Monday, December 16, 2013

Some praise, but also criticism of Enrique Peñalosa's TED talk about cities

I have a few comments about Enrique Peñalosa's recent TED talk, which has been making the rounds. I'll say up front that he makes a lot of very good comments and it is worth watching. He's right: a bus with 80 people ought to have dedicated lanes. The streets should be safe for children and families. We have to rein in sprawl. But I have some criticism of the example pictures that he shows. I think that they undermine his message.

(all images are from the TED talk)

Here we see his example of a bus rapid transit system in his country. I'll say that from a technical perspective, it's pretty neat. Cheap labor costs, reuse of existing roadway, good ridership all make this a cost-effective project. The big problem is the location. This is highway-median transit. Look at what the people have to walk across just to get to the station. The 3 lanes of cars that you can see are only a fraction of the total. There's even more lanes of cars outside the frame of the picture, on the other side of the pointlessly landscaped median. This is about as anti-urban a scene as you can get. For all that he talks about equity for people, this is an example of a highway where the vast majority of the space is dedicated to private automobiles. And I think that this is a miserable example and it won't inspire anyone to want to copy it. It puts the transit in a very inhospitable location.

This is an example from Guangzhou, China. It's slightly better, in that there's only 6 lanes of traffic surrounding the busway. But it's still a hideously wide, Hypertrophic corridor: there's even grade separation of pedestrians. You can't cross this street without going up and over. It's a highway. If you want to talk about superiority over subways, then you need to show us a corridor with urbanism that matches what a subway can provide. This is not that. This is not human scaled. This is a massive hole in the city.

A terrifying vision, Enrique Peñalosa's "Radiant City"
Now, at this point in the TED talk is where Peñalosa goes off the deep end. The last 2 pictures could have been dismissed as just unfortunate choices. But this here is his "vision" of the city of the future. And it is HORRIFYING!

The picture painted here is one that might make Le Corbusier's heart warm. Towers in the park. Giant highways. Grade separated pedestrian ways. It's practically a compendium of what-not-to-do in urban areas. This is a vision which was implemented in 20th century cities and turned out to be a complete disaster (e.g. 1950s public housing projects). What in the world is Enrique thinking?

I picked out these two examples of streets with large pedestrian/bikeway areas. Actually, it's not clear if the top example even has a roadway for cars. The major problem with these two pictures is: where's the life? Both are in almost completely dead scenes. There's no street life. There's no "eyes on the street" because they're both set in desolate places. Instead there's tons of "buffering greenspace." These would make fine recreational paths through a park. But for a city street, these violate the first principles of safety in numbers. Perhaps this is not his vision for a city street, he's not entirely clear. I hope not. This is not the vision of a vibrant city. At best, it is the vision of an outlying suburb. Compare that to this picture he also shows:

Isn't that obviously much better? Life, people, a human-friendly street and neighborhood. Simple. But not from Colombia.

I don't know what to make of Enrique Peñalosa in this talk. On the one hand, he says lots of good things, and even shows some nice pictures from places like Amsterdam. On the other hand, he highlights examples from his own country that look like the worst of the tragic mistakes from 1950s urban planning.


  1. Strongly agree -- why on earth does he want to replicate Soviet atrocities when a highly dense form suitable for pedestrian/transit living, that is also beautiful, vital, and human, is already existing in Colombia?

    What also really jumps out at me from all those 'Radiant City' drawings these guys put out is the strong atomization inherent in their environments. You have totally private space in the apartments of those towers, and then totally public, exposed space in the form of massive carriageways/green space. They've lost the intermediary vocabulary of courtyards, alleys, lanes, streets, patios, balconies that provide a healthy intermediary space that allows human beings to merge in and out of public and private. You can see Peñalosa maybe grasping at this a little with those shared patios on top of his Radiant Towers, which is a minor improvement.

  2. Jarrett sent me this comment, because he was unable to get past the filter:

    ```Do not project developed-world urbanist expectations onto the developing world! And do not let your aesthetic preferences make you an opponent of liberation.

    Developing world BRT is about moving great masses of people who already have a disincentive to driving, and the standard busy-road-median, while not ideal urban design, is an overwhelming, transformative improvement in millions of lives.

    The developing-world reality includes weak law enforcement, so transit lanes need to be self-enforcing. That's why you see the emphasis on infrastructure that protects transit speed and reliability.

    Yes, developing world people love the small town ambience too. But they must work with the cities and right-of-way opportunities that they have. And the must move masses of people that just don't fit through a cute urban design where everything is slow and mixed.

    Efficiency is abundance!

