Finding Exurbia: America's Fast-Growing Communities at the Metropolitan Fringe

Metropolitan Policy Program Finding Exurbia: America’s Fast-Growing Communities at the Metropolitan Fringe

Alan Berube, Audrey Singer, Jill H. Wilson, and William H. Frey

Findings This study details a new effort to locate and describe the exurbs of large metropolitan areas in the “Not yet full- United States. It defines exurbs as communities located on the urban fringe that have at least 20 percent of their workers commuting to jobs in an urbanized area, exhibit low housing density, and have relatively high population growth. Using demographic and economic data from 1990 to 2005, this fledged suburbs, study reveals that: ■ As of 2000, approximately 10.8 million people live in the exurbs of large metropolitan areas. This represents roughly 6 percent of the population of these large metro areas. These exurban but no longer areas grew more than twice as fast as their respective metropolitan areas overall, by 31 percent in the 1990s alone. The typical exurban census tract has 14 acres of land per home, compared to 0.8 acres per home in the typical tract nationwide. wholly rural, ■ The South and Midwest are more exurbanized than the West and Northeast. Five million people live in exurban areas of the South, representing 47 percent of total exurban population nationwide. Midwestern exurbs contain 2.6 million people, about one-fourth of all exurbanites. South exurban areas are Carolina, Oklahoma, Tennessee, and Maryland have the largest proportions of their residents living in exurbs, while Texas, California, and Ohio have the largest absolute numbers of exurbanites.

undergoing rapid ■ Seven metropolitan areas have at least one in five residents living in an exurb. These metro areas include Little Rock (AR), Grand Rapids (MI), and Greenville (SC), as well as areas like Poughkeepsie (NY) that serve as “satellites” to nearby larger metro areas. Both fast-growing and change in popu- slow-growing metropolitan areas have developed exurbs.

■ Nationwide, 245 counties have at least one-fifth of their residents living in exurban areas. lation, land use, The Louisville metro area has the highest number of exurban counties (13), followed by Atlanta, Richmond, and Washington, D.C., which each have 11. These exurban counties grew by 12 percent overall between 2000 and 2005, faster than population growth in urban, inner suburban, or outer suburban counties (like Loudoun County, VA). However, outer suburban counties added 4.5 and economic million people in the last five years, exceeding the 1.8 million-person gain in exurban counties.

■ Residents of the “average” exurb are disproportionately white, middle-income, homeowners, function.” and commuters. Yet exurbanites do not conform to all popular stereotypes. For instance, they do not appear to telecommute, work in the real estate industry, or inhabit super-sized homes at higher rates than residents of other metropolitan county types. Middle-income families’ “drive to qualify” for more affordable new homes that are in limited supply elsewhere fuels growth in many metropolitan exurbs.

Despite their popularization by political analysts, media, and local growth activists, the “exurbs” do not abound nor fit a single, neat stereotype. Just 6 percent of large metro area residents live in an exurb, and these exurbs vary from affordable housing havens for middle-class families, to “favored quarters” for high-income residents, to the path of least resistance for new development. While they may continue to capture interest among political observers, the real test for exurbia lies in how our nation accommodates future growth. Will exurbs remain exurbs or become the suburbs of tomorrow?

October 2006 • The Brookings Institution • Living Cities Census Series 1 Introduction rural in nature, these exurban areas torsky’s 1955 rendering, Fairfield are reportedly undergoing rapid County (CT) and Rockland County eyond the suburbs, at the far change in population, land use, and (NY) housed exurban commuters to edges of metropolitan areas, economic function. New York City. After five decades of communities both new and Notwithstanding what they will continuous metropolitan decentraliza- old are developing the become in future generations, exurbs tion, places like Greenwich (CT) and Bcapacity to house large flows of incom- are important places to understand in New City (NY) are nothing if not con- ing residents. their contemporary form. They lie at ventional New York-area suburbs. Popularized in books like David the forefront of important local Some common threads run through Brooks’ On Paradise Drive, and in debates around growth and develop- popular descriptions of the exurbs. numerous news articles analyzing the ment issues. As such, they help “set They generally lie within the orbit of a aftermath of the 2004 Presidential the table” for future metropolitan big city, but have weaker economic election (Bai 2004; Brooks 2004; growth, and their prevalence may and social ties to the urban core than Brownstein and Rainey 2004), these serve as an important indicator of suburbs, consistent with their more areas have come to be known as emerging social trends or the effective- remote locations. They are more resi- “exurbs.” Exurbs are drawing the ness of various policies to shape met- dential with newer housing stock, and attention of journalists, pundits, poli- ropolitan development. less commercial than suburbs. They cymakers, and sociologists attempting Yet exurbia is hardly a new concept. have population densities that are to understand how Americans live It first gained popular attention when lower, more like their rural neighbors, today and how different community A.C. Spectorsky described its residents and their economies may still depend types matter to politics and policy. In in his 1955 book, The Exurbanites. somewhat on agricultural employ- many exurbs, battles over roads, land, Despite this long history, the concept ment. sewers, schools, quality of life, and of the exurbs is enjoying greater atten- Perhaps most importantly, present- “loss of small town character” are tion now than ever before. The terms day exurbs seem to be defined by fast- reaching a fever pitch. exurbs, exurban, and exurbia appeared paced, low-density growth and But where exactly are the exurbs? in U.S. newspapers twice as many development. If that type of develop- And who lives there? times in 2005 as they did just two ment continues, new housing and In most metropolitan areas, a cen- years ago, and at four times the rate commerce could fill in the space on tral city (or cities) still lies at the core, they did ten years ago.2 the far fringes of metropolitan areas, containing a cluster of jobs and resi- All the talk about exurbs, however, turning them into more “mature” sub- dents. Outside the core lie suburbs, has yielded little methodical informa- urbs and potentially creating new areas that developed somewhat later, tion or consistent definitions regarding exurbs even farther out. More immedi- comprised today of both residential their location and their residents. In ately, however, the development pat- and commercial space. Suburbs are addition, it is unclear what might dis- terns typically identified as “sprawl” the dominant landscape in contempo- tinguish exurbs from the suburbs in intersect closely with those found in rary America, and their speed and which Americans have lived for gener- the exurbs. This paper does not scale of change has reconfigured the ations. Some researchers have tackled attempt to define “sprawl,” nor does it way Americans live (Hayden 2003). By these questions in the academic litera- equate sprawl with exurbia. But the Census Bureau’s definition, about ture. But many recent observers have because public policies designed to 53 percent of the nation’s residents resorted to an “I-know-it-when-I-see- curb (or promote) sprawl in metropoli- live in suburbs.1 Whereas most people it” approach for identifying the exurbs tan areas often affect the exurbs, it is used to work in cities and their imme- and their residents. Many counties, for important to understand more about diate environs, today more than half of instance, may have a more mixed char- where these communities are located, all metropolitan jobs are located at acter and are probably not well- who lives there, and why. least 10 miles from a traditional city described by alternating and Sparked by new popular interest in downtown (Mieszowski and Mills overlapping classifications. the subject, this paper describes an 1993; Berube forthcoming). A lot of journalists, and some aca- effort to “find exurbia,” based on a Exurbs, it is argued, lie somewhere demics, argue that exurbs represent a review of the academic literature and beyond the suburbs. At the urban- path-breaking form of development, analysis of data from Census 2000 and rural periphery, outer suburbs bleed defined in part by the extreme dis- beyond. We begin by examining previ- into small-town communities with an tances from these areas to the metro- ous attempts to identify exurban terri- agricultural heritage. Not yet full- politan core. Yet today’s exurbs may tory and its residents, and discuss the fledged suburbs, but no longer wholly just be tomorrow’s suburbs. In Spec- implications of those inquiries. We

2 October 2006 • The Brookings Institution • Living Cities Census Series then describe a new methodology for bia, Spectorsky’s 1955 book The labeled “exurbs.” Teixeira (2006) fur- identifying exurbs and the assump- Exurbanites, sought to describe not ther investigates the politics of exur- tions that underpin it. Applying that so much these far-flung (at the time) bia using Lang and Sanchez’s methodology, we proceed to describe places, but the people who inhabited classification system. the incidence and characteristics of them. To Spectorsky, these people These studies that attempt to char- exurban places nationwide, in states, were a new breed of commuters, acterize exurbia and exurbanites form and in selected large metropolitan seeking a semi-rural lifestyle to part of an even larger literature areas. Having pinpointed the exurbs as which they could retreat from their devoted to classifying the “hierarchy” geographies, we then profile their 9-to-5 jobs in Manhattan’s “ideas of urban, suburban, and rural places demography, asking who lives there, industries”—advertising, media, and and their residents. Some of these whether they differ from their coun- the arts. Davis and Nelson (1994) studies classify metropolitan counties terparts elsewhere in the metropolis, and Nelson and Sanchez (1997) take and places by their degree of “urban- and what their reasons might be for a similar approach, identifying exur- ness” and the residential/commercial locating there. The paper concludes ban households in the Portland area functions they serve (Orfield 2002; with some suggestions for future and asking what—if anything—dif- Lang and Gough 2006; Lang and research and possible implications for ferentiates them from other subur- Sanchez 2006). Puentes and Warren public policies that aim to slow the ban households. (2006) focus on older “first suburbs” rise of exurbia by promoting balanced • What are the development impacts that were the first to develop outside metropolitan growth patterns. of exurbs? Several other studies big cities. Mikelbank (2004) uses the identify the exurbs as part of an characteristics of both suburban resi- effort to understand the extent of dents and their locations to derive a Background “sprawling” development patterns in typology of suburban places. Frey different parts of the U.S., and the (2004) advances a typology of places his study hardly represents implications for public policy. within metropolitan areas based on the first attempt to define Theobald (2001), Irwin and Reece their population and employment lev- the exurbs. Planners, geogra- (2002), Nelson and Sanchez (2005), els, and their distance from the urban phers, sociologists, and— and Wolman and colleagues (2005) core. USDA’s Rural-Urban Continuum Tmost recently—election analysts have all use methodologies based in part Codes (2004) have for three decades all devoted time and energy to examin- or in whole on density (of population provided a particularly influential clas- ing new forms of development occur- or housing) to measure exurban sification of counties that span the ring at the metropolitan fringe, some development. entire U.S. settlement hierarchy. of which posit the existence of quasi- • What are the electoral impacts of We make no attempt here to sum- urban, quasi-rural places that are exurbs? Finally, very recent analyses marize the findings from this rich liter- sometimes referred to as exurbs. by a collection of political analysts, ature, since the studies—spurred by Different research questions have including Brownstein and Rainey different inquiries—adopt such a wide motivated this literature, with little (2004), Texeira (2004, 2005), Gersh range of methodologies and come to consensus emerging on the following (2004), and Lang and Dhavale very different conclusions. key questions (relevant studies are (2005) use various definitions of Given such a rich literature, then, listed in Table 1): exurbs to look at voting patterns in what can this study possibly add to the • Where are the exurbs? Several different types of communities. information and interpretation that studies have advanced one or more These analyses were fueled in large already exists? For one thing, because ways to define exurbia as a geo- part by stories from the 2004 presi- the exurbs are generally acknowledged graphic or demographic concept. dential election documenting the as existing at the urban-rural fringe, Lamb (1983), Blumenfeld (1986), success of Republican voter-mobi- they likely change rapidly and older Lessinger (1987), Nelson (1992), lization efforts in new, outlying com- studies may not reflect the contempo- Beale and Kassel (2005), and Clark munities in “swing” states like Ohio rary situation. Furthermore, our and colleagues (2006) use varying and Florida. To study the impact of review of the literature suggests to us techniques to identify the scope of different types of suburbs on the a few core principles for this effort exurbia (and similar geographic con- 2000 and 2004 presidential elec- that distinguish it from past efforts to cepts) nationwide, and in selected tions, Lang and Sanchez (2006) “find exurbia:” areas of the U.S. develop an urban/suburban typology • Place before people. We choose to • Who lives in exurbia? The first for the 50 largest metropolitan areas identify exurbs based on their char- work to describe the notion of exur- in which the outermost counties are acteristics as places, rather than on

October 2006 • The Brookings Institution • Living Cities Census Series 3 Table 1. Studies Defining Exurbia and Related Concepts

Study (by year) Concept measured Geographic Unit Description Spectorsky (1955) Exurbanites Counties and places in NY Outer edges of the NYC commuter shed: Fairfield County, metro area CT; Rockland County, NY; Bucks County, PA Lamb (1983) Exurban sprawl Counties and places Within 50 miles of urbanized area (UA) > 250k people; growth rate > 5% in 1960s; outside UA in 1970 Blumenfeld (1986) Metropolitan fringe Counties Outside Standard Metropolitan Statistical Area (SMSA) but: (a) within 70 miles of large SMSA (> 2M people) central city; or (b) within 50 miles of mid-sized SMSA (500k to 2M people) central city Lessinger (1987) Penturbia Counties Below-average population growth 1950–1970; far above- average growth 1970–1985 Nelson (1992) Exurbs Counties Within 50 miles of central city boundary in mid-sized Met- ropolitan Statistical Area (MSA) (500k to 2M people); within 70 miles in large MSA (> 2M people); not central county or otherwise in metro area in 1960 Davis and Nelson (1994) Exurban movers Block groups and places, Recent movers to portions of Portland and Salem, OR Portland and Salem, MSAs located outside urban growth boundary; at least OR MSAs 10% commuting to MSA Nelson and Sanchez (1997) Exurban population Households in American Recent movers to parts of metropolitan areas outside UA Theobald (2001) Exurban areas Housing Survey, 22 metro Housing density of 10–40 acres per unit areas Irwin and Reece (2002) Exurban areas Block groups Housing density between 5 and 40 acres per unit Beale and Kassel (2005) Exurbs Block groups in Ohio Population density < 500 people per square mile; scores Census tracts on 4 population characteristics: % of adults with college degrees; % employed in arts, entertainment, recreation, information industries; median household income; average commute time Nelson and Sanchez (2005) Exurban areas Block groups in 35 largest Population density of 300 to 999 per square mile metro areas Brownstein and Rainey (2004); Exurbs Counties Various combinations of metropolitan location and popula- Teixeira (2004, 2005), Gersh tion characteristics (2004), Lang and Dhavale (2005) Wolman et al. (2005) Extended Urban Census tracts in six large UA plus adjoining tracts with minimum housing density Area metro areas (60 units per square mile) and minimum commuting to UA (30 percent) Lang and Sanchez (2006) Exurbs (and other Counties within 50 largest Lower population density, less urbanized population, suburban types) metro areas higher non-Hispanic white population Clark, McChesney, Munroe, Exurbia Small grid cells (170 acres) Population density of 100 to 1,000 and Irwin (2006) from LandScan population model Source: Brookings Institution analysis

the characteristics of the people who Because exurbs are at their root ing patterns)—and then seek to live there. If we were seeking to find places, however, we identify the understand whether the people who the “exurbanites” as Spectorsky did, attributes of a geographic area that live there are demographically dis- we might then consider demographic might label it an exurb—largely inde- tinct from people elsewhere.3 characteristics like race/ethnicity, pendent of the social or demo- • Places we know. Advances in Geo- income, homeownership, or occupa- graphic characteristics of its graphic Information Systems (GIS) tion as part of our definition. residents (aside from their commut- technology have enabled researchers

4 October 2006 • The Brookings Institution • Living Cities Census Series to describe and classify very small Methodology and Data difficult to pinpoint.7 Some geographic areas. Many have used researchers have used distance from, GIS and small-area data to identify he principles above guide or adjacency to, a city or its estab- as “exurban” places as small as a how we translate our model lished suburbs to define exurbs. But couple of acres. However, such tiny of exurbia from concept to the underlying assumption is that geo- geographies, in isolation or in the real-world geography. Based graphical proximity to an urban area aggregate, have limited relevance to Ton research, popular literature, field implies some kind of “connectedness.” policymakers and the public. At the work, and interviews with experts, we Though those connections may take same time, defining exurbs at too observe three measures that define various forms—where people shop, large a geographic scale glosses over exurban places: connection, density, where their friends live, what newspa- the very different development pat- and growth. We first identify exurbs per they read, or what sports teams terns that may characterize parts of using census tracts—small areas with they root for—commuting ties offer that area. This is a particular prob- an average of 4,000 people—and then one way to measure economic connec- lem in the Western U.S., where the aggregate these areas to the county tion, and one for which data are average county covers 2,825 square level for further analysis.6 To qualify as widely available. miles. We attempt to strike a balance exurban, a census tract must meet all In our view, an exurb must be physi- by first pinpointing exurban territory three of the criteria described below. cally located outside a large city and its using census tract-level (i.e., neigh- close-in suburbs, yet have some pro- borhood-level) data, and then aggre- 1. Economic connection to a large portion of its residents working in that gating those tracts up to the county metropolis. Implicit in the term area. To identify these areas, we use level to connect our classification to “exurb” is some relationship with an data compiled by the USDA’s Eco- widely-known geographies. urban area. The definition of its prefix, nomic Research Service (ERS) from • Useful across time and place. Just as “ex,”—meaning “outside of”—is more the Census 2000 tract-to-tract com- yesterday’s rural area is today’s exurb, the exurbs of prior decades— like Greenwich and New City—are today’s mature suburbs. In light of Figure 1. Exurban census tracts send at least 20% of their these dynamics, we base our exurban workers to urbanized areas in large metropolitan areas identification on geographies and data that are likely to be available at regular intervals over time. Certainly, Census tract from one-time surveys of land use, sewer- which at least 20% Urbanized area age lines, or night lighting could of workers commute PC (densely settled to an urbanized area territory of at least form the basis for a methodology to in a large metro 50,000 people) identify exurban territory. However, these unique data sources are difficult to collect regularly or may be unevenly available locally. Our selec- Principal City tion criteria derive from decennial Metro area of at PC Micropolitan area least 500,000 people census data, with the expectation in 2000 (88 total) that very similar, if not identical, data will be available across time to PC chart the movement and changing characteristics of exurbia. These data also have the advantage of cov- Urban cluster (densely settled ering the entire United States. territory of less than 50,000 people)

Source: Brookings Institution

October 2006 • The Brookings Institution • Living Cities Census Series 5 muting file.8 We identify as potentially composed of only one county than within a large metropolitan area, yet exurban those tracts for which ERS their large-metro counterparts.11 lie outside that urbanized area. Thus, identifies that at least 20 percent of Future research should consider many exurbs lie in areas that already workers commute to an urbanized area whether and how exurbs might be fail to meet the population density within a large metropolitan area.9 defined differently in the small metro- threshold necessary to qualify as Urbanized areas are the basic census politan context; in this paper we focus “urbanized” under the Census geographies that define the core of on the incidence of large metropolitan Bureau’s definition.15 Because some metropolitan areas (Frey et al. 2006), exurbs only.12 tracts that meet our commuting crite- and they typically represent a central rion do lie in smaller urbanized areas city and its built-up suburbs.10 The 20- 2. Low housing density. All measures (see the commuting discussion above), percent commuting threshold is some- of urbanity and rurality take density the housing density criterion ensures what lower than the 30-percent into consideration. Since the most vis- that we capture areas with a more threshold used by Wolman et al. ible component of an area’s settlement spread-out feel than is typical in sub- (2005) to define the extended urban type is the built landscape, we use urbia today. area, and also below the 25-percent housing density rather than population county-level threshold used by OMB density to identify exurban areas. 3. Population growth. Exurbs are to differentiate outlying and central Using Census 2000 data, we rank all emerging growth centers, where peo- counties in metropolitan areas. In this census tracts nationwide on housing ple have been moving in large num- way, our minimum commuting thresh- density (units per square mile). We bers recently. Some of these people old seeks to capture places that may then distinguish as potentially exurban may have come from outside the met- still be evolving toward fuller inclusion those tracts which, beginning from the ropolitan area, but an even larger in the metropolis. lowest density tract, collectively con- number may have relocated from We further limit qualifying tracts by tain one-third of the nation’s housing denser parts of the same metropolitan identifying as exurban only those units.13 In 2000, these tracts had a area, in search of lower housing costs, tracts with commuting ties to an maximum housing density of roughly more open space, and/or a small-town urbanized area within a Metropolitan 2.6 acres per unit. “feel.” To qualify as exurban, a census Statistical Area that had at least Why the bottom third? First, this tract must have experienced popula- 500,000 residents in 2000 (see Figure approach ensures that we capture tion growth between 1990 and 2000 1 for a visual representation of this cri- areas in which housing is significantly that exceeded the average for its terion). Exurban tracts may be located more spread out than is typical in the related metropolitan area.16 In addi- inside a metropolitan area, either in a U.S. today. Second, it includes a size- tion, the tract must have grown by at smaller urbanized area or outside able proportion of the typical residen- least 10 percent in the 1990s (thus urban areas altogether; or they may lie tial development style at the excluding neighborhoods with very lit- outside metropolitan areas, in some metropolitan fringe today, which tle population growth located in cases within micropolitan areas. occurs on lots of at least one acre. declining metropolitan areas). Simi- By using a minimum metropolitan When aggregated to the census tract larly, tracts that grew by at least 3 population threshold, we do not mean level, most developments at this den- times the national rate in the 1990s to indicate that smaller metropolitan sity likely fall within the range we (at least 39.6 percent) are considered areas lack exurbs. Rather, we expect define as exurban.14 Notably, our range to have satisfied this criterion, even if that applying similar criteria to smaller of acceptable housing densities for their metropolitan area grew some- metro areas would yield a very differ- exurbia includes more dense settings what faster (as was the case in metro ent sort of exurb. For instance, com- than those used by Theobald (2001) areas such as Austin, Las Vegas, muting distances from the exurbs of and Irwin and Reece (2002), but simi- Phoenix, and Raleigh). Binghamton, New York (assuming lar settings to those considered exur- Some may argue that exurbs need there are any) are likely much shorter, ban by Nelson and Sanchez (2005; not be growing at all, and that any and housing prices much lower, than equivalent to a range of 1.6 to 5.3 semi-rural community with an impor- in the exurbs of Washington, D.C., acres/per unit). tant economic relationship to an San Francisco, or even Tulsa. Addi- Our use of housing density inter- urban area should be considered exur- tionally, for purposes of this study, sects to a certain degree with density ban. But most popular accounts of aggregating tracts in order to identify measures that underlie our commut- exurbia focus on communities grap- some counties as “exurban” would sac- ing criterion. Specifically, potentially pling with change, often facing rapid rifice refinement in small metro areas, exurban tracts must have an economic growth as they try to preserve their which are much more likely to be connection to an urbanized area rural character. The profile of these

