How much did highways really matter for suburbanization?

Following up on my earlier post, Dan Klein points my attention to the following piece by Wendell Cox, Peter Gordon, and Christian Redfearn, from Econ Journal Watch.  Excerpt:

Suburbanization has, for a long time, been a trend based on consumer preferences and larger trends, notably rising wealth and transportation and communications improvements (including the highways Baum-Snow investigates). Jackson (1985) finds U.S. suburbanization began at the end of the 19century. Indeed, he refers to “streetcar suburbs.” The 20th century U.S. experience is shown in Figure 1, which shows the percentage of US population living in metropolitan areas, and breaks that percentage down into central cities population and suburbs. The growth of the suburbs relative to the central city is seen well before 1950. Moreover, in the figure the relative decline of the central cities is understated because central cities have been annexing suburbs for many years.

The simple broad narrative is that, by and large, suburban living expanded throughout the twentieth century. Around the world, as incomes rise, people choose the mobility of the automobile; they overwhelmingly prefer the range and choice of personal transportation. As they choose automobility, origins and destinations disperse; and as these disperse, the attraction of the auto grows. It is a self-reinforcing cycle that is facilitated by better highways. But as with most public sector infrastructure developments, these usually follow rather than lead.

This article offers some striking facts.  Before the interstate highway system, the percentage of the U.S. population living in suburbs went from 7.1 percent in 1910 to 23.3 percent in 1950.  From 1950 to 2000 it had a smaller proportional increase, namely from 23.3 percent to 50 percent.

The central city of Copenhagen reached its population peak in 1950 and by 1990 had lost nearly 40 percent of its population; that is comparable to some of the highest losses in the U.S. Rust Belt.  The central city of Paris reached its peak population in 1920 and has lost one-quarter of its population by 1990.

There are many other interesting points in this piece.  I am not suggesting that highways do not matter, but the extent of the influence is maybe not as large as many people think.


And the idea that suburbs wouldn't exist or be popular without government subsidy is also put to test by similar density declines in Europe.

I would also like to point out that the building of commuter rails further and further outside the city most certainly has led to a rapid increase in some area's property values (Salem comes to mind).

Cities are so 20th century. The 21st century will be dominated by suburbs, AROUND the world. The fact that the US got there first is just a testament to the level of development the US reached long before the rest of the world.

I would also like to point out that many suburbs have central business districts in them that form little "downtowns" including boutique shops, restaurants, etc. And they are getting nicer and nicer. Those cookie cutter malls are getting nicer and nicer too. They are becoming pleasant spaces with seating, ambiance, etc. Modern advancements in transportation, telecommunication, and the internet make cities obsolete.

It is funny, city dwellers think suburbanites only like suburbs because they need somewhere to put their burdensome SUVs and city parking is terrible.

Naturally, Suburbanites think urbanites only like cities because they don't have cars and have to take mass transit everywhere.

People select into one or the other based on their preferences, there isn't a right or wrong answer, it is a question of individual values. Personally I like both.

I wonder when the major cities of Beijing, Shanghai, and Hong Kong will reach their peak populations?

That is, of course, assuming that they follow a similar pattern...

If I remember correctly, Wall Street used to mark the edge of Manhattan. Then as transportation improved, New Yorkers started moving up the Island. Horse drawn tram and streetcars preceded the subways. So Chris is right--you trade space and status for a long commute, using the best available means of transport and, since you have money, being a force for improving transport.

There's a certain amount of equivocation going on between city and suburb here. What was a suburb in the earlier period may well be regarded as part of the core city in the later period (or even vice versa). Think of Alexandria and Del Ray. Del Ray was a streetcar suburb of the city of Alexandria (the streetcar ran down what is now the grassed median of Commonwealth Ave). Now they're both either parts of the urban core of the Washington-Arlington-Alexandria MSA, or both parts of an inner suburb of Washington DC, depending on your point of view.