    Jarrett Walker,'''

    1. My response:

      I don't disagree with your comment, because my main critique was not
      on the BRT, but on the urban design element of Peñalosa's talk. I
      think it's great that he and the others were able to take lanes and
      put together a world class BRT system, given the scenario. The mistake
      was building a highway like that in the first place, especially in a
      place where car ownership is so low. You and I know that BRT can be
      built in all sorts of contexts, but the general public could have used
      a better example.

      But there is a second part to his talk where he pushes the idea of a
      particular urban design, and he starts by talking about how many new
      cities and neighborhoods will be developed in the coming decades. He
      correctly identifies the problem: too much space given to cars, and
      too little given to people. But his proposed solution is almost
      identical to the Radiant City ideology that we already know is a
      massive failure, and which has produced many of the problems that he
      so eloquently speaks about.

      He seems to be on board with the superblocks, giant
      towers-in-the-park, the massive highways, parking lots, and the
      isolation created by Corbusier-style urban planning. According to him,
      it can all be fixed by just adding green ped/bike ways. I think he's
      wrong on that, massive amounts of automobile infrastructure is
      unaffordable for the developing world, and by pushing on this
      particular urban design scheme he is detracting from the very good
      work on BRT that he helped bring about.

    2. Jarrett writes: ```I think he's working with the realities of Latin America, and I'll be reluctant to second-guess him about those things. I believe in seeing not just the beauty of urban design but also the beauty of mass-liberation, and that's the context of most developing world thinking: Liberate people from poverty and create a more civilized city, partly by helping them not buy a car or motorbike. And liberation requires abundance, which requires speed and frequency and reliability, which in the developing world means clear running ways for buses somewhat divorced from urban design features that would slow down or entangle them.'''

    3. My response:

      I agree that he knows more about Latin America than either of us. But
      it's when he started to talk about future cities, and in other regions
      of the world, that I grew concerned. The Transmileno system may serve
      well for accessibility through high mobility given the current state
      of Bogotá. But when talking about a clean slate, new cities, new
      neighborhoods, his proposals would perpetuate or worsen the local
      accessibility problems that push people into purchasing personal
      vehicles. That's because, despite his rhetoric, his proposals seem to
      be for new cities to be built at automobile-scale. It won't matter how
      good the transit is -- if it sucks to walk, then people will aspire to
      buy cars, even if they cannot afford it.

    4. Jarrett writes: ```I do think Penalosa is making most of his mark in the developing world, which, we must remember, is most of the world. The Latin American BRT standard is having a huge impact across many countries that I follow, including all of low-income Asia and Africa.

      I agree that superblocks convey "automobile scale" to us in the developed world, but they are also highly efficient development blocks for getting lots of stuff built fast. The developing world is in a desperate hurry to do a lot with not much money, so no authority is going to say we should have less development so that we can have an intimate Portland street grid. Development is wealth for the political right and money for alleviation of poverty for the left. Nobody is going to say no to that anytime soon. If we get superblocks reasonably pierced by bike and ped ways, that's going to have to count as success. That's what I'm seeing in most developing cities I follow.'''

    5. My response:

      I'm glad that he is getting some good stuff done.

      Regarding superblocks, it's not just that it's automobile-scale in
      distance, it's also automobile-scale in price. Maintaining the
      infrastructure to support superblocks is much more expensive than
      traditional, fine-grained blocks. I think it's telling that
      pre-automobile cities were developed with small blocks. They were not
      wealthy like we are today, local accessibility was paramount, so there
      was no money nor motivation to build large-scale blocks with
      super-sized streets. When such streets were created, it was usually to
      please some aesthetic preference of an elite (e.g. Paris).

      But you are probably right about the political situation with regard
      to developers. They seem to prefer when you hand over large swathes of
      land, much like what was done with Title I housing projects in the
      U.S. I worry that the end result will be similar to the 1950s housing
      projects, even with ped/bike pathways added: total disconnection from
      the urban context around them, followed by crime, fear, and eventual

  3. The discussion on superblocks was interesting. Unfortunately, I think the upfront appeal of the superblock in the developing world that Jarrett described - its purported ease in accommodating rapid urbanization - leads to its downfall as a supportable (or "sustainable" if you will) typology later on.

    Yes, as Jarrett said, it's possible to have superblocks frequently parsed by walkways and bikeways, and there are not a few examples of this - even among the few surviving US housing projects. But people don't use walkways or bikeways just because they're there. If there's a walkway connecting two miles' worth of superblocks with nothing but housing and a few convenience stores on them, will anyone use them? Perhaps in the developing world, where there's less transportation choice, people might be compelled to travel on these walkways/bikeways further than anyone in the developed world would, but isn't this still extraordinarily poor design: doesn't it put *unnecessary* burdens on residents?