6 October 2006 • The Brookings Institution • Living Cities Census Series areas is probably quite distinct from character of recognizable geographies. muter areas. Because they must, by slow-growing or stagnant fringe com- In Findings D and E, we develop an definition, grow faster than their sur- munities with either recent or long- “urban hierarchy” of metropolitan rounding areas, the typical exurban standing commuting ties to the urban counties that permits us to compare tract experienced rapid population core. (See Box 1 below for an exami- exurban counties with other types of growth in the 1990s. While the nation nation of “slow/no-growth exurbs.”) urban and suburban counties. We grew by 13.2 percent over the decade, Thus, by including population growth explore the demographic, economic, median population growth in exurban in our conceptual model, we seek to and housing characteristics of exurban census tracts was 31.4 percent, more identify exurbs that have undergone versus other metropolitan residents, than double the national rate. Simi- significant recent development and using data derived from Census 2000, larly, housing in exurban areas is built change, where the urban-rural fringe the Census Bureau’s Population Esti- at low densities; with the typical exur- is shifting rapidly. mates Program, the Internal Revenue ban census tract possessing nearly 14 We apply these three criteria to all Service, and the 2004 election results. acres of land per housing unit.19 By census tracts nationwide, using data contrast, the median tract nationwide from the Geolytics® Neighborhood had just 0.8 acres of land per housing Change Database to hold tract bound- Findings unit in 2000. And exurbs are clearly aries constant between 1990 and 2000 bedroom communities, evidenced by to measure population growth. A. In 2000, approximately 10.8 mil- the 52 percent of workers in the typi- Throughout the paper, we refer to lion people lived in the exurbs of cal exurban tract who commute into census tracts that meet these criteria as large metropolitan areas. the urbanized core of a large metropol- “exurbs” or “exurban areas,” and to their Applying the three criteria described itan area. residents as “exurban population” or above to all U.S. census tracts in Most exurbs are located within the “exurbanites.” Certainly, not every resi- 2000, we find that just 2,127 tracts outer reaches of metropolitan areas. dent of these census tracts lives an (3.3 percent of all tracts) can truly be Yet in 2000, nearly 14 percent of exur- “exurban” lifestyle, to the extent that the labeled “exurban” (Table 2). They con- banites lived outside metro areas alto- term denotes commuting long dis- tained total population of 10.8 million gether. A significant share of these tances, living on a large-acre lot, and that year, a little under 4 percent of nonmetropolitan exurbanites (44 per- participating in significant recent national population. Viewed against cent) could be found in newly-desig- growth. However, because both new the backdrop of large metropolitan nated micropolitan areas, smaller and old residents of these areas are areas, to which our conceptual model communities that often adjoin metro- exposed to the growth pressures, devel- affixes exurbs, the population of these politan areas. For instance, LaSalle opment impacts, and economic changes exurban areas equaled a little over 6 County, part of the Ottawa-Streator, taking place in these areas, we feel com- percent of total large metropolitan IL Micropolitan Statistical Area, bor- fortable assigning exurban status to the area population.18 ders the Chicago metropolis and entire population of these tracts. The most limiting of the three crite- houses significant numbers of com- For a portion of the analysis, we ria relates to the census tract’s eco- muters to that urban area. Similarly, aggregate our tract results to the nomic connections. Of the more than Monroe County in eastern Pennsylva- county level to describe the exurban 65,000 census tracts nationwide in nia, which forms the East Strouds- character of counties based on the 2000, just 5,828 (9 percent) qualified burg, PA Micropolitan Statistical Area, proportion of their residents living in under our commuting criterion, which sends commuters to the urban core of census tracts identified as exurban. requires that at least 20 percent of the the nearby Allentown-Bethlehem-Eas- We start from census tracts rather tract’s workers commute to an urban- ton area, as well as to urban portions than counties because the latter vary ized area within a large metro area. Of of the greater New York area. greatly in size across the country; these tracts, 3,831 (66 percent) met If exurbanites represent only 4 per- small counties in the Southeast may the density standard, and 2,952 (51 cent of all Americans, are they such a be largely characterized by one type of percent) expanded sufficiently in the big deal? First, recall that under our development, while massive counties 1990s to meet the population growth classification system, many parts of in the West might span the distribu- criterion. In the end, 2,127 tracts met the nation fail to qualify because they tion from older cities, to mature sub- all three of these criteria. lie at a great distance from a large urbs, to exurbs, to wholly rural Because of the way we define metropolitan area. As such, a better territory.17 exurbs, it comes as no surprise that way to view the incidence of exurbia is County-based analysis does, how- exurban communities are, on average, as 6 percent of large metropolitan pop- ever, allow us to identify the exurban fast-growing, low-density, high-com- ulation. Smaller metropolitan areas

October 2006 • The Brookings Institution • Living Cities Census Series 7 Box 1. Slow/No-Growth Exurbs

ur method for defining exurbia embraces fast growth as a defining characteristic of the exurban experience. This is consistent with most popular accounts of exurbs, which tend to focus on rapidly changing areas near the urban-rural fringe, and our methods help to identify the locales that are making perhaps the greatest impacts on metropolitan development patterns. Moreover, using this additional growth criterion enforces Ogreater similarity across the various communities that we ultimately label “exurbs.” What if instead we had decided that exurbs need not grow fast—or even grow at all? Many past efforts to find exurbia have not included growth among their definitional criteria (see Table 1). In these studies, exurbs are areas of semi-rural character (defined most often by density) that lie within the orbit of big cities and their metropolitan areas, but they may not be growing at all, or they may even be losing population. How might our exurban map change if we were to include these slow/no-growth areas that met our commuting and density criteria only? As indicated in Finding A, there are roughly 1,700 additional census tracts nationwide that would qualify as exurbs if we dropped the growth criterion. In 2000, 6.8 million people lived in these slow/no-growth exurbs. Including these areas would thus boost our overall exurban population by roughly two-thirds. Moreover, it would alter somewhat the geographic distribution of exurbia. The table below shows that, compared to the exurban tracts identified in the text, slow/no-growth exurbs are tilted more heavily towards the Middle Atlantic states, and away from the interior South and the Mountain states. This makes sense, given the slower growth prevailing in the northern part of the United States. Within the Mid- Atlantic, metropolitan areas such as Albany, Buffalo, Harrisburg, Pittsburgh, Rochester, and Syracuse all possess large populations in these types of slow/no-growth locales.

Regional Location and Population of Slow/No-Growth Exurbs, 2000

Slow/No-Growth Exurbs Exurbs (Table 2) Share of National Share of National Division/Region Population 2000 Exurban Population Population 2000 Exurban Population New England 361,552 5.3 494,084 4.6 Middle Atlantic 1,459,919 21.5 1,005,709 9.3 Northeast Subtotal 1,821,471 26.8 1,499,793 13.9

East North Central 1,227,699 18.1 1,790,439 16.6 West North Central 361,445 5.3 835,705 7.8 Midwest Subtotal 1,589,144 23.4 2,626,144 24.4

South Atlantic 1,456,057 21.4 2,235,117 20.8 East South Central 298,290 4.4 922,158 8.6 West South Central 751,144 11.0 1,874,031 17.4 South Subtotal 2,505,491 36.8 5,031,306 46.7

Mountain 134,669 2.0 554,883 5.2 Pacific 748,120 11.0 1,051,152 9.8 West Subtotal 882,789 13.0 1,606,035 14.9

U.S. Total 6,798,895 100.0 10,763,278 100.0

Source: Brookings Institution analysis of Census 2000 data

Collectively, these slow/no-growth exurban census tracts grew by only 5 percent in population between 1990 and 2000, compared to 31 percent growth in the exurbs defined in the text. As such, they seem distinct from the fast-growing metropolitan communities explored throughout this study. Nonetheless, the significant number of people living in these slow/no-growth exurbs across the country suggests that they merit further study and consideration of their unique impacts on metropolitan form.

8 October 2006 • The Brookings Institution • Living Cities Census Series Table 2. Regional Location of Exurban Census Tracts

Ratio of Share of Percent of Total Exurban to National Number of Total Number Tracts that Population in Total Large Metro Large Metro Exurban Census Division Exurban Tracts of Tracts are Exurban Exurban Tracts Population Population Population Population New England 99 3,207 3.1 494,084 13,922,517 10,260,511 4.8 4.6 Middle Atlantic 212 9,974 2.1 1,005,709 39,671,861 32,090,895 3.1 9.3 Northeast 311 13,181 2.4 1,499,793 53,594,378 42,351,406 3.5 13.9

East North Central 382 11,358 3.4 1,790,439 45,155,037 26,895,804 6.7 16.6 West North Central 169 5,106 3.3 835,705 19,237,739 8,070,198 10.4 7.8 Midwest 551 16,464 3.3 2,626,144 64,392,776 34,966,002 7.5 24.4

South Atlantic 381 10,793 3.5 2,235,117 51,769,160 30,144,612 7.4 20.8 East South Central 167 3,941 4.2 922,158 17,022,810 5,446,570 16.9 8.6 West South Central 359 7,108 5.1 1,874,031 31,444,850 18,726,322 10.0 17.4 South 907 21,842 4.2 5,031,306 100,236,820 54,317,504 9.3 46.7

Mountain 127 4,285 3.0 554,883 18,172,295 9,886,618 5.6 5.2 Pacific 231 9,565 2.4 1,051,152 45,025,637 34,716,459 3.0 9.8 West 358 13,850 2.6 1,606,035 63,197,932 44,603,077 3.6 14.9

Nation 2,127 65,337 3.3 10,763,278 281,421,906 176,237,989 6.1 100.0

Source: Brookings Institution analysis of decennial census data may have their own forms of exurbia, rules and regulations that guide (or suburban growth generally far out- but we suspect that they are quite dis- fail to guide) development, and the paced city growth in the 1990s, and tinct from the large-metro exurbs iden- location and number of their large few natural barriers stand in the way tified here. Second, although 6 metropolitan areas. of continued metropolitan decentral- percent may not seem like a large por- Overall, the South and the Midwest ization (Berube 2003, Fulton et al. tion of metropolitan population, there are the most exurbanized regions, 2001). are many regions of the country, and based on the ratio of population in The West and Northeast show a far metropolitan areas within those their exurban tracts to population in lower incidence of exurbia than the regions, where exurbs capture much their large metropolitan areas. In the other regions. Exurbanites in these greater shares of the population. The South, these exurbanites represented regions represent between 3 and 4 next section explores variations across 9.3 percent of large metro area popu- percent of large metropolitan popula- the United States in the prevalence of lation in 2000, and the comparable tion. In particular, the Mid-Atlantic exurbs. proportion in the Midwest was 7.5 (New York, Pennsylvania, and New Jer- percent (Table 2). Within these sey) and Pacific (California, Oregon, B. The South and Midwest are more regions, the East South Central divi- Washington, Alaska, and Hawaii) divi- exurbanized than the West and sion, comprising Kentucky, Tennessee, sions, though they each have more Northeast. Mississippi, and Alabama, exhibited an than 1 million exurban residents, have The 2,127 census tracts we identify as even more highly exurbanized popula- relatively small shares of their large- exurban in 2000 do not distribute tion than the larger South, with nearly metro populations in such areas. Envi- evenly across the country. Some one in six residents living in an exurb. ronmental constraints to outward regions of the country seem to “exur- The same was true of the West North growth in the West, such as deserts banize” more than others, depending Central division (the Dakotas, and mountains, the older urbanized on factors such as historical and con- Nebraska, Kansas, Missouri, Iowa, and development of the Northeast, and temporary settlement patterns, their Minnesota) compared to the overall more similar growth rates between natural constraints to development, Midwest. In these areas of the nation, cities and suburbs make for smaller

September 2006 • The Brookings Institution • Living Cities Census Series 9 shares of population in these regions living in far-flung, low-density exurbs. Table 3. States Ranked by Total Exurban Population, 2000 The South as a whole contains almost 47 percent of the nation’s exur- State Total population Exurban population Percent exurban ban population, far more than any 1 Texas 20,851,820 1,241,472 6.0 other region. The South Atlantic states 2 California 33,871,648 725,921 2.1 alone (coastal states from Maryland to 3 Ohio 11,353,140 465,585 4.1 Florida) contain one-fifth of all exur- 4 Michigan 9,938,444 457,184 4.6 banites. Roughly one-quarter of exur- 5 New York 18,976,457 455,833 2.4 ban population resides in the Midwest, 6 Tennessee 5,689,283 438,600 7.7 with about one in six living in the East 7 Virginia 7,078,515 417,721 5.9 North Central division (the “Rust 8 Maryland 5,296,486 395,084 7.5 Belt” states from Ohio to Wisconsin). 9 South Carolina 4,012,012 379,369 9.5 The rest of exurbia splits fairly evenly 10 Wisconsin 5,363,675 375,881 7.0 between the Northeast and the West. 11 Florida 15,982,378 373,092 2.3 Drilling down to the state level, the 12 Pennsylvania 12,281,054 368,000 3.0 more exurbanized nature of the South 13 North Carolina 8,049,313 360,292 4.5 and Midwest continues to stand out 14 Missouri 5,595,211 353,722 6.3 (Figure 2). South Carolina, Oklahoma, 15 Minnesota 4,919,479 307,425 6.2 Tennessee, and Maryland—all South- 16 Oklahoma 3,450,654 307,354 8.9 ern states—lead all others in the share 17 Washington 5,894,121 261,709 4.4 of their populations living in exurban 18 Georgia 8,186,453 260,191 3.2 communities. The Midwest also places 19 Indiana 6,080,485 255,979 4.2 three states (Wisconsin, Missouri, and 20 Illinois 12,419,293 235,810 1.9 Minnesota) among the 10 most exur- 21 Alabama 4,447,100 224,129 5.0 banized. Due to the happenstance of 22 Massachusetts 6,349,097 218,592 3.4 political boundaries, exurban residents 23 Colorado 4,301,261 199,548 4.6 in some of these states actually com- 24 Kentucky 4,041,769 190,125 4.7 mute to adjacent states for work. For 25 Arizona 5,130,632 187,584 3.7 instance, much of Maryland’s exurban 26 New Jersey 8,414,350 181,876 2.2 population commutes to Washington, 27 Louisiana 4,468,976 180,877 4.0 D.C., and its Northern Virginia sub- 28 Arkansas 2,673,400 144,328 5.4 urbs. Many Wisconsin exurbanites 29 Kansas 2,688,418 132,235 4.9 commute to Minneapolis-St. Paul and 30 Connecticut 3,405,565 127,699 3.7 its Minnesota suburbs. 31 New Mexico 1,819,046 104,302 5.7 Also striking is the rank-order of 32 New Hampshire 1,235,786 83,851 6.8 states by their total exurban popula- 33 Mississippi 2,844,658 69,304 2.4 tions. Texas and California place num- 34 Oregon 3,421,399 52,435 1.5 ber one and number two, respectively, 35 Rhode Island 1,048,319 47,750 4.6 not surprising in light of their overall 36 Nevada 1,998,257 40,912 2.0 size. Yet Ohio, the seventh-largest state 37 Nebraska 1,711,263 29,748 1.7 by overall population, places third on 38 Delaware 783,600 24,697 3.2 total exurban population. Michigan, 39 West Virginia 1,808,344 24,671 1.4 the eighth-largest state, comes right 40 Utah 2,233,169 22,537 1.0 behind in fourth place (Table 3). 41 Maine 1,274,923 16,192 1.3 Notably, seven states and the Dis- 42 Iowa 2,926,324 12,575 0.4 trict of Columbia lack exurbs alto- 43 Hawaii 1,211,537 11,087 0.9 gether. These states—including 44 Alaska 626,932 0 0.0 Vermont, Montana, and Alaska—pos- 45 District of Columbia 572,059 0 0.0 sess no large metropolitan areas, and 46 Idaho 1,293,953 0 0.0 are located at great enough distances 47 Montana 902,195 0 0.0 from other large metropolitan areas 48 North Dakota 642,200 0 0.0 that they do not have any tracts that 49 South Dakota 754,844 0 0.0 50 Vermont 608,827 0 0.0 51 Wyoming 493,782 0 0.0 Total 281,421,906 10,763,278 3.8 10 Source: Brookings Institution analysis of decennial census data metro area altogether. Figure 2. Percentage of Population Living in Exurban Census • A second group, consisting of Tracts by State, 2000 Poughkeepsie and Worcester, represents “satellite metros” linked closely to the adjoining metro areas of New York and Boston. Many residents of these areas live in fast-growing communities where the majority of people work within the county, even as a significant minority commute long distances to adjacent metropolises. (See Box 2.) • A third set of large, Southern metropolitan areas (Birmingham, None Knoxville, Nashville, and Austin) < 2% reflects fast-growing regions where 2% to 3.99% homebuilding seems to have spread 4% to 5.99% quickly into the rural hinterlands. 6% and up At the other end of the spectrum Source: Brookings Institution analysis of Census Bureau data lies a group of areas where exurban Note: Alaska has no exurban population, and less than 1 percent of Hawaii’s population lives in residents amount to fewer than one in exurbs 40 metropolitan residents. These metro areas, too, divide into three groups. qualify on our commuting criterion A second revealing finding is that each • Some face natural barriers, such as (the District, on the other hand, lies at and every one of the 88 large metro mountains, oceans, deserts, and wet- the fully-urbanized heart of the Wash- areas has at least some associated lands, that inhibit the outward, low- ington, DC metro area). Still, these exurban population—from just 4,600 density development characteristic states probably have growing, com- in the New Haven-Milford, CT area to of exurbs (Salt Lake City, San Diego, muter-dominated areas with low-den- 445,000 in the Dallas-Forth Worth San Jose, Honolulu, and Miami).21 sity housing that might form the area (details on exurban populations • New York and Los Angeles anchor subject of future “small-metro exurbs” associated with all 88 large metro dense mega-regions that each have a research. areas can be found in Appendix considerable number of exurbanites. Table A). However, those exurban populations C. Seven metropolitan areas have at Among the top 10 metro areas in are quite small viewed against the least one in five residents living in the ratio of exurban to metropolitan backdrop of total metropolitan popu- an exurb. population, seven actually have exur- lation. Exurbia, at least by our rendering, is a ban-to-metropolitan population ratios • Finally, Youngstown, Pittsburgh, and metropolitan phenomenon. Therefore, of at least 20 percent. This represents New Haven are growing so slowly— examining the incidence of exurbia at more than three times the national or losing population—that relatively the metropolitan level can offer per- average. The top 10 comprise three few of their communities (even at haps the best insights into where somewhat different types of “heavily the metropolitan fringe) have gained exurbs are most common, and what exurban” areas: enough population recently to qual- drives their development.20 • One set contains relatively smaller ify as exurbs. As a first observation, large metro- metro areas with populations Ranked by absolute exurban popula- politan areas vary greatly in their num- between 500,000 and 750,000 (Lit- tion, the Dallas, Minneapolis-St. Paul, ber of exurban residents, and the ratio tle Rock, Grand Rapids, Greenville, and Washington, DC metro areas lead of exurban residents to total metropol- Madison, and Knoxville). These the way. These regions feature promi- itan residents (Table 4). From the metro areas seem to contain signifi- nently in journalistic accounts of exur- most exurban metro area (Poughkeep- cant rural territory within their met- ban change; Frisco (TX), Carver sie, NY) to the least exurban (Miami, ropolitan boundaries, and have large County (MN), and Caroline and King FL), there exists a roughly hundred- numbers of workers commuting into George counties (VA) have all served as fold difference in the exurban ratio. their urban cores from outside the subjects of recent news stories about

October 2006 • The Brookings Institution • Living Cities Census Series 11 Table 4. Top and Bottom Metropolitan Areas by Percent Exurban, 2000

Population, Total Exurban Percent Exurban, Exurban Metropolitan Area 2000 Population, 2000 2000* counties

1 Poughkeepsie-Newburgh-Middletown, NY 621,517 200,728 32.3% 2 2 Little Rock-North Little Rock, AR 610,518 144,328 23.6% 6 3 Grand Rapids-Wyoming, MI 740,482 168,523 22.8% 3 4 Greenville, SC 559,940 123,734 22.1% 1 5 Madison, WI 501,774 110,127 21.9% 3 6 Birmingham-Hoover, AL 1,052,238 224,129 21.3% 5 7 Knoxville, TN 616,079 129,497 21.0% 6 8 Worcester, MA 750,963 149,104 19.9% 0 9 Nashville-Davidson—Murfreesboro, TN 1,311,789 253,100 19.3% 8 10 Austin-Round Rock, TX 1,249,763 221,611 17.7% 5

79 Salt Lake City, UT 968,858 22,537 2.3% 2 80 San Diego-Carlsbad-San Marcos, CA 2,813,833 60,331 2.1% 0 81 Youngstown-Warren-Boardman, OH-PA 602,964 12,870 2.1% 0 82 Pittsburgh, PA 2,431,087 51,035 2.1% 0 83 San Jose-Sunnyvale-Santa Clara, CA 1,735,819 28,185 1.6% 0 84 Honolulu, HI 876,156 11,087 1.3% 0 85 New York-Northern New Jersey-Long Island, NY-NJ-PA 18,323,002 219,667 1.2% 3 86 Los Angeles-Long Beach-Santa Ana, CA 12,365,627 111,511 0.9% 0 87 New Haven-Milford, CT 824,008 4,593 0.6% 0 88 Miami-Fort Lauderdale-Miami Beach, FL 5,007,564 13,478 0.3% 0