So what's your issue with auto-dependent suburbs, David?

One major difference between US cities and those in most of the "Old World" are that in general you
have higher income people living in the downtowns of the latter, whereas higher income are more likely
to be in the suburbs in the US, although most major cities will have some areas of wealthy people
downtown, with such areas reappearing in the gentrification movement.

Yes Barkley, has it right: In many European cities, living in the suburbs shows a lower status than living in the city. As buildings that used to house apartments become office buildings, prices go up, and those who can't pay are pushed further out. The middle class has to move away from the center in cities like Madrid, because they just can't afford million dollar condos. Nobody goes to live to Tres Cantos or Mostoles because it raises their status.

Compare that to downtown St. Louis, where nobody wants to live and few even want to work at, and the high prices come from suburbs that have good communications and huge luxury single family homes.

ample parking has its dark side, as it imposes huge costs on society through the externalities it causes. In the US, only 1% of all parking (nonresidential) is payed for through parking fees, leaving 99% of parking costs internalized. And in most US cities, parking accounts for about 1/4 of all land area, creating a hugely wasteful system, which drives - and has driven through many decades - the vicious down circle for public transportation.

This seems highly dubious. Homeowners and renters pay for their private garages, driveways and parking spaces in the form of higher housing prices or rents. Employers pay for their employee parking lots, leaving less money for wages and benefits. Businesses pay for their customer parking lots and pass the costs on to the customers. And unpriced public parking is funded through taxes paid mainly by drivers, including taxes exclusive to drivers like gas taxes and vehicle registration fees. Yes, there's obviously some inefficiency, but to what extent are drivers really subsidized for the costs of parking, rather than just paying those costs indirectly?

With respect to public transportation, direct subsidies alone are about 72 cents of every dollar. A bus ride for which the rider pays $1 typically costs more like $4 to provide. And more than that, if you count externalities. I don't think you can seriously argue that subsidies work to favor of driving over using public transportation.

The issue of suburbs β€” besides boredom β€”in which no other means of transport except cars are available is so profound, obvious and widely discussed/debated that I assume you are making an attempt at levity. You are joking, aren't you? :)

OK, so you have aesthetic objections to them. But other people seem to like them. Why do you insist on imposing your preferences on others? I don't want a big blank green lawn (though I'd like one with natural areas and trees), but I don't care if other people want them. I walk around my suburb all the time, and there are buses, although of course they're most useful for commuters on a perfectly regular schedule. But I live much closer to work (which is in the suburb) than all the people who take the bus to the Metro to DC but use more energy commuting than I do.

Cars are not subsidized anywhere near what mass transit is. (Well, mass transit intercity buses aren't subsidized any more than cars.) The gas tax taxes people who drive more, too, though there could be yet fairer taxes.

I know that in Denver, suburbs (coincidentally?) became more popular around the time of forced racial integration in schools. They called it "white flight".

This type of motivation, not exclusively racial but cultural as well, seems important to the debate.

To say that Paris isn't as big as it was in 1920 is only true depending on how you define Paris. The Paris metro region is much larger today than 1920.

The population of the city of Paris (the 105 sq km Ville de Paris) peaked in 1921 and has since declined dramatically. Over the same period, the population of the Paris metropolitan area has grown dramatically. The proportion of the metro area population that lives in the city keeps going down. This has been the typical postwar pattern for cities in both the US and Europe. Here's how Robert Bruegmann describes the postwar population change in the Paris area, in his book Sprawl: A Compact History:

