of destinations all around the city forced Paris to build an extremely complicated subway system to everywhere in the city. The result is that they ran out of money to expand transit far beyond the core because they had to build so many lines to properly serve the core. As a result, few Paris subway lines extend beyond the 6-km radius of the old city. Fortunately, they’ve built the RER network to connect to suburbs.
    The dispersal of destinations is actually terrible for a city, it’s the pattern of sprawl, not of a proper city. Paris does well because the entire old city is extremely dense, but apply that pattern to American cities and you’d get… well, Oklahoma City or some similar car-dependent metro area.
    One of the biggest issues of dispersion is that it makes walking almost useless. If stores and offices are set apart from each other, walking from one to the next may be so long and difficult people may opt to drive or be driven instead. A city in which every trip requires a bus ride is no better than a city in which every trip requires a car. The traditional single downtown pattern was great for small cities because once downtown, everything was at walking distance. Even if people took a bus to go downtown, they didn’t need to take a bus until they had to go home, no matter how many things they had to do. Of course, the idea of having one single walkable downtown for million-plus metro area is a pipe dream, but dispersal doesn’t seem like a good idea because it involves the motorization of many, many trips.
    The best pattern for big metro areas in my opinion is undoubtedly the Japanese pattern where “downtowns” exist at train/subway stations, but as you get away, 10-15 minutes on foot you fall into suburban-like areas (well, suburban for Japan, which is still 4 times denser than in North America). That way, you have very dense commercial nodes at stations and residential areas some distance away. This allows walking to complement transit perfectly rather than competing with it. The result is very low bus ridership, but very high walking and biking, which is a very good thing.
    Midsize Japanese cities also build secondary downtowns at the terminus stations of their subway networks (see Shinsapporo in Sapporo and Nagamachi station in Sendai), this allows suburban feeder buses to the terminus to also serve to go to a high-density commercial node.

  3. Danny Howard November 20, 2015 at 1:16 pm #

    I always thought that downtown was South of Houston.

  4. Alon Levy November 20, 2015 at 6:45 pm #

    Which Asian cities are you thinking about when you say any point can be called downtown? Because it’s pretty much 100% false for the major high-income East Asian cities. Central Tokyo has the same number (and density) of jobs as Midtown Manhattan; Ikebukuro, Shinjuku, and Shibuya are also major business centers, but are clearly smaller, while other parts of Tokyo are yet less central. In Singapore, there’s a commercial center around Orchard and Paterson, a CBD around Raffles Place, and a new CBD that the government’s building around Marina Bay; all have plenty of neighborhoods within a few km radius that are identifiably not the CBD. I am not very familiar with Seoul, but my understanding that the Seoul CBD is as dominant as Tokyo’s, with Gangnam having the same role as Shinjuku.
    What of Europe, then? Its cities are if anything more monocentric. Randstad and the Ruhr are exceptions, as urban areas that grew out of several independent cities that became intertwined. Most cities have an identifiable center, with a concentration of employment and retail: Hötorget, Part-Dieu, the City of London, the first 4 arrondissements. Sometimes there’s also a second CBD where they built skyscrapers without ruining city center’s value as a museum of how cities looked in 1900, for example Kista or La Defense; it has the same relationship to the primary CBD that Shinjuku has to Central Tokyo. I live 3 km away from T-Centralen/Hötorget, and you will never confuse my neighborhood with the CBD, even if the buildings are approximately the same height.
    The reason commuter trains in New York and Chicago run one-way has nothing to do with employment peakiness. After all, polycentric LA has commuter trains with barely any reverse-peak service in the first place, while CBD-centric Tokyo has commuter trains that run full in all directions at all times of day. Hell, the subway trains in New York aren’t empty in the reverse-peak direction, despite the Manhattan core’s dominance. Rather, the issue is that American commuter rail is terrible at serving any kind of employment except the CBD at rush hour. Plus, suburban employment centers are practically never transit-oriented: Stamford’s jobs and retail are a lot less walkable from Metro-North than Richmond and Metrotown’s are from SkyTrain – and Stamford is still far better than Metropark, Edison, and Woodbridge.