Total 176,237,989 10,763,278 6.1% 245

* Ratio of total exurban population to total metropolitan population Source: Brookings Institution analysis of decennial census data the rise of exurbia (Lyman 2005; Curry system fueling it. Notably, Atlanta—a testament to the denser development 2004; Cohn and Gardner 2006). popular example of the problems that characterizes the Pacific coast Analyzing the metro-area results by associated with lower-density devel- (Fulton et al. 2001). As one heads region further illuminates the variable opment patterns—shows up with into California’s interior, however, forms and drivers of exurban develop- only an average degree of exurban- exurbs become more prevalent. ment across the United States. ization. Roughly 85 percent of the Riverside-San Bernardino, Fresno, • South. Consistent with the findings region’s population in 2000 actually Stockton, and Bakersfield possess by region, the vast majority of South- lived inside the massive Atlanta, GA more exurban development relative ern metropolitan areas (26 of 32) urbanized area (mostly at suburban to their populations than their exhibit above-average degrees of densities), implicitly limiting the coastal counterparts. In the Moun- exurbanization. The trend is particu- extent of exurban development. tain West, natural barriers seem to larly pronounced in the Carolinas, Additionally, the region grew by a limit the extent of exurbia in places which place six metro areas among staggering 38 percent in the 1990s, like Salt Lake City, Phoenix, and Las the 20 most exurban. Metro areas in mounting a high growth bar for any Vegas. Yet the same does not hold for the interior South, including those community to qualify as exurban.22 Denver, Tucson, Colorado Springs, in Tennessee, Alabama, Arkansas, • West. In contrast to the South, 12 of and Albuquerque, all of which and Kentucky also rank high, with 20 Western metro areas have a exhibit above-average degrees of few natural barriers standing in the below-average ratio of exurban popu- exurbanization. way of their outward development, lation to metro area population. Sev- • Northeast. As the “top 10” results and an extensive interstate highway eral in California rank very low, a suggest, significant exurban develop-

12 October 2006 • The Brookings Institution • Living Cities Census Series ment can be found in some smaller Cleveland), show fairly low degrees exurbanization has pushed north- “satellite” metropolitan areas in the of rapid exurban development. ward into southern New Hampshire, Northeast. Beyond Poughkeepsie and These findings emphasize that pop- and westward into Worcester Worcester, a host of eastern Pennsyl- ulation growth alone does not deter- County. Within Massachusetts, most vania metro areas (Scranton–Wilkes- mine the degree to which a communities inside Interstate 495 Barre, Allentown-Bethlehem-Easton, metropolitan area develops exurbs. In are either too dense, or too slow- Harrisburg-Carlisle) exhibit above- slow-growth regions of the Northeast growing, to qualify as exurbs. Just average ratios of exurban-to-total- and Midwest, rapid homebuilding at beyond the interstate, however, lies a metropolitan population. Households the fringe may drain population away series of rapidly growing commuter at the fringe of each of these metro from cities and close-in suburbs, fuel- towns, like Lancaster, Sterling, areas commute to their urban cores ing the rise of exurban communities. Bolton, and Harvard, where lower- for work, but others drive to larger In Buffalo, Scranton, St. Louis, and density development predominates. neighboring regions (e.g., New York, Detroit, the pace of homebuilding far Meanwhile, the smaller city of Philadelphia, Baltimore) for their outstripped household growth in the Worcester itself has spawned exur- jobs. While none is growing 1990s (Bier and Post 2003). Alterna- ban development to its west, in quickly—the Scranton area actually tively, proximity to an economically towns like Barre, Hubbardston, and lost population during the 1990s— dynamic region, coupled with a slow Westminster. Several tracts in and each contains neighborhoods where rate of homebuilding in that adjacent around Worcester qualify on our fast development occurred over the region, may force growth into exurbs commuting criterion (yellow areas) previous decade.23 Even more striking in satellite metro areas. The Boston because they still maintain signifi- is the collection of upstate New York and New York areas, in particular, have cant commuting ties to the larger metro areas, including Albany, Buf- experienced declining rates of home- Boston urbanized area, but they are falo, Rochester, and Syracuse, which building in recent decades (Glaeser, too high-density and slow-growing to despite very slow metropolitan Gyourko, and Saks 2005). To gain be exurban. growth or even population loss still access to affordable housing, house- • Chicago (Map 2) has sprouted a contain significant exurban popula- holds seem to have shifted outward to considerable number of exurban tions. exurbs in areas like Worcester, Hart- communities in former farmlands • Midwest. The Midwest places a cou- ford, Poughkeepsie, and Allentown. throughout the metro area, and in ple of metropolitan areas (Grand Similarly, in Southern and Western some instances, beyond. Outside the Rapids and Madison) very high on metro areas, fast growth alone has not metro area’s older counties of Cook the exurban scale. Several more dis- dictated the extent of exurbia. Where and DuPage, exurbs are evident in play high rates of exurbanization, natural barriers to metropolitan decen- the remaining eight counties—most such that 11 of the 18 Midwestern tralization are limited—such as in the notably in McHenry County to the metro areas exhibit above-average Southeast, some areas of the Inter- city’s northwest, where nearly one- exurban ratios. The Detroit, Kansas mountain West, and California’s quarter of residents live in exurban City, and St. Louis regions, in partic- Inland Empire and Central Valley— census tracts. The region’s exurbs ular, each possess more than rapid metropolitan expansion has extend into southern Wisconsin and 200,000 exurban residents. All are brought significant exurbanization. In western Indiana; more than half the losing population in the urban core, other fast-growing metro areas near residents of Jasper County, IN live in but are experiencing rapid, low-den- oceans, deserts, wetlands, and moun- fast-growing, low-density areas with sity growth at the exurban fringe. tains, by contrast, fewer exurbs have significant commuting ties to the Columbus (OH), Cincinnati, Mil- sprung up (e.g., Phoenix, Las Vegas, Chicago area’s urban core. waukee, and Cleveland lag not far coastal California, and Florida). In • The map of the greater Los Angeles behind. The Chicago area, like the many cases, rapid suburban develop- area (Map 3) highlights the region’s Twin Cities, has more than 300,000 ment in these regions has assumed a polycentric nature. The dark green exurban residents, but that repre- denser “boomburb” form than is typi- areas on the map represent exurbs sents a relatively small fraction of cal elsewhere (Lang and Simmons not only to the city of Los Angeles the metro area’s 9 million-person 2003). and its immediate environs, but to reach. Finally, Dayton, Toledo, and Maps of selected regions help to other large cities throughout the Youngstown, largely by virtue of their illuminate the variable number and region, such as Long Beach, Ana- stagnant overall population growth, location of exurban areas across the heim, Thousand Oaks, Riverside, as well as their proximity to other United States: and San Bernardino. The extent of struggling regions (Cincinnati and • In the greater Boston area (Map 1), the dense urban core (grey areas of

October 2006 • The Brookings Institution • Living Cities Census Series 13 Map 1. Exurban Character of Census Tracts in the Greater Boston Area Based on Commuting Ties, Population Growth, and Housing Density

Grafton Carroll Cumberland

Belknap Win d s or I91 MAINE S ulliva n York

I89 Merrimack Strafford

I393

NE W HAM P S HIR E Criteria Met by Tract

None

I293 Rockingham Commuting, not density Cheshire Hillsborough Density, not commuting, not growth I95 Commuting and density, not growth Exurban: All three

Essex

I93 F ranklin Middlesex Boston-Cambridge-Quincy, MA-NH I190 Metropolitan Statistical Area Wo rc e s te r Worcester, MA Metropolitan Statistical Area MAS S ACHUS E TTS I495 S uffolk Hamps hire I290

I90

Norfolk Hampden

I84 I295 Plymouth

Providence Tolla nd Win d h a m I395 Bristol CONNE CTICUT RHODE ISLAND Barnstable

Kent I195 Hartford New L ondon Newport Wa s h i n g to n Dukes

04812162 Miles Source: Brookings Institution analysis of Census 1990 and 2000 data

14 October 2006 • The Brookings Institution • Living Cities Census Series Map 2. Exurban Character of Census Tracts in the Greater Chicago Area Based on Commuting Ties, Population Growth, and Housing Density

Da ne J efferson Waukesha Milwaukee

Racine Criteria Met by Tract Rock Wa l w o rth None I43 Kenosha WIS CONS IN Commuting, not density Density, not commuting, not growth ILLINOIS

I94 Commuting and density, not growth Win n e b a g o Boone McHenry Lake Exurban: All three

I90 Chicago-Naperville-Joliet, IL-IN-WI Metropolitan Statistical Area

Og le

I190 Kane

DeKa lb I88 I290 DuP a g e Cook I355 MICHIGAN Lee I294

Kendall

I80 LaPorte

Will Porter Bureau I39 Lake

La S alle

Grundy I55 Starke

Putnam I57

Kanka kee I65 ILLINOIS INDIANA Pulaski Marshall Jasper

Newton

Livingston

Iroquois Woodford Ford Wh ite

McLean Benton Carroll Ta z e w e l l

03.5 7 14 21 28 Source: Brookings Institution analysis of Census 1990 and 2000 data Miles

October 2006 • The Brookings Institution • Living Cities Census Series 15 I40 Riverside SanDiego SanBernardino Source: Brookings Institution analysis of Census 1990 and 2000 data I10 I215 I15 On ta rio, CA MS A Rivers ide-S an B ernardino- Ora ng e I605 I105 I710 Los Angeles I5 I405 I110 I210 atAaC MSA SantaAna,CA Los Angeles-Long Beach- Based on Commuting Ties, Population Growth, and Housing Density Population Ties, Based on Commuting aesil,CA MSA Bakersfield, Miles Map 3. Exurban Character of Census Tracts in the Greater Los Angeles Area Angeles in the Greater Los Map 3. Exurban Character of Census Tracts Kern Ventura Oxn a rd - None Commuting, not density Density, not commuting, not growth Commuting and density, not growth Exurban: All three Los Angeles-Long Beach- Santa Ana, CA Metropolitan Statistical Area Thous a nd Oa ks - etr,CA MSA Ventura, Criteria Met by Tract 3.75 0 7.5 15 22.5 30

16 October 2006 • The Brookings Institution • Living Cities Census Series Map 4. Exurban Character of Census Tracts in the Greater Washington Area Based on Commuting Ties, Population Growth, and Housing Density

Criteria Met by Tract PENNSYLVANIA None

Commuting, not density I68 Density, not commuting, I83 not growth y Commuting and density, Ha rford not growth I70 Exurban: All three Baltimore Washington-Arlington- Frederick Alexandria, DC-VA-MD-WV I795 Metropolitan StatisticalI81 Area I695 Baltimore-Towson, MD e Metropolitan Statisitical Area Baltimore City

I895 I270 I195

Win c h e s te r MARYLAND I97 I370

Loudoun Montgomery

I495 Anne Arundel

VIRGINIA I595 DC

Arling ton I295 h Prince George's I66 Alexa ndria

Fairfax

Rappahannock P rince William

Ca lvert Fauquier Cha rles Culpeper

Stafford MARYLAND

I95 St.Mary's Fredericksburg VIRGINIA

King George

Spotsylvania We s tm o re l a n d

ville Albemarle Ca roline I64 Louisa Essex

Northu m b e rla n d 03 6 12 18 24 Source: Brookings Institution analysis of Census 1990 and 2000 data Miles

October 2006 • The Brookings Institution • Living Cities Census Series 17 Box 2. Whose Exurb?

n this section, we associate exurban census tracts with the metropolitan areas in which they are located, and in the case of tracts outside metro areas, with the metro area to which they have their most significant commuting ties. As a result, satellite areas like Poughkeepsie, NY and Worcester, MA, rank among the most highly exurban metro areas, in part because many of their residents commute to nearby large metro areas. Relating exurban tracts to metro areas Iin this way maintains consistency with other parts of our analysis, which assign tracts to the regions and states (Findings A and B) and counties (Findings D and E) in which they are located. Under an alternative approach, one could associate each exurban tract not with its “home” metro area, but with the metro area to which it has qualifying commuting ties. In this view, some exurbs in Worcester County, MA would be assigned to the Boston metro area; likewise, exurbs in Dutchess and Orange counties, NY would relate to the New York, rather than Poughkeepsie, metro area. How would a commuting-based assignment change our rankings? The vast majority of metro areas would not change their relative position very much. Only 13 out of the 88 large metro areas change by at least six positions within the list in either direction under this alternative allocation scheme (see table below). Those shedding significant amounts of exurbia under the commuting approach include Worcester and Poughkeepsie, as well Riverside and Oxnard (satellites to the Los Angeles metro area) and Allentown (a satellite to both New York and Philadelphia). Gaining are regions that draw in commuters from surrounding metro areas, including Milwaukee (from Madison and Chicago), Los Angeles, and Boston. Some gains represent the addition of only one census tract (e.g., Dayton, Stockton, and Virginia Beach). In the majority of places, the two methods produce minor differences, or none at all. In the end, viewing exurbs through the lens of their own metropolitan areas puts the issues associated with their growth and development squarely at the feet of policymakers who must contend with these issues on a daily basis. Nonetheless, it is useful to examine how “exurbanized” metro areas like Boston, New York, and Los Angeles appear through the commuting kaleidoscope.

Difference in Percentage Exurban Rank, Location-Based versus Commuting-Based Allocation of Exurban Tracts, Large Metropolitan Areas

Location-Based Commuting-Based Exurban Exurban Metro Area* Population % in Exurbs Rank Population % in Exurbs Rank Exurban Population Increases Boston, MA-NH 130,785 3.0 75 248,363 5.7 52 Milwaukee, WI 116,780 7.8 38 148,708 9.9 28 Stockton, CA 26,432 4.7 64 29,101 5.2 56 Bridgeport, CT 18,985 5.3 57 51,696 5.9 50 Dayton, OH 21,242 2.5 77 28,492 3.4 70 Virginia Beach, VA 95,292 6.0 52 100,654 6.4 45 Los Angeles, CA 111,598 0.9 86 277,066 2.2 80

Exurban Population Decreases Springfield, MA 35,795 5.3 58 28,312 4.2 64 Poughkeepsie, NY 200,728 32.3 1 70,489 11.3 23 Allentown, PA 80,705 10.9 25 44,485 6.0 49 Oxnard, CA 47,920 6.4 48 17,864 2.4 78 Riverside, CA 217,750 6.7 44 87,216 2.7 76 Worcester, MA 149,104 19.9 8 34,428 4.6 62

* Metro area names abbreviated; full names displayed in Appendix A Source: Brookings Institution analysis of Census 2000 data

18 October 2006 • The Brookings Institution • Living Cities Census Series the map) is striking, reflecting the grained view of where, and how much, the number of counties that are 15 to natural barriers to low-density devel- exurban development occurs in and 20 percent exurban (54), and the opment that have limited the overall around metropolitan areas. But ques- number that are 20 to 25 percent degree of exurbanization in the tions regarding who lives in exurbia exurban (37), suggesting that a sort of region. The yellow areas, which are just as important for understand- natural break exists at this threshold maintain significant commuting ties ing what drives the growth of these (see Appendix Figure A). We also com- with more than one of the region’s communities, and what, if anything, pared the list of counties that Lang urbanized areas, occur frequently as differentiates them from the rest of and Sanchez (2006) identify as exur- well. Nodes of exurban development suburbia. ban to our county rankings, and while are nonetheless evident in several our methodological approaches differ, parts of greater Los Angeles, includ- Identifying Exurban Counties we find significant overlap between ing: southern Orange County around Unfortunately, the decennial census is their exurban counties and counties Laguna Niguel; northern Los Ange- practically the only data source that that exceed our 20-percent threshold.24 les County outside of Lancaster and provides tract-level information on Picking 20 percent as our lower- Palmdale; the outskirts of Simi Val- household social and economic char- bound threshold, we identify 245 ley and other areas surrounding acteristics. For some statistics pre- counties (7.8 percent of all U.S. coun- Thousand Oaks; and inner Riverside sented in this and subsequent ties) that have significant exurban County near Corona and Ontario. sections, Census 2000 provides the character. Many of these counties Yet the map may make exurbaniza- most recent information available to cluster around a handful of metropoli- tion in the region look slightly more describe these characteristics. To take tan areas (these counties and some of prevalent than it is, since many dark advantage of newer data, we aggregate their characteristics are listed in green tracts are actually sparsely census tracts to the county level, and Appendix Table B). Louisville lays populated desert and mountain profile so-called exurban counties and claim to the largest number of exurban areas. their inhabitants using a variety of counties, with 13 in its metropolitan • Measured as a proportion of metro- national demographic and economic orbit. Three other metropolitan areas politan population, Washington’s data sources. in the South—Richmond, Washing- exurbs are the largest among these As noted in the Methodology sec- ton, and Atlanta—possess 11 exurban four regions (Map 4). The tentacles tion, census tracts are small enough counties each. Minneapolis-St. Paul of the Washington region reach far areas that designating a whole tract, and Dallas-Ft. Worth both claim 10. south into Spotsylvania County, VA, and 100 percent of its residents, as Together, more than one-quarter of all where they begin to bleed into Rich- “exurban” probably does not create exurban counties lie in and around mond’s northern exurbs; west into considerable error. By contrast, label- these six metro areas. Warren County, VA, at the base of ing a whole county as “exurban” is This clustering reflects not only the the Shenandoah Mountains; north more difficult, since counties are exurban character of certain metro into Frederick County, MD; and to much larger entities, especially in the areas, but also the typical size of coun- the southeast in Calvert, Charles, Western U.S. and New England. ties in their home states. All else and St. Mary’s counties, MD. Balti- To determine a threshold for identi- equal, smaller counties at the metro- more, pictured partially on the map, fying exurban counties, we ranked all politan fringe are more likely than very also exhibits significant exurban U.S. counties on the percentage of large ones to reach the 20-percent development, and its exurbs meet their populations living in exurban exurban threshold. Birmingham, for Washington’s in fast-growing west- census tracts (Appendix Figure A). instance, exhibits a larger exurban ern Howard County, MD. Unlike in Overall, 574 counties contained at population than Louisville (Table 4), Boston, Chicago, and Los Angeles— least one exurban census tract. Of but has only five exurban counties to regions bounded by water on one these, 329 counties had less than 20 Louisville’s 13. Birmingham’s exurban side—development in the Washing- percent of their populations living in counties, however, boast 64,000 peo- ton region is occurring on all four exurban areas, containing 5.1 million ple on average, more than double the sides of the urban core. people (47 percent of total exurban size of the average exurban county population). A lower-bound threshold around Louisville. Notably, the D. Nationwide, 245 counties have of 20 percent to identify exurban Worcester metro area—ranked eighth at least one-fifth of their residents counties, then, captures a slight overall on the ratio of exurban-to-total living in exurban areas. majority (53 percent) of all people liv- population—possesses no exurban Analyzing exurban population at the ing in exurban areas. Furthermore, counties, because it is composed of census tract level provides a fine- there exists a significant drop between only county (Worcester County, MA)

October 2006 • The Brookings Institution • Living Cities Census Series 19 that fails to meet the 20-percent of their life cycles, and are closer to counties in 2000. Furthermore, at threshold. Thus, the number of exur- what Lang and Sanchez (2006) term least 95 percent of residents in these ban counties in any one metro area “emerging suburbs” (see below). counties lived in an urbanized area depends on its geographic extent and In identifying exurbs, we rely on in 2000. Several are urban city- governance arrangements, as well as decennial census data. Six years on counties such as Philadelphia, Balti- the degree of local exurbanization. from Census 2000, however, it is more, and San Francisco. Others are Our list of exurban counties con- worth inquiring whether exurbs Sunbelt counties with large, “elastic” tains several of the “usual suspects” remain growth centers in their respec- central cities such as Houston, identified as such in the popular tive metropolitan areas. To examine Charlotte, and Albuquerque that press—Jackson County north of this and several additional questions maintain urban population densities Atlanta, Scott County outside of Min- regarding the demographic and eco- (Rusk 2003). neapolis-St. Paul, and Fauquier nomic profile of exurbs using more • Inner suburban counties (82 in County outside of Washington, D.C.25 recent data, we divide jurisdictions total) meet either of the criteria to Yet other counties widely cited as within the 88 metropolitan areas into be considered urban, but not both. exurbs, such as Loudoun County, VA; four types of county-based geogra- Some have high population densities Forsyth County, GA; Delaware County, phies:26 but do not contain significant cen- OH; and Pasco County, FL, fail to • Urban counties (40 in total) have tral-city population. Many of these qualify. Each of these counties actu- significant central-city populations, are the “first suburbs” just outside ally has significant population living in and a highly urbanized character. large cities identified by Puentes and exurban census tracts (and significant Specifically, one or more of the 123 Warren (2006), such as Nassau land area consumed by exurban devel- metropolitan central cities identified (NY), Montgomery (MD), and opment) but not quite enough to rep- in Living Cities Census Series publi- Orange (CA). Others contain one or resent 20 percent of total county cations, based on city size and met- more central cities, but have a population. In fact, these counties ropolitan area name, accounted for majority of their populations living have largely passed the exurban stage at least half the population of these outside them, such as Los Angeles

Table 5. Exurban Population Change, 2000 to 2005

Population Change and Components of Change by County Type Population % Change by Component County Type Number 2000 2005 % Change Natural Increase Immigration Internal Migration

Urban 40 44,279,015 45,581,910 2.9 4.3 4.0 -5.8 Inner Suburban 82 69,143,556 71,912,902 4.0 3.6 3.4 -2.9 Outer Suburban 211 50,637,945 55,178,398 9.0 3.0 1.5 4.6 Exurban 245 14,454,037 16,226,633 12.3 2.7 0.7 9.0

Total 578 178,514,553 188,899,843 5.8 3.5 2.8 -0.5

Metro Areas with Largest Absolute Change in Exurban County Population Population Change 2000–2005 Metro Area Exurbs Metro Area % in Exurbs

Washington-Arlington-Alexandria, DC-VA-MD-WV 139,681 409,300 34.1 Orlando-Kissimmee, FL 121,572 276,867 43.9 Minneapolis-St. Paul-Bloomington, MN-WI 113,401 166,333 68.2 Atlanta-Sandy Springs-Marietta, GA 108,500 649,632 16.7 Houston-Sugar Land-Baytown, TX 91,653 538,657 17.0 Dallas-Fort Worth-Arlington, TX 88,814 629,290 14.1 Chicago-Naperville-Joliet, IL-IN-WI 79,397 323,634 24.5

Source: Brookings analysis of Census 2000 and Census Population Estimates Program data

