We know that there's a strong correlation between urban density and energy efficiency. Communities packing 12 or more households per acre are more efficient than less-dense communities built with the latest Energy Star appliances and materials. When planned and executed well, high-density residential areas can be appealing even to those reluctant to give up the space of single-family-home suburbia. When high density is the result of lack of planning or poor decisions, however, the result can be bad -- very bad.
The article in Tuesday's Washington Post talking about urban density patterns has generated a bit of attention because of its seemingly counter-intuitive ranking (based on US Census data) of the Los Angeles metropolitan area as having the highest density of any urban region in the US. Indeed, this is surprising to those of us accustomed to thinking of density as meaning skyscrapers and closely-packed townhouses. But the story that leapt out at me from the numbers was just how close the San Francisco-Oakland metropolitan region was to LA in terms of density. Based on the 2000 census, Los Angeles has 7,068 people per square mile, while SF-Oakland has 7,004 people per square mile.
Flying into Los Angeles International Airport at night, it's easy to believe that density figure. There are points during the approach when, from thousands of feet in the sky, all an airplane passenger can see is a sea of lights stretching from horizon to horizon, hundreds of square miles of human occupation, stopped only by the Pacific Ocean to the west. From the ground, during the day, it doesn't have quite the same character (except perhaps when trying to drive the 405), due to the "polycephalic" character of the city. Los Angeles doesn't have a single center or downtown, it has dozens. San Francisco, conversely, has the trappings of the conventionally dense city: strong core, abundant pedestrian and mass transit traffic, mixed-use neighborhoods and residential streets with houses pushed (sometimes literally) right up against one another.
Los Angeles and San Francisco, neck-and-neck at #1 and #2 on the population per square mile list, represent two very different models for the urban density future -- and both versions face important challenges.
I've lived most of my life in one or the other of these two metropolitan regions, and feel very comfortable and at home in both. There are aspects of life in each that I find marvelously appealing; the diversity and abundance of culture and activity in LA, the street life and complexity of San Francisco. (Perhaps that's why I find London so attractive -- the sheer volume of Los Angeles coupled with the traditional core density of SF.) I offer this aside not to personalize this post so much as to ground my observations: these two locations are very much part of my world.
As the Post article discusses, the engine of the Los Angeles density model is "infill." Because Los Angeles cannot, for all intents and purposes, expand its geographic area, its relentlessly growing population settles interstitially, whether in quick developments on previously waste-strewn areas (such as Signal Hill, which required bioremediation efforts to make the land clean enough to build upon) or simply by packing more people per house. The latter is startlingly common phenomenon in the poorer, and largely Hispanic, parts of Los Angeles. Garages are converted into tiny homes; "single-family" houses come to hold two or even three families. More traditional high-density housing can't be built in volume, as there are few places left upon which to build, and tearing down homes to put up apartment and townhouse complexes simply removes residential space even as more people arrive.
But if the legendary Los Angeles sprawl transformed into a high-density setting, the infrastructure has not kept up. Roads, water and sewage systems, schools, parks and myriad day-to-day services are hard-pressed to provide for a population many times that expected by planners in the 1950s and 1960s. And trying to build new services faces the same unpleasant dilemma as trying to build more housing: construction of new schools or businesses (let alone parks) means tearing down inhabited homes. The Post notes that a single elementary school can require 200 houses be destroyed -- and given the residency patterns, this can mean far more than 200 families displaced. This helps to understand the greater traffic problems in the region -- if businesses can't be built near residential communities, the citizens will have to travel to the businesses to work.
This hyperdense sprawl clearly came as a surprise to planners, and Los Angeles was caught flat-footed in its response. There was no planning, just reaction, and often not even enough of that. And while Los Angeles is the first major region to see the infill density engine kick in, it won't be the last: any metropolitan region facing geographic limits to continued sprawl but continuing to see a population influx will start to see the dense sprawl that Los Angeles now experiences.