during much of the mid-20th century, as Paris was losing population, the ring of suburbs immediately around it was gaining. However, since the early 1970s even these inner suburbs have started to decline in population and density as the outer suburbs and exurbs have boomed. Between 1962 and 1990, as the city of Paris slipped steadily in population from 2.79 million to 2.15 million, the inner suburbs first gained in population, overtaking the population of the city and reaching just over 3 million by 1975, but then declined again, slipping back to 2.94 million by 1990. During the same period, the outer suburbs witnessed an accelerating growth rising from 1.66 million to 2.62 million. Beyond that an "exterior zone," including the rest of the large Parisian region, the Ile-de-France, with its comprehensively planned new towns, exurban developments and still-rural areas, grew from 1.2 million to 2.9 million. By 1999 the Ile-de-France had nearly 10 million people, meaning the city of Paris accounted for fewer than a quarter of all Parisians. ... the outer Parisian suburbs and exurbs, with their low-density subdivisions of single-family houses, shopping centers, industrial parks, and freeways, function and look increasingly like those of the United States.

Well, whatever you wanna call it, it's unnatural and not market-like to require land owners to build X amount of parking.

This is, generally, what annoys me so much about this entire debate – people acting like subsidies are the only form of government intervention. What about the ubiquitous outright mandates, mandating that a building not rise above two stories, that houses not be attached, that parking be provided, etc? They're not subsidies, but they're definitely intervention in favor of the low density suburban form.

Yes, zoning and land use restrictions are bad, definitely including those. I've said that. Tyler has said that. All the libertarians have said that. But, if you notice, central cities adopt those sorts of restrictions on their own, and as just as much as suburbs. Indeed, if only suburbs adopted those restrictions, that would just make city centers grow much faster, since the suburb growth would be slowed but not the city.

So there's intervention to reduce density, but both cities and suburbs engage in it, and they do it at the local level. I certainly complain about it, as you should, but realize that the cities are doing it to themselves and hastening their own decline. There is a reason to make a distinction between one area being forced to subsidize another, and an area creating its own decline, even if both are bad.

What generally annoys me about the whole debate is people pretending that libertarians are only upset about certain subsides, when I tend to think that they're projecting and focusing on the end they want (aesthetically pleasing to them urban areas) and judging all means by that.

The National Academy of Sciences says greatly increased suburban density wouldn't help CO_2 much at all. The objection to suburbs, as I see it, is really just aesthetic.

I live in a streetcar/early automobile suburb in Denver.
We are all of 4 miles from the Capitol which is very close to zero/zero as we grid-loving people would say.
The trolley ended in May 1950.
My house was built in 1926.
My lot is a huge 5K square feet.

My wife and I have looked at smaller houses (ours is a massive 2500 sqft) to live in. And, we have ruled out apartment-style buildings as we enjoy our gigantic back yard (complete with grill) and dining table (complete with umbrella). We also want a place for our cats to roam and be eaten by foxen.

Benny Lava,

You're wrong. See my response to your comments about Paris. Yes, municipal boundaries sometimes change, but there is no doubt that population has been massively decentralizing. As have jobs. Sprawl is a real phenomenon, not an artifact of changes in definitions.

The suburbanization that started in the 19th century is nothing like the surbubanization of the late twentieth century. "streetcar surburbs" had walkable centers. Queens, the Bronx, Cambridge and Daly City are "streetcar suburbs". Compare that to your vast parking lot sprawl you see today.

Using the 19th century word and confusing it with its 21st century meaning is completely dishonest.

Benny Lava,

Then you don't have a point. Come back when you have something useful to contribute. Tyler is talking about the dispersal of population away from historical urban cores. Not about whether a particular location is formally classified as part of a city or a suburb. That dispersal is typically referred to as suburbanization or sprawl. The fact that some suburbs are denser or more city-like than others, and that city boundaries sometimes expand to include areas previously outside the city limit, does not change the fact that suburbanization is real.

Brooklyn was originally a suburb of New York before it was rolled up into Greater New York. Should it still be considered a suburb? Is it still a part of the decentralization argument?

