Walkability

Posted by on . 20 comments.

Atrios writes:

Suburban development is inevitably going to be automobile-centric….However, being automobile-centric and being designed in a way which almost entirely excludes the potential for other modes transportation are very different things. The car and the light rail can coexist. Sidewalks can run to areas with retail. One could even allow a corner store and a pub within a residential neighborhood! Maybe, just maybe, there can be small corridors of street level retail without giant parking lots, small town style. Places like this do exist, mostly but not just in older suburbs.

And Kevin Drum responds with a description (and map!) of his neighborhood:

I live in a subdivision of Irvine, California, called Woodbridge. It’s actually fairly famous as one of the original master planned communities of the 60s, and believe you me, it’s master planned to within an inch of its life. This has its drawbacks (lots and lots of beige houses), but there are also benefits. The main one is that it really was planned as an integrated community of sorts.

To get an idea of what I mean, here’s a Google Earth picture of Woodbridge. It’s the piece inside the yellow oval loop plus the strip just outside it, and the total population is about 30,000. There are houses and apartments on the north and south, with the central section reserved mostly for shopping, churches, schools, medical offices, parks, and so forth. There are sidewalks everywhere, of course, and also bike lanes.

The central section is actually pretty handy. There are six separate areas designed for shopping (outlined in red), and those areas include four supermarkets, a couple dozen restaurants, three department stores (though one is shutting down), a bookstore, two movie theaters, two drugstores (with one more about to open), several banks, a hardware store, two Blockbusters, and lots of other miscellaneous shops. Every single one of these places is safe, easily accessible, brightly lit, and a maximum of 1.5 miles from every single point within Woodbridge. Short of being downtown, this is about as walkable as it gets.

And walk it I do. All the time. (This isn’t out of environmental altruism, it’s because I shop for food daily as a way of forcing myself to get out of the house and get some minimal exercise.) And here’s the thing: aside from occasional dog walkers, I have the place to myself. Despite the fact that it’s about as pedestrian friendly as a suburb can be, nobody walks anywhere. They don’t bike either — the only cyclists I see are biking for exercise. Woodbridge is, as near as I can tell, about 99.9% car-centric despite having a design that’s about as pedestrian friendly as you’ll find in a suburb.

Like I said, I don’t have any big axe to grind here — except to say that as important as pedestrian-friendly design is, it’s also possible to overstate that importance. Something more has to happen to reduce our dependence on cars. Maybe the price of gas just needs to double a couple more times. Maybe better mass transit is the key. Maybe something else. But here in Woodbridge, anyway, we built it and they did not come. Not on foot, anyway.

Let’s start with a few basic points. One, driving is easy. If you’re watching television and a commercial comes on and you want a bag of chips from the store, and the store is a half mile away with a big parking lot, you’re going to get in the car and drive. Distance+parking lot=drive, even if the sidewalks are studded with diamonds.

Two, Woodbridge, as described by Drum, packs about 30,000 people. Based on Googlemap distances, that works out to a population density of somewhere in the neighborhood of 10,000 to 13,000 people per mile. Compare that to walkable neighborhoods in the District like Georgetown and Adams Morgan, where population densities are a bit over 20,000 people per square mile. And those places are built at nothing like the density of downtown, or of a place like Manhattan (where pop. densities are something like 80,000 people per square mile).

Even so, Brookland (my hood) is fairly walkable in places (and is used as such) despite having a population density considerably lower than Drum’s Woodbridge. The difference is mixed-uses. It’s common to have retail directly abutting the residential properties in Brookland. The farther you get from the main drag the less this is true (and the fewer pedestrians you see), but there are parts of the neighborhood which are pretty well designed. Of course, Brookland is an old streetcar suburb.