San Francisco-Oakland faces a different problem, albeit with a similar underlying cause. Like LA, the SF Bay Area region continues to see a net gain in population (SF lost population shortly after the dot com bubble popped, but has since started again to grow). Unlike Los Angeles, however, SF-Oakland doesn't have a significant lower-density sprawl in which to "infill," either with new construction or multiple families in one home. Instead, the continuing population demand without significant increase in supply means that housing prices in the area have skyrocketed. The median sale price of a single-family home in San Francisco stands at over $700,000 currently, up from just over $560,000 in 2003; Alameda County (in which Oakland is the biggest city) has a median price of $620,000.
Where is everyone going who is priced out of the SF market? The US Census statistics give one clue: outside of LA, SF and San Jose, the remaining three California cities in the urban density top ten -- Davis, Vallejo and Tracy -- are all distant "commuter" cities in the broader San Francisco region (each about 60 to 90 minutes away by car... if there's no traffic.) And infill is happening here too, with the suburban cities between SF/Oakland and these distant commuter centers seeing both rapid sprawl and increased prices. If the population of the SF region continues to grow -- and no planning is done -- eventually these distant low-density communities will start to see the "dense sprawl" effect hit, too.
If these are the two models for very-high-density urban environments, what can we do?
Both the LA infill density and the SF surreal estate density models beg for planning relief, but they ask very different planning questions.
One very interesting aspect of this dilemma is that many of the more obvious steps that would retrofit sprawl for density -- multi-family housing complexes, better mass transit, higher-efficiency systems for water and power -- are also those needed to make these communities more closely resemble the "bright green" urban ideal.
This question is not easy to answer, however, as the "more obvious steps" can also be among the more costly steps. This, in turn, has political implications. As useful as efficient mass transit may be when serious infill density hits, for example, the limited use it would see early on would make it unable to pay its own way, and lead to questions about the wisdom of the endeavor. Not many political figures would have the guts to hold on in such a situation.
Solutions here are, if anything, even more difficult than in the LA model. One possibility that comes to mind is the encouragement and development of traditional high-density centers in the peripheral area outside of the city. To an extent, this has happened with Oakland, and arguably was a trigger for the growth of San Jose; there are few if any other cities in the greater SF area that are growing in a traditional high-density model. (This concept has some relationship to the "Edge City" model proposed by Joel Garreau in his book of that name.)
Such a solution would be risky for the original core, however, as businesses and taxpayers may end up moving out of the city to the new location, as it presumably offers similar -- or better -- amenities at a lower price. While this has a positive effect on prospective residents (it decreases demand, and lowers prices), it can have a starkly negative effect on the ability of the core city to support the services its citizens demand. Cutting services and raising taxes would both further contribute to the flight.
So, what are the solutions? How can we build higher-density, more efficient, and still attractively livable cities without triggering either the LA "infill density" problem or the SF "hyperexpensive density" problem? Are there examples of high-density, high-efficiency, livable cities that have managed to avoid both fates?
If you live in a city that hasn't quite reached these heights of density, but is well on its way, what can your city do to avoid either the Los Angeles or the San Francisco model?
Can you turn a city into a Bright Green Urban Dream and break neither the bank nor the people?
Comments (13)
Does this energy efficiency take into account the energy costs of transporting food and energy into an area instead of producing them locally? That is, is this compared to rural areas or is this an increase in efficiency when compared to other suburban and urban areas?
With a few acres of land per household, you could produce most of your veggies locally, keep a chicken coop, irrigate using a local well, and have a solar or wind farm for at least part of your energy needs. A large number of people in the US lived this way well into the 1940s, using wind to pump water and solar to heat water.
Posted by jet | August 12, 2005 10:13 PM
Posted on August 12, 2005 22:13
I know the science fiction solution to the density problem: arcologies.
http://en.wikipedia.org/wiki/Arcology
Unfortunatly they fail your Bright Green Urban Dream because they break the bank :-(
Posted by Eric Boyd | August 12, 2005 10:19 PM
Posted on August 12, 2005 22:19
I don't know of any cities, but I'd be interested how places like Chicago, Toronto or NY fall on the spectrum according to you, Jamais.
