skip to main content

→ Top Stories:
Safe Chemicals
Defending the Clean Air Act

Kaid Benfield’s Blog

"Cities" may not matter as much as we think - regions and neighborhoods are where things actually happen

Kaid Benfield

Posted November 1, 2010 in Living Sustainably, U.S. Law and Policy

, , , , ,
Share | | |

I’m going to start by being deliberatively provocative:  “cities” don’t matter nearly as much as we sometimes suggest.  I surround the word with quotation marks, because I am talking about cities as municipalities with legal boundaries – which is not, by the way, how most people use the word.  And that’s where we can get into trouble when it comes to policy.

I know that’s a bit over the top.  Of course, cities matter.  But in what sense?  It is important to be careful about vocabulary and about statistics.

  Can you tell where the city limits are? Didn't think so (Cincinnati region, courtesy of Payton Chung)

Let me give an example:  a friend in the business of growth management recently returned from a meeting in which there was discussion about so-called “shrinking cities.”  This is a very sharp, aware guy.  At that meeting someone reportedly said “29 US cities have grown recently and the rest have shrunk.”  This made my friend wonder if we should rethink “smart growth” as being about growth management, since so few places are actually growing.

That didn’t sound right to me at all, so I checked the facts:  Of 366 metropolitan areas tracked by the US census, 324 (89 percent) grew between 2000 and 2009; 161 of them, roughly half, grew by ten percent or more in only nine years.  Now, as we all know, central-city population in America was in decline for most of the last several decades, although the trend is clearly toward recovery.  I didn’t check the central-city facts.  But my guess is that the speaker had a point he or she wanted to make and chose a very select set of "facts" to support it - probably looking only at a subset of the places tracked by the census, and probably looking only at what was happening inside the central-city jurisdictional boundaries.  Those city limits date back to the early 20th or even 19th century in many cases and bear little relation to how places really function today.

Consider the image below, which shows the city limits of Atlanta drawn on a satellite photo of the real Atlanta:

  the jurisdictional city of Atlanta drawn on the functional city of Atlanta (via Google Earth, boundary by me)

Whatever the statement's origin, the impression generated - that few places are growing - could not be more false.  It misled my friend and it could mislead policy-makers into some very ill-informed decisions.  Managing growth is still highly relevant to any serious policy for sustainability, and I find it shocking that someone could come away from a growth management meeting, of all places, thinking that growth is less relevant than we may have thought. 

Here’s how the problem occurs:  Merriam-Webster defines a city as “an inhabited place of greater size, population, or importance than a town or village.”  That’s the way most of us use the word, most of the time.  But Merriam-Webster also defines a city as “a usually large or important municipality in the United States governed under a charter granted by the state.”  Two different things:  using the first definition, one might say that “Atlanta” is a sprawling metropolis and powerful economic engine with a population of 5.4 million people; but, using the second, Atlanta becomes a much smaller area confined within an artificially drawn boundary containing only some 540,000 residents.

The smaller, jurisdictional Atlanta may mean something to candidates for city office and cartographers, but it has very little to do with economic or environmental reality. Let's zoom in.  On the image below, the bright line shows the city limits of Atlanta.  If there is a meaningful distinction in the real world to be made between the areas inside and outside the "city," it sure eludes me:

  close-up of Atlanta city limit (via Google Earth)

In truth, it has been a long time since economies operated within a jurisdiction’s municipal boundaries, if they ever did.  As the Brookings institution puts it, metropolitan regions “are our hubs of research and innovation, our centers of human capital, and our gateways of trade and immigration. They are, in short, the drivers of our economy, and American competitiveness depends on their vitality. “

Or, as put by one Barack Obama during his presidential campaign:

“That is the new metropolitan reality and we need a new strategy that reflects it – a strategy that’s about South Florida as much as Miami; that’s about Mesa and Scottsdale as much as Phoenix; that’s about Stamford and Northern New Jersey as much as New York City.”