20 October 2006 • The Brookings Institution • Living Cities Census Series (CA), Hamilton (OH), and Allegheny Bureau 2006). Outer suburban coun- exurban population gain. The Wash- (PA). Still others have a large central ties grew by the largest absolute ington area contains three urban juris- city, but are built at lower, suburban- amount, adding 4.5 million people in dictions, six inner suburban style densities, such as Pima (AZ), 5 years, compared to 1.8 million peo- jurisdictions, 11 exurban counties, and Guilford (NC), and Travis (TX). On ple added to exurban counties. five outer suburban counties (Figure balance, the suburban character of In fact, within metropolitan areas, 3). Between 2000 and 2004, there was these counties outweighs their urban population growth in outer suburban a clear hierarchy of migration within character. counties sometimes outpaced that in the region. Urban counties—specifi- • Outer suburban counties (211 in exurban counties. In the Dallas metro cally, the District of Columbia, Arling- total) include the residual counties area, for example, the outer suburban ton, VA, and Alexandria, VA—lost within the 88 large metropolitan counties of Collin, Delta, Denton, and considerable population on net to areas that do not meet the criteria Rockwall combined for a nearly 30 inner suburban jurisdictions. The for any of the other three county percent growth rate between 2000 and inner suburbs, in turn, also experi- types (including exurban counties— 2005, compared to 15 percent in Dal- enced net migration losses within the see below). Lang and Sanchez las’ exurbs. The Austin and Jack- region as they re-distributed migrants (2006) term most of these either sonville areas exhibited similar growth to outer suburbs and exurbs. The “mature suburbs” or “emerging sub- advantages “inside” of their far-flung largest net movement was from the urbs,” and they include growing exurbs. Of the 58 large metro areas inner suburbs to the outer suburbs, counties such as Anne Arundel (MD, that had associated exurban counties, particularly from Fairfax County, VA, outside Baltimore), Collin (TX, out- 25 had their fastest population to the nearby Virginia counties of side Dallas), Pinal (AZ, outside increases in other county types. This Loudoun and Price William.28 Phoenix), and Forsyth (GA, outside pattern suggests that in many metro- But exurbs gained from these inner Atlanta). These counties do not politan areas, exurbs remain at the suburban counties, too, to the tune of always lie at the outer edges of their edge of the growth envelope, where 43,000 people from 2000 to 2004. metropolitan areas, but the title con- rapid smaller-lot, suburban-style devel- They also drew residents on net from notes their position in the metropolis opment has yet to take root. In the outer suburban counties, over 22,000 relative to inner suburban counties, Twin Cities, Orlando, and Washing- during that time period. What is more, and the fact that they are distinct ton, D.C. areas, however, exurbs have Washington-area exurbs actually expe- from exurbs, which lie outside the captured a very significant proportion rienced a small net loss of residents to suburban range. of recent metropolitan population other parts of the United States during • Exurban counties (245 in total) are growth (Table 5). this period, further highlighting the those described above, both inside What is the source of exurbs’ con- importance of the inner parts of the and outside large metropolitan areas, tinuing growth? In large measure, Washington metropolis as a source of that have at least 20 percent of their these counties are growing thanks to its exurbs’ population gains. populations living in census tracts domestic in-migration. Between 2000 that satisfy the three exurban and 2005, they grew by 9 percent due E. The “average” exurb is dispropor- criteria. to internal migration alone, far out- tionately white, middle-income, stripping comparable rates in the other homeowner-dominated, and com- Recent Exurban County Growth county types (Table 5). By contrast, muter-oriented. Exurban areas are defined in part by they experienced little migration from Among those acquainted with exurbia, their rapid population growth between abroad. And natural increase—the theories and assumptions about its 1990 and 2000. County-level analysis excess of births over deaths—has profile abound. Some commonly suggests that their growth advantage added relatively less population to advanced notions hold that exurbs are has continued into the current decade. exurban counties than to other county characterized by: Among the four county types, exurban types in recent years. • Large non-Hispanic white popula- counties grew fastest overall between Census Bureau population esti- tions; 2000 and 2005, by 12.3 percent mates do not provide information on • Mostly married-couple family house- (Table 5). This was more than double the exact geographic source of county holds, especially with children; the rate for the 578 counties com- in-migrants. But IRS county-to-county • High-income newcomers mixing bined. Thirty-eight (38) exurban coun- migration data, applied to the Wash- with moderate-income longtime resi- ties actually ranked among the 100 ington, D.C. region as a case-study dents; fastest-growing in the United States example, offer a picture of within-met- • Workers making long-distance during this period (U.S. Census ropolitan migration’s importance to “super-commutes” back into the

October 2006 • The Brookings Institution • Living Cities Census Series 21 urban core, and workers telecom- muting and working from home; Figure 3. Net migration patterns within the Washington, D.C. • A real estate-driven economy, with metropolitan area, 2000 to 2004 large numbers of workers employed 16,107 in construction, landscaping, real U)rban Counties (2005 pop: 881,823 7,496 estate brokerage, and home-related D)istrict of Columbia, Arlington (VA), Alexa ndria city (VA retail; 61,355 • Enormous cookie-cutter “McMan- 3,261 sions” abutting old farmhouses; and I)nner Suburban Counties (2005 pop: 2,862,170 • Reliably Republican voters Montgomery (MD), P rince George’s (MD), F airfax (VA), F airfax 15,731 city (VA), F alls Church city (VA) Having defined and identified exurbs, we explore in this section the 72,592 accuracy of these hypotheses about O)uter Suburban Counties (2005 pop: 618,311 exurbs and their residents. Table 6 dis- Clarke (VA), Loudoun (VA), P rince William (VA), Manassas city 3,712 plays the characteristics of the “aver- (VA)*, Manassas P ark city (VA)* age” exurb, one exhibiting the 23,067 population-weighted mean values on several of the indicators explored E)xurban Counties (2005 pop: 925,043 Calvert (MD), Charles (MD), F rederick (MD), Culpeper (VA), below, compared to large-metropolitan Fauquier (VA), King George (VA), Orange (VA), S potsylvania 2,329 averages. In many respects, the aver- 43,272 (VA), S tafford (VA), Warren (VA), J efferson (WV) age exurban area seems to live up to Migration into/out of region Migration within region popular expectations. Of course, there is no one “average” Source: Brookings Institution analysis of IRS data. *The city-county equivalents of Manassas and Manassas Park, VA have population densities that qualify them as “inner suburban,” but are labeled exurb, and the more detailed explo- as “outer suburban” here because they lie wholly within outer suburban Prince William County, VA rations below suggest that exurbanites do not always differ dramatically from their suburban counterparts. It also Hispanic whites than other large- points overall, and dropped in 187 of bears emphasizing that not all resi- metro county types. In 2004, over 83 245 exurban counties.29 dents of exurbs exhibit exurban percent of exurban-county residents “behavior.” That is, many people in were non-Hispanic white, compared to Household types—slightly more these places lived there long before 61 percent of large-metro residents “nuclear” families new homebuilding and fast growth overall (Table 7). With the exception The results from Census 2000, the lat- began to occur. And while some exur- of American Indians, each race and est available for examining household ban counties have the majority, or ethnic group—including people of two types at the county level, confirm the even all, of their residents living in or more races—comprised a lower hypothesis that exurbs do contain exurban census tracts, others merely share of population in exurbs than in higher proportions of “nuclear” fami- meet the 20-percent threshold. Thus, any other county type. Moreover, the lies than other types of places. The dif- the overall characteristics of exurban- vast majority of exurban counties— ferences are not dramatic, however. county residents explored here con- 211 of 245 (86 percent) had a white Overall, about 28 percent of exurban- flate the profiles of newcomers, population share that exceeded the county households are married-couple long-time residents of exurbs, and national average of 67 percent. Blacks families with children, about 2 per- adjoining areas that do not exhibit (7.4 percent) constituted a slightly centage points higher than in outer exurban character. Using these county higher share of exurban-county popu- suburban counties, and 4 percentage proxies, however, enables us to com- lation than Hispanics (6.5 percent). points higher than the large-metropoli- pare a much wider, up-to-date set of Although whites are still over-repre- tan average (Figure 4). The exurban demographic and economic indicators sented in exurban counties relative to advantage is somewhat larger, actually, in exurbs versus other metropolitan their share of the nation’s population, with respect to childless married cou- locations. exurban counties actually grew slightly ples. These may represent a mix of more diverse in the 2000s. Between younger professional couples who have Race and ethnicity—predominantly 2000 and 2004, they gained roughly 1 not yet had children, and empty-nester white communities million new white residents, but over- and retiree households whose children Exurban counties do have a signifi- all non-Hispanic white share of popu- have left home.30 cantly higher representation of non- lation declined by 1.6 percentage People living alone do, notably,

22 October 2006 • The Brookings Institution • Living Cities Census Series make up a far smaller proportion of The income profile of exurban coun- from those in the poorest exurban households in exurban (21 percent) ties largely mirrors that in other subcounty (Torrance County, NM) in than urban (30 percent) counties. urbs, with one notable difference. 2003 (Table 8). In the counties at the Not surprisingly, then, the average Exurbs tilt somewhat more towards top of the list, roughly one in six exurban household (2.67 persons) is middle- and upper-middle-income households broke the $100,000 bar- slightly larger than that in other types households than their suburban neigh- rier, compared to roughly one in 40 at of counties. bors, with a greater percentage of their the bottom of the list. Large metro residents in the $40,000 to $60,000, areas such as Washington, New York, Income—largely middle-income and $60,000 to $100,000 AGI cate- Chicago, and the Twin Cities tend to Exurban areas have households at all gories. Inner and outer suburbs had have wealthy exurbs, while low-income income levels, but retain a somewhat relatively more high-income house- exurbs ring somewhat smaller regions more middle-income profile than sub- holds with AGI of at least $100,000. in the South such as Knoxville, Mem- urbs. The most recent evidence avail- Of course, these statistics mask sig- phis, Charlotte, and Tulsa. able on household incomes by county nificant variation in the income pro- comes from tax return data reported files of exurban counties themselves. Commuting—more early risers and by the Internal Revenue Service (IRS), Nearly $27,000 separated the typical “super-commuters” which contains information on tax- incomes of households in the richest Commuting patterns form part of our payer Adjusted Gross Incomes (AGI).31 exurban county (Stafford County, VA) definition of exurbia. At the census tract level, at least 20 percent of workers must commute to a large urbanized Table 6. “Average” Exurban County vs. Large Metropolitan Area area in order for the tract to be consid- Characteristics ered exurban (in addition to its meeting density and growth criteria). Thus, Exurban counties (%) All large-metro counties (%) we might expect that commuting pat- % population change (2000-05) 12 6 terns in exurban counties are distinct % non-Hispanic white (2004) 83 61 from those in other types of metropoli- % married-with-children (2000) 28 24 tan counties. % with AGI over $100,000 (2003) 8 11 This turns out to be the case in % working outside county of residence (2000) 51 29 some, though not all, respects. A large % commuting > 1 hour each direction (2000) 11 9 proportion of workers from exurban % in manufacturing or construction (2000) 26 19 counties worked outside their county % homeowners (2000) 78 63 of residence in 2000—roughly half % of new houses with 9+ rooms (2000) 14 20 overall (Table 9). This is a far higher % voting for Bush (2004) 63 47 share than in other metropolitan counties (reflecting in part the smaller Source: Brookings Institution analysis of Census 2000, Population Estimates Program, average geographic size of exurban Internal Revenue Service, and 2004 election data counties, as well as our method for identifying exurbs). Only 11 percent of

Table 7. Race/Ethnic Profile by County Type, 2004

Urban Inner Suburban Outer Suburban Exurban Total (race/ethnicity) Population % Population % Population % Population % Population %

White 21,237,954 47.0 39,615,756 55.3 40,187,525 74.0 13,236,748 83.4 114,277,983 61.1 Black/African American 10,549,626 23.3 9,033,773 12.6 5,114,065 9.4 1,177,314 7.4 25,874,778 13.8 American Indian 240,034 0.5 495,159 0.7 271,221 0.5 104,351 0.7 1,110,765 0.6 Asian/Pacific Islander 2,991,678 6.6 5,718,352 8.0 1,699,035 3.1 165,837 1.0 10,574,902 5.7 Two or More Races 568,986 1.3 1,207,137 1.7 699,923 1.3 163,183 1.0 2,639,229 1.4 Hispanic 9,624,373 21.3 15,538,624 21.7 6,324,603 11.6 1,029,743 6.5 32,517,343 17.4

Source: Brookings Institution analysis of Population Estimates Program data

October 2006 • The Brookings Institution • Living Cities Census Series 23 exurban workers commute into an urban county, however; most only go Figure 4. Household Types by County Type, 2000 as far as the inner or outer suburbs.32 To reach their jobs, exurban residents are more likely than workers from urban or inner suburban counties to drive alone to work, though they are very similar in this respect to residents of outer suburban counties. They are slightly more likely than commuters from other county types to carpool. Only a small fraction—less than 1 percent—use public transit, compared to 16 percent in urban counties, reflecting that most exurban areas have no viable public transportation options. Contrary to the notion that exurbs are hotbeds for “telecommuters,” there Urban Inner Suburban Outer Suburban Exurban is no evidence that exurbanites work from home at higher rates than other metropolitan workers.33 The vast majority who work outside the home do, however, face longer commutes— Source: Brookings Institution analysis of Census 2000 data; roughly one in nine exurban com- “Children” refers to own children under the age of 18 muters travels more than an hour in each direction to reach his/her job.34 Park and Pike counties, at the far edges of the Denver and New York Figure 5. Taxpayers by Adjusted Gross Income and metro areas, respectively, have the County Type, 2003 highest proportions of these “super- commuters” among exurban counties (Table 10). Exurban commuters also hit the road earlier than other metropolitan workers, with roughly one in seven starting his/her commute before 6 AM. “Early riser” exurbs include a mix of counties at the edge of traffic- snarled metro areas like Washington and Houston, where exurbanites might depart early to beat the rush, and counties where agricultural jobs remain important, around metro areas like Louisville.

Employment—more manufacturing, construction, and farming jobs Urban Inner Suburban Outer Suburban Exurban In On Paradise Drive, David Brooks portrays the archetypal exurban household headed by a married couple, “Patio Man” and “Realtor Mom.” Not only is one member of the couple Source: Brookings Institution analysis of IRS data employed in the real estate industry,

24 October 2006 • The Brookings Institution • Living Cities Census Series Table 8. Top and Bottom Exurban Counties by Median Adjusted Gross Income, 2003 but their personal endeavors revolve County Metro Area* Median AGI ($) % over 100k around the home—having the per- 1 Stafford, VA Washington, DC-VA-MD-WV 47,214 17.5 fectly manicured lawn, a spacious 2 Scott, MN Minneapolis-St. Paul, MN-WI 46,218 15.3 backyard deck, the most powerful gas 3 Sussex, NJ New York, NY-NJ-CT-PA 45,673 16.4 grill available on the market. 4 Calvert, MD Washington, DC-VA-MD-WV 45,531 17.5 Brooks’ scene is mostly caricature, 5 Carver, MN Minneapolis-St. Paul, MN-WI 45,215 17.5 of course. But it hints at what many 6 Fauquier, VA Washington, DC-VA-MD-WV 44,907 18.2 consider to be powerful home-related 7 McHenry, IL Chicago, IL 44,698 15.8 economic drivers in exurbia, such as 8 Livingston, MI Detroit, MI 44,618 16.5 homebuilding, home sales, and home 9 Elbert, CO Denver, CO 44,080 14.3 improvement. 10 Frederick, MD Washington, DC-VA-MD-WV 42,804 15.6 While they do not allow us to look at detailed industries like home- 236 Sevier, TN Knoxville, TN 22,074 4.0 related retail, or architecture and land- 237 Atascosa, TX San Antonio, TX 22,059 3.1 scaping, Census 2000 data do reveal 238 Waller, TX Houston, TX 21,938 4.6 that exurban-county residents are 239 Marshall, MS Memphis, TN-AR-MS 21,880 2.5 more likely to work in construction 240 Saluda, SC Columbia, SC 21,799 2.5 than other metropolitan residents. In 241 Anson, NC Charlotte, NC-SC 21,348 2.3 2000, 9 percent of workers from exur- 242 Okmulgee, OK Tulsa, OK 21,189 2.2 ban counties were employed in the 243 Union, TN Knoxville, TN 21,125 2.0 construction industry; compared to 6 244 Washington, MO St. Louis, MO-IL 21,110 1.3 percent of workers from other metro- 245 Torrance, NM Albuquerque, NM 20,429 2.0 politan counties (Figure 6). Construc- tion accounted for an especially All Exurban Counties 31,605 8.4 significant portion of jobs in boom- county exurbs outside of Virginia Source: Brookings Institution analysis of IRS data Beach (Currituck County, NC); Den- *Names abbreviated; full names displayed in Appendix Table A ver (Park and Elbert counties, CO); and Atlanta (Pickens and Walton counties, GA). The proportion of exurban-county workers employed in real Table 9. Commuting Characteristics by County Type, 2000 estate, however, appeared no different than in other metropolitan counties. Percentage of Workers (unless otherwise noted) Only in Summit County, UT (east of Urban Inner Suburban Outer Suburban Exurban Salt Lake City) were more than 5 per- Counties (number) 40 82 211 245 cent of workers employed in that industry. Work Location Besides construction, two other Work outside county of residence 22.0 22.6 35.4 50.8 industries appear to exhibit a loca- Work in an: tional preference for exurban counties. Urban county 87.1 8.6 10.7 11.7 Farming, not surprisingly, is more Inner Suburban county 6.3 85.1 12.1 14.8 important in the exurbs, though less so Outer Suburban county 5.0 4.5 72.7 13.9 than in prior decades—only about 2 Exurban county 0.7 0.7 1.4 53.1 percent of exurban residents were Other county 1.0 1.1 3.1 6.5 employed in agriculture and mining in 2000. Manufacturing looms largest Commuting Mode among exurban industries, accounting Drive alone 63.5 74.4 80.8 80.6 for a much higher proportion of jobs Carpool 12.2 12.2 10.9 12.7 (17 percent) than in the remainder of Public transit 15.7 6.2 2.1 0.8 large metropolitan areas (13 percent) Work at home 2.8 3.3 3.2 3.2 (Figure 6). This makes sense, given the land intensity of manufacturing, Commuting Time and the decreasing price of land as At least 60 minutes (in one direction) 10.3 8.9 7.7 11.5 Leave before 6 AM 9.8 9.8 10.8 14.4

Source: Brookings Institution analysis of Census 2000 data 25 Table 10. “Super-Commuter” and “Early Riser” Exurbs, 2000

Super-Commuter Exurbs Workers with 60+ Super-commuter County Associated metro area* Total commuters minute commute share (%)

1 Park, CO Denver, CO 7,110 2,344 33.0 2 Pike, PA New York, NY-NJ-CT-PA 18,643 5,712 30.6 3 Warren, VA Washington, DC-VA-MD-WV 14,994 4,346 29.0 4 San Jacinto, TX Houston, TX 7,891 2,156 27.3 5 Charles, MD Washington, DC-VA-MD-WV 60,032 15,225 25.4 6 Calvert, MD Washington, DC-VA-MD-WV 36,187 9,127 25.2 7 Amelia, VA Richmond, VA 5,326 1,334 25.0 8 Van Zandt, TX Dallas, TX 18,759 4,675 24.9 9 Elbert, CO Denver, CO 9,863 2,429 24.6 10 Culpeper, VA Washington, DC-VA-MD-WV 15,469 3,804 24.6

Early Riser Exurbs Workers departing Early riser commuter s County Associated metro area* Total commuters before 6 AM hare (%) 1 Crawford, IN Louisville, KY-IN 4,377 1,234 28.2 2 Juniata, PA Harrisburg, PA 9,915 2,733 27.6 3 Nye, NV Las Vegas, NV 11,660 3,211 27.5 4 Washington, MO St. Louis, MO-IL 8,331 2,260 27.1 5 Warren, VA Washington, DC-VA-MD-WV 14,994 4,050 27.0 6 Liberty, TX Houston, TX 25,409 6,525 25.7 7 San Jacinto, TX Houston, TX 7,891 2,003 25.4 8 Bibb, AL Birmingham, AL 7,754 1,923 24.8 9 East Feliciana, LA Baton Rouge, LA 7,280 1,784 24.5 10 Breckinridge, KY Louisville, KY-IN 7,518 1,805 24.0

Source: Brookings Institution analysis of Census 2000 data * Names abbreviated; Full names displayed in Appendix Table A one moves outward from the urban communities. More than three-fourths with long-time residents of these com- core (Glaeser and Kahn 2001). of exurban-county residents in 2000 munities who inhabit homes of far These job figures provide a snapshot owned their home, a significantly more modest value. of exurban employment as it existed in higher proportion than in any other This stereotype, however, overlooks 2000. As these areas continue to metropolitan county type. the evidence that exurbs serve as develop residentially, some may One type of exurbanite homeowner, sources of affordable homeownership acquire more diversified, higher-value the sort described by Spectorsky in many metropolitan areas. In these industries and occupations that reflect (1955), Nelson (1992), and Brooks regions, middle-income families may the growing demands and skills of (2004), has relocated to the urban be “driving to qualify” for a home in their populations. In the end, some fringe to enjoy a countryside or small- their price range, one that does not may shed their exurban status and town atmosphere, while retaining exist in closer-in suburbs. Lacking morph into outer suburban employ- access to metropolitan amenities such detailed information on new home val- ment centers, even as others remain as shopping, culture, and high-paying ues at the county level, we examine more rural and single-industry jobs. These newcomers might live in average incomes for owners of new dependent in character. enormous new houses in gated subdi- homes in Census 2000—those built visions, or in restored farmhouses on between 1995 and 2000—as a proxy Housing—more affordable, modestly- large tracts of land that enable them to for the value of the homes purchased sized homes engage in “hobby farming.” They may by those households. Exurbs are, by and large, homeowner find themselves living side-by-side Overall, owners of new exurban

