Monday, September 02, 2013

An alternate view of sprawl

I sorta had an insight this week I thought I'd share with my readers and see what you think.  Sprawl is pretty universally characterized as a bad thing - that we used to have these tight, dense urban cities that then exploded with car-oriented suburban sprawl.  But I think that perspective needs to be re-framed.

Nothing is universal, but traditionally Americans have preferred small, even rural, communities.  Small town America has always appealed to our psyche.  In the 19th-century agricultural age, our economy needed those small towns spread pretty evenly across the country.  Then in the early to mid 20th-century industrial age, the population coalesced into more medium-sized towns oriented around factories.  Factories were still pretty widely spread, and their economics work best when they have a loyal workforce with modest wages and low turnover - thus a preference for small to mid-sized towns where they could be the employer-of-choice (vs. bigger cities where their employees might always be quitting to earn another couple bucks an hour at another employer down the road).

Next came the information age, and the economic shift from manufacturing to services.  All of a sudden, clustering together with others was very important.  With the economy concentrating, it made a lot more sense to live in a small community with easy access to a big one (i.e. a suburb), rather than an isolated small community.  People get the best of both worlds: they live in a small community, but then get access to the big city amenities when they need them, like airports, museums, sports teams, commercial and retail services, and, most importantly, employers and customers.

What I'm saying here is that it's not so much that cities sprawled out, but more that rural/small town population was drawn in towards cities as we moved from the agricultural to the industrial to the information age.  Rather than cities growing out, the national population of small communities coalesced around major metros - the population clustered together from being more decentralized to more centralized as the economics of that became more important.  It's not so much that people rejected living in urban cities to live in the suburbs, but that they rejected living in isolated small towns to live in the suburbs around big cities.   And, btw, those small towns have pretty much always been car-oriented, no matter where they were located.

If you could animate a population map of the country over the last century, what you'd see is the population slowly coming together and concentrating around the big metros (while the total population also dramatically increased, of course).  This is the opposite of the traditional sprawl story, which says that we've been decentralizing.  Yes, many people wish that those suburbanites would make the full move to the urban city (and many are, especially the youngest generation), but in any case we're better off having them in the suburbs than in isolated small and medium sized cities, like they used to live.  Sprawl is an improvement over traditional population distributions, not a regression.

Thoughts appreciated in the comments.

Labels: ,


At 9:06 AM, September 03, 2013, Anonymous Anonymous said...

Interesting theory, but it does not seem to be backed up by any of the census data, which shows a hollowing-out of the population in urban cores at the time of the development of suburbs, or with other data showing a movement from northern centralized urban cities to sunbelt suburbs. How do you square your theory with the data?

At 9:33 AM, September 03, 2013, Blogger Tory Gattis said...

Oh, there certainly was also some movement from the urban cores outward (which is mostly reversing now), but I think it was relatively small compared to the overall national consolidation. It would be interesting if someone could find the stats on what percent of the U.S. population was in the Top 20(?) metros over the last century.

At 4:09 PM, September 03, 2013, Anonymous awp said...

Here is the urbanization rate of the United States.

At 4:21 PM, September 03, 2013, Anonymous Anphang said...

So, I definitely agree with a thesis of demographic convergence with the national average occurring in the country's population centers, motivated over time by different technologies and economies. And I honestly hadn't ever thought to compare and contrast the impact of cars on dense/large metros versus rural/small towns over time - that'll be some interesting fuel for my wiki-trawling over the next few weeks!

My thoughts: I've always felt the divide on development and density issues wasn't so much Montrose vs. The Woodlands as it was the Tomballs of the country versus the Liberties, or maybe even versus the Lumbertons: places that fall just inside or rather outside of a big metro area like Houston's sphere of influence, or that are just as close, but to a smaller economic cluster, or one that isn't as vibrant. So, having someone else acknowledge the distinction is nice. :)

That being said, I would add an emphasis on form to your thesis. Barring some kind of corruption or incompetence, growth in tax base solves or puts off most problems.

But the details of a connected street grid, liberal zoning ordinances, a good street interface, transportation capacity, an integrated metropolitan economy, regional cooperation - some "details", I know! - are what determine what happens when the growth levels off, be it in a single neighborhood or in an entire city. I'd wager that the City of Detroit would be much better off than it is today if it didn't have an infamously awful relationship with its suburbs. Or, closer to home, that DART would be more efficiently and effectively laid out if some of Dallas's biggest job centers weren't incorporated into neighboring suburbs quite opposed to paying in. I'd also wager that growth in Austin would be far more equitably apportioned to East Austin if a double-decker highway hadn't discouraged people to go east from downtown than north, south, or west, until very recently.

At 4:31 PM, September 03, 2013, Anonymous awp said...

also, I think what you want to see is Austin Contrarian's weighted density for the U.S. by census year.

At 4:34 PM, September 03, 2013, Blogger Michael said...

Sprawl and urbanization both exist side by side. Which one is having a greater overall effect is up for debate, but sprawl certainly exists and is not to be confused or conflated with general trend toward urbanization.

For an example of sprawl, you only need to look at the population of St. Louis, a metro area that continues to grow today (albeit much more slowly than Houston) - but whose population in the core city was 850k people of roughly 1.4 million people in the metro area as of ~1950, yet in 2010 had roughly 300k people of the roughly 3 million in the metro area. So - while the metro area has more than doubled in population over the past 60+ years, the core has lost 60-70% of its population. (Sources: wikipedia entries for St. Louis and greater St. Louis

And no - it is not "mostly reversing" - the city is still 500k people below 1930 levels. I think to some extent you see this pattern in every major city that existed and experienced a great deal of growth before the automobile - if not in losing outright population, then by losing the core share of the population within the metro area that it held before.