  5. Jeff Wegerson November 20, 2015 at 8:18 pm #

    What you wrote immediately clicked with me. But then reading the comments I found myself getting a bit confused. I think, Jarrett, that you are on to something here but you need to clarify it and potentially reframe it.
    What makes cities vibrant are walkable and bikeable neighborhoods where everything you need in your life is right there: your house/apartment, your job, your shopping and your recreation. And since it is all right there you don’t want to have to have a car to get to your friends neighborhood and its different and therefore exotic set of the same basic neighborhood needs. Voila, transit.
    In that sense, the big “central business district” downtown is just one more neighborhood. So are the airports and the high-speed rail terminals and the grand parks.
    I immediately grasped your usage of Paris as an example of a downtown-less metro, but then I lost it at New York having a single big downtown. The problem for New York is that they have too many big downtowns jammed right up against each other. Asian style as you say.
    Chicago, (me), is evolving a string of downtowns along the lake. Evanston on the north is the obvious new-comer. But the Gold Coast, Lincoln Park, Lakeview and Edgewater all sport high-rises on the lake and are slowly filling in the nearby streets with the shopping and recreation of downtowns. Even the aborted downtown of Uptown is coming back. Not to mention the goings on in the south direction.

  6. Dylan November 20, 2015 at 8:34 pm #

    A few counter arguments:
    – The suburban shopping mall terminal may be fine for people in Houston, but you’re failing to consider the people coming in from Dallas. Downtown Houston is a larger destination for business travellers and tourists than any other part of the city.
    Also, Downtown has the largest concentration of hotels that HSR passengers could walk to when they get off the train.
    – Transferring to a commuter rail or bus will add a lot of time for passengers heading downtown.
    – The primary purpose of transit should be to take people to their destinations, not urban redevelopment.
    As far as I’m concerened, this line is a waste of money if it doesn’t connect both downtowns.

  7. Michael November 20, 2015 at 10:11 pm #

    You are treading into unchartered waters with the idea that we do not need to focus on downtown, for a number of reasons:
    – North American cities have allowed so much decentralization, that the last thing they need, is to encourage it even more, regardless of if it is in subcenters or sprawl.
    We have let our downtowns, even the vibrant ones, lose so much of their dominance, with the corresponding sprawl, transit usage declines, lack of business growth, etc.
    – Large cities are of course going to have subcenters. However, we should always encourage and preserve one true heart of the metro region, which works in synergy with these subcenters. And we should not be promoting subcenters in cities where the main downtown still has a significant amount of room to grow (most North American downtowns).
    – The strength of downtown and transit are connected. Cities with strong downtowns, clearly have higher transit usage rates. Cities with significant subcenters or decentralization, still struggle to get high transit mode shares to the suburban centers. Part of this reason is that most subcenters are just not central enough to allow transit to be a good option for people coming from all different directions. This is one of the reasons Calgary, until recently, had a higher transit mode share than Greater Vancouver. Because Calgary has a stronger downtown than Vancouver.
    Some cities have had minor success with subcenters, like Toronto, where the suburban centers attract 20 – 30% of work trips (and in some cases shopping trips) by transit. But this still pales in comparison to downtown.
    – Your examples don’t really hold up. European cities tend to be very centralized, with a historic CBD, and often a high-rise office cluster a little further out.
    Paris does have a city centre with almost 1 million jobs in it, and the corresponding main retail area, etc. I don’t have the stats in front of me at the moment. But many European cities have as much as half or more of their metro jobs in the centre and immediate surrounding neighborhoods. A huge contrast from North American cities, where this stat can below 15% or even below 10%.
    European cities also have policies to maintain the dominance of the city centre. My friend was just living in Amsterdam, where laws prevent almost any kind of suburban retail. Instead, the policy is to provide rapid transit to outlying areas, so they can access city centre shopping areas.
    -New York and Chicago are not centralized cities, compared to their European counterparts. Manhattan, for example, accounts for less than 20% of metropolitan NYC jobs. Even with NYC city limits, Manhattan is not the dominant workplace for outer borough residents.
    Chicago’s downtown accounts for even less jobs, and Chicago is the second most decentralized region in the USA for workplaces. East Side Access will not have empty trains in the off peak direction, because of Manhattan office workers. It will have empty trains in the off peak direction, because the MTA continues to operate their suburban rail network as a commuter rail system, and not a metropolitan rail network like in Paris, Sydney, and other cities.
    -LA, as much as it is known for sprawl, actually has a pretty centralized office employment area, stretching from downtown along Wilshire Blvd. Downtown and Wilshire account for something like 40% of jobs in the region.
    So of course, some subcentreing is okay. But North American cities do not need to be encouraged to decentralize any more than they have. If anything, they need to bring downtown back, and encourage the dominance of downtown again.