Heck, even in my own life, my wife and I live in Washington, DC, “the city,” but she works in suburban Virginia; I rode my bike yesterday and pretty much every weekend from my house into Maryland and back without even noticing where the jurisdictional boundary is.  I’m not exactly sure whether my doctor’s office is in DC or Maryland.  I just know it’s close to the line, on one side or the other; the street and buildings look exactly the same on both sides of the boundary.  My NRDC colleagues go home in the evenings to a dozen different municipalities.

  Chesapeake Bay watershed (by: U. of Virgina Department of Environmental Health and Safety)

The environment doesn’t respect political boundaries, either:  the Potomac and Anacostia Rivers receive runoff from Virginia and Maryland as well as from the jurisdictional city of Washington; the Chesapeake Bay watershed (above) includes parts of seven states.  Regional transportation patterns completely ignore jurisdictional boundaries.  The air in Chicago moves freely around the seven counties and 284 separate communities just within the Illinois part of the region, to say nothing of those in nearby Wisconsin and Indiana.  Very little of the energy consumed within the jurisdictional limits of the city of San Francisco is generated there.  And so on.  Statistics about only what is happening inside city limits very seldom tell us much about what is relevant environmentally.

I would submit that the other scale (besides regions) that matters most is the neighborhood.  Neighborhoods become especially relevant when they are complete, when there are multiple amenities, shops, and conveniences within walking distance or, at worst, a short rather than long drive away.  Neighborhoods are where we eat and sleep and where, if we are lucky, our kids play and go to school; where we shop for food, take our dry cleaning, and maybe grab a bite to eat; where we have chance encounters with others; if we’re really lucky, the neighborhood will even have a library and a hardware store.

  block party, Portland (by: JAGwired, creative commons license)  block party, Brooklyn (by: Adrian Miles, creative commons license)

While those of us who live in metro areas – and that’s 83 percent of all Americans – zip all around them to visit friends, conduct business, and shop, we’re usually going to other neighborhoods when we do.  If the region represents the economic scale of real cities, the neighborhood represents the human scale.

Neighborhoods are also the scale at which land development takes place, where new buildings and facilities are proposed, debated, and constructed.  They are where development decisions actually occur, and where we must pay attention if we want to have influence.  In fact, one of the best ways to reduce regional emissions is to revitalize older neighborhoods, because their relatively central locations reduce transportation emissions and they require little if any increase in runoff-causing impervious surface.  Not that you would ever guess that by the miniscule number of professionals in national environmental organizations who work directly on making it happen.

  a complete street with green infrastructure, imagined in Portsmouth VA (courtesy of Steve Price, Urban Advantage)

So where does that leave cities (in the jurisdictional sense of the word)?  They certainly remain very important politically.  In big cities, we can address problems at scale when we lack the legal and regulatory tools to do so regionally.  So that's where we frequently advocate stormwater regulation, building standards, complete streets, zoning reform, and even climate policy.  It's where we can get things accomplished.

But that's partly because cities are the low-hanging fruit of environmental standards and regulation.  We often look to mayors, for example, for leadership on environmental issues, even though their authority is limited to the parts of their regions that are almost always already the most sustainable, on a per capita basis.  It frequently pays off:  urban mayors tend to be more innovative and progressive than their suburban counterparts.

CO2 per capita from transportation, Cincinnati region (by: Center for Neighborhood Technology, HTA Index)But, when we focus inside the city limits instead of on the region, we’re missing most of the problem, and we shouldn’t pretend otherwise.  Stormwater runoff per capita is much worse in suburban sprawl, as are emissions of all sorts (CO2 per capita from transportation, left).  

One can even make the case that we should be going easier on cities than on sprawling places:  to paraphrase David Owen, why put skinny people on diets?  My personal view is that our environmental framework absolutely should be tougher on sprawling places than urban ones, but that urban ones should also do their fair share to heal our ecosystems, through appropriate standards, safeguards and mitigation.  Unfortunately, I think we remain relatively less attentive to the suburbs, largely because our crazy patchwork of municipalities makes them legally so diffuse and with very rare exceptions there simply is no regional authority to address them as a group.  This can sometimes lead to perverse results, where the most inherently sustainable places become subject to the toughest regulation.