26 October 2006 • The Brookings Institution • Living Cities Census Series homes had significantly lower average Figure 6. Employment by Industry, Residents of Exurban incomes ($68,790) than owners of Counties versus Other Metropolitan Counties, 2000 new urban ($93,627), inner suburban ($100,010), and outer suburban ($88,618) homes. Thus, while outer suburbs seem slightly more affordable overall than urban or inner suburban 5% 2% 5% 9% counties, new homes in the exurbs were occupied by families with 7% incomes on average $20,000 lower than in outer suburbs. 17% In theory, this pattern could derive from the disproportionate number of 10% exurban counties located in the South, where incomes tend to be lower. But even within metro areas, high-income households tended to buy nearer the 4% 7% core, while middle-income households bought toward the fringes. In fast- Agriculture & Mining growing areas like Denver and Wash- 8% 12% Construction ington, DC, incomes for owners of new homes are highest in the urban 6% 2% 6% Manufacturing core, and decrease with county dis- Wholesale Exurban Counties tance from the core (Figure 7).35 In Retail slower-growing areas, especially those Transp/Utilities with a struggling urban core, new Information owner incomes tend to be highest in Finance/Real Estate inner and outer suburbs, while exurbs Education appeal to a middle- and upper-middle- Professional/Admin income market. In the Richmond 5% 1% 6% 5% Health metro area, new exurban homeowners Arts/Rec/Hospitality have incomes nearly equal to those in 8% 13% the outer suburbs; in the Philadelphia Public area, the exurban market is somewhat Other more middle-income. 4% Statistics on the size of new homes 11% that exurbanites buy are consistent with their more middle-income status. Based on the fact that the median- 11% sized home nationwide constructed between 1995 and 2000 contained 11% just under six rooms, we categorize “super-sized” homes as those contain- 5% ing nine or more rooms. One in six 4% 8% 8% homes built during that period were that large. In exurbs, these super-sized Other Metropolitan Counties homes made up one in seven recently- constructed houses in 2000. The proportion of homeowners in inner (21 percent) and outer suburbs (23 per- Source: Brookings Institution analysis of Census 2000 data cent) that occupied “super-sized” new homes in 2000 was roughly 50 percent higher than in exurbs. Thus, it seems

October 2006 • The Brookings Institution • Living Cities Census Series 27 that exurbs generally offer more inter- mediate, affordable new home types Figure 7. Average Household Income, Owners of than do metropolitan suburbs.36 Newly Built Homes (1995–2000), Selected Metro Areas Housing affordability, of course, is a by County Type, 2000 function of both homebuyer incomes and home prices. Viewed through this lens, exurban homes do appear somewhat more affordable to their inhabitants than homes elsewhere in metropolitan areas. In 2000, just under one-quarter of all exurban homeowners with mortgages spent

more than 30 percent of their income Average Income ($) on housing, a standard metric for assessing housing-cost burdens (Fig- ure 8). This was a lower proportion than in urban, inner suburban, and outer suburban counties. About 45 percent spent a modest amount (less than 20 percent) of income on hous- Denver Philadelphia Richmond Washington, D.C. Metropolitan Area* ing costs. Interestingly, inner suburban counties appeared to have the highest housing costs relative to resident Source: Brookings Institution analysis of Census 2000 data incomes, with almost one-third of * Not all metro areas contain all county types households spending over 30 percent of income on those costs. separated the two candidates (this Summary Politics—heavily Republican, but small county category overlaps most closely This section demonstrates that in sev- vote total with Lang and Sanchez’s “emerging eral ways, exurban-county residents Among all topics related to exurbia, suburbs”). So reliably Republican were can indeed be distinguished from their perhaps the most ink has been spilled the exurbs that only 19 of the 242 metropolitan neighbors. They are analyzing their political leanings (see counties cast a majority of their ballots more likely to: be non-Hispanic white; Background). In particular, Teixeira for Kerry. live in a “nuclear” family; face daily (2006) acknowledges that Republican As Teixeira (2004, 2005, 2006) has “super-commutes” of an hour or more voters dominate exurbia, but points to noted, however, the exurbs still form a each way; work in construction or Lang and Sanchez’s (2006) “emerging small piece of the electoral pie. Our manufacturing; and vote Republican. suburbs” as the political battleground analysis lends further support to this In other important ways, however, of tomorrow. view. Of the more than 75 million exurbanites do not look that different, Our method for identifying and votes cast in large metropolitan coun- or actually confound some popular classifying exurbs and other county ties for either of the candidates in stereotypes. For instance, the exurbs types differs from Teixeira’s (ours also 2004, fewer than 7 million (9 percent) do not appear to be overrun with real applies to a greater number of metro- originated in exurban counties.38 Elec- estate agents. Nor are they generally politan areas), but our basic conclu- toral dynamics in the inner and outer destinations for high-income home- sions echo his.37 In 2004, 63 percent suburbs, the source of more than two- owners living in super-sized new man- of votes cast in exurban counties for thirds of large-metro votes, will likely sions. In many metropolitan areas, either of the two major-party presiden- play a much more significant role in though, the growth of exurbs may be tial candidates went to George W. shaping future election results than related to middle-income families’ Bush. This formed the mirror image of shifts in voting behavior occurring at “drive to qualify” for more affordable the situation in urban counties, where the exurban fringe. Differing demo- new homes that are in limited supply John Kerry captured a little under 63 graphic drivers of growth in and elsewhere. percent of the vote. The margin was among these county types may influ- As to individual families’ reasons for closest in outer suburban counties, ence their future political path.39 locating in the exurbs, and their aspi- where less than 6 percentage points rations for life there, these nationwide

28 October 2006 • The Brookings Institution • Living Cities Census Series data are merely suggestive. There is Figure 8. Percentage of Income Spent on Housing Costs, clear variation among exurban coun- Homeowners with Mortgage, by County Type, 2000 ties themselves on many of the indicators described here, which forms the subject of the next section.

Exurban Discussion

n this paper, we have attempted Outer Suburban to “find exurbia” in an effort to inform not just news stories about the exurbs, but the very Inner Suburban Ipolicies that shape metropolitan development patterns. We find that exurbs emerge to varying degrees in different regions of the country, and serve Urban somewhat different functions depending on their location. Consequently, it is difficult to generalize about the policy factors that drive exurbanization, or curb it, and we have only begun to scratch the surface on these issues. As Source: Brookings Institution analysis of Census 2000 data a next step for research, we would rec- ommend “unpacking” the various types of exurban areas in order to dig deeper into the variable causes of their Figure 9. Presidential Vote Totals by Candidate and County growth, and the consequences of that Type, 2004* growth for efforts to achieve sustain- able metropolitan development. This might involve, among other strategies, interviewing new households in different types of exurbs to discover their 55.1% motives for relocating (see Davis and Nelson (1994)). 44.9% Thus far, we have been studiously 52.9% neutral in our descriptions of exurbs, 47.1% 62.6% in order to identify and describe them as objectively as possible. Yet evidence and observation suggest that exurbs Total votes Total represent a sub-optimal form of devel- 37.4% opment.

63.1% To be sure, some exurbanites prefer the acreage and semi-rural lifestyle 36.9% that living “out there” offers, and are more than willing to endure long commutes to their jobs to secure it. For Urban Inner Suburban Outer Suburban Exurban some, exurbs satisfy an ever-growing demand for new, large housing. When offered a range of conventional hous- *Democratic and Republican candidates only ing options, most potential homebuy- Source: Brookings Institution analysis of Census 2000 data ers surveyed (from 71 to 83 percent) prefer single-family, detached

October 2006 • The Brookings Institution • Living Cities Census Series 29 dwellings (Fannie Mae Corporation 1997; National Association of Home Affordable Exurbs—Examples Builders 1999). Over the past 45 Boston Strafford County, NH years, the size of the average new sin- Dallas Henderson and Hunt counties, TX gle-family home increased from 1,200 New York Dutchess County, NY and Pike County, PA to 2,343 square feet (U.S. Census Philadelphia Cecil County, MD Bureau 2005). Portland Yamhill County, OR But given a greater range of viable housing options, many exurban residents (perhaps with the exception of those living in “recreation exurbs,” rates to vary greatly. Policymakers and expensive, and perhaps out of many described below) might prefer to live planners have only recently begun to families’ reach. Considering that most elsewhere in the metropolis, closer to consider housing affordability in the of these metro areas are growing at a jobs, in safe communities with good context of the transportation costs moderate pace or faster, they have a schools and nearby amenities. Prefer- associated with neighborhoods in need for new housing. Yet a range of ence surveys reflect continued which new homes are located. policy or planning factors may limit demand among homebuyers for large The profile of exurbs developed in the supply of housing generally, and lots (57 percent preferring more than this paper points toward the existence affordable housing specifically, in one acre), but also the ability to walk of at least a few different exurban closer-in jurisdictions, pushing some- to stores and restaurants (51 percent) types, explored below. A full review of what affordable development to the and have reasonable daily commutes the policy levers that may promote or urban fringe. Some local governments, (79 percent desiring 45 minutes or curb exurbanization is beyond the including many in the greater Boston, less) (National Association of Realtors scope of this descriptive analysis. For Denver, and San Francisco Bay areas, 2004). Much exurban development each of the exurban types described, use exclusionary tools like low-density- seems to be satisfying the first prefer- however, we identify policy areas that only zoning and permit caps that ence in lieu of the other two. deserve further scrutiny from state, severely limit the supply of apartment Moreover, a significant body of local, and regional leaders seeking to buildings and affordable housing (Pen- research evidence demonstrates that understand the differing drivers of dall, Puentes, and Martin 2006). In such low-density development patterns exurbanization. This is by no means an other cases, building codes in cities are fiscally inefficient, and forego the exhaustive account of all types of and first suburbs may drive up the productivity gains that smarter growth exurbs that exist today, nor do we costs of rehabilitation and re-use of development patterns can generate attempt to place every exurban county older housing stock, or effectively pre- (Muro and Puentes 2004). It is not into one of these categories. This dis- vent redevelopment altogether (Burby, clear that these negative externalities cussion instead aims to illustrate the Salvesen, and Creed 2006). are fully captured in the prices that diverse range of exurbs and their resi- Inclusionary zoning requirements new exurban residents pay for their dents, and seeks to provoke additional and special subsidies for key public- homes and land.40 research into why these places have sector workers have emerged as Nor are the transportation costs emerged and how public policy should responses to the lack of affordable associated with exurban living readily respond. housing in cities and older suburbs, apparent to many residents. New but these tools are still used relatively research is demonstrating that the Affordable exurbs. infrequently across the United States. increased transportation costs that In a number of large metropolitan In addition to considering programs encumber households in exurban loca- areas, particularly those on the East such as these, policymakers contend- tions may severely mitigate the bene- and West coasts, exurbs appear to ing with rapid growth in “affordable fits that accrue to those households offer new housing that is considerably exurbs” should also update local meas- from the less expensive housing found more affordable to middle-income ures of housing affordability to reflect there. In a recent Brookings publica- homebuyers than new housing else- the full range of location costs, includ- tion, the Center for Transit-Oriented where in the metro area. This does not ing transportation. Development and Center for Neigh- mean that the cities or older suburbs borhood Technology (2006) found that in these areas have no affordable Recreation exurbs. fewer transportation options and housing stock whatsoever, but rather Access to natural amenities, such as longer commutes in outer parts of the that most owner-occupied housing lakes, mountains, forests, and temper- Twin Cities area cause affordability closer to the core tends to be quite ate climate, has long been recognized

30 October 2006 • The Brookings Institution • Living Cities Census Series part of the “favored quarter” for met- Recreation Exurbs—Examples ropolitan residents. This may owe in Austin Burnet and Blanco counties, TX (Texas Hill Country) part to the decentralized nature of Denver Park and Gilpin counties, CO (Rocky Mountains) these metropolitan economies, as the Providence Washington County, RI (Narragansett Bay and location of a majority of their jobs far Long Island Sound) from the central-city downtown Seattle Island County, WA (Puget Sound) reduces commute times for exurban Stockton Calaveras County, CA (Sierra Nevada foothills) residents (Glaeser, Kahn, and Chu 2001). They may also offer a perceived refuge from the economic and social distress that afflicts their far-away cen- as a major driver of growth in rural future growth in these types of exurbs; tral cities. areas (McGranahan 1999). As most the test for state, regional, and local In most of the metro areas that fea- exurban areas lie at the intersection of planners will be to accommodate that ture these favored-quarter exurbs, rural and urban America, many have growth in a balanced and efficient way. housing unit growth outpaced house- experienced rapid growth precisely hold growth significantly over the because they are proximate to these Favored-quarter exurbs. 1990s (Bier and Post 2003), and the attractive settings. Indeed, about one- In a minority of metro areas, exurbs trend appears to have continued into tenth of the exurban counties identi- are not exactly the affordable-housing the current decade. Thus, exurban fied in this report score high on haven for middle-income homebuyers. growth is occurring largely at the McGranahan’s Natural Amenity Index. They feature more upscale homes, expense of other parts of the region, as In many of these locales, including occupied by high-income households jurisdictions compete to acquire the those listed below, second homes—for seeking the semi-rural lifestyle enjoyed tax base that high-income residents recreational or seasonal use—account by Spectorsky’s exurbanites some 50 bring. Residents of every region, even for at least one-fourth of the housing years ago. Many large and prosperous those with no net household growth, stock. metro areas noted above—like New will exhibit some demand for new and Thus, even with a healthy local York, Washington, and the San Fran- better housing, particularly if their recreation-based economy, these cisco Bay Area—have these types of incomes are rising (Gottlieb 2002). A places exhibit important economic exurban communities. lack of coordinated local land use connections to their nearby large met- But these exurbs are somewhat planning, however, may drive new ropolitan areas. While other exurbs more prominent in slow-growth areas housing growth to the urban-rural have grown due to land-use conver- of the country that have underper- fringe. In these metropolitan areas, sions—from agriculture to housing formed the rest of the nation economi- first suburbs and central cities that are development—the recreation exurbs cally in recent decades. In these losing higher-income households to have undergone more of a “housing regions, new exurban homeowners are the exurbs might consider joining conversion,” from seasonal to year- very nearly as wealthy as their coun- forces to push for statewide and round living. Advances in technology terparts in inner or outer suburbs, and regional growth management reforms and changes in the way people work generally much better-off than those that require greater local collabora- have enabled more workers to estab- in the city. Their houses are typically tion, and bring the pace of develop- lish their primary residences in these quite large, with significant acreage. ment into line with projected growth locales. As an educated, high-income These exurbs have effectively become (Puentes and Warren 2006). Baby Boom generation moves into its pre-retirement and semi-retirement years, development will likely continue to proceed apace in these counties, Favored-Quarter Exurbs—Examples and their metropolitan connections Baltimore Carroll County, MD may grow even stronger. New chal- Cleveland Geauga County, OH lenges will undoubtedly arise for Detroit Livingston County, MI efforts to preserve the resource sys- Hartford Tolland County, CT tems that make these recreation New Orleans St. Tammany Parish, LA exurbs such valued locations in the Richmond Goochland County, VA first place. Overall, growing incomes and natural beauty virtually guarantee

October 2006 • The Brookings Institution • Living Cities Census Series 31 Many other exurbs do not fit neatly efficient manner as these places con- References into any of the categories explored tinue to add housing and jobs. Other- above. Exurbs around cities like wise, low-density “build-out” could Bai, Matt. “Who Lost Ohio?” The New Louisville, Nashville, and Tulsa, for fuel rapid growth farther out in the York Times Magazine. November 21, instance, seem to provide new homes region, creating even more severe eco- pp. 67–74. that are affordable, but not signifi- nomic and environmental challenges. cantly less expensive than those in the Beale, Calvin and Kathleen Kassel. inner or outer suburbs of those metro- 2005. “Perspectives on Rural Popula- politan areas. Counties such as DeS- Conclusion tion.” Presentation to Brookings Insti- oto (MS), Wagoner (OK), and Nelson tution Metropolitan Policy Program, (KY) are not blessed with considerable fter so much analysis, it is Washington, May 26. natural advantages. Neither do they worth returning to a ques- form part of the “favored quarter” in tion posed much earlier in Berube, Alan. 2003. “Gaining But Los- declining regions, since they are the paper: Are exurbs that ing Ground: Population Change in located in areas of the country that are Abig a deal? We find that the roughly 11 Large Cities and Their Suburbs.” In B. growing. million residents of exurbia amount to Katz and R. Lang, eds., Redefining Still, several metropolitan areas fea- just 6 percent of the population of Urban and Suburban America: Evi- ture exurbs like these that simply lie large metropolitan areas to which they dence from Census 2000, volume 1. within the path of rapid residential attach. As such, exurbs are probably Washington: Brookings. development. For instance, the exurbs receiving a slightly outsized share of of Virginia Beach and Norfolk con- attention nationally. Politically, these ––––––––. Forthcoming. “Job Sprawl: tained 11 percent of that region’s pop- jurisdictions were not alone responsi- An Update.” Washington: Brookings ulation in 2000, but have captured 26 ble for the outcome of the 2004 presi- Institution. percent of recent housing growth. dential election (Teixeira 2005). Some Around Memphis, exurban counties might be tempted to joke that exur- Bier, Thomas and Charlie Post. 2003. contained 21 percent of metropolitan banites are nothing more than “fringe “Vacating the City: An Analysis of New population in 2000, but have reaped elements” in their metropolitan areas, Homes vs. Household Growth.” Wash- 38 percent of new housing since then. and hardly represent the norm. ington: Brookings Institution. The pattern is repeated in several If exurbs signal the possible shape other metropolitan areas, especially of things to come, however, that Blumenfeld, Hans. 1986. “Metropolis those in the Midwest and South where makes them all the more important to Extended: Secular Changes in Settle- few natural barriers impede continued understand today. Spectorsky’s The ment Patterns.” Journal of the Ameri- outward growth. Over the past two Exurbanites is a fascinating period can Planning Association 52: 346–348. decades, these areas have consumed piece, but it failed to foresee a con- significant amounts of formerly agri- temporary New York metropolis in Brooks, David. 2004. On Paradise cultural lands to accommodate popu- which Poughkeepsie—another 50 Drive: How We Live Now (And Always lation growth. The Nashville metro miles north of Rockland County— Have) in the Future Tense. New York: area, for instance, urbanized an aver- forms the exurban frontier. Absent Simon and Schuster. age of nearly one acre of land between continued research and policy focus as 1982 and 1997 for each additional to the causes and consequences of ––––––––. 2004. “For Democrats, resident (and even more per house- exurban growth, we might find our- Time to Meet the Exurban Voter.” The hold) of the region (Fulton et al. selves wondering in 2050 how New New York Times. November 10, p. E9. 2001).41 York’s exurbs arrived in Albany. We Local and regional leaders in areas hope that the foregoing analysis pro- Brownstein, Ronald and Richard like these should recognize the distinct vides a useful baseline for further Rainey. 2004. “GOP Plants Flag on possibility that these exurbs are well inquiry, and helps state and local lead- New Voting Frontier.” Los Angeles on their way to graduating from exur- ers wrestle with some of the economic, Times, November 22, p. 1. ban status into what Lang and environmental, fiscal, and quality-of- Sanchez (2006) term “emerging sub- life issues posed by their fast-growing urbs.” As such, political and corporate fringe communities. leaders must consider how future metropolitan growth can be accommo- dated in a more compact, fiscally

32 October 2006 • The Brookings Institution • Living Cities Census Series Burby, Raymond J., David Salvesen, Fannie Mae Corporation. 1997. Gottlieb, Paul. 2002. “Growth With- and Michael Creed. 2006. “Encourag- National Housing Survey. Washington out Growth: An Alternative Economic ing Residential Rebuilding with Build- (www.fanniemae.com/global/pdf/med Development Goal for Metropolitan ing Codes: New Jersey’s Experience.” ia/survey/survey1997.pdf [July Areas.” Washington: Brookings Insti- Journal of the American Planning Asso- 2006]). tution. ciation 72 (2): 183–196. Frey, William H. 2004. “The Fading of Hayden, Dolores. 2003. Building Sub- Center for Transit-Oriented Develop- City/Suburb and Metro/Nonmetro urbia: Green Fields and Urban Growth ment and Center for Neighborhood Distinctions in the United States.” In 1820–2000. New York: Pantheon. Technology. 2006. “The Affordability T. Champion and G. Hugo, eds., New Index: A New Tool for Measuring the Forms of Urbanization: Beyond the Irwin, Elena and Jason Reece. 2002. True Affordability of a Housing Rural-Urban Dichotomy. Burlington, “Urbanization Trends in Ohio: Track- Choice.” Washington: Brookings Insti- VT: Ashgate. ing Ohio’s Urban Growth and Land tution. Use Change.” College of Food, Frey, William H. et al. 2006. “Tracking Agricultural, and Environmental Clark, Jill, Ron McChesney, Darla K. American Trends into the Twenty-First Sciences, Ohio State University. Munroe, and Elena G. Irwin. 2006. Century: A Field Guide to the New (http://aede.osu.edu/programs/ “Exurban Settlement Pattern and the Metropolitan and Micropolitan Defini- exurbs/reportex4.htm [December Exurban Condition: A Typology of tions.” In A. Berube, B. Katz, and R. 2005]). U.S. Metropolitan Areas.” Paper pre- Lang, eds., Redefining Urban and Sub- sented at the 53rd Annual North urban America: Evidence from Census Lamb, Richard. 1983. “The Extent American Meetings of the Regional 2000, volume 3. Washington: Brook- and Form of Exurban Sprawl.” Growth Science Association International, ings Institution Press. and Change January, 40–47. Toronto, CA. Fulton, William, Rolf Pendall, Mai Lang, Robert and Patrick Simmons. Cohen, Darryl T. and Keith G. Deb- Nguyen, and Alicia Harrison. 2001. 2003. “Boomburbs: The Emergence of bage. 2004. “Research Note: Toward “Who Sprawls the Most? How Growth Large, Fast-Growing Suburban Cities.” an Effective Subcounty Settlement Patterns Differ Across the U.S.” Wash- In B. Katz and R. Lang, eds., Redefin- Classification: Comparative Density ington: Brookings Institution. ing Urban and Suburban America: Evi- Revisited.” Urban Geography 25 (7): dence from Census 2000, volume 1. 675–684. Gersh, Mark. 2004. “Battlefield Ero- Washington: Brookings. sion.” Blueprint Magazine, December Cohn, D’Vera and Amy Gardner. 2006. 13. Lang, Robert and Dawn Dhavale. “Three Virginia Exurbs Near Top of 2005. “The 2005 Governor’s Race: A U.S. in Growth.” The Washington Post, Glaeser, Edward L. and Matthew E. Geographic Analysis of the ‘Four Vir- March 16, p. A1. Kahn. 2001. “Decentralized Employ- ginias.’” Alexandria, VA: Metropolitan ment and the Transformation of the Institute at Virginia Tech. Curry, Tom. 2004. “Bush Hopes for a American City.” Working Paper 8117. Minnesota Breakthrough.” MSNBC Cambridge, MA: National Bureau of Lang, Robert and Meghan Zimmer- Interactive, October 11 Economic Research. man Gough. 2006. “Growth Counties: (www.msnbc.msn.com/id/6221915/ Home to America’s New Metropolis.” [April 2006]). Glaeser, Edward L., Matthew E. Kahn, In A. Berube, B. Katz, and R. Lang, and Chenghuan Chu. 2001. “Job eds., Redefining Urban and Suburban Davis, Judy S. and Arthur C. Nelson. Sprawl: Employment Location in U.S. America: Evidence from Census 2000, 1994. “The New ‘Burbs: The Exurbs Metropolitan Areas.” Washington: vol. 3. Washington: Brookings Institu- and their Implications for Planning Brookings Institution. tion Press. Policy.” Journal of the American Plan- ning Association 60 (1): 45–59. Glaeser, Edward L., Joseph Gyourko, Lang, Robert and Thomas Sanchez. and Raven Saks. 2005. “Urban 2006. “The New Metro Politics: Inter- Downs, Anthony. 2004. “The Costs of Growth and Housing Supply.” Discus- preting Recent Presidential Elections Sprawl Revisited.” Speech before the sion Paper Number 2062. Cambridge, Using a County-Based Regional Typol- Urban Land Institute District Council MA: Harvard Institute of Economic ogy.” Alexandria, VA: Metropolitan Meeting, Washington, April 15. Research. Institute at Virginia Tech.