  8. Dave November 21, 2015 at 8:06 am #

    Is there merit to having a short headway ring route around a downtown that stays out of the heart itself, with a small number of circlator routes into the heart itself, so that all major routes don’t terminate at the same place and get congested on one transit mall for example?

  9. The_ararar November 21, 2015 at 4:19 pm #

    If high-distance transit gets in the city center, you have eliminated the last mile problem for its users, without buses, thanks to the walkability. This increases its usefulness.
    This applies very much to small euro cities which only have buses with lowish frequency to offer to people as a last-mile transportation mean.
    Jarret has a point when it comes to bigger euro cities though: where the main train station for long distance trains is is kinda irrelevant because most people except tourists will have take the underground anyway to reach their final destination, even if they work in the retail centre, due to how big it is (and the ground floors are all occupied by shops for kilometers).
    Taking the example of Milan, the station is not in the city center but right outside its borders. Even the old one was so, although closer.
    There are 2 buts:
    1. it’s well-linked to the center with in-station underground links
    2. the high-rise office buildings where built AFTER the rail was built, and they were built where the city underground rail passes and where the two biggest stations are, not in the city centers.
    So it may not be applicable to the US.

  10. John November 21, 2015 at 10:49 pm #

    I suspect that some of the disagreements in this thread arise from not being clear about your frame of reference.
    Take Paris. If you zoom out and look at the Paris region, it’s a metropolis with a single very dominant centre – Paris Ville, the area within the Boulevard Peripherique. I suspect that transit in the region as a whole, dominated by the suburban train lines (the RER), is very tidal-flowish towards the centre.
    But this ‘centre’ is a circle with a diameter of about 10km. If you zoom in and stand inside the circle, transit looks much more like an anywhere-to-anywhere business.

  11. P November 22, 2015 at 2:10 am #

    PETULA CLARK “Downtown” (Excerpt)
    When you’re alone and life is making you lonely
    You can always go downtown
    When you’ve got worries, all the noise and the hurry
    Seems to help, I know, downtown
    Just listen to the music of the traffic in the city
    Linger on the sidewalk where the neon signs are pretty
    How can you lose?
    The lights are much brighter there
    You can forget all your troubles, forget all your cares
    So go downtown
    Things will be great when you’re downtown
    No finer place for sure, downtown
    Everything’s waiting for you


  12. David Bickford November 22, 2015 at 5:01 am #

    The argument that many cities are polycentric must be used with caution. Often it is misused as an argument against investing in transit of any type. Houston, Dallas, Phoenix, and Los Angeles have all been stereotyped as “so spread out” in attempts to subvert any type of transit funding. Despite the cliches, those are also cities where new transit projects have yielded ridership in excess of expectations. Transit networks should reflect the needs of multiple nodes within the same city, but it’s important not to play into the hands of anti-transit voices by devaluing traditional urban cores.

  13. The_ararar November 22, 2015 at 6:42 am #

    thanks for making me discover where the melody of this cringe-inducing supermarket ad comes from:
    (paraphrasis of the speech: he wants to thank the staff for the tips about what to cook for dinner the night before)
    @David Bickford
    I think that John’s argument solves this issue: maybe Jarret is talking about city centres that are already dense enough to have rapid transit on their own and too big to provide everything at walkable distance at the same time.
    I don’t know the city in question but looking at a satellite map it looks like the station is so far out that there’s an suburby-looking neighbourhood of houses with gardens between it and the city centre.
    So I don’t think it’s optimal.
    But it seems to me that Jarrett is suggesting to ignore the issue of the station being too far out as a political salami slice tactic.
    You start by getting the high speed rail and its funding approved, and then promote a costly rapid transit link across the city later, otherwise the whole thing won’t get done at all.