Similarly, when we throw statistics around about “cities” that are limited to a fraction of a place's actual developed area and population, we’re frequently being arbitrary and missing the economic and environmental points that matter most.  To return to the "shrinking cities" phenomenon, if your area’s economy is in general decline leading to regional as well as central city population loss, that’s one thing, leading to one set of appropriate responses; but, if regional centers are hollowing out while the economy remains viable enough to support continued sprawl on the fringe, that’s quite another – and one that must be addressed by looking outside the city’s jurisdictional borders as well as within.  Failure to do so is, in my opinion, simply irresponsible.  But a statistic only about the jurisdictional central city won't tell you which of the two differing circumstances is actually occurring. 

Finally, and deviating slightly from topic, a personal beef:  there is also another way in which these otherwise artificial jurisdictional lines matter.  Tomorrow, most of you will have the opportunity to cast a vote for a Congressperson and perhaps a Senator.  As a resident within the city limits of Washington, I’m allowed to do neither.

Move your cursor over the images for credit information.

Kaid Benfield writes (almost) daily about community, development, and the environment.  For more posts, see his blog's home page

Share | | |


Dave MossNov 1 2010 09:06 AM

Yes, but we can't have regional entities crossing state boundaries to control individual jurisdictions that are given their powers by the states themselves, right?

I've never seen it written anywhere, but is there a difference between the existence of state governments and an organization that redistributes the economic development of urban regions into rural regions while allowing urban regions to allocate to themselves the resources of the rural areas?

Kaid @ NRDCNov 1 2010 09:21 AM

Dave, your first point is mostly correct, with rare inroads made by compacts among states, as in the case of the Chesapeake. In addition, MPOs routinely cross state boundaries in allocating federal transportation funds. But both examples are fundamentally weak.

Still, I'd settle for more regional entities within states, as there is around Portland. That would be major progress. In multi-state regions, one still could get to two or three entities with meaningful authority rather than hundreds, a frequent situation today.

I'm not sure about your second point.

Jon ReedsNov 1 2010 12:01 PM

I see what you're saying about local communities, but I'd be a bit wary about placing too much emphasis on "neighbourhoods" - in a metropolitan setting at any rate. The concept of neighbourhoods of 5,000 or so has been a key idea for the sprawl lobby for decades, big enough to support a secondary school and therefore obviously the unit to design in terms of cul de sacs, circular distributor roads and all the rest of the car dependent sprawl that's been going on for much of the last 100 years.
The other issue is "identity" of course. Do people in large metropolitan areas usually identify with their locality, or their city? I suspect it's usually the latter.
But your point about the chaos of administrative boundaries and lack of effective regional oversight is just as true here in the UK as in the US.
Mrs Thatcher abolished overarching authorities in the seven largest English metropolitan areas in 1986 (Gtr London, Gtr Manchester, Merseyside, South and West Yorkshire, the West Midlands and Tyne & Wear) and, so far, only Greater London has got an authority back. The other largest metropolitan conurbation, Strathclyde, lost its overarching authority in the 1990s. The justification was the usual "unnecessary and expensive tier of bureaucracy" one, although I suspect it was mostly because these large and significant authorities were politically opposed to the Government of the day.
Since then, of course, Scotland, Wales and Northern Ireland have got their own regional assemblies, leaving England as one of the very few places in Europe lacking elected regional bodies. (Check out the Smart Growth elements in the recent Planning Policy Wales incidentally). The new government has even now abolished the weak joint boards and quangos that were supposed to cover planning and economic development in the English regions. Everything is supposed to be "localist", but the chaotic patchwork of local authorities is having its funding cut drastically and there seems little appetite to do much even at the city-region level (with a couple of exceptions) on anything except economic growth, and there isn't much money for that.
But what is the ideal unit of governance for a city-region?

Kaid @ NRDCNov 1 2010 12:19 PM

Jon, having devoted the last ten years of my career to the building of sustainable neighborhoods, I obviously differ with you on that point. I think it is critical that we get them right.

Also, perhaps it is different US vs UK, but here suburbanites identify more with their subdivision (not the same as a neighborhood, BTW) or the region than with their municipalities. City dwellers in Washington say, "I live in Mount Pleasant" or "I live in Capitol Hill" every bit as much as "I live in Washington."