    To expand, it's theoretically possible to build a superblock so fine-grained in its mixture of residences/work/shopping/entertainment (and pathways connecting these things) that it would essentially function like a traditional city block(s). But accommodating this richness would hamper the very development "efficiency" that Jarrett mentioned! The superblock is desirable in the developing world precisely because it allows poor municipalities to bypass the "fine-graining" process in their rush to urbanize. So there's really no advantage in asking for thoughtfully-composed superblocks if you want to save time and money - you might as well ask for traditional city blocks! There's ultimately every incentive to compose sloppy, single-use tower districts, interspersed with a couple shopping centers.

    To that end, doesn't the very "efficiency" of this superblock urbanization in the developing world lead to the inefficiency of their municipal infrastructure? I.e. how many BRT lines were induced among all the dreary tower districts of Shanghai *because* the walking that is possible on a traditional urban pattern has been rendered impractical? (The transit version of induced traffic!) And once those superblock dwellers become rich enough, they'll even abandon the BRT in favor of cars, because that's what superblocks inherently encourage.*

  4. *I was in Poland a few years ago, and during the communist era they built many such superblock housing districts interspersed with shopping centers. Of course, during the communist era most people were too poor to have cars, so there was a LOT of transit ridership in and around these superblocks - again to travel to things that were too far away to walk to thanks to the superblock pattern!

    But now car ownership is *exploding* in Poland, and guess where it's rising fastest? Apart from small towns and rural areas, it's ballooning among the superblocks: many former "green spaces" in between the towers are turning into parking lots. Walking among superblocks is unpleasant, but doable if you have no choice. It gets a bit easier if you can take a bus. It gets a LOT easier if you have a car! It's no different from suburbia, really.

    Finally, I think the economic arguments for the proliferation of superblocks in the developing world are compelling (the developers' desire to control and quickly build-up ever-larger plots of land cannot be overestimated!), but I still wonder if simple outdated planning/design inertia is also responsible. (As an arch school grad, I can confirm that the superblock is still often the default design strategy.)

    I've yet to see compelling data that the superblock pattern is the most "efficient" in comprehensive terms other than simplistic *ease of development.* I think developing nations will be surprised to discover that accommodating rapid urbanization won't be the hard part: *maintaining* and supporting that development financially and infrastructurally over the decades will be the hard part! So a preference for traditional urbanism goes beyond mere "aesthetic" predilections; indeed I'd argue that perhaps the developing world's focus on superblocks is at least partially aesthetic: what better/faster way for an unstable government like the CCP to instill a sense of pride and rapid achievement than to create huge tower block districts? (There seems to be a historical connection between the totalitarian impulse and the Radiant City pattern.)

    Let's also not forget that much developing-world urbanization is occurring in societies and housing markets that aren't exactly free, open, and private - except maybe on paper - so the superblock pattern may also be a product of giant, top-down development agencies that simply don't/can't produce anything else. Outside their reach, developing world urbanization continues to follow the traditional pattern Matthew described...

    ...the shantytown! They begin as squalid slums, and over the generations they molt into hospitable and even desirable neighborhoods (assuming there is enough political/social/economic stability to support the evolution). This is, in fact, how all the first world's "aesthetic" urbanism started, and I'm not convinced that the top-down stamping-out of superblocks has the same ability to age, adapt, and support itself (All the old superblock examples keep dying, all the new examples are just that: too new.) Has a tower block dormitory ever been able to transition into anything else without top-down "renewal?" Or an office park? Or a shopping center?

  5. Another point I forgot...

    Almost all the discussion on making superblocks "walkable" focuses on the theoretical possibility of infusing their interiors with enough walkways and bikeways so the superblock feels like an agglomeration of smaller blocks.

    But where's the even-more-important discussion on how people are supposed to walk *between* superblocks? Due to their form and function, superblocks induce giant arterial roadways between them, and it's even harder to cross these than it is to traverse the superblocks themselves! (This, I argue, is one more reason behind the "induced transit" - and later induced auto travel - in such districts.)

    Right now, perhaps because they're still self-evidently necessary for so many people, developing nations are still putting pedestrian tunnels (sometimes connecting to metro stations) or pedestrian bridges (sometimes connecting to BRT medians) under or over these arterials to facilitate inter-superblock pedestrian travel.

    But what happens when these societies reach their wealth tipping-points and their DOTs, under the incessant pressure to value engineer and rededicate resources to the rising crush of cars, stop putting in these meager pedestrian connections altogether?

    The earliest first world superblock proposals looked so seductive on paper, partly because they contained so much elaborate pedestrian infrastructure exquisitely protected from cars. But public finances are always limited - especially in the developing world! - and it wasn't long before the pedestrian accommodations in first world superblocks vanished. How long before a financially-strapped developing world, under the rising pressure to spend on cars first, does the same?


Note: Only a member of this blog may post a comment.