  14. Michael November 22, 2015 at 7:06 am #

    There is a difference between locating a high speed rail station just beyond the borders of a historic centre (but still within the main downtown/central area of the city) like in Europe. And locating a high speed station 20 minutes out of downtown at some suburban mall or park and ride lot.
    Cities set their own priorities, and the health and vibrancy of downtown is tied more to public policy than anything else. If a city wants a strong centre, then they set policies to achieve that.
    Deciding to locate rail stations, libraries, city halls, and other destinations outside of a downtown, shows the city is not interested in maintaing a strong downtown.
    I find my fellow planners can sometimes get stuck in utopic ideals way too much. No matter how much you want it, we are never going to achieve this sort of idea that people are going to live and work all within 2 minutes of their home, and that sub centers will allow that.
    Will some people live and work within close proximity? Yes they will. But in today’s economy, with many two income families, and people changing jobs often. You just are not going to get everyone in self contained neighborhoods. And when you move to many jobs to sub centers or in just plain sprawl, you reduce the access to jobs for a large percentage of the metropolitan workforce, whether by car or transit.
    Young people a side, I think one of the reasons we may be seeing more and more office move back downtown, is because of location advantage. A downtown workplace can draw a metropolitan wide workforce, you can get to work in a reasonable time by transit.
    If you are in a sub centre on one side of the region, you start to have access issues, or really really long commute times.
    Again, we can’t have everything downtown. But I think we have to understand the location advantage of having many jobs back in an easy to get to downtown or central area.
    I remember at one of my previous jobs, having to help people choose their commute options, because their offices were moving out of downtown to a suburban job centre. It was hell, and most of them who had been taking transit for 20 or 30 years, all of sudden found themselves having to drive, or face a 2 or 3 hour commute from one end of the metro area to another. The common comment was “I just wish my job stayed downtown”.
    In closing, downtowns are just cool :). You do not get the synergy of uses, excitement, the wide assortment of offerings, anywhere else but in a healthy downtown. And you can’t have it anywhere else, because you have to have that vibrancy that comes from hundreds of thousands of people coming to a common centre.

  15. Jarrett November 22, 2015 at 2:29 pm #

    Thanks for all the great comments. Let me hit some key points in response that I may not have been clear enough about.
    First, I’m making an critical distinction between polycentricity and sprawl, which are fatally confused in a lot of pubilc discussions and in some of the comments here. The Dutch Randstad is polycentric but not sprawling. It has all the features of compact, sustainable urbanism despite not having any single dominant center. Many older US metros, especially in the midwest and northeast, are sprawling and yet still relatively monocentric.
    Second, I’m pointing out that if you control for the degree of sprawl, polycentricity yields much more efficient and effective transit markets than monocentricity, because two-way flows of demand yield more effective use of demand.
    Third and most important, several of the comments recognize downtown as meaning either “the oldest part of the city” or “the part of the city where unique destinations of citywide interest are most likely to be concentrated.” I should have separated out these meanings, neither of which necessarily make all that much difference to transit when considered by themselves. So in Paris the innermost four arrondissements are certainly the most historic, but unique citywide destinations are all over the inner ring of arrondissements and increasingly all over the city, and again, the biggest clusters of towers, both jobs and housing, are in the innermost suburbs. When I was thinking of Asian cities, certainly all have a downtown in terms of having a place where the city began and a large area where most destinations of citywide interest are, but the high uniform density across a much larger area, including a lot of secondary mixed use, gives this a different significance for transit than a typical US downtown has. I will admit too that I was thinking more of fast-replicating Chinese cities, and of Hong Kong, rather than Japanese ones.
    Fourth, of course “downtown” is a relative term. Zoom out on the New York region and Manhattan is a single downtown. Zoom into Manhattan and there are clearly two, Midtown and Lower Manhattan. So a lot of arguments about how many “downtowns” an area has are relative. Be conscious of this, and those disagreements become simple matters of different frames of reference.
    As for the Houston case, the consensus among Houston transit leaders who responded to my private query on this is striking. This is a great example of why “downtown” is just a misleading concept for discussing Houston. If you think most folks from Dallas are going to a point that is that so far east of the centroid of Houston, just because it’s the historic center, then I’d respectfully suggest that you study Houston a little more. There are highrise concentrations of jobs and activity centers all around Houston, with a general pattern of westward growth, and the new station is not far from the centroid of that. It would be great to see a Dallas-Houston HSR line have two or more Houston stations, but regardless of whether you’re a public or private entity, it can be too expensive, and too dangerous to the whole project, to take this on in Phase 1.