Yes, it's still part of the decentralization argument. New York's population used to be much more centralized than it is today. The central borough of Manhattan had the largest share of the population. Manhattan's population peaked around 1910, and has been in long-term decline since then. The outer boroughs, including Brooklyn, continued to gain population after 1910. By 2000, Manhattan had fewer people than both Brooklyn and Queens, and only slightly more than the Bronx. Staten Island, which had only a negligible share of the population until around 1900, has grown steadily since then. The change in the population distribution in New York is a clear illustration of the decentralization phenomenon.

Detroit was one of the originators of sprawl, mostly contained within the city limits at the time because it was surrounded by townships that could pretty much be annexed at will, and originally it had nothing to do with highways. The reason was as the car industry grew, they needed larger assembly plants, and the only way to accumulate enough land for the plants economically was to go to the edges of the city. Then the workers moved outward to be near the plants. At the same time, middle class streetcar and rail transit suburbs developed in Oakland County (Royal Oak, Ferndale). Now, arbitrarily, the areas that were annexed would be classed as central city and those that were not would be classed as suburban, but looking at the houses that were built in any period, you'd be hard pressed to tell them apart. Detroit never had much of a high-rise or even mid-rise apartment culture - it's all single-family houses with small yards.

Now, arguably, the building of the freeways in the 40's through 60's facilitated yet more sprawl, but equally responsible would be building of more plants (and the GM tech center, which required a full square mile) further out where land was still cheap, and then the infamous school busing order of 1972, which caused the opening of the first ring of suburbs beyond the order.

Yes, white flight was part of it (especially after the riots in 1967), and now black flight, as blacks who can afford to leave for suburbs such as Southfield, but to argue highway building is the only cause, or even the primary cause, seriously misstates the case.

Benny Lava,
You really are completely confused. The phenomenon Tyler is talking about is the decentralization of population away from traditional urban cores or "central cities." It has nothing to do with the "annexation of suburbs." It is a change in the spatial distribution of the population. It may involve either an absolute loss of population in the central city, which seems to be the usual case, or merely a lower share of the population in the central city, as outlying areas grow faster than core areas. Paris is in the first category. The population of the city of Paris, the 105 sq km Ville de Paris, has declined dramatically in both absolute number and as a share of the metropolitan area. The center lost population, and the periphery gained population. I don't understand why you're having such a hard time understanding the concept.

And yes, weighted density is a better measure of sprawl at the level of Urbanized Areas than standard density. Weighted density captures population concentrations within a UA that are missed by standard density. That's why the New York UA, with its hyperdense concentrations of people in the lower half of Manhattan and parts of Brooklyn and Queens, has a higher weighted density than the Los Angeles UA, but a lower standard density. I'm not sure why you think this in any way undermines the decentralization phenonmenon we have been discussing.

regarding copenhagen, there is a fundamental difference between using trains and cars to get in from the suburbs.

I think cars give a really nasty externality to residents in the city. My young kids are on their bikes all the time in the center of the copenhagen - vesterbro - and i dread any increase in car traffic - especially road ragers coming in from the suburbs.

What I'd like to see is a graph (a histogram, really) where the x-axis is population density and the y-axis is the number of people in the US (or the percent) who are living in a place with that population density (where "place" is defined to be as small as practicable - neighborhood, census tract, or something like that). Ideally with an animation to show how this distribution has changed over time.

I'm not sure if highways caused or simply enabled suburbanization.

One commenter expressed confusion about why suburbs seem to be hated. I'm not sure suburbs themselves are hated or the design of them for the last 40 years is hated.

From ages 30-38, I lived in a suburb. Prior to that I lived in a variety of places but mostly urban areas. What I came to dislike greatly about suburb design is the need for a car to do simple things like buy a pint of half and half on Saturday morning for coffee. It seemed absurd that I couldn't walk out to a convenience store and be back in 5 minutes to enjoy my freshly brewed coffee.

What often develops in a suburban area is the next irritant: strip malls. They seem like a godsend to a new suburb. But, as traffic increases, a simple Saturday afternoon errand - e.g. buying groceries for the night's dinner - can easily take an hour due to wading through mall traffic.