And where do cities in Europe stand, such as london?
Posted by andrew | August 12, 2005 10:20 PM
Posted on August 12, 2005 22:20
There are ways to live at high density without living in a large city. Look at most traditional farming villages in places like Spain, France, Italy, Greece, and so on.
There are ways to live at high density without high-rise buildings. Many cities, especially in Middle Eastern countries, build 3 to 5 story courtyard buildings that achieve densities comparable to skyscrapers. One difference, however, is that a lot of the open space is privately held within building compounds.
If you still have a 1-hour commute from home to workplace, a lot of opportunity for energy efficiency is lost. Dense, *mixed-use* settlements work best.
Posted by David Foley | August 13, 2005 7:23 AM
Posted on August 13, 2005 07:23
Metro area statistics like this aren't very useful for determining the actual character of urban places. They are mostly an artifact of how boundaries are drawn and how the Census Bureau defines "urbanized."
For instance, the urbanized area densities of L.A. and S.F. work out to 11 persons per acre, or around 4 dwellings per acre (at the average occupation rate of 2.5 persons per dwelling). You can see what that looks like here, here and here. Swimming pools and two-car garages -- not exactly what I would call "hyperdense."
According to the Trust for Public Land, New York City has more parkland than any other U.S. city "with more than one-quarter of its land as protected public open space." Obviously, factors like that affect the overall statistics in a major way.
Even so, the density of Kings County (Brooklyn) is 35,000/sq mi. with development generally of detached houses, rowhouses and low- to medium-rise apartments. In comparison, the density of Los Angeles County is 2300/sq mi. Even in downtown and south central L.A., the densities are around 15,000/sq mi (and 7 dwellings/acre), less than half of Brooklyn's. See gCensus for easy-to-navigate statistics.
When it comes to the costs of infill and suburban retrofits, it's important to be aware of the current web of regulations, policies, subsidies and inertia that comprise the U.S. system of real estate development. The result is the comprehensive directing of growth into suburban patterns and locations. It's not just federal, state and local governments that play a role in this, but the finance and insurance industries, professional associations, political/activist groups, and NIMBY residents.
Posted by Laurence Aurbach | August 13, 2005 8:23 AM
Posted on August 13, 2005 08:23
http://www.pbs.org/newshour/newurbanism/index.html
New urbanism seems to take a lot of these issues into account. After peak oil, many think we'll naturally head along those lines of thinking. Anywa, see what you think :)
Posted by Syniel | August 13, 2005 8:28 AM
Posted on August 13, 2005 08:28
Jamais,
Your post focused on design and planning solutions to these issues rather than governmental. In Pittsburgh, where I live, we are having a very different problem of dramatically decreased density resulting from falling population and continued sprawl. The proposed solutions, rather than focusing on urban design and planning, are looking at the structure of municipal governance and taxation in Allegheny County. (Bear in mind there are 130 separate municipalities in the county) Of course we are also looking at better transit and better development, but governance is *the* core issue.
I think the catch-22 you identified in the competition between edge cities and core cities can only be resolved at the level of intercity government. For more info on the environmental implications of regional government check out sustainablepittsburgh.org.
Eli Pousson
Posted by Eli Pousson | August 13, 2005 9:51 AM
Posted on August 13, 2005 09:51
Jet, it's my understanding that, yes, the "density=efficiency" measurements do take into account energy costs of bringing in food. While dense urban settings may not be as self-sufficient & efficient as the model you describe (growing your own food, generating your own power, etc.), that model isn't common in the US any longer, and hasn't been for quite awhile (it exists, but isn't common). It's also unclear what portion of the population could move into that sort of setting simply in terms of land availability.
Eric, arcologies are interesting from a science fiction perspective, but the real-world arcology movement has, um, issues. Search "arcology" here on WC for some relevant posts.
Andrew, I don't know enough about Chicago and Toronto to make anything other than unsubstantiated assertions. I'm somewhat more familiar with NYC, and -- from what I've seen, heard and read -- it appears to fall into the "SF model" (or, perhaps more properly, SF is the latest manifestation of the NYC model).