  16. Rick Robinson November 22, 2015 at 2:37 pm #

    New to Human Transit and these excellent discussions!
    Transit planners have to work with the cities they have, not the ones they might like, but I can see a major transit-specific advantages to a single downtown: it makes commuter service practical even from distant exurbs. Commuter service always has a dispersion problem at the outer end, in the burbs, but at least with a single downtown core the commuter lines have just one downtown end, so frequent service (even if peak-only) is far more attainable.
    Why does this matter? Commuter service has a something of a bad rap, for good reasons, but it pushes back against further sprawl in several ways. It makes working downtown a more practical option – giving people in the burbs a reason to leave the car at home, or at least in the park & ride lot, instead of driving 20+ miles each way to an office park in some other burb. This in turn encourages companies to locate downtown instead of those isolated office parks; even in the burbs it can encourage denser development around commuter stations.
    Commuter service also gives its users a personal stake in the core city and in transit – for that matter, a simple awareness of transit – that is otherwise lacking. For most suburban residents an hourly coverage bus is useless and all but invisible; to them, transit is for Other People, if they notice it at all. But commuter service is something that they, their friends and neighbors actually use. Transit becomes a part of their lives. And politically, modern downtown commuters tend to be middle to upper middle class: They are influential, and service for them is not stigmatized as “welfare.” Which creates a constituency for transit, even in the exurbs.
    Finally, a thriving commuter operation has at least the potential for increasing its span with frequent offpeak service and thus becoming true regional rapid transit, making downtown transit-accessible all day. Yes, there are technical constraints, at least for rail. Commuter rail is usually the least expensive form of rail transit to build, since it can use existing tracks – at least for service to the historic downtown where rail lines converge – but on single-track lines it is almost impossible to provide two-way frequent service. (Commuter buses don’t have this problem, though they are less conspicuous in the burbs than trains are, and currently have less political cachet.)
    To be sure, none of this is a given! In the SF Bay area, Caltrain did not prevent the growth of Silicon Valley office parks; we now have the silly paradox of private commuter buses taking San Francisco residents to suburban jobs. And like most commuter rail, at least in the US, it has minimal offpeak service even though it has a double track. On the other hand, BART does provide true regional rapid transit in the East Bay, making San Francisco transit-accessible for both work and play. But coming back around to the original point, a single strong downtown core instead of several “downtowns” gives commuter service a much better chance of expanding transit, and a transit culture, out to the burbs and exurban fringe.

  17. Christof Spieler November 23, 2015 at 9:41 am #

    Also worth noting: When we’re talking about intercity rail (as opposed to light rail or a subway) the “Downtown” station will rarely be in a place that is actually within walking distance of many Downtown destinations. Dallas is a good example: Union Station is “Downtown”, but it’s a mile from the center of the highrise office district. LA, Denver, Portland, Seattle all have similar setups; even Philadelphia, Baltimore, and DC have stations that are at the periphery of the core and more than 10 min walk from most of it. Thus, even a “Downtown” station involves making a connection to a “last mile” mode. That can change over time as Downtown expands (e.g. the new development around the station in Denver) but unless the station is in the center of Downtown to start with that issue will never go away. We’re not talking transfer vs. no transfer; we’re talking about where the transfer is and what mode it is to.