When my wife and I visited Sonoma County 5 years back we were drawn to the setup of the towns like Sonoma and Healdsburg. They were setup to have a town center that was totally walkable and acted as a community center. If suburbs were designed more like small towns, with mixed residential and commercial use, I think they would be more palatable.

Tyler is talking about the dispersal of population away from historical urban cores. Not about whether a particular location is formally classified as part of a city or a suburb. That dispersal is typically referred to as suburbanization or sprawl. The fact that some suburbs are denser or more city-like than others, and that city boundaries sometimes expand to include areas previously outside the city limit, does not change the fact that suburbanization is real.

My point, and I think Benny's, is not that suburbanization is not real. It is that the measurements being used seem to be poor, because of definitional issues. Therefore statistical analyses and statements based on those measurements are highly dubious, at least if the Census Bureau's data is used.

The cited article says:

Before the interstate highway system, the percentage of the U.S. population living in suburbs went from 7.1 percent in 1910 to 23.3 percent in 1950. From 1950 to 2000 it had a smaller proportional increase, namely from 23.3 percent to 50 percent.

But that's only true under a certain definition of "suburb" which does not necessarily match what we think of as suburban sprawl. Indeed, to the degree suburbs are city-like the figures become not just imprecise but misleading. I live in Cambridge, population around 100,000, one of the most densely populated cities in the country. Yet in the statistics I live in the suburbs.


Annexation is not the issue. The definition of "suburb" is. As I noted above, I live in what the Census Bureau considers a suburb - Cambridge -because it is not politically a part of the largest city in the area - Boston. But Cambridge is an urban environment - dense population, etc. So the numbers are unrealistic WRT my area at least. (There are other areas of "suburban" Boston that have an urban character - Brookline, Somerville). As for NY, the outer (non-Manhattan) boroughs are not suburbs by the Census Bureau's definition because they are part of NYC.

Incidentally, the Bureau's definition of a "metropolitan area" has changed over time, as they themselves say. It is not consistent at all.

I really don't understand what point you are trying to make.

As for NY, the outer (non-Manhattan) boroughs are not suburbs by the Census Bureau's definition because they are part of NYC.

No, but we can analyze the change in population distribution within NYC in an equivalent way. The city as a whole represents the "Metropolitan Area" and the most populous borough represents the "central city." As I discussed in an earlier post, the population of NYC has decentralized in a similar way to the decentralization of MAs. The population distribution within NYC has become much flatter. I suspect this is true of most old, dense cities. The large cities of the south and west - Los Angeles, Phoenix, Las Vegas, San Antonio, Houston, Atlanta, etc. - which have grown mostly during the era of mass ownership of automobiles, never had a reason to become as centralized as old cities like New York, Boston and Chicago, and hence had a much flatter population distribution from the start.

"A cost internalized by the business providing and paying for the parking is not, by definition, an externality. The business has to pay the opportunity cost of using urban land that could be used for something else, in addition to higher property tax. It's not, by and large, an externality. OTOH, city-run subsidized parking is a subsidy."

Those businesses and employers are, and have for a long time been required to provide standard parking supply through public regulation, which in effect forces them to supply one group of people with the benefit of parking and financing it (both construction, maintenance and opportunity cost) through other (almost always) internalized means. But it is true that those same businesses and employers may very well have the freedom to charge the users of the parking lots as they please, but just choose not to (for usually obvious reasons).

But it cannot be regarded as fair towards the relatively (and understandably) small group of non-drivers that they help finance automobile infrastructure this way can it, be it public or private "conspiracy"? They do get some of it back through public transportation subsidies yes, but does it fully repay the loss of vibrant urban walkable communities, environmentally sustainable cities and being perceived as second class citizens?