October 2006 • The Brookings Institution • Living Cities Census Series 33 Lessinger, Jack. 1987. “The Emerging ––––––––. 2005. “The Effectiveness U.S. Census Bureau. 2006. “Popula- Region of Opportunity.” American of Urban Containment Regimes in tion Estimates for the 100 Fastest Demographics (June): 32–66. Reducing Exurban Sprawl.” DISP 160: Growing Counties with 10,000 or 42–47. more Population in 2005: April 1, Lyman, Rick. 2005. “In Exurbs, Life 2000 to July 1, 2005 (CO-EST2005- Framed by Hours Spent in the Car.” Orfield, Myron. 2002. American Met- 09).” March 16. New York Times, Dec 18, ropolitics. Washington: Brookings Institution Press. U.S. Department of Agriculture. 2004. McGranahan, David. 1999. “Natural “Measuring Rurality: Rural-Urban Amenities Drive Rural Population Pendall, Rolf, Robert Puentes, and Continuum Codes.” Economic Change.” Agricultural Economic Jonathan Martin. 2006. “Guiding Research Service. Available at Report 781 (October). Growth: Landscapes of Regulation in www.ers.usda.gov/Briefing/Rural- U.S. Metropolitan Areas.” Washing- ity/RuralUrbCon/ [December 2005]. Mieszowski, Peter and Edwin S. Mills. ton: Brookings Institution. 1993. “The Causes of Metropolitan ––––––––. 2005.”Rural-Urban Com- Suburbanization.” Journal of Economic Puentes, Robert and David Warren. muting Area Codes.” Economic Perspectives 7(3): 135–147. 2006. “One-Fifth of the Nation: Amer- Research Service. (www.ers.usda.gov/ ica’s First Suburbs in Transition.” Data/RuralUrbanCommutingArea- Mikelbank, Brian A. “A Typology of Washington: Brookings Institution. Codes/ [December 2005]). U.S. Suburban Places.” Housing Policy Debate 15 (4): 935–964. Rusk, David. 2003. Cities Without Wolman, Harold et al. 2005. “The Suburbs. Washington: Woodrow Wil- Fundamental Challenge in Measuring Muro, Mark and Robert Puentes. son Center Press Sprawl: Which Land Should Be Con- 2004. “Investing in a Better Future: A sidered?” The Professional Geographer Review of the Fiscal and Competitive Spectorsky, A.C. 1955. The Exurban- 57 (1): 94–105. Advantages of Smarter Growth Devel- ites. Philadelphia: J.B. Lippincott opment Patterns.” Washington: Brook- Company. ings Institution. Texeira, Ruy. 2004. “The Spatial Dis- National Association of Home tribution of Bush’s Vote Gains.” Builders. 1999. Housing Survey. Wash- (www.emergingdemocraticmajority- ington. weblog.com/donkeyrising/archives/0 00942.php [December 2005]). National Association of Realtors. 2004. American Community Survey. ––––––––. 2005. “More on Exurbia.” Washington (www.realtor.org/ (www.emergingdemocraticmajority- PublicAffairsWeb.nsf/files/ weblog.com/donkeyrising/archives/0 charts-release.pdf/$FILE/ 01007.php [December 2005]). charts-release.pdf [July 2006]). ––––––––. 2006. “The Next Frontier: Nelson, Arthur C. 1992. “Characteriz- A New Study of Exurbia.” Washington: ing Exurbia.” Journal of Planning Lit- New Politics Institute. erature 6 (4): 350–368. Theobald, David. 2001. “Land-Use Nelson, Arthur C. and Thomas W. Dynamics Beyond the American Sanchez. 1997. “Exurban and Subur- Urban Fringe.” Geographical Review ban Households: A Departure from 91 (3): 544–564. Traditional Location Theory?” Journal of Housing Research 8 (2): 249–276.

34 October 2006 • The Brookings Institution • Living Cities Census Series Appendix Figure A. Number of counties by percentage exurban, and cumulative share of exurban population, 2000

160 110%

Number of counties Cumulative share of exurban population 142 100% 140

90%

120 80%

70% 100

60%

80 70 50% 63 Number of counties

60 54 40% Cumulative share of exurban population of exurban share Cumulative

30% 37 40 35 29 26 20% 20 20 17 14 14 12 11 10% 7 7 66 2 2 0 0% 0 to 5 to 10 to 15 to 20 to 25 to 30 to 35 to 40 to 45 to 50 to 55 to 60 to 65 to 70 to 75 to 80 to 85 to 90 to 95 to 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100%

Share of population in exurban tracts

Source: Brookings Institution analysis of decennial census data

October 2006 • The Brookings Institution • Living Cities Census Series 35 Appendix Table A. Large Metropolitan Areas by Percent Exurban, 2000

Exurban Exurban Total Pop Pop Exurban Percent Population Inside Outside Pop, Exurban, Exurban Metropolitan Area Region 2000 Metro Metro 2000 2000* Counties 1 Poughkeepsie-Newburgh-Middletown, NY NE 621,517 174,747 25,981 200,728 32.3% 2 2 Little Rock-North Little Rock, AR S 610,518 121,049 23,279 144,328 23.6% 6 3 Grand Rapids-Wyoming, MI MW 740,482 110,777 57,746 168,523 22.8% 3 4 Greenville, SC S 559,940 91,553 32,181 123,734 22.1% 1 5 Madison, WI MW 501,774 80,537 29,590 110,127 21.9% 3 6 Birmingham-Hoover, AL S 1,052,238 208,935 15,194 224,129 21.3% 5 7 Knoxville, TN S 616,079 85,341 44,156 129,497 21.0% 6 8 Worcester, MA NE 750,963 141,883 7,221 149,104 19.9% 0 9 Nashville-Davidson—Murfreesboro, TN S 1,311,789 236,654 16,446 253,100 19.3% 8 10 Austin-Round Rock, TX S 1,249,763 209,366 12,245 221,611 17.7% 5 11 Louisville, KY-IN S 1,161,975 190,029 12,821 202,850 17.5% 13 12 Columbia, SC S 647,158 96,191 13,875 110,066 17.0% 6 13 Tulsa, OK S 859,532 140,503 5,060 145,563 16.9% 6 14 Raleigh-Cary, NC S 797,071 117,303 14,585 131,888 16.5% 2 15 Greensboro-High Point, NC S 643,430 83,346 21,252 104,598 16.3% 1 16 Charleston-North Charleston, SC S 549,033 75,323 10,319 85,642 15.6% 1 17 Oklahoma City, OK S 1,095,421 154,160 7,631 161,791 14.8% 4 18 Harrisburg-Carlisle, PA NE 509,074 61,989 11,152 73,141 14.4% 2 19 Richmond, VA S 1,096,957 153,501 0 153,501 14.0% 11 20 Charlotte-Gastonia-Concord, NC-SC S 1,330,448 95,481 73,235 168,716 12.7% 3 21 Minneapolis-St. Paul-Bloomington, MN-WI MW 2,968,806 342,512 27,900 370,412 12.5% 10 22 San Antonio, TX S 1,711,703 211,624 0 211,624 12.4% 7 23 Albuquerque, NM W 729,649 74,276 10,804 85,080 11.7% 2 24 Kansas City, MO-KS MW 1,836,038 190,225 20,453 210,678 11.5% 9 25 Allentown-Bethlehem-Easton, PA-NJ NE 740,395 64,992 15,713 80,705 10.9% 1 26 St. Louis, MO-IL MW 2,698,687 259,470 22,618 282,088 10.5% 8 27 Memphis, TN-MS-AR S 1,205,204 125,307 0 125,307 10.4% 5 28 Baton Rouge, LA S 705,973 72,071 0 72,071 10.2% 4 29 Colorado Springs, CO W 537,484 50,133 4,536 54,669 10.2% 1 38 Milwaukee-Waukesha-West Allis, WI MW 1,500,741 88,582 60,126 148,708 9.9% 1 30 Orlando-Kissimmee, FL S 1,644,561 135,910 22,863 158,733 9.7% 2 31 Columbus, OH MW 1,612,694 128,261 17,127 145,388 9.0% 5 32 Cincinnati-Middletown, OH-KY-IN MW 2,009,632 163,406 13,231 176,637 8.8% 8 33 Dallas-Fort Worth-Arlington, TX S 5,161,544 337,807 106,710 444,517 8.6% 10 34 Baltimore-Towson, MD S 2,552,994 197,819 17,884 215,703 8.4% 2 35 Wichita, KS MW 571,166 45,043 2,693 47,736 8.4% 2 36 Tucson, AZ W 843,746 58,956 11,012 69,968 8.3% 0 37 New Orleans-Metairie-Kenner, LA S 1,316,510 108,806 0 108,806 8.3% 2 39 Jacksonville, FL S 1,122,750 80,126 6,105 86,231 7.7% 4 40 Washington-Arlington-Alexandria, DC-VA-MD-WV S 4,796,183 317,066 39,743 356,809 7.4% 11 41 Scranton—Wilkes-Barre, PA NE 560,625 12,901 27,742 40,643 7.2% 1 42 Houston-Sugar Land-Baytown, TX S 4,715,407 330,153 2,870 333,023 7.1% 5 43 Seattle-Tacoma-Bellevue, WA W 3,043,878 142,489 68,721 211,210 6.9% 1 44 Riverside-San Bernardino-Ontario, CA W 3,254,821 217,750 0 217,750 6.7% 0 45 Denver-Aurora, CO W 2,179,240 100,237 44,642 144,879 6.6% 4 46 Cleveland-Elyria-Mentor, OH MW 2,148,143 124,039 15,993 140,032 6.5% 2 47 Detroit-Warren-Livonia, MI MW 4,452,557 244,192 44,469 288,661 6.5% 4 48 Oxnard-Thousand Oaks-Ventura, CA W 753,197 47,920 0 47,920 6.4% 0 49 Fresno, CA W 799,407 32,962 16,764 49,726 6.2% 0 50 Atlanta-Sandy Springs-Marietta, GA S 4,247,981 236,041 24,150 260,191 6.1% 11 51 Syracuse, NY NE 650,154 34,498 5,170 39,668 6.1% 0 52 Virginia Beach-Norfolk-Newport News, VA-NC S 1,576,370 89,483 5,809 95,292 6.0% 7 53 Hartford-West Hartford-East Hartford, CT NE 1,148,618 50,782 18,000 68,782 6.0% 1

36 October 2006 • The Brookings Institution • Living Cities Census Series Appendix Table A. Large Metropolitan Areas by Percent Exurban, 2000 (continued)

Exurban Exurban Total Pop Pop Exurban Percent Population Inside Outside Pop, Exurban, Exurban Metropolitan Area Region 2000 Metro Metro 2000 2000* Counties 54 Rochester, NY NE 1,037,831 53,608 5,675 59,283 5.7% 1 55 Omaha-Council Bluffs, NE-IA MW 767,041 42,323 0 42,323 5.5% 3 56 Portland-Vancouver-Beaverton, OR-WA W 1,927,881 85,987 16,947 102,934 5.3% 1 57 Bridgeport-Stamford-Norwalk, CT NE 882,567 18,995 28,108 47,103 5.3% 0 58 Springfield, MA NE 680,014 35,795 0 35,795 5.3% 0 59 Albany-Schenectady-Troy, NY NE 825,875 30,980 11,255 42,235 5.1% 0 60 Indianapolis, IN MW 1,525,104 68,138 7,534 75,672 5.0% 4 61 Buffalo-Niagara Falls, NY NE 1,170,111 52,789 4,803 57,592 4.9% 0 62 El Paso, TX S 679,622 13,285 19,222 32,507 4.8% 0 63 Akron, OH MW 694,960 22,643 10,179 32,822 4.7% 0 64 Stockton, CA W 563,598 14,600 11,832 26,432 4.7% 1 65 Toledo, OH MW 659,188 28,046 0 28,046 4.3% 0 0 66 Tampa-St. Petersburg-Clearwater, FL S 2,395,997 94,285 0 94,285 3.9% 1 67 Bakersfield, CA W 661,645 25,326 0 25,326 3.8% 0 68 Providence-New Bedford-Fall River, RI-MA NE 1,582,997 57,922 0 57,922 3.7% 1 69 Phoenix-Mesa-Scottsdale, AZ W 3,251,876 109,171 8,445 117,616 3.6% 0 70 Chicago-Naperville-Joliet, IL-IN-WI MW 9,098,316 273,226 44,632 317,858 3.5% 7 71 Sarasota-Bradenton-Venice, FL S 589,959 20,325 0 20,325 3.4% 0 72 Philadelphia-Camden-Wilmington, PA-NJ-DE-MD NE 5,687,147 156,823 33,670 190,493 3.3% 1 73 Sacramento—Arden-Arcade—Roseville, CA W 1,796,857 50,636 5,380 56,016 3.1% 0 74 McAllen-Edinburg-Mission, TX S 569,463 7,802 9,610 17,412 3.1% 0 75 Boston-Cambridge-Quincy, MA-NH NE 4,391,344 102,722 28,063 130,785 3.0% 1 76 Las Vegas-Paradise, NV W 1,375,765 16,176 24,736 40,912 3.0% 1 77 Dayton, OH MW 848,153 16,786 4,456 21,242 2.5% 0 78 San Francisco-Oakland-Fremont, CA W 4,123,740 90,795 11,929 102,724 2.5% 0 79 Salt Lake City, UT W 968,858 18,880 3,657 22,537 2.3% 2 80 San Diego-Carlsbad-San Marcos, CA W 2,813,833 60,331 0 60,331 2.1% 0 81 Youngstown-Warren-Boardman, OH-PA MW 602,964 12,870 0 12,870 2.1% 0 82 Pittsburgh, PA NE 2,431,087 51,035 0 51,035 2.1% 0 83 San Jose-Sunnyvale-Santa Clara, CA W 1,735,819 16,808 11,377 28,185 1.6% 0 84 Honolulu, HI W 876,156 11,087 0 11,087 1.3% 0 85 New York-Northern New Jersey-Long Island, NY-NJ-PA NE 18,323,002 178,384 41,283 219,667 1.2% 3 86 Los Angeles-Long Beach-Santa Ana, CA W 12,365,627 111,511 0 111,511 0.9% 0 87 New Haven-Milford, CT NE 824,008 4,593 0 4,593 0.6% 0 88 Miami-Fort Lauderdale-Miami Beach, FL S 5,007,564 7,261 6,217 13,478 0.3% 0

Total 176,237,989 9,306,451 1,456,827 10,763,278 6.1% 245

* Ratio of total exurban population to total metropolitan population Source: Brookings Institution analysis of decennial census data

October 2006 • The Brookings Institution • Living Cities Census Series 37 in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area B. Profile of Exurban Counties, Listed by Metro Appendix Table County NMTorrance, NMValencia, NJWarren, GABarrow, Albuquerque, NM GABartow, Albuquerque, NMArea Related Metropolitan Butts, GA Allentown-Bethlehem-Easton, PA-NJCoweta, GA Atlanta-Sandy Springs-Marietta, GADawson, GA Atlanta-Sandy Springs-Marietta, GAJackson, GA GAJasper, Atlanta-Sandy Springs-Marietta, GA Atlanta-Sandy Springs-Marietta, GALumpkin, GA Atlanta-Sandy Springs-Marietta, GANewton, GA population Atlanta-Sandy Springs-Marietta, GA 36,220Pickens, GA areas (%) GAWalton, 14,673 Atlanta-Sandy Springs-Marietta, GA Atlanta-Sandy Springs-Marietta, GA 2005 (%)Bastrop, TX 2004 (%) 35,111 Atlanta-Sandy Springs-Marietta, GA 35.4 2000 (%)Blanco, TX 14,694 Atlanta-Sandy Springs-Marietta, GA 2003 ($) 10,888 31.8Burnet, TX 40,221 34,282 2000 (%) Atlanta-Sandy Springs-Marietta, GA 15,999Caldwell, TX 46.2 2000 (%) 7.2 Austin-Round Rock, TXHays, TX 86.9 14,148 2000 ($) 28.8 55.8 45.1Carroll, MD Austin-Round Rock, TX 51.8 100.0 16.3 4,481Queen Anne’s, 5,367 89.4MD Austin-Round Rock, TX Austin-Round Rock, TX 79.8 34.0 LA 12,996 3.1East Feliciana, 21.9MD Baltimore-Towson, 6.8 21.2 83.8Livingston, LA 14,603 4.4 27.9 MD Baltimore-Towson, Baton Rouge, LA 39.2 Austin-Round Rock, TX 25.5 Baton Rouge, LAWest 16,048 Baton Rouge, LA 30.7 57.2 24.7 76.1 21.0 69.4 LA Feliciana, West 40,136 95.1 29.7 39.3 63.5 31,280 Baton Rouge, LABibb, AL 14.3 14.9 Baton Rouge, LA 86.9 26.7 29,546 26.4Blount, AL 30.9 24.6 37.8 27.4 18.8 27.5Chilton, AL 20,429 21.7 14.6 72.4 91.2 36,622 27,438 ALShelby, 27.8 32,458 47,340 24,359 22.8 13.1 65.2 ALSt. Clair, 72.8 94.4 30,790 AL Birmingham-Hoover, 3,873 21.2Strafford, NH 75.5 24.9 25.5 AL Birmingham-Hoover, 10.1 13.0 81.4 16.1 AL 13,837 SC Birmingham-Hoover, 8,372 93,963 Berkeley, 27.4 82.0 75.2 8.2 27,330 26,797 59,601 16,542Anson, NC 24.3 AL Birmingham-Hoover, 11.5 83.9 30,285 AL Birmingham-Hoover, 46.0 64,171 78.0Iredell, NC 76.6 MA-NH 78,338 48,372 Boston-Cambridge-Quincy, 29.5 81.4 19.9 32,144 43.0 24.5 SCYork, 83.9 44,203 18.5 13.0 Charleston-North Charleston, SC 77,155 40.8 51,505 34,405 74.9DeKalb, IL 71,271 13.6 7.6 8,104 46,152 13,839 62.6 IL 51.9 49.6Grundy, Charlotte-Gastonia-Concord, NC-SC 12.5 20.7 17.1 64,566 79.1 72.3 INJasper, Charlotte-Gastonia-Concord, NC-SC 11.9 17.2 36,846 77.7 80.5Kenosha, WI 27.1 9,615 11.1 25.7 37.9 70,792 47.5 81.0 58,300 64.1 Charlotte-Gastonia-Concord, NC-SC 82.1 ILMcHenry, Chicago-Naperville-Joliet, IL-IN-WI 25,298 64,228 89.1 29,847Newton, IN 13,031 76.5 Chicago-Naperville-Joliet, IL-IN-WI 40.1 41,351 24.0 63,795 93.8 62.8 26.4 22.9 15,992 -2.6 63.6 Chicago-Naperville-Joliet, IL-IN-WI Chicago-Naperville-Joliet, IL-IN-WI 0.3 46,777 69,133 25,996 25.8 22.5 36,031 Chicago-Naperville-Joliet, IL-IN-WI 6,412 16.9 24,086 26,403 62.6 59,907 18.0 26,735 81.0 33.0 26.6 52.9 41,455 40.4 62.2 Chicago-Naperville-Joliet, IL-IN-WI 0.4 32.8 18.5 54,904 42,565 28,997 80.3 93.0 25.1 5.6 13.9 16.0 25.4 31,236 92.5 7.9 21.8 23.6 8.8 18.6 25.0 49.6 4.9 58,820 8,591 78.6 6.0 16.0 94.9 24,282 32.1 33.4 18.6 17,467 69.6 78.4 54,059 9.4 24,002 75.6 11.1 35.1 1.0 14.0 91.1 61,289 83.2 55,022 26.8 84.8 28,736 54,765 61,136 22.9 65.8 86.1 23.3 82.0 14.7 14.7 64.9 89.2 80,836 25.8 49.6 58.1 26,670 36.1 5,851 80.1 28.3 3.0 9.4 26.7 23.6 30,728 86,903 11.9 24,020 27.7 69,444 16.3 26,793 31.0 75.4 82.4 27.6 18.5 24.9 24,904 10.7 40.2 83.6 78.8 5.6 7.0 25,768 39,572 16.2 83.8 57,794 92.8 27,780 21,348 6.9 29,719 11.9 25.6 15.2 57,297 74.5 15.3 95.2 83.2 25.0 51,755 -0.7 86.6 29,602 6.3 27.6 6.9 64.4 10.0 80.2 69,794 32,709 9.8 83.5 7.8 29.8 25.0 82.2 95.3 37,261 36.6 73,012 40,563 47,294 74.2 7.1 80.9 76.0 31,634 32,401 45,959 83.7 75.3 44,698 8.8 27.4 54,786 10.6 83,491 44,384 51,980 71,773 73.1 27,376 12.3 8.6 16.8 59.6 72.3 65,663 71,650 77.5 14.1 73,806 69.1 83.1 57,668 79,786 93,900 80.0 54,951