  18. Nathan Landau November 23, 2015 at 10:52 am #

    Christof, we’re also talking about how long the “last mile” trip really is, how long it takes, and how frequent service is. LA Union Station, for example, is definitely on the periphery of Downtown LA and not on Downtown’s prime growth path. But an Amtrak passenger can get on a subway (with 6 minute daytime headways) and take about a 5 minute ride to the center of downtown. Is it really going to be that convenient with Houston’s station?
    Is there any transit system in the United States that servces secondary downtowns as well as the regional CBD? I know that systems like Houston are trying to improve their service to those downtowns, but it seems like the CBDs have more service.
    The areas around secondary downtowns tend to be more suburban and sprawling, in some cases the growth of the secondary downtown has stimulated suburban growth further out. Downtown Walnut Creek and Downtown Concord in the Bay Area stimulated growth further out. Downtown Walnut Creek mitigates this some with housing development whose residents could theoretically walk to work, Downtown Concord has much less. Some would say BART facilitated this sprawl, others argue it would have happened anyway.
    I realize that we can’t roll the clock back to the primate downtowns of 1915, but like many others above, I’m very leery of the notion that public policy should encourage further deconcentration.

  19. Alon Levy November 23, 2015 at 11:14 am #

    Jarrett, the job density in those cities is very non-uniform. It’s not quite about historicity – I have no trouble identifying Part-Dieu as Lyon’s primary CBD but relegating La Defense to secondary status. La Defense has a lot of tall buildings, but the big concentration of jobs is where the busiest Metro and RER lines meet around Les Halles. For a quick sanity check, look at population density figures: the first four arrondissements have much less residential density than the rest, even as the buildings are the same height. (This is similar to the low residential density of Chiyoda and of community boards 1 and 5 in Manhattan.)
    Where Paris differs from Tokyo is that the CBD is geographically big, whereas in Tokyo, it’s only a couple subway stations wide, from Marunouchi to Hibiya.
    The importance of this is that we have a precedent for a high-speed rail line that stopped far short of the primate city’s CBD: the Tohoku/Joetsu Shinkansen. In 1982, these lines opened to Omiya, 30 km from Central Tokyo, and in 1985 they were extended to Ueno, about 4 km north of Tokyo Station. They were considered basket cases: unprofitable examples of JNR’s overexpansion and of excessive political interference. Well, in 1995 they were extended to Tokyo Station, and nowadays their combined ridership is almost as high as that of the Tokaido Shinkansen, and JR East is making money on them. Judging by train schedules, they are especially well-utilized by long-distance commuters working in Tokyo.

  20. The Overhead Wire November 23, 2015 at 2:58 pm #

    This is the quote that got me to disagree…
    “the terminal station area is massively redevelopable. You could easily build yet another downtown there, and if HSR is built, they probably will.”
    That’s not how development works. Though I wish it did! Dense buildings and creating another edge city are dependent on land values and from the building stock that is where the line would terminate now it looks like they are all low cost industrial uses. Once the train gets there it’s unlikely to skyrocket in value and develop into a new center. The only hope for that would be a massive employer developing a new campus. (ie Exxon outside of The Woodlands) And even then they are going to demand gobs of parking in a place like Houston.
    Otherwise you won’t be able to command rents from office tenants of the magnitude to get to 50k+ workers in such densities in new buildings. Especially in Houston where rents are probably cheaper than other places.
    This is the same argument we use to try to justify for crappy commuter rail corridors on freight rights of way all the time. It’s cheaper and TOD will develop on the “developable” land. Sure there is land, but just because it’s there doesn’t man it has the land values to support densification like we would want to see from a second downtown. If anyone knows of a place where a new edge city has developed from nothing recently let me know. Would love to see the economics play out.
    In the meantime, I realize this is a private project and they are trying to avoid spending money and getting into fights, but value of the line itself from a land use perspective is a function of and land value gravity well that is created from downtown and radiates out to other centers. If this is just an airport copycat, expect to get the same development that airports get on the periphery…car rental lots and industrial uses.
    The Netherlands commentary bothers me too in that those cities have regional and high speed rail stops right in their office cores and transit connections from those centers. None of those cores were created later.
    I also agree with Nathan that the connection time matters. 5 minutes is much different than 30-45. Even if your terminal is on the outskirts, it’s much more viable than creating a whole new land value market from scratch. Transit doesn’t create value, it extends it. And from talking with Eric Eidlin about his work on HSR in France I think this has borne out with this technology as well.
    Folks might be interested in hearing what he has to say about his research in France and Germany. Two very different systems.

  21. The Overhead Wire November 23, 2015 at 3:12 pm #

    You can find the specific commentary in the podcast at minute 20.