Bernard Yomtov,

I don't know why you keep going on about Cambridge. As you say, it doesn't really fit the suburb stereotype. That's because it's not a typical suburb. Cambridge is an old city that was established long before the rise of the automobile. It's an inner suburb, close to the central city of Boston. It's a university town with lots of young, unmarried people. And it has extensive transit services. The existence of Cambridge, and similar suburbs around old cities, does not challenge or undermine the city-to-suburb decentralization trend that the Census Bureau analysis shows.


You keep telling me things I already know.

I understand Cambridge. I live here. The only reason I go on about it, as you put it, is to illustrate that there are problems with the census bureau definitions. Cambridge is not an anomaly. There are many places that qualify as suburbs that nonetheless do not fit the mold. Similarly, there are central city areas that anyone would define as "suburban" in nature.

I'm not arguing that there is no decentralization going on. I never said that. Please read my comments. What I have said, and the only thing I've said, is that using Census Bureau definitions might lead one to incorrect conclusions, especially if one is trying to do careful quantitative analysis. It's one thing to argue for general trends and quite another to start with regressions and so on.

The city is not, nor will it ever be, obsolete, despite technological innovations for transportation and communication infrastructure. It is a manifestation of the deeply rooted human need for community, and along with great people watching, the city provides opportunities for commerce, for culture, for recreation, for strolling, the sum total of which is called "living."

The Internet gives us the appearance of being connected; for instance, I feel connected to you who are reading this comment I've posted. But I am not connected. I have no clue who you are, where you are, or what you do. Wouldn't it be nicer to have this conversation in a pub, the way Jane Jacobs did in New York and Toronto?

Most suburbs give the appearance of connection. The church, the school, the homes, and the businesses may appear to be close and integrated, but in reality they are separated. They are their own pods that connect only to a main "feeder" road, so that to get to the store from the home 1/8 of a mile away, you have to hop in the car and drive. You can't walk down a sidewalk. There are no sidewalks. The highway, too, gives the appearance of connection: everyone all crammed close together, sitting in their own isolation booths.

If suburbs are starting to get better, as some posters here have claimed, it is only because they've started to adopt and adapt some tried-and-true methods for creating community that cities of the 20th century used to great effect (central parks, mixed-use developments, town centers, etc). These new suburban developments show us just how vital the city is. But the difference is that these suburbs bring the pleasures of the city to the relative few who can afford to live there, whereas a great city is something that all residents can partake in and share. A good downtown is for everyone.

Continuing with what Bernard and Benny say, both sides will continue to talk past each other until we come up with a clear definition of suburb. And I would expect no such clarity coming from Wendell Cox, who serially abuses terms and definitions to advance his arguments. Kenneth Jackson persuasively argues in Crabgrass Frontier, in accordance with the authors, that the allure of the suburbs - those areas with access to the city without the drawbacks of the city - goes back to the Babylonians. The first American suburb in the modern sense was Brooklyn, which underwent explosive growth thanks to the steamships that ferried commuters across the East River. Suburbanization only accelerated throughout the 19th century with the adoption of the omnibus, railroad, and streetcar. As the metro regions grew, those suburbs, whether they were officially annexed or not (Brooklyn was, Cambridge MA was not) were absorbed into a contiguous urban fabric that functioned as a city. Because those "suburbs" were founded before America's love affair with zoning, those places had the freedom to grow to accommodate the growing demand for living in the metro areas. That, in my mind, is what's wrong with more modern suburbs today as compared with the older suburbs. Metro regions that experienced suburban growth in the post-WWII years often found that their suburbs' zoning restrictions precluded the type of density and mixed uses necessary to keep the suburbanites close enough to the core to use transit to get to work and shopping. In this way, suburban growth became detrimental to the vitality of core cities in a way that was not seen before. This was certainly not a free market outcome. Plus, we should not forget that government intrusion into the free market known as FHA mortgage insurance, which heavily subsidized ownership of new suburban homes vis-a-vis renovation of urban rental stock.

Comments for this post are closed