In any event, I raise the point not so much to debate drivers of sprawl or ideal forms of governance as to highlight a flaw in the use of statistics to describe the character of a place, and to suggest that a focus only on jurisdictional "cities" ignores the places that are most unsustainable. I stand by those points.

When it comes to governance, stronger regional entities would absolutely help here in the US. But we have a long way to go to get there and, as I see it, very little political will to do so.

Mike HicksNov 1 2010 04:32 PM

Yes, it's very true that the term "city" is misapplied all over the United States. In Minnesota, we had a number of townships incorporate as cities in the latter half of the 20th century. Considering that townships are usually 5 or 6 miles on a side, large amounts of rural area suddenly came under the banner of being a "city" just because the residents managed to erect a simple government. It's pretty crazy when you consider how small many old cities were. I've heard it said that Renaissance Florence would fit within the same space as a large shopping mall -- mostly due to the massive amount of land used for parking.

Arguably, most of us in the United States have little or no idea of what a true city is -- even our densest cities are much more spread out than what you'd find in Europe. Even the Canadians have done a significantly better job than Americans.

But yes, we really need to find better ways of measuring and describing the types of development we do have in this country if we are going to find the right way forward. You can find statistics stating that the New York metropolitan area is less dense than that of Los Angeles, for instance. Relying on municipal boundaries can really screw things up.

I'm confident that strong mathematical models could be developed to define "logical city" boundaries from census tract data which would be more useful for analysis purposes than currently-defined municipal or metropolitan boundaries. This would be important for the Twin Cities to remain a major player in the U.S. economy -- many organizations use the list of 50 largest cities to compare things ("Which cities have the best/worst [noun]?"). Minneapolis is in danger of falling out of the top 50 cities by population, so it wouldn't be included in nearly as many analyses as it is today. Of course, the studies it does get into are often skewed because of the slightly weird municipal geography we have.

E M RisseNov 1 2010 06:46 PM

Dear Kaid:

You have opened the door to an extremely important field of inquiry.

This is a field to which we devote a lot of attention in The Shape of the Future,


Your work has inspired us to spend a fair amount of additional time on the topic in the past six months. PRIMER, the product of that work, will be on the way to you soon. At least the Foreword and Table of Contents.

In the meantime, congratulations. You have opened the topic using only three of the eight Core Confusing Words related to human settlement patterns. Just shows that abstinence is the best policy when it comes to dysfunctional Vocabulary.

Two quick notes for now:

1. A good deal of the northern boundary of the Commonwealth of Virginia is the Potomac River. The Anacostia River runs into the Potomac from the North. I think you will find there is no natural drainage from the Virginia into the Anacostia. In general your geopolitical observations are sound, but someone may bop you over that one.

2. Our recent work on settlement pattern components was triggered by your work on LEED ND. To get you in the right frame of mind to consider our work product:

Jon Reeds suggested a ‘neighborhood’ population of 5,000. What is your ideal, average, minimum, maximum ‘neighborhood’ population?

Keep up the good work.


E M RisseNov 1 2010 06:52 PM

Mr. Hicks:

Check out the work of Prof William H. Lucy, UVA

Confronting Suburban Decline (2000) with Dave Phillips, and

Foreclosing the Dream (2010)

both employ data aggregation strategies that overcome most problems with municipal borders when combined with Radial Analysis.