  22. M1EK November 24, 2015 at 7:07 am #

    Seems to me like some people with vested interests in pretending that this will work due to a position on transfers that defies reality are pretending that this will work, while most of us in reality land are observing that this is going to attract zero to-Houston business this way (maybe some from-Houson business if the Dallas terminal is not so stupidly sited).

  23. Rick Robinson November 24, 2015 at 11:25 am #

    Regarding Houston, how does the proposed HSR terminal stack up for business convenience compared to the airport? Presumably HSR will be competing mainly with air travel (or driving 240 miles on I-45), not with whatever other options are available today.

  24. Eric November 25, 2015 at 9:12 am #

    Taxi fare from Houston airport to downtown and other job centers is $50-$80. From the proposed HSR station to the same job centers would be $20-$50. That’s a significant difference.

  25. mrsman November 25, 2015 at 10:21 am #

    On a historical note in NYC, Lower Manhattan was the original Downtown. When railroads were first coming on the scene, there was a law that prohibited steam trains south of 42nd street, which was at the time the limits of the developable city. So Grand Central Station was placed there.
    Now of course, when talking about NYC’s two “downtowns” the Midtown job center is centered around Grand Central and actually has more jobs than the job center around Wall Street.
    So something like this is definitely possible in Houston – although it may take decades for the development to play out.

  26. mrsman November 25, 2015 at 10:29 am #

    Does anyone know why the HSR cannot utilize the existing railroad tracks at conventional railroad speeds between Northwest Mall and some point in Downtown Houston (like the current Amtrak station or maybe the University of Houston)?
    I believe that most of the US proposals for HSR plan on using conventional railroad right of way in the centers of cities but utilizing a new straight corridor in suburban and rural areas where the trains can pick up speed.

  27. Nathanael November 26, 2015 at 11:31 am #

    I think the key issue in Houston is lack of connectivity.
    They picked a station location which is not connected to Houston Metro rail, and is not planned to be connected to it for a long time, and even then is not planned to be connected to it in a straightforward manner. Getting from this station to downtown will require taking three different rail lines.
    So… you do lose a lot by having the station in a location with *poor connections*.

  28. JRC November 30, 2015 at 6:58 am #

    While the use of centroid is interesting it is nonetheless incorrect. The is nomass or indeed any uniform density in that location only a developer in need of a sale at a high price.
    I am aware that some study from NYU in 2009 states that there are approximately 100,000 commuters between Dallas/Ft. Worth and Houston. How many of those are truly commuters? i.e. daily? How many are just business people or visitors who travel once a week? How much will it cost to build this train and how long before it is paid for by these “commuters: and its maintenance is pair for by those same “commuters”?
    Bravo to Nathanael for his astute observation.
    More taxpayer dollars handed over to city/urban planners, consultants and especially developers.

  29. Vin December 1, 2015 at 11:30 am #

    Midtown Manhattan isn’t really “centered” around Grand Central. If there is a “center” to Midtown, it’s Times Square, or maybe 5th/6th Aves. in the 50s. Grand Central is actually near the eastern edge of Midtown’s office cluster, just as Penn Station is near its southwestern edge. Inasmuch as Midtown development followed transit, it followed the subways – for which Times Square is a more important convergence point than either Grand Central or Penn Station – rather than commuter or intercity rail.
    Actually, the edge of the center – not the center of the center – is where major intercity/commuter rail stations usually go. Think of the ring of stations around central London, or Union Station in Chicago, plus Grand Central and Penn. But these stations are still within walking distance of a number of “downtown” destinations, and quick transit ride away from the rest. Connections are the key.

  30. Vincent December 1, 2015 at 2:08 pm #

    As a non-expert, but interested citizen of Houston, I’d like to say that Jarrett is correct in his assessment of the city. There are seven “downtowns” in Houston and the location of the HSR terminal will not change that for the better or worse. Most importantly, the Northwest Mall location is more conveniently located in relation to the mass of area residents and commerce than downtown — or either airport for that matter.
    If an obvious connection is made to the Northwest Transit Center 1.5 miles away, that gives passengers access to the frequent express bus service into downtown with this year’s New Bus Network. (Good move, Jarrett and Metro.)

