City Growth, and "What's a City?"

[Originally written by Jason King] The constantly shifting definition of 'City' or 'Urban' makes for some ambiguity in talking about our aggregations (or is that agglomerations?) of development.  Grist mentions the perceived dilemma with contradictions that the 2010 census shows that cities are adding population faster than suburbs for the first time in the last century, while also finding documentation from Atlantic Cities that exurbs are the fastest growing areas in the US.  Are we talking about the same things?  Is it just semantics, or does this difference between suburb, exurb, fringe, metro area actually make sense?  Is a suburb and an exurb the same thing?  Or is one a low-density, non-urban entity (suburb) with it's parallel, urban counterpart of low-density development (exurb).  Is it purely based on density and type of development, or is proximity to center make a difference?  Does a center exist, and if so, does it matter, or are there multiple centers?

A typology of regional-scale development patterns. Image credit: Peter Newton, “Reshaping Cities for a More Sustainable Future” (1998). (image recovered from

I talk a lot about clarity as the main weapon we have in having rational conversations about cities, design, landscape, planning, and all forms of discourse.  As mentioned in an essay in Atlantis magazine from last year (focused on the definitions of landscape and urbanism) the interpretation can not be fuzzy.  As I mention, we should demand:

"...a call for more clarity in writing about these terms, specifically the need for clear definitions when discussing terms.  We are too loose with terminology today, and the overall impact and reach of our discussion suffers from this. Whichever way you choose to interpret and intervene the urban conditions, there needs to be shared understanding of fundamental issues, because, as I mention: "In the end, no discussion or argument (binary or otherwise) is worth much if it happening around vague language..."

So what is the clear picture, and how does the exurb fit into the discussion of city growth?  We need to know what we mean by terms.  A quick web search brings up a definition that falls in line with my understanding:.  Webster's defines exurb as "...a region or settlement that lies outside a city and usually beyond its suburbs and that often is inhabited chiefly by well-to-do families."   This is contrasted with suburb, again from Websters, which is variously defined as "... an outlying part of a city or town, a smaller community adjacent to or within commuting distance of a city or, the residential area on the outskirts of a city or large town."

Without getting too far into the rabbit hole of etymology, that at least makes sense - with a heirarchy of city and non-city.  Thus a city includes a suburb, but not exurb.  I'm sure a more thorough accounting of definitions would yield many contradictory definitions.  On that note, linked in the Grist article, is a bit more of a statistical treatment of the data is from the Urban Institute where the authors Todd Gardner and Matthew C. Marlay use the census data to compare and contrast growth from 2000-2010.  One conclusion is that  "... in most of the 100 largest metros, growth remains higher in the exurbs than in older, denser areas closer to the core."

In this relationship - the exurb as i read it, is a part of the urbanized area, i.e. distinct from the suburbs but part of the city, which is contrary to the above condition.  If the implication is that urbanized area = city, then it goes to show that urban growth has outpaced suburban growth, not as much from an influx of people moving out of suburban areas to the city, but rather from redefinition of what components can be considered urban.  But if we take the meaning of 'city' to include suburb, fringe, etc. - the whole discussion changes - because we are talking not about a distinct urban/suburban divide, but a more broad urban/rural divide to make the distinction - with exurbs being part of the rural and suburbs being part of the city.  Confused yet!?

Anyway, back to Gardner and Marley, where their study includes a graph, below, shows the overall breakdown - which includes the afforementioned 'urbanized' areas, without, from what i can find, is a clear definition of what this means (or more specifically, how urbanized is differentiated from non-urbanized).  They are clear in defining some parameters for the categories.  For instance, the authors mention that they use a specific definition of exurbia "... using three characteristics of the built environment: (1) housing unit density, (2) age of the housing stock, and (3) commuting links to neighboring economic centers."   The assumption being that each of these categories (high-density, moderate density, low-density, fringe, and exurban), are distinctly defined.


The graph is telling - but it's still unclear what exurb is in relation to the city (is the city high, moderate and low density, and fringe is the suburbs).  Is exurban an actual 'urbanized' condition?   The issue of, and need for this clarity, is that the authors use it and the associated data to make a strong point about the strength and role exurban development, and it's place in the metropolis (which again, doesn't necessarily, mean the city!).  As noted in the concluding comments:

"These results illustrate the continued importance of the exurbs. Although land-use planners, environmentalists, and others concerned about sprawl and resource use decry their very existence, it is clear that the exurbs continue to house a significant share of the metropolitan population, and in fact account for much of the metropolitan population growth over the past decade. These areas have weathered the economic storm of the Great Recession; whether they are poised to continue their amazing growth of the past decade is as yet undetermined."

A caveat at the end of the summary mentions a bit of their methodology in definition exurbia, and also contrasts the results with another study.  Their study:  "... combines data from the 2000 and 2010 decennial censuses with aggregated American Community Survey (ACS) data from 2005-2009. The 2005-2009 ACS data averages to 2007, produces results roughly comparable to those using the single-year 2007 ACS data, and enables us to perform tract-level analyses for every metropolitan area. To ensure comparability across time, we used consistent census tract boundaries for all data points."


They contrast this to a recent NY Times article by William Frey on the decline of the exurbs, noting that "... contrasting results arise in part from methodological differences. For example, our study uses census tracts as the basic geographic unit, while Frey uses counties. Also, his article includes the period from 2010 to 2011, while this commentary focuses on the previous decade."

Methodological differences will occur, and clarity of a particular approach is a good way to parse the differences, but most folks won't get the difference between census tract data resolution versus county level data - and to me, it shows that the two studies aren't actually talking about the same thing when they refer to exurbs.  I think it's also a big ambiguous to relate density (as noted in the definitions of some each of these categories) with location. This also brings up issues of location theories - and the mortgage crisis diminishing the mobility of people to relocate (due to being underwater) or to locate in a more desirable location (due to inflation of housing values and lack of access to lending).  But that's a topic for another time.

It does seems to echo the clean divisions of the urban-to-rural transect, with a monocentric model of radiating density that combines both spatiality and density - each level (or ring) becoming less dense and less 'city'.

A diagram of the urban to rural transect concept - from Center for Applied Transect Studies (retrieved from:

It would be interesting to see where the distinction between the categories by Gardner and Marley ((high-density, moderate density, low-density, fringe, and exurban) relate to the T-zones of a typical urban to rural transect - and if exurb would equal T3 (suburban) or a hybrid of General Urban (T4) or be an outer version - an exclusive enclave of T2 (rural reserve)?  I realize these are models and abstractions, but that should make definitions easier because they lack the messiness of reality and can be defined in concrete terms.  This can translate to at least a common parlance for terms, as an idea of some shared way of  discussing a city isn't too much to ask, particularly when we are using this to develop policy and allocate resources.

A fascinating topic worth future exploration - and i hope to discuss more - with comments, as always, welcome.