The point is this–walkability isn’t just about ample sidewalks, it’s about design. If you build a lowish density neighborhood, separate the homes from the retail, and surround the retail by roads and parking lots, well, you’re not going to get walkers. If you try to make the development more like a small town, however, with residences over retail on the main strips, distributed retail throughout the neighborhood, and a fairly compact design, then you can get real walkability, to a certain extent, even with single-family homes and yards. And you can also follow the model in New England (and old England, actually) and drop that development onto a commuter rail line or transit connection, and then you’ve reduced driving still more, even as most or all residents of the neighborhood own and frequently use cars.

It is possible.

Comments

  1. Dave Murphy says:
    September 10, 2008 at 12:09 am

    I did a post about this recently as it pertains to low income housing in the suburbs, as did Just Up the Pike.

    Packing garden apartments in isolated clusters with strip malls nearby may be dense-ish, but it’s not walkable, and for low income housing, it’s not exactly opening the door to the folks who live in places like White Oak.

    Kevin Drum fails to mention anything about pedestrian safety and ease of street crossings. My neighborhood in Laurel is set up so that many homes are geographically close to retail, but if it means walking along (or God forbid, across) Route 1 to get there, forget it.

  2. monkeyrotica says:
    September 10, 2008 at 6:50 am

    All the more reason why dense transit corridors like Route 1 (MD and VA side) need streetcars. You used to be able to take a streetcar from Mount Vernon to the foot of the Capitol, transfer near Union Station, and take another streetcar to Annapolis. It’ll never happen again, but if it did, you’d definitely have more dense development along that transit corridor.

  3. BeyondDC says:
    September 10, 2008 at 10:50 am

    I have long contended that for good urbanism you need three Ds:

    DENSITY to provide users, DIVERSITY to give users places to go (mixed use), and DESIGN for walkability.

    If you only have 2 out of the 3 Ds then mediocre urbanism is the best you can hope for.

  4. BeyondDC says:
    September 10, 2008 at 10:51 am

    Transit, by the way, supports all the Ds, making them easier to accomplish, but isn’t strictly necessary on its own terms.

    Transit is a tool to create good urbanism, not necessarily an end in and of itself.

  5. Laurence Aurbach says:
    September 10, 2008 at 11:42 am

    Drum has some misconceptions about walkability. 1.5 miles is not a distance most people are going to walk for everyday shopping and activities. Bicycle maybe, but not walk, no matter how ample the sidewalk. One-quarter mile is a more realistic distance, give or take some, depending on a variety of factors like weather, hills, demographics, etc.

    Design does make a difference for pedestrian safety, comfort, convenience and interest. Hiking through acres of parking, or past block-long blank facades, is no one’s idea of fun. In commercial areas, active facades are correlated with greater pedestrian activity. Woodbridge also has low street connectivity (lots of dead ends), which makes neighborhoods less walkable in variety of ways.

  6. Dan Staley says:
    September 10, 2008 at 12:07 pm

    Good planning programs should study SoCal’s development, and esp Drum’s area. Mine did.

    Point being, there are many examples out there to build off of. Do we do it? No. Why not? is the question and the one to be worked on, not whether 12 DU/ac should have granny flats and 3.25 pocket parks/1000. That is: stop obsessing over tiny details and get the d*mn code so that something like this can happen in the first place.

    [/pet peeve]

  7. Mixner says:
    September 10, 2008 at 2:09 pm

    Presumably, Washington DC is, overall, a far more “walkable” city than Houston, Texas.

    Yet Ed Glaeser found that households in Washington do almost as much driving as households in Houston. Or, at least, that the former produce almost as much CO2 from driving as the latter. With respect to combined CO2 emissions from driving and transit, the two cities are virtually identical.

    This supports Drum’s view that just because a place is more “walkable” doesn’t mean people who live there will do more walking. Or, at least, that they won’t do any less travelling by motorized transportation.

  8. Cavan says:
    September 10, 2008 at 2:55 pm

    You have to remember what Ed Gleaser defines as Washington, DC. He defines it as the 5 county census region. As we know, the outlying areas of those five counties contain some hideous sprawl.