London, as well, even though there's tremendous geographic spread to that city. The underlying issue is how much it costs to live in a traditional dense core city, and NYC and London are near the top (SF being slightly more than NYC, apparently, and neither coming close to London).
David, I think you're on to something there; the "dense village" concept may be a good working model. I was a bit surprised to see a post at Cascadia Scorecard on density, however, claiming that "density of employment" is more important than "density of residence" when it comes to efficiency, at least based on Seattle-area numbers. I'm not sure how generalizable that is.
Laurence, I think you're correct here, and note that the density figures in the Post for NYC-Newark also include "Connecticut Suburbs." I wonder how the numbers would have looked without the Conn. 'burbs included. As I mention above, what I call the SF model in the post is likely more properly a NY model (in US terms), as the same general issues -- cost of living, demand for core access, etc. -- apply.
Syniel, there are lots of fans of New Urbanism here, along with a few critics. I'd like to see an updated NU model, though, that has a more forward-looking approach to sustainability (e.g., integrating microgeneration and neighborhood power, better water management, etc.).
Eli, thanks for the info about Pittsburgh. Some very good friends of mine are moving to Pittsburgh later this year, and have been telling me some stories about the municipal structure there. The need for transmunicipal governance is clear; I was going to link this post to an earlier one on "megapolitan" approaches for more detail, but it was getting too long already.
Posted by Jamais Cascio | August 13, 2005 11:40 AM
Posted on August 13, 2005 11:40
The fatal flaw with density is with water and tax money limited as they are you simply cant fuel even the existing big cities much less new ones.
Cities cost LOADS of money and require oodles of water.. we dont have either.
Posted by wintermane | August 14, 2005 12:11 PM
Posted on August 14, 2005 12:11
Jamias, there are quite a few people concerned with an explicitly sustainable model for new urbanism. The LEED-ND committee (half of which is composed of new urbanists) will soon release a draft of neighborhood environmental assessment standards. They'll be building on the work of some pathbreaking developments.
Civano was one of the first thoroughly sustainable new urban projects (although its recent sale to a new developer has brought into question the sustainable quality of future construction there). In B.C., UniverCity incorporates a sustainable vision while Clayton Village (East Clayton) implements innovative stormwater strategies, both aided by designers at UBC. Solar Village Prospect features active and passive solar along with district water heating. Loreto Bay is a model sustainable resort in Baja California. West Hyattsville is a TOD with ambitious Low Impact Development (LID) techniques.
As to Wintermane's comment, studies consistently show that higher-density development has lower per capita water usage. This report by Western Resource Advocates found that if cities in the western U.S. had water efficiency on par with Civano, water savings would range from 52 to 77 percent.
Posted by Laurence Aurbach | August 14, 2005 1:44 PM
Posted on August 14, 2005 13:44
You didnt understand. While the homes may use less water because of course they do have less to water the very simple fact is a city cant get its water localy and no one is giving up thier water for a nearby city these days. In fact they are taking back the water cities stole from them.
So again where are you gona get the water not just for the homes but for everything else a big city needs water... Over a certain density and you go over the limit of water your area can maintain.
And the money.... alot of wanabe new yorks and whatnot are capped simply because they cant get the money localy and no one outside that area will let em have more. Without that money they cant afford the services and systems needed to manage a dense population.
Posted by wintermane | August 14, 2005 9:45 PM
Posted on August 14, 2005 21:45
One thing you got wrong: Davis, number 6 on the Post's list, is not a commuter town for the Bay Area. It's density is home-grown and represents a different model altogether. However, the SF surreal estate market certainly applies there. Prices for what are essentially tacky little tract houses are pushing the million dollar mark. Go Figure.
Posted by veronica snowbank | August 23, 2005 2:21 PM
Posted on August 23, 2005 14:21
Veronica, I live in the area, and while Davis isn't a major commuter city for SF, there is still quite a bit of traffic back and forth from up there every day.
Posted by Jamais Cascio | August 23, 2005 2:29 PM
Posted on August 23, 2005 14:29