This fact may be lost on Houstonians since Houston did not really boom until after the automobile so Houston does not share this trajectory. But imagine - if Houston had followed this same trajectory something like 1 million people inside the western portions of loop 610 by 1940, and now 350k. (As it is downtown Houston has a pretty pitiful population of something like 10-20k. But expecting that to grow soon as new developments get off the ground)

Has greater St. Louis continued to urbanize and grow? Yes. Has it sprawled just as most every other major American city that boomed before the prevalence of the auto / air conditioning? Yes. But they are two separate things.

At 8:21 PM, September 03, 2013, Blogger Jackalope said...

Regarding your blog post of September 2, "An alternate view of sprawl", you make the argument that so-called sprawl is less a result of people moving out from central urban areas and more a result of people moving in from outlying rural areas (just not all the way into the center), which leads you to the justifiable conclusion that urban sprawl is an improvement over the more diffuse urban-rural mix that came before. I agree with you that in-migration from rural areas to urban areas probably is an under-appreciated mechanism for the growth of urban areas, but I think that giving one primacy over the other or over other mechanisms (such as sheer population growth) is an open question, especially when viewed over a long expanse of time.

Certainly, many rural areas and isolated communities have been losing population, a process that must have begun at the start of industrialized agriculture with the pre-war (Civil War) invention of the mechanical reaper. However, without the facilities of a farm to grow and make much of one's own needs and without facilities to stable horses, people had to rely on their own feet for most day-to-day travel, and the result was dense towns. With the growth of large-scale industrial enterprises, the follow-on result was extremely dense cities.

The pre-war (World War I) development of affordable, easy-to-care-for (relative to a horse), transportation in the form of the automobile provided extraordinary flexibility in personal travel options. The subsequent pre-war (World War II) advent of universal telephone service provided by the Communications Act of 1934 provided extraordinary flexibility in personal communication. With access to these technologies, the common person had the wherewithal to live just about anywhere that person wanted. However, the resulting impact was delayed roughly a decade-and-one-half by the Great Depression and World War II.

Of course, after the war (WWII), many phenomena took effect almost all at once: the automobile, the telephone, economic growth, and population growth. As you say, people continued to move in from rural areas, but people left dense urban areas (e.g. Pittsburg, Detroit), also. Furthermore, more people needed places to live; the population of the United States more than doubled from 1950 to 2010 ( All these people had to go somewhere, and many of them populated suburbs.

I agree with you that Americans want to combine the traditional preference for small, sparse communities with access to amenities that come with large, dense cities. They have seen the big city and do not want to (or cannot) return to the farm; however, the city is too much. Suburban communities are a means of achieving a balance, a "thesis-antithesis-synthesis" result. They are a compromise, and no compromise satisfies everybody; some will find them too sparse and too far from the "real city" while others will find them too dense and too far from the "real country". However, so many people live in suburban communities that they must be providing an acceptable balance to many.

I do think that urbanists should refrain from seeing suburbs and suburbanites as "enemies" to be vanquished or "brought into line" with a vision that many urbanists think is "best"; this turns into a kind of religious war that cannot have a good end. I think that urbanists instead should see suburbs as a resource that adds to the flexibility and lifestyle diversity of metropolitan areas, thereby evolving cities past a model that existed from the dawn of civilization until the opening of the 20th century. That is change; that is life.

At 9:19 AM, September 04, 2013, Anonymous Anonymous said...

For those of us who connect sprawl with immigration, it doesn't follow from your interesting thumbnail history of domestic migration that there is an ongoing American referendum on the desirability of urban sprawl; less still, that that debate was ever settled. Except by fiat, perhaps.
(Another Anonymous)

At 3:14 PM, September 05, 2013, Anonymous Anonymous said...

"And, btw, those small towns have pretty much always been car-oriented, no matter where they were located."

Not really, especially not when you set your timeframe earlier as the founding of America. Cars are post-war thing, and many pre-war small town streets were widened ("improved") to meet post-war car-first street standards. I highly recommend the Strong Towns blog by fiscal conservative midwesterner Chuck Marohn for more on this.

I think you could do with defining your terms, or sharpening up your own internal definition/understanding. There's no problem with metro areas growing: it's a truism that big cities have a bigger surface area than small towns, but nobody says most of Paris or London or New York are sprawl. The edges are probably ugly, but it's not like they have a little downtown surrounded by tract housing and strip malls. They're an agglomeration of multiple complete communities.

Sprawl is about the use of the land, whether it's paved and empty or whether it has a building on it, and what form that building takes. Residential suburban sprawl is taking the suburban form of single family housing and arranging it on an unwalkable car-first dendritic street network, as opposed to a high-intersection density network of pedestrian-first thin streets. Commmercial sprawl is boxes in a sea of parking.

Take a look at these images . T3 is the suburban housing form, and you'll see from the second image of "community units" that they recommend three transects per five minute walk. This is obviously just a rule of thumb - Paris has T5 for ages - but the point about suburban sprawl is that it has something like the T3 built form for ages, on a disconnected street network, and with a total ban on stores, workplaces etc. within walking distance of residences.


Post a Comment

<< Home