38 October 2006 • The Brookings Institution • Living Cities Census Series in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area (continued) B. Profile of Exurban Counties, Listed by Metro Appendix Table County INPorter, Bracken, KYBrown, OHClinton, OHDearborn, INArea Related Metropolitan Chicago-Naperville-Joliet, IL-IN-WI Cincinnati-Middletown, OH-KY-IN INFranklin, Cincinnati-Middletown, OH-KY-INGallatin, KY Cincinnati-Middletown, OH-KY-INGrant, KY Cincinnati-Middletown, OH-KY-IN KYPendleton, Cincinnati-Middletown, OH-KY-INGeauga, OH Cincinnati-Middletown, OH-KY-INMedina, OH 39,796 COTeller, Cincinnati-Middletown, OH-KY-IN Cincinnati-Middletown, OH-KY-IN 1,864 populationCalhoun, SC 31,002 OH Cleveland-Elyria-Mentor, areas (%) SCFairfield, 27.1 8,596 OH 2005 (%) Cleveland-Elyria-Mentor, 20,965 SCKershaw, 2004 (%) 22.5 2000 (%)Lexington, SC Colorado Springs, CO 73.3 Columbia, SC 5,542 2003 ($) SCNewberry, 21.2 7.1 Columbia, SC 45.5 7,870 2000 (%)Saluda, SC 4.6 Columbia, SC 2000 (%) 9,500 22,384Madison, OH Columbia, SC 4.3 25.0 2000 ($) 90.4 OHMorrow, 100.0 Columbia, SC 4.6 5.9 98.5 OHPerry, 100.0 97.8 66.0 OHPickaway, 35,720 Columbia, SC 27.6 Columbus, OH 3.9 95.0 97.5Union, OH 47,108 3.5 25.0 Columbus, OH 35,678 28.1Ellis, TX 9.1 98.6 26,328 4.3 39.3Henderson, TX 25.4 Columbus, OH 95.8 29.2 Columbus, OH 26,679 31.2Hood, TX 97.5 28,208 8.3 33,426Hunt, TX 98.2 Columbus, OH 31.0 15,514 20.4 TX Worth-Arlington, 27.9 Dallas-Fort 4.3Johnson, TX 18.5 10.0 28,077 TX Worth-Arlington, Dallas-Fort Kaufman, TX 29.5 26,380 76.6 30.5 7.4 TX 76.9 75.5Parker, 8.6 96.7 5,200 TX Worth-Arlington, Dallas-Fort 96.1 27,762 79.5Rains, TX 27,231 77,416 12.8 5,289 TX 48,401 TX Worth-Arlington, Dallas-Fort Worth-Arlington, Dallas-Fort Zandt, TXVan 10.5 21,879 68.9 TX Worth-Arlington, Dallas-Fort 51,923 32.0 78.6 53,821 6.0 TXWise, 31.6 34.2 11.5 17.6 81.4 59,911 Clear Creek, CO 9,395 35,156 TX Worth-Arlington, 71,019 Dallas-Fort 22.6 77.0 36,473 41.6 TX Worth-Arlington, Elbert, CO Dallas-Fort 92.0 18,931 24.9 18,828 TX Worth-Arlington, Dallas-Fort 60,027 74.1 -0.9Gilpin, CO 4,694 CO Denver-Aurora, 47,245 77.9 24,651 12,211 26.0 COPark, 5.9 2.1 54,695 TX 27.2 5.8 Worth-Arlington, Dallas-Fort 6.9 MI 54,802 Lapeer, 25.8 36,893 46.8 17,930 52.2 8.5 12,727 24.5Livingston, MI 27,505 CO Denver-Aurora, 47,374 22.1 33,861 38.6 40.4 87.3 37,382 3.4Sanilac, MI CO 70.8 Denver-Aurora, 14,382 81.2 81.0 20.4 89.8 MI 105,596 St. Clair, 8.7 34.0 2.7 37.3 -1.5 CO Denver-Aurora, 39,180 18.6 MI Detroit-Warren-Livonia, 11.6 35.9 83,792 37.4 MI 19.1 Detroit-Warren-Livonia, 61.0 22,840 7.9 16,365 24.0 52.4 35.2 26.2 83.5 15.5 3,941 90.9 22,252 60.0 MI Detroit-Warren-Livonia, 0.3 69.2 26,691 3.3 80.9 44.3 14.3 MI Detroit-Warren-Livonia, 19.0 29,978 97.7 7.2 23.5 23,189 34.0 8.1 11.0 88.9 21.2 26.6 74,106 24,125 43.1 91.4 21.9 33.3 98.3 7.9 80.4 15.1 28.1 78.4 8.3 23,113 74.2 31,760 5.3 94.5 84.3 21,799 47.5 31,512 22.3 8.5 4,791 26.7 28,956 22.6 27.7 77.5 31.2 7.4 88.1 50,519 24.0 30,653 82.0 30.6 20.6 31.3 77.2 30,975 87.4 15.9 14.7 4.6 26,213 11.4 29,923 17,992 75,149 42,995 61,095 88.0 51.4 24,553 11.3 49,995 29,134 30.4 76.7 36,426 60,347 4,757 18.6 80.0 84.1 13,892 25.0 80.6 72.3 18.1 76.2 31,595 7.2 13.2 46,681 21.7 46,379 90.5 47.9 9,203 69.5 17.6 -1.3 82.2 20.9 57,143 24,917 100.0 46,216 63,961 81.2 4.3 74,536 23,930 30.9 54,767 31.2 79.4 74.6 13.1 78.9 93.9 13.2 14.5 28.2 71.4 73,844 63.4 79.2 5.7 30,586 24.9 77.5 49,294 2.8 64,689 61,430 23.1 59,687 67,544 92.0 95.8 80.6 21.3 0.5 94.0 15.3 74,433 4.1 17.6 80.9 91.9 30,124 71,247 82.7 37.4 95.5 34.0 54,264 91.7 31.2 93.3 53,796 81.3 20.5 44,080 44,618 13.2 34,381 25.5 59,535 25.1 31,524 25.8 23,731 24.6 10.6 75.9 35,958 30,440 19.8 19.5 87,851 13.7 89.4 88.1 33.0 13.5 85.0 78.5 85,579 93,863 74,383 81.9 87.8 73,405 79.6 52,862 70,192 71,951

October 2006 • The Brookings Institution • Living Cities Census Series 39 in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area (continued) B. Profile of Exurban Counties, Listed by Metro Appendix Table CountyAllegan, MI MIBarry, Newaygo, MIRandolph, NC MI Grand Rapids-Wyoming, Laurens, SCArea Related Metropolitan Juniata, PA MI Grand Rapids-Wyoming, MI Grand Rapids-Wyoming, PA NCPerry, Greensboro-High Point, CTTolland, Greenville, SCAustin, TX TXLiberty, Harrisburg-Carlisle, PA TXMontgomery, Harrisburg-Carlisle, PA Hartford-East Hartford, CT Hartford-West San Jacinto, TX population TXWaller, Houston-Sugar Land-Baytown, TX Houston-Sugar Land-Baytown, TX areas (%) 27,939Hancock, IN Houston-Sugar Land-Baytown, TX 2005 (%) Houston-Sugar Land-Baytown, TXMorgan, IN 2004 (%) 21,037 19,238 33,602 2000 (%) 28,121Owen, IN 26.4 2003 ($)Putnam, IN Houston-Sugar Land-Baytown, TX Indianapolis, IN 2000 (%) FLBaker, 43.9 33.9 2000 (%) 25.8 20.6 Indianapolis, INBradford, FL 6.6 2000 ($) 137,150 FLClay, Indianapolis, IN 5,179 Indianapolis, IN 42,488 8,624 4.2Nassau, FL 5.2 18,036 5.6 7.8 90.5Bates, MO 16,745 17,647 46.7 Jacksonville, FL Jacksonville, FLCass, MO 92.2 22.0 96.0 60.6 37.8 83.7 90.1Clinton, MO 28,521 29.0 81.1 24.1 40.5 Jacksonville, FL 27.1 KSFranklin, Jacksonville, FL 26.4 31,427 27.5Johnson, MO MO-KS Kansas City, 9.9 25.5 26.6 6.3 2.8 10.5 87.3Lafayette, MO 78.0 26,561 MO-KS MO-KS Kansas City, Kansas City, 32,633 0.9 2.5Leavenworth, KS 26,036 42,784 MO-KS 70.4 Kansas City, 4.2 73.5 96.7Linn, KS 79.5 MO-KS Kansas City, 32.5 6.0 MO-KS 14.7 Kansas City, Miami, KS MO-KS 70.5 Kansas City, 97.8 12,661 6.0Blount, TN 27.1 4.7 5.2 34,616 29.1 26.8 82.9 22.5 17,121Jefferson, TN 50.8 84.5 21.1 26.2 27,100Loudon, TN 26,478 26,660 85.9 66,006 22,463 22.9 76.6 73.5 15,235 MO-KS 7,534 Kansas City, 15.6Roane, TN MO-KS 22,949 50,717 30,278 Kansas City, 25.7 25.8 68,280 103,871 TNSevier, Knoxville, TN 16.2 54,150 Knoxville, TN 22.9 20.3 13.4 27.3 6,105 6,422Union, TN 21,938 78.2 42.3 34.6 Knoxville, TN 11.6Nye, NV 4.5 4.3 77.2AR 34,048 9,115 27,464Faulkner, 95.9 88,202 Knoxville, TN 79.0 77.7 87.9 23.4 28.9 28,166 19.4 10,098ARGrant, Knoxville, TN 79.6 2.3 56,291 4.2 77.4 97.7 52,312 50,558 12,321ARHot Spring, 53,578 Knoxville, TN 30.4 12,001 24.2 54.7 47.6ARLonoke, 64,432 72.5AR Little Rock-North Rock, 48,462 19,974 34.3 53.2 7.8 9.7 9,434 94.4AR NV 97.7 Las Vegas-Paradise, Perry, 38,581 29.4AR Little Rock-North Rock, 49.7 63,712 20.8 11.7AR Little Rock-North Rock, 24.9 1.8 34,110 74.1 82.8 14.0 26.8 29.1 25.9 28.6 8.8AR Little Rock-North Rock, 4.5 82.3 16,809 4,922 89.0 5.5 29,435 26,101 96.3 22.0 30.3 4.8 92.4AR Little Rock-North Rock, 95.5 8.0 6.0 0.2 81.4 93.0 30.3 24,350 27,141 25.0 59.3 51.4 37,779 10.8 25.1 18.1 88.9 34,748 29.9 13,598 26.8 82,496 79.7 82.8 32,903 31,637 95.1 24,005 7,279 16,007 26.0 12,763 34,095 14.4 28,508 78.6 9.9 81.6 64,621 27.1 43.9 7.0 3.2 32.8 30.7 11,807 22,489 30.3 28,884 11.7 25.6 24,736 51,371 19.5 8.3 25,873 48,303 14,779 24.0 81.3 41.0 77.5 7.5 33,129 11.3 79.0 94.7 97.0 12.4 25,973 17,808 4,931 8.7 22.7 42.6 77.9 8.5 48,422 8.7 76.1 75.0 80.7 46,377 20.8 10.6 9.5 2.9 86.4 29.2 23.5 79.6 79.0 5.1 100.0 66,260 13.0 3.8 93.7 48.3 49,063 78,386 94.4 14.1 73.5 1.8 31,972 22.9 25,308 66,412 62,874 94.5 86.2 10.6 61.5 94.2 27.2 75.4 67.0 23.3 6.7 54,885 23.6 89.4 2.2 94.6 83.8 56,165 27,853 22.2 21.8 96.0 23.5 64,099 8.1 27,898 74,621 29.0 24,078 97.6 31.4 29,234 94.9 23,372 21.6 19.2 27,831 82.5 23.0 5.6 78.5 28,786 4.1 27.5 25,935 27,983 6.9 51,247 25.5 22,074 79,021 4.4 11.2 9.0 21,125 68.6 23,761 75.9 22.8 5.4 77.9 5.3 59,891 4.9 79.1 80.2 78.0 54,471 10.5 47,460 15.9 76.4 75.9 64,593 57,690 44,038 77.5 73.3 80.9 48,402 55,849 53,379 82.1 46,867 38,086 55,816

40 October 2006 • The Brookings Institution • Living Cities Census Series in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area (continued) B. Profile of Exurban Counties, Listed by Metro Appendix Table CountyARSaline, Breckinridge, KYClark, INCrawford, IN Louisville, KY-INHarrison, INAreaAR Related Metropolitan Little Rock-North Rock, KYHenry, Meade, KY Louisville, KY-IN Louisville, KY-INNelson, KY Louisville, KY-INScott, IN KYShelby, Louisville, KY-IN KYSpencer, Louisville, KY-IN KYTrimble, Louisville, KY-IN 30,750 population INWashington, areas (%) Louisville, KY-INColumbia, WI Louisville, KY-IN 2005 (%) Louisville, KY-INGreen, WI 36.8 2004 (%) Louisville, KY-IN Louisville, KY-IN 2000 (%)Jefferson, WI 2003 ($)DeSoto, MS Madison, WI 4,783 2000 (%) TNFayette, 8.6 2000 (%)Marshall, MS Madison, WI Madison, WI 2000 ($) MS 25.6Tate, 27,279 2,787 93.6 Memphis, TN-MS-AR TNTipton, 21,303 Memphis, TN-MS-AR WIWashington, Memphis, TN-MS-AR 27.3 11,921 3.2 MN 28.3Carver, 25.9 18,303 62.1Chisago, MNAllis, WI Milwaukee-Waukesha-West 30,286 30,875 Memphis, TN-MS-ARIsanti, MN Memphis, TN-MS-AR 95.4 79.2 5.0Pierce, WI 3.9 69.5 5,251 18,621 MN-WI Minneapolis-St. Paul-Bloomington, 6.7Pine, MN 11,766 MN-WI Minneapolis-St. Paul-Bloomington, 82.4 4.7 23.9 WIPolk, 11,052 88.4 98.1 5.3 5,034Scott, MN MN-WI 97.3 Minneapolis-St. Paul-Bloomington, 23,421 7.1 22.9 55.9 100.0Sherburne, MN 80.7 MN-WI Minneapolis-St. Paul-Bloomington, 64,497 9.0 18,504 21.9 24.0 40.6 92.8 22,748St. Croix, WI 35,967 MN-WI Minneapolis-St. Paul-Bloomington, 62.0 28.2 28,853 61,282 90.9 16.9Wright, MN 13.8 30.3 28,464 24,317 3.3 MN-WI Minneapolis-St. Paul-Bloomington, 92.1 MN-WI 21,490 Minneapolis-St. Paul-Bloomington, Cannon, TN 54.9 20,631 9,098 29,262 MN-WI 17,670 Minneapolis-St. Paul-Bloomington, 26.4 24.7 9,379 43.4 2.2 87.5 11,713Cheatham, TN 10.4 33.2 82.4 MN-WI 26.9 Minneapolis-St. Paul-Bloomington, 81.9 96.0 98.1 27,279Dickson, TN 16.6 27.5 3.1 19,820 6,360 68.7 MN-WI Minneapolis-St. Paul-Bloomington, 27,373 19.7 59.0 31.6 39,878Hickman, TN 6.9 23.9 98.0 39,527 7.1 52,101 97.4 27.9 18.8 12,593 Nashville-Davidson—Murfreesboro, TN 27,839 31.8 5.2 26.5 Nashville-Davidson—Murfreesboro, TN 31.9 26.1 28,576Robertson, TN 25.5 82.9 70.0Rutherford, TN 8.6 38,681 92.9 19.4 96.2 78.1 31,343 14.5 24.0 35,133 18.3 26.9 Nashville-Davidson—Murfreesboro, TN 94.8 77.8 84.1 26,809 27.7 1.7 TNTrousdale, 96.3 4.5 80.9 79.1 Nashville-Davidson—Murfreesboro, TN 51,306 10.0 30.5 4.3 70,595 63,732 6.0 31.9 Nashville-Davidson—Murfreesboro, TN TNWilson, 25,554 27,566 96.3 64,241 38.2 77.4 Nashville-Davidson—Murfreesboro, TN 30.2 66.8 61.2 73.9Plaquemines, LA 32.3 48.2 93.1 4.3 6.8 5.6 25.3 6.7 25.2 97.6 29.8 8.6 6.7 78.0 20,049 96.9 Nashville-Davidson—Murfreesboro, TN 3,085 45,215 78.5 60,086 6.9 31.5 39,042 53,397 39,080 29.5 31,136 LA New Orleans-Metairie-Kenner, 9.5 22.8 32,575 21.2 8.9 59,687 21,296 67.3 20.8 93.1 25.9 95.2 Nashville-Davidson—Murfreesboro, TN 72.7 77.5 26.1 82.6 12,726 75.8 28.0 33,585 96.9 88.6 55.8 22.0 26,236 24.1 21,248 21,880 4.6 12.7 31,547 3.6 79,813 75,631 30,189 81.1 96.5 69,984 80.6 35,261 54,571 25.5 22.8 6.0 36.8 3.7 49.3 28.8 95.7 24.5 37.9 57.1 18.0 54,523 14.5 4,800 6.9 23,929 49,900 26,046 39.0 83.5 3.3 10.5 87.0 42,568 76.0 27,991 30.8 4.8 43.8 30,904 46,218 74.9 5.6 79.2 29,190 103,696 7.3 5.9 34.3 65,212 77,751 85.2 40,574 94.9 80.3 6.0 95.6 66.1 10.1 10.7 17.9 68,656 6,453 80.5 11.9 65,379 12.8 71.7 19.0 38,733 14.7 73.7 64,639 73.1 92.3 60,917 3.2 32.9 54,774 93.3 31.3 85.7 25.4 70,681 7.5 78.3 83.7 67,383 84.0 24.1 4.8 76.2 81.3 73,303 10.6 80.1 31,374 26.2 24,889 86.6 12.6 57,270 52,122 70,732 25.9 60,580 28.5 76.4 61,162 86.0 26,976 28.5 88,846 8.4 84.3 23,871 29,007 89.8 13.1 9.0 81,332 31,241 71,325 24.1 15.4 68.4 21.2 29.1 78.5 23,823 83.6 7.6 7.3 32,643 28.7 76.1 53,933 63,002 80.2 14.8 76.5 24,625 54,678 69.8 43,034 7.1 64,721 65,512 76.3 8.5 81.4 51,698 71,471 78.9 66,402

October 2006 • The Brookings Institution • Living Cities Census Series 41 in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area (continued) B. Profile of Exurban Counties, Listed by Metro Appendix Table County LASt. Tammany, Monroe, PA LAPike, PA New Orleans-Metairie-Kenner, Sussex, NJ OKGrady, Area Related Metropolitan New Jersey-Long Island, NY-NJ-PA31,308York-Northern New Lincoln, OKLogan, OK New Jersey-Long Island, NY-NJ-PA28,639York-Northern New McClain, OK New Jersey-Long Island, NY-NJ-PA40,918York-Northern New 22.6Cass, NE OK Oklahoma City, OK Oklahoma City, 80,139Mills, IA 61.9 28.4 NE 16.7Washington, OK Oklahoma City, OK Oklahoma City, Lake, FL 41.9 populationOsceola, FL 20.7 77.9 Omaha-Council Bluffs, NE-IA Omaha-Council Bluffs, NE-IA areas (%) 5.8Cecil, MD 2005 (%) Omaha-Council Bluffs, NE-IA ORYamhill, 14.6 2004 (%) 87.5 28.6 2000 (%)Dutchess, NY 91.5 2003 ($) Orlando-Kissimmee, FL Orlando-Kissimmee, FLOrange, NY 30,821 83.3 2000 (%) 27.7 RIWashington, PA-NJ-DE-MD Philadelphia-Camden-Wilmington, 2000 (%) 33.9 OR-WA Portland-Vancouver-Beaverton, NCFranklin, NY Poughkeepsie-Newburgh-Middletown, 32,161 2000 ($) 49,072 30.6 21.2Johnston, NC 45,673 RI-MA River, Providence-New Bedford-Fall NY Poughkeepsie-Newburgh-Middletown, 18,233Amelia, VA 11,238 30,938 6,152Caroline, VA 30.6 8,059 NC Raleigh-Cary, 78.3 57.1 24.5 14,811 VACharles City, 26,492 NC Raleigh-Cary, 2,925 88,959 40.1 19,764Cumberland, VA 14.8 71,206 35.0 31,584 84.8 32.8Goochland, VA 13.1 33.1 Richmond, VA 85,788 Richmond, VA 82.7 Richmond, VA 43.7 95.5 VAHanover, Richmond, VA 20.1 72,738 31.8 23.3 111,145 80.5 8.3 VAKing William, 72,984 0.5 42,338 25.6 91.5 5.2 Richmond, VALouisa, VA 25.1 5.5 75,464 8.7 8.1 85.7New Kent, VA Richmond, VA 4.9 5.0 85.8 8.1 Richmond, VA 20.1 27.2 42.3 97.2 VAPowhatan, 3.6 96.5 80.2 8.7Sussex, VA 84.6 35,111 26.5 96.6 78.6 82.8 26.0Orleans, NY Richmond, VA Richmond, VA 30.2 30.1 32.9 93.9 28.5Summit, UT 25,808 Richmond, VA 73.5 25.8 23,434 28.4 UT 33,570 26.7Tooele, 26.6 28.2 32,833 9.4 80.9 51.9 24,655 24.7 23,444Atascosa, TX 28,683 Richmond, VA 67,939 32,781 37,524 NY 30.7 Rochester, 29,783Bandera, TX 8.3 13.9 36,868 UT Salt Lake City, Comal, TX 16.7 74.9 2.9 25.8 35,070 4.7 52.2 UTGuadalupe, TX 6.0 Salt Lake City, 6.6 13.7 55.7 11,400Antonio, TX San 28,217 75.7 80.1 3.6 70,612 23,025 8.2Kendall, TX 2,498 9,665 5.6 77.3Antonio, TX San 18.3 79.7Medina, TX 14.3 5,740 56,034 47,773 78.4 81.3Antonio, TX San 100.0 68.9 18.8 TXWilson, 12,500 79.5Antonio, TX San 81,395 69.6 9.1 73,274 6.6 36.1 72.9 43.7 59,667 64.6 58,176 67.0 95,744 8,322Antonio, TX 63.7 62,093 San 24,299 64,888 73.0 95,045 6.9Antonio, TX San 74.1 93,588 81.5 67.8 15.5 24.3 2.7Antonio, TX San 13,462 10,483 63.3 26.9 28.1 4.1 62,933 60,240 73.7 28,330 14.3 22,377 64.8 43.9 28,970 100.0 62.1 11.5 12.0 40.9 25.3 74.4 100.0 4,419 14.9 7,721 22.6 18.8 9,651 11,159 28,607 9.9 19.0 75.5 20.4 87.1 16.6 29,503 28,088 24.4 20,378 17.6 77.8 35.3 26.0 24,293 21.8 15,193 81.4 25.0 38,378 73.4 29.2 27.4 32.7 77.6 51,271 19,162 16.5 82.3 37,209 52.8 8.4 16.7 -3.3 33,953 60,001 42,559 21.7 28.4 -1.8 86.1 82.0 23.1 22.8 19,331 7.7 31.4 82.0 21.5 11.1 26,207 37,695 87.3 84.9 37.4 47.7 60,435 10.6 31,252 77.2 87.4 4.5 41,448 85.2 12.1 19,251 60,081 86.7 81.4 60,466 14.6 38.2 46,485 34.7 16.6 21.9 66.7 107,083 85.1 8.3 14.4 24.5 81.3 38.6 84.3 11.4 59.4 35,062 23,476 19.3 58.8 30.9 69,191 25,742 73.6 34,835 90,784 88.8 9.0 81.4 22.4 88.9 22,059 14.7 79.5 10.8 29.1 5.4 84,679 26.3 58,818 26,500 14.4 7.5 72,375 50.2 29,085 59.3 14.2 30,878 30.4 69.5 75.5 16.8 78.3 75.6 30.3 34,038 118,591 40,738 32.5 78.5 6.8 56,394 8.5 23,426 53,056 82.9 27,626 53,191 8.9 77.0 59,017 77.2 11.6 13.4 65,231 79.6 73,496 79.7 85,896 85.0 51,169 56,584