  31. Jason December 9, 2015 at 5:23 am #


    1. While the Tohoku and Joetsu shinkansen extends into Tokyo Station, the segment from Omiya to Tokyo is no faster than the regular train at all – takes almost 30 minutes to travel 30km – due to noise pollution concerns (they were sued back in the 1980s). The point of it is to reduce transfers and improve ride quality, as well as connecting to the Tokaido shinkansen. Most people going to Shinjuku get off at Omiya.

    2. Tokyo CBD is definitely bigger than that. You can extend the area you described a station or two in almost all 4 directions – makes it almost twice as big.

    In any case, I feel like CBD in the strict sense isn’t what Jarrett is talking about. It’s about density and mixed use. While New York CBD may be comparable to Tokyo CBD the non-CBD areas of Tokyo is so big and yet so dense and so full of office buildings, retail establishments, restaurants, etc, that you might as well say there are centers everywhere in the North American sense. There is certainly a significant mixed use “center” around most train and subway stations even if a significant percentage of central Tokyo land-wise is purely residential. Not everyone works at Tokyo CBD or shops at Shibuya (or Ikebukuro or Shinjuku or Ginza, or whatever), so the metro has not exploded and ground to a halt – and that’s exactly the point of having many centers, big and small, spread out.


    The part I disagree with is “building another new center.” It can and probably will be redeveloped, but I wouldn’t be surprised if it doesn’t get as significant as other centers for a long time. From Japan, we can see many examples of shinkansen stations lying on the outskirts (built there instead of in the old center due to cost, curvature, etc) never getting developed anywhere close to existing centers. Shin-Yokohama is not Yokohama, Shin-kobe is not Kobe, even Shin-Osaka, which has been developed quite well (as it’s been around for 50 years, waiting for the slow expansion of the city), pales in comparison to Osaka proper. Move farther out, and you have sad things like Shin-Aomori which is basically a park-and-ride with absolutely nothing around it and no intention to build anything whatsoever for at least 20 years. Shin-Hakodate in Hokkaido, to be opened next march as the new terminus up north, will probably end up in the same fate.

    If the HSR station is getting built outside of a city center, there’d better be fast and direct connections to the centers.

    • Jason December 9, 2015 at 5:30 am #

      Also I would like to add that shinkansen commuting is rather uncommon nowadays, because the monthly pass is VERY expensive. The commuter’s pass from Omiya to Tokyo is around 55,000 yen / month, more than three times as expensive as regular train. From Kumagaya or Odawara, where the regular train (1-1.5h) is long enough to be painful, the month pass for shinkansen is 70,000+ yen, twice as expensive as regular train, and well out of the reach for many people unless their company pays for it.

  32. Heyward Dixon December 9, 2015 at 7:09 am #

    The idea of a new “Downtown” springing up because of the construction of a rail station is a real head-scratcher. It’s much more likely to be a giant parking lot or garage. It wouldn’t advance local mobility needs. And wouldn’t spur local action to link our transit lines to the station. Metro doesn’t to my knowledge have any plans to extend lines to either airport, why would they do so for a private rail operator?
    Let’s assume that all of METROs (published) dreams come true and there is a connector added to link to the HRS. Passengers are still presented with a tortured path to to the places they they are presumably trying to reach. Major points like: Business, Conventions, Sports, Tourism, Theater, Music, Museums and Health care institutions are concentrated along the spine of our existing transit system. They would have at least one mode change, and likely two or more line transfers and numerous stops before they reach their destination. All while managing their own luggage. A fifty dollar cab fare starts to sound pretty good.
    No we can’t move downtown, the train needs to go to the one we have or we need a local line that gets people there quickly.

  33. WAN HEMPEL March 2, 2016 at 12:14 am #

    Practical commentary . I Appreciate the facts ! Does someone know where my business might get access to a blank a form document to complete ?

    • Maira Salter March 3, 2016 at 8:28 am #

      Greetings WAN HEMPEL. my business partner got ahold of a template a form form using this

  34. Louis February 12, 2018 at 3:25 am #

    Downtow where I live is a place dangerous for drivin safely. Means no rullers.

  35. Louis February 12, 2018 at 3:26 am #

    Muito obrigada senhor por compartilhar. Muito bom seu site.