    Perhaps a better question would compare inside the Beltway to inside some comparable perimeter in Houston. Or perhaps inside a 10 mile radius of Metro Center to the corporate limits of Houston, since Houston has been able to annex its inner suburbs. I think the answer would be quite different.

  9. Mixner says:
    September 10, 2008 at 3:16 pm

    Glaeser also estimates the difference in CO2 emissions between suburban and inner-city households for each metropolitan area. The difference is larger for Washington than for Houston, but only moderately so.

  10. Richard Layman says:
    September 10, 2008 at 4:13 pm

    Ryan has problems with Ed G. So do I. I haven’t looked at the paper, but there is no way that his estimates can be correct. See this diagram for some reasons why:

    http://www.flickr.com/photos/rllayman/107880025/

    Given that energy use comprises most of the CO2 emissions we generate per capita, how, if the urban average of BTU use is 7/12 of the suburban average, could Glaeser’s estimates be correct? Plus, a region like Washington has far more transit usage than Houston, which should reduce BTU usage significantly at the core of DC.

    2. Monkeyerotica, streetcars to annapolis would be too slow… BeyondDC in the past has suggested railroad service. I incorporated that idea into a paper on creating a regional (DC-MD-VA) railroad system as opposed to having two center city commuting oriented railroads.

  11. Mixner says:
    September 10, 2008 at 5:12 pm

    Richard Layman,

    Glaeser’s findings are presented in his paper The Greenness of Cities published by the Kennedy School of Government at Harvard. He provides detailed references to his data sources from the Census Bureau, the Department of Energy, etc. and detailed descriptions of the methodology he used to calculate the figures he presents.

    I have no idea where the figures in your flickr image come from. There are no references to data sources and no description of methodology. The only attribution I see is “Jonathan Rose Companies LLC,” whatever that is.

  12. Mixner says:
    September 10, 2008 at 5:14 pm

    FYI, here’s a link to the Glaeser paper:

    http://www.hks.harvard.edu/rappaport/downloads/policybriefs/greencities_final.pdf

  13. Dan Staley says:
    September 10, 2008 at 6:47 pm

    I’m a Kahn fan but not a Glaeser fan.

    At any rate, DC rates #32 in the aforementioned paper (that Matt pointed to on his blog some time ago), Houston 62 out of 66. I’m not sure why Mixner would cite this paper as supporting his assertion. Note also Table 2, that supports the above assertions regarding suburb-city differences.

  14. Laurence Aurbach says:
    September 10, 2008 at 6:58 pm

    This supports Drum’s view that just because a place is more “walkable” doesn’t mean people who live there will do more walking.

    Many very solid studies have found that walkable neighborhoods do in fact have more walking, less greenhouse gas emissions and less driving. Sometimes much less driving.

    For example, the Urban Land Institute book Growing Cooler reviewed dozens of studies and found that,

    Depending on several factors, from mix of land uses to pedestrian-friendly design, compact development reduces driving from 20 to 40 percent, and more in some instances

    Another recent study is by the Transit Cooperative Research Program is TCRP 128: Effects of TOD on Housing, Parking and Travel. It looked at the actual transportation performance of 17 transit oriented developments and found driving was 50 percent lower than the standard traffic engineering models had predicted.

  15. Mixner says:
    September 10, 2008 at 8:00 pm

    Laurence Auerbach

    Households in “walkable” cities do appear to drive less than households in “sprawl” cities. But, in general, not a whole lot less. As reported by Glaeser, CO2 emissions from driving are 24,992 in Washington vs. 26,294 in Houston. And households in “walkable” cities tend to emit more CO2 from public transit use than do households in sprawl cities, which narrows the differences from driving. Transit is often no cleaner per passenger-mile of travel than driving.

    Overall, the difference in household CO2 emissions attributable to transportation between old-style “walkable” cities and new-style “sprawl” cities just isn’t very large, and could be completely offset by a modest increase in average automobile fuel-efficiency. See the sample in Glaeser’s Table 1.