42 October 2006 • The Brookings Institution • Living Cities Census Series in change Hispanic with in change Gross super-home home new county Share married Median Share HH Median married county Share Share of Share Average population Population population Population non- couples Adjusted with Share income, Exurban exurban 2000– white, children, Income, commute*, owners, owners**, owners, commute*, Exurban exurban 2000– Income, children, white, Appendix Table B. Profile of Exurban Counties, Listed by Metro Area (continued) B. Profile of Exurban Counties, Listed by Metro Appendix Table Source: Brookings Institution analysis of decennial census data * Share of workers commuting at least one hour in each direction ** Owners of homes built between 1995 and 2000 County PAWayne, Island, WA MOFranklin, Jefferson, MOLincoln, MO PA Scranton—Wilkes-Barre, Area Related Metropolitan Monroe, IL WA Seattle-Tacoma-Bellevue, St. Louis, MO-IL MOMontgomery, St. Louis, MO-ILSte. Genevieve, MO St. Louis, MO-IL St. Louis, MO-IL MOWarren, St. Louis, MO-IL MO St. Louis, MO-ILWashington, Calaveras, CAHernando, FL St. Louis, MO-IL St. Louis, MO-IL populationCreek, OK areas (%) 19,375Okmulgee, OK Stockton, CA 27,578 2005 (%) FLOsage, OK Petersburg-Clearwater, Tampa-St. 2004 (%) 2000 (%) OKPawnee, 40.6 2003 ($) OK Tulsa, Rogers, OK OK 38.5 Tulsa, 2000 (%) OKWagoner, 2000 (%) 58,645Currituck, NC OK Tulsa, 2000 ($) 56,451 OK 4.6 Tulsa, 10.3Gates, NC 42,309 OK VA 34,767 Tulsa, Gloucester, OK Tulsa, 62.5 2,549 Beach-Norfolk-Newport News, VA-NC 95.2 VA Virginia Isle of Wight, 5,909 86.2 28.5 8,908Middlesex, VA 32.3 89.3 Beach-Norfolk-Newport News, VA-NC Virginia VASuffolk city, Beach-Norfolk-Newport News, VA-NC Virginia Beach-Norfolk-Newport News, VA-NC Virginia 22.9 5.3 21.0 10,096 26.0 11,722 33.1 VASurry, 24,525 7.5 32.3Calvert, MD Beach-Norfolk-Newport News, VA-NC 25,292 Virginia 31,252 20.5 21.6 18,520 Beach-Norfolk-Newport News, VA-NC Virginia 96.9 17,741Charles, MD 43.2 3,295 100.0 64.4 0.3 96.3 1.6 11,832 VACulpeper, 11.8 87.1 94.9 15.7 VAFauquier, 8.3 2,514 28.1 53.2 14,673 59.7 Beach-Norfolk-Newport News, VA-NC Virginia DC-VA-MD-WV Washington-Arlington-Alexandria, 95.8 16.3 26.0 MDFrederick, 29.5 97.7 31.3 2.6 DC-VA-MD-WV Washington-Arlington-Alexandria, 97.7 29.2 29,069 35,012 15.4Jefferson, WV 29.9 DC-VA-MD-WV 70.1 Washington-Arlington-Alexandria, 80.5 32,642 10,322 26,135 25.3King George, VA 23.0 11.8 22,902 8.3 94.5 23.6 89.4 DC-VA-MD-WV Washington-Arlington-Alexandria, 27.9 94.6 24,536 30,433 3,263 DC-VA-MD-WV Washington-Arlington-Alexandria, 31.7 69,937 Orange, VA 9,949 6.8 15.1 52,726 14.0 26,392 47.0 23,060Spotsylvania, VA 11,264 DC-VA-MD-WV 28,497 Washington-Arlington-Alexandria, 63,278 DC-VA-MD-WV Washington-Arlington-Alexandria, 72.6 86.0 26.0 39,624 23.0 21.7 26.6 25.7 9.7 5.3 34.0 4,262Stafford, VA 26.4 13.6 13.3 61.0 47.8 24,671 86.1 29.0 48,715 VA 7,256 78.0Warren, 20,240 DC-VA-MD-WV 29,495 30,089 16.9 Washington-Arlington-Alexandria, 12.9 47.9 21,110 13.2 25.5 54.1 25.3 DC-VA-MD-WV 27.3 KS Washington-Arlington-Alexandria, 32.4Butler, 14.5 80.1 83.4 0.2 49,932 8.2 68,300 25.7 86.5 80.8 25.1 1.7Kingman, KS 19.3 58.5 23.3 2.5 34,709 31,158 82.9 69.0 DC-VA-MD-WV 7,750 Washington-Arlington-Alexandria, 43.2 12.6 15.7 35.2 63,587 78.7 16.9 24.3 25.0 82.3 59.8 57,138 25,486 56,494 12.3 16.4 32,266 68.9 1.9 DC-VA-MD-WV Washington-Arlington-Alexandria, 20,850 80.2 81.3 55.2 77.4 47,728 48.3 1.0 15.9 32,071 50,855 33.4 15.3 27,622 18,026 13.2 KS 22.0 85.4 Wichita, KS Wichita, 83.1 81.5 29.9 11.2 7.8 79.9 84.8 76,259 29.2 19.9 66.8 22.2 16.7 45,531 26.0 22.6 27.3 82.1 89.2 26.0 65,350 22.4 82,206 42,048 78.8 39,186 77.6 81.4 78.3 29.8 22.1 21,189 5.4 57.1 16.3 80.9 24,712 30.8 24.8 82.0 33,723 26,511 78.7 78.1 25.2 26.0 44,907 61,435 25.5 23.7 71,859 42,804 30.5 28.9 25.4 23,331 29.2 83.8 50,052 83.0 10.8 78,586 72.2 12.1 23,672 34,944 5.2 24.6 18.9 29,244 74.4 39,953 33.9 85.2 29,731 22.7 64,834 65,223 17.4 78.2 21.5 90.5 72.6 7.3 41,037 84,083 23.3 78.0 70.5 77.0 38.2 8.6 14.8 83,960 32,624 76.2 4.3 45,885 5.4 75.8 24.2 47,775 74,957 56,851 47,214 80.5 23.3 87,389 75.9 80.0 94,638 71.8 33,388 15.9 81.1 81.0 56,571 69,803 24.5 47,824 82.2 70,880 60,556 63,455 29.0 77.1 30,735 2,693 74,106 80.6 57,908 74.1 51.7 86,190 31.1 67,188 4.5 -5.9 93.1 96.3 30.0 26.3 32,701 23,980 10.1 3.0 78.1 77.7 61,128 67,747

October 2006 • The Brookings Institution • Living Cities Census Series 43 Endnotes burgh); and Rock Hill, SC (outside Char- lotte). Urbanized areas are, by definition, part 1. This figure is the sum total of metropolitan of metropolitan areas. area residents who live outside Census Bureau-defined central cities. Source: 10. An urbanized area consists of an incorporated www.census.gov/hhes/www/poverty/hist- or census-designated place(s) and adjacent pov/hstpov8.html [accessed July 2006]. territory with a general population density of at least 1,000 people per square mile of land 2. Brookings Nexis search of U.S. Newspapers. area, that together have a minimum residential population of at least 50,000 people. 3. Though commuting and population growth are arguably “people-based” measures that 11. Of the 88 metropolitan areas with 2000 popu- serve to define exurbs in this paper, they lations of at least 500,000, 13 are composed closely reflect the relationship between people of one county, versus 38 of the 78 metropoli- and their surrounding places, unlike indica- tan areas with populations of 250,000 to tors such as race/ethnicity and education 500,000. (which have been employed in other definitions of exurbia). 12. Accordingly, in most cases we express exurban population as a percentage of relevant large 4. This figure excludes Alaska. Counties in the metropolitan population, rather than national other three regions average between 600 and or regional population. 750 square miles in land area. Analysis of Census Bureau data. 13. We also tried adjusting these rankings by Cen- sus region and division to reflect the variabil- 5. We first attempted to apply our methodology ity in typical development density across the to county subdivisions rather than census U.S. (using a technique described in Cohen tracts, but were constrained by a lack of con- and Debbage 2004), but those adjustments sistent data for these geographies across time did not significantly change our results. and in all states. 14. Because new residential development at the 6. Each census tract is associated with one and urban-rural fringe typically occupies only a only one U.S. county. portion of a census tract, and often occurs in a “leapfrog” fashion that leaves open space 7. An alternate view might hold that the “ex” in between developments, when aggregated to the term “exurban” refers to the “ex” (i.e., for- the tract level we expect that much of this mer) urban dwellers who relocated to the far development would qualify as exurban (with reaches of the metropolitan area in which the tract densities of at least 2.6 acres per unit). exurbs are located. Fieldwork in the Washington, D.C. area con- firmed that many developments with one-acre 8. ERS used this file to assemble Rural-Urban lots were situated in census tracts that quali- Commuting Area (RUCA) codes for all U.S. fied as exurban. census tracts. Thanks to John Cromartie at ERS for making these data available. 15. For Census 2000, the Census Bureau classi- fies as “urban” all census block groups that 9. A qualifying tract may be located within an have a population density of at least 1,000 urbanized area if at least 20 percent of its persons per square mile, plus surrounding commuters travel to a larger urbanized area census blocks that have a population density for work. Such tracts are often located near of at least 500 people per square mile. Urban- smaller city/town centers in outlying parts of a ized areas are those urban areas with a popu- larger metropolitan area, such as Antioch, CA lation of at least 50,000. (outside Oakland); Frederick, MD (outside Washington); Monessen, PA (outside Pitts-

44 October 2006 • The Brookings Institution • Living Cities Census Series 16. Because some exurban tracts are located out- 23. Pike County, PA, long the prototypical New 26. Other than exurbs, we derived the counties side metropolitan areas altogether, we com- York City exurb, has actually been integrated types described here—urban, inner suburban, pare the tract growth rate to the growth rate into the New York-Northern New Jersey-Long and other suburban—primarily for purposes for the large metropolitan area to which the Island, NY-NJ-PA metropolitan area. Richard of comparing exurbs to other geographies, and tract sends the largest number of commuters. D. Lyons, “Lure for Long Commute: Cheaper not necessarily to advance a new method for In addition, tracts designated as exurban must Homes.” The New York Times, September 20, categorizing metropolitan counties. Other sys- have had total population of at least 100 in 1987; Anthony DePalma, “New York Suburbs tems are more refined, but were not appropri- 2000. Spilling Westward.” The New York Times, Feb- ate for use here. The Lang/Sanchez typology ruary 14, 1988; Nick Ravo, “Commuting at applies only to the 50 largest metropolitan 17. See Lang and Sanchez (2006). the Edge of the Exurb Belt.” The New York areas (and includes some residential demo- Times, December 11, 1994; Matthew Purdy, graphic information in its classification sys- 18. Technically, exurban population is not itself a “Across Two Rivers and Into the Poconos: tem), while the Puentes/Warren typology is component of large metropolitan population, City Folk Go Country.” The New York Times, not county-based, but relies on central-city since 14 percent of exurban population falls March 18, 2001; Mark Morrison, “Living Too data not available between decennial cen- outside the boundaries of large metropolitan Large in Exurbia.” BusinessWeek Online, suses. Lang and Sanchez (2006); Puentes and areas. See Finding B. October 17, 2005. Warren (2006).

19. This does not indicate that houses in the typi- 24. The overlap between our methodology and 27. Specifically, we treat as a central city all cities cal exurban tract were sited on 14-acre lots, that of Lang and Sanchez was actually some- named first in the metropolitan area name, because our housing density calculation takes what greater among counties with 20 to 25 plus additional cities within the name that into consideration both developed and unde- percent exurban populations than among had populations of at least 100,000 in 2000. veloped land. those with 25 to 30 percent exurban popula- In the New York-Northern New Jersey-Long tions. The 20-percent exurban population Island, NY-NJ-PA metro area, we treat New 20. Here, we consider the extent of exurban pop- threshold is also analogous to the 20-percent York, NY and Newark, NJ as central cities. ulation by metro area based on the location of lower-bound commuting criterion we use to This yields 123 central cities within the 88 census tracts we identify as exurban. This is identify qualifying census tracts. Rather than large metro areas. Recognizing the city-county consistent with our approach above, which adopt the approach described here, we could consolidation that occurred in 2003, associates exurban census tracts with their have identified as exurban only those 81 Louisville, KY is treated as the Louisville-Jef- respective regions, divisions, and states. That counties in which a majority of the population ferson Consolidated City post-2000. noted, our results would be somewhat differ- lived in exurban tracts. However, these coun- ent if we identified exurban census tracts with ties would have captured only 19 percent of 28. Note that these figures capture moves within the metro areas to which they had their quali- U.S. exurban population. Further, because the United States alone, and do not account fying commuting ties. In particular, very large exurbanization is an evolutionary process, a for other sources of population gain/loss metro areas along the Eastern seaboard, such lower threshold identifies as exurban those including immigration and natural increase. as Boston, New York, and Washington, would counties that may in future years (and per- acquire additional exurban tracts that actually haps as of today) have even faster growth and 29. Interestingly, in the extended region encom- lie within adjoining metro areas. See Box 2. stronger commuting linkages with the metro- passing Washington, D.C., and Richmond, Where an exurban census tract lies wholly politan core. In addition, we found that the exurban counties such as Charles County, outside metropolitan areas, however, we demographic characteristics of residents of MD, and Stafford County, VA, have seen their assign that tract’s population to the metro counties that were majority exurban did not white population shares drop, even as more area(s) to which it has its largest qualifying differ significantly from those of residents of rural exurbs such as Virginia’s Caroline and commuting tie. counties that were 20 to 50 percent exurban. King William counties saw those shares increase. The effects of these migration and 21. A few metro areas, such as San Diego, CA; 25. See, e.g., Janet Frankston, “The Birth of a population dynamics on exurban diversity may McAllen, TX; and Detroit, MI, face interna- City: Whetting Growth.” The Atlanta Journal- merit deeper study in other selected large tional barriers to continued outward expan- Constitution, May 2, 2005, p. 1E; Larry metropolitan areas. sion. Werner, “New City Manager Ready to Remake Burnsville.” Star-Tribune, April 12, 2003, p. 30. The proportion of exurban-county households 22. As the next section demonstrates, however, a 1D; Sandhya Somashekhar, “Town Reaps headed by an individual age 65 or over in large number of the small counties that ring Hard Cash for New Houses.” The Washington 2000 was actually the same as the metro-wide the Atlanta region have significant exurban Post, July 16, 2006, p. A1. average (19.6 percent). character.

October 2006 • The Brookings Institution • Living Cities Census Series 45 31. Adjusted gross income differs from total 38. Note, however, that 33 of the 242 exurban income, in that it excludes earnings directed counties lie outside large metropolitan areas. to a retirement account, or used to pay health insurance premiums. It may also exclude cer- 39. This forms the subject of a forthcoming Met- tain other household expenses such as inter- ropolitan Policy Program analysis by William est paid on student loans, job relocation costs, H. Frey. and tuition and fees paid for post-secondary education. 40. To ensure that developers help pay for existing, new, or expanded infrastructure needed 32. The high percentage of workers commuting to to serve development, some municipalities the inner suburbs for work partly reflects the impose impact fees on the owners of new fact that major urban centers in some metro- development. However, fewer than half of politan areas, including Los Angeles, Phoenix, jurisdictions in large metropolitan areas in the Miami, and San Antonio, are located within Northeast, Midwest, and South impose such counties classified as “inner suburban” fees (Pendall, Puentes, and Martin 2006). because they are large and include lower-den- Nor do such fees generally compensate for sity development. other externalities such as increased traffic congestion and loss of open space. 33. This does not necessarily indicate that exurbanites do not work from home; some may 41. The term “urbanized” here connotes devel- alternate between commuting and telecom- oped land, rather than land meeting the mini- muting. The question on the census long form mum population density requirements to be asks the respondent how he/she “usually” got considered “urbanized” by the Census Bureau to work in the previous week. (Fulton et al 2001). These statistics on land urbanization are derived from the Natural 34. The share of commuters in urban counties Resources Conservation Service’s Natural traveling at least an hour in each direction is Resources Inventory (NRI) at the U.S. relatively high primarily due to large numbers Department of Agriculture. Unfortunately, the of workers in New York City’s outer boroughs data from the 2002 NRI have not yet been traveling into Manhattan, or to inner suburbs, publicly released. As such, 1997 data are the via public transit. latest national available.

35. The county with the highest average income for owners of new homes in 2000 was Man- hattan, at $243,000. The county with the lowest average was Union County, TN, an exurban county outside of Knoxville, at $38,000. In only six exurban counties—Sum- mit, UT; Sussex, NJ; Goochland, VA; Geauga, OH; Tolland, CT; and Carver, MN—did average incomes for owners of new homes exceed $100,000.

36. Stafford County, VA, the Washington, DC exurb with the highest median adjusted gross income in 2003, had the highest proportion of new super-sized homes in 2000 among exurban counties, at 45 percent.

37. Voters in the large-metro counties analyzed here cast 62 percent of the popular ballots cast nationwide for either Bush or Kerry.

46 October 2006 • The Brookings Institution • Living Cities Census Series Acknowledgments The authors are grateful for the insights offered by participants in an expert roundtable in March 2005, including: Bill Bishop, John Cromartie, Todd Gardner, Royce Hanson, Robert Lang, Ge Lin, Chris Nelson, Paul Over- berg, Marc Perry, Michael Ratcliffe, Andrea Sarzynski, and Hal Wolman. They also benefited tremendously from comments on early drafts of the methodology and paper provided by John Cromartie, Anthony Downs, Richard Forstall, Amy Liu, Chris Nelson, Robert Puentes, Michael Ratcliffe, and Hal Wolman. Thanks go to Dorn McGrath for his insights and first- hand tour of Loudoun County, Virginia; and to officials in the planning departments of Fauquier and Clarke counties, Virginia.

For More Information: Alan Berube Fellow The Brookings Institution Metropolitan Policy Program (202) 797-6075 [email protected]

For General Information: The Brookings Institution Metropolitan Policy Program (202) 797-6139 www.brookings.edu/metro

October 2006 • The Brookings Institution • Living Cities Census Series 47 About the Living Cities Census Series

Sweeping U.S. demographic changes alter the context in which urban and metropolitan policies take shape. With support from Living Cities: The National Community Development Initiative, the Brookings Institution’s Metropolitan Policy Program continues the Living Cities Census Series, a major effort to illustrate how recent demographic, social, and economic trends affect the vitality of cities and metropolitan areas. Building upon prior work focused on the results of Census 2000, the series is paying special attention to changes occurring during the current decade.

Living Cities: The National Community Development Initiative is a partnership of leading foundations, financial institutions, nonprofit organizations, and the federal gov- ernment that is committed to improving the vitality of cities and urban communities. Living Cities supports the work of community development corporations in 23 cities and uses the lessons of that work to engage in national research and policy development. Visit Living Cities on the web at www.livingcities.org.

55 West 125th Street, 11th floor • New York, New York 10027 Tel: 646-442-2200 • Fax: 646-442-2239 www.livingcities.org

The Brookings Institution

1775 Massachusetts Avenue, NW • Washington D.C. 20036-2188 Tel: 202-797-6000 • Fax: 202-797-6004 www.brookings.edu

Direct: 202-797-6139 • Fax/direct: 202-797-2965 www.brookings.edu/metro