JM SchillingNov 2 2010 10:56 AM

Kaid: as always I appreciate your effort in expanding our thinking about these issues. As one immersed within the shrinking city arena, I agree that in order to address these intractable problems of disinvestment and depopulation we have to deploy policies and programs at different scales and at different points in time as communities move through a temporal cycle of decline, stabilization, abandonment, etc. Think of the three tiered Vulcan chess board that Spock used in Star Trek as it illustrates the complexities of shrinking cities involving multiple challenges at different dimensions. Unfortunately, our current policy chess board is still flat and one dimensional.
Confronting these difficult challenges at the city or even neighborhood scale may bring temporary relief to the residents who live and work in such distressed places (which is an essential short term strategy to improve the conditions and health of the residents that remain), but I agree with you that long lasting solutions will demand regional approaches that can only come from state and federal government leadership as they control the policy levers at the regional scale, not the cities or the neighborhoods.
Your points call for a true placed-based approach that adapts classic smart growth strategies for slow growth regions with no growth cities. The challenge right now is we don't really have any successful policy models that simultaneously operate through these scales--that coordinate actions and policies from the regional, through the city and down to the neighborhood level. I think there are some places, such as Cleveland, that have the potential, but they too are searching for new frameworks that could guide them. This was also the thinking that underlies the Community Regeneration Sustainability and Innovations Act (CRSI) that was proposed by congressmen Ryan (Youngstown) and Higgins (Buffalo)—a comprehensive, regional planning pilot program that would infuse sustainability and smart growth programs and policies within a community regeneration context for cities that have substantial losses in population.
Finally, while the statistical lens of coming up with Top 20 lists of shrinking cities only tell part of the story, I do see some benefit in examining the conditions and characteristics of these shrinking cities and drilling down to identify those neighborhoods within them that are shrinking and growing as it calls attention to a special set of challenges that have been far too ignored by most policymakers and practitioners. Keep us thinking Kaid.


Kaid @ NRDCNov 2 2010 11:11 AM

Mike - great points, thanks for adding to the mix.

Ed - The LEED-ND Reference Guide states:
"Size is a defining feature of a neighborhood and is typically based on a comfortable distance for walking from the center of the neighborhood to its edge; that suggests an area of 40 to 160 acres." (LEED-ND is designed to evaluate projects from small neighborhood fragments up to developments twice that size and may start to become less useful as the size exceeds 320 acres or so.) The population of such a "pedestrian shed" will vary, depending on how intensely developed it is.

Joe - Fair, thoughtful points, though now you have outed yourself as a Trekkie. ;)

PaulNov 2 2010 12:53 PM

@Dave: We do have entities, or rather a set of closely coordinated entities that transcend state boundaries. They're called DOTs, and they basically act as the planning proxy for regions (with or without MPOs).

Blooming RockNov 3 2010 01:05 PM

Kaid, here in the Phoenix area, there is a new regionally-focused plan called the Discovery Triangle focusing on connectivity and development between the cities of Phoenix, Tempe and Scottsdale.

I totally agree with you about focusing on neighborhoods and making them stronger. I think this is the key to making our sprawling city of Phoenix, which is 500 square miles, a livable place. And yes, 500 square miles is the jurisdictional boundary, so in our case, our boundary is too big and we should perhaps shift our focus inward to various neighborhoods as you mention.

However, because of the strong connection, both environmental and cultural between Phoenix, Tempe and Scottsdale, I believe a regional approach is also appropriate here.

David CrossleyNov 4 2010 03:53 PM

As the guy who shocked Kaid by seeming to be misled into thinking that growth is less relevant than we thought, I need to clarify. The meeting in question was in Detroit, so there was quite a bit of conversation about that place and some others, like Buffalo and Cleveland. I confess the half-overheard comment someone made about 29 cities growing was just a factoid thrown to Kaid to ask about a question other than the one being addressed in this column.

As a longtime student and supporter of sustainable development, as opposed to growth, I have always worried that "smart growth" was not really a fully realized ecological concept. Basically, the term is the last two words in a sentence that goes something like "If we are going to have growth, then it should be smart growth." I'm totally on that and have been a smart growth advocate, where there is growth, for a long time. But that leaves a question about whether we need growth. That's all I was putting out there. And of course, it's interesting to think about how smart growth works in a place that is not growing.

But all this stuff Kaid has given us is, as usual, huge food for thought, so I promise to read it three times and think about it.

Comments are closed for this post.


Switchboard is the staff blog of the Natural Resources Defense Council, the nation’s most effective environmental group. For more about our work, including in-depth policy documents, action alerts and ways you can contribute, visit

Send Me Updates About: Kaid Benfield

As new content on your chosen topic gets posted, you'll receive an automated email via FeedBurner. You can unsubscribe at any time.

Feeds: Stay Plugged In

Feeds: Kaid Benfield’s blog