  16. Mixner says:
    September 10, 2008 at 8:14 pm

    At any rate, DC rates #32 in the aforementioned paper (that Matt pointed to on his blog some time ago), Houston 62 out of 66. I’m not sure why Mixner would cite this paper as supporting his assertion.

    I explained why. And citing the rankings is misleading, because those rankings refer to average total household CO2 emissions, not average transportation CO2 emissions. And also because the rankings cover only a relatively small range. Yes, Washington is at #32 and Houston at #62 in average total household CO2 emissions. But the difference in CO2 emissions cost is relatively small: $1,180 for Washington vs. $1,334 for Houston.

    And essentially all of that difference comes from differences in residential heating and electricity use, which seem to be primarily driven by differences in climate. Virtually none of it is differences in CO2 emissions from transportation.

  17. Dan Staley says:
    September 10, 2008 at 10:07 pm

    The paper is Green Cities, not Green automobiles. If you are a sprawling city in a hot, humid climate, one expects your city to use much more resources to create habitable microclimates in boxes and to move around by automobile. One would also expect, in a better world, that policy-makers would take steps to decrease the carbon footprint, knowing you are a sprawling city in a hot & humid climate. Nonetheless,

    When one looks solely at MSAs for Houston vs DC, one finds that the HOU VMT per capita is 36, whereas the VMT for the DC MSAs is mid-20s, and the public transportation gap can be, as you say, completely offset by a modest increase in average transit fuel-efficiency (Table HM72 – Urban Areas Selected Characteristics 2006).

  18. Mixner says:
    September 10, 2008 at 10:38 pm

    One would also expect, in a better world, that policy-makers would take steps to decrease the carbon footprint, knowing you are a sprawling city in a hot & humid climate.

    Depends on the nature of the steps. Of course, there are lots of things policymakers can do to help reduce the carbon footprint, such as promote the development and sale of cleaner and more fuel-efficient cars, promote carpooling and telecommuting, promote more energy-efficient residential appliances and heating/cooling systems, and promote cleaner sources of electricity.

    Given the apparently modest potential for reducing carbon footprints through realistically-achievable alternative forms of urban design, the huge practical, political and economic obstacles to large-scale changes in current laws, practises and consumer behaviors, and the timescale required to achieve significant benefits, I think focusing on making housing greener (rather than smaller and denser) and making cars cleaner and more fuel-efficient is a much more promising way to address the issue of CO2 emissions.

    Your second paragraph is a bit mystifying. What document are you referring to?

  19. Dan Staley says:
    September 11, 2008 at 11:05 am

    Your second paragraph is a bit mystifying. What document are you referring to?

    This one.

    think focusing on making housing greener (rather than smaller and denser) and making cars cleaner and more fuel-efficient is a much more promising way to address the issue of CO2 emissions.

    Large yards contribute negatively to the Urban Heat Island and to eutrophication of receiving waters. This is well-known.

    Nonetheless, allowing the market to determine an optimum level of density in a particular time period is better than stating density is unnecessary. Of course there is a segment of the population that wants density. Why prevent them from having it (via single-use zoning, for example)?

  20. BruceMcF says:
    September 14, 2008 at 2:12 pm

    Laurence Aurbach Says:
    September 10th, 2008 at 11:42 am

    Drum has some misconceptions about walkability. 1.5 miles is not a distance most people are going to walk for everyday shopping and activities. Bicycle maybe, but not walk, no matter how ample the sidewalk. One-quarter mile is a more realistic distance, give or take some, depending on a variety of factors like weather, hills, demographics, etc.

    Precisely. This is why in retrofitting suburbs around stops on dedicated transport corridors, I’d call for the mixed-used easement permitting ground-floor sidewalk commercial and professional, with shop-top housing, to be a 1/4 mile radius around the stop, and the residential infill-development easement, the easement for stacked twin townhouse, to be a 1/2 mile radius.

    Given the existence of those suburban village centers, there’s a two to three mile bike-and-ride radius around the stop, without any redevelopment required to the existing sprawl suburban development.