Journal of the American Planning Association
For much of the 20th century, the terms, urban, suburban, and rural, could be used as shorthand for designating different "ways of life" in the United States, reflecting local areas' racial and demographic profiles. Moving from a city to a suburb, for example, meant escaping the dense, heterogeneous urban polyglot in favor of a more child-oriented, middle-- class lifestyle among mostly White neighbors with similar backgrounds and values. Rural populations, except in the South, were also largely White, but began to age sharply as rural youth followed jobs to the cities. Back then, well-recognized boundaries separating these very different worlds could be traversed with just a local move. Travel between them for commuting, shopping, visiting relatives, recreation, or the like could and did occur quite easily and often.
Results from the 2000 Census show a fading of these local cultural boundaries in favor of increasingly sharp regional ones. Each region is taking on its own cultural and demographic personality-a development that flies in the face of the conventional view that we are a "single melting pot" nation. These new regional divisions are being shaped by very different immigration and domestic migration flows that are creating distinctions among a suburb-like "New Sunbelt" region, an increasingly diverse "Melting Pot" region, and an aging, slow-growing "Heartland" region. Figure 1 shows the composition of these three new regions based on my analysis of data from the 2000 Census.1
The Three Regions
The New Sunbelt
This region might be characterized as "America's suburbs" because of the demographic dynamics that are creating its growth. Its 13 states, located primarily in the Southeast and West, contain about a fifth of the nation's total population and include the fastest growing states outside of the Melting Pot. Their collective population grew by 24% over the 1990s, compared to only 13% for the nation as a whole. While most of the nation's growth relies heavily on new immigrant minorities, the New Sunbelt states grew mostly by domestic migration of Whites and Blacks. Over the 1990s, domestic migrants to the New Sunbelt outpaced immigrants by a ratio of five to one; and 79% of the nation's White population gain was absorbed by these 13 states.
Contributing to these gains are today's suburbanites-young Gen-Xers, especially those forming families, and new retirees, a group whose numbers will explode in the next decade. Although Ozzie-and-Harriet-style families (White married couples with children) are declining nationally, 9 of the 10 states that gained such families in the 1990s are located in the New Sunbelt, led by Nevada where their number grew by 25%.
At first blush, this phenomenon might seem to be an extension of the old Frost Belt-to-Sunbelt migration. However, it is important to make a distinction between these New Sunbelt states and the Old Sunbelt juggernauts, California, Texas, and Florida. The latter states contain some of the nation's largest urban immigrant gateways that now contribute significantly to their population gains. Their growth, while still substantial, has peaked. New Sunbelt states draw domestic migrants with more "suburban" characteristics, and their growth trajectories are still on the rise. Figure 2 is an analysis of data from the 2000 Census showing these changes in selected states. The congressional reapportionment based on the 2000 Census awarded seven new seats to New Sunbelt states, compared to only five for California, Texas, and Florida combined. In contrast, after the 1990 Census, 14 new seats went to the latter three states, compared with only 5 to the New Sunbelt.
It is, in fact, the New Sunbelt's suburb-like character that is attracting Whites and Blacks in large numbers. They are trading the pricey, congested, commuting towns of more urbane metropolises in California and the Northeast for what they see as the more peaceful, family-friendly communities in this region. Within the New Sunbelt, the fastest growth is occurring in outer suburban areas, exurban rural counties, and smaller metropolitan areas. Figure 3 shows that during the 1990s, White population growth was greatest in South and West nonmetropolitan areas and in smaller metro areas that were not the large immigrant ports of entry.2 In fact, the fastest growing counties in the U.S. are largely White and White-- gaining counties on the peripheries of New Sunbelt metro areas. Figure 4 illustrates this pattern for counties in the Atlanta metropolitan area. The creation of this new region is being shaped by migrants from other parts of the country. In their exodus from the largely cosmopolitan, liberal-leaning urban areas, the participants in this new suburban flight are sharpening the differences-cultural and political, as well as demographic-between the New Sunbelt and Melting Pot regions.
The Melting Pot
While it is true that America is becoming more racially and ethnically diverse, this diversity is hardly spread evenly across the country. The nine states that comprise the Melting Pot region (see Figure 1) are home to 74% of the nation's combined Hispanic and Asian populations but only 41% of its total population. These states include the six with greatest immigrant gains in the 1990s (CA, NY, TX, FL, IL, and NJ), as well as New Mexico, Hawaii, and Alaska-states with large and varied ethnic minorities. Collectively, Melting Pot states grew by 13% in the 1990s. This growth was dominated by immigrants and immigrant minorities, with Asians and Hispanics accounting for 76% of the gains and other non-White, non-Black races (including American Indians, other races, and mixed races), contributing an additional 17%. As a group, these states have lost White population over the 1990s. Individual states that gained Whites (FL, TX, NM, and AK) were nonetheless dominated by increased minorities. Population growth in the Melting Pot region is overwhelmingly attributable to immigrants (and their children), for these states collectively registered a net domestic out-migration over the 1990s.
While it is true that over the 1990s most counties in the United States gained Hispanics or Asians, these groups were heavily clustered in the Melting Pot region. These states contain 70% of the U.S. foreign-born population and 76% of all Americans who speak Spanish at home, compared with only 37% of the nation's native-- born population and only 34% percent of those who speak only English at home. In addition, 55% of the nation's mixed-race married couples reside in these states.
The attraction and retention of immigrant minorities in this region is caused, in part, by a national immigration policy that emphasizes family reunification and encourages migration to occur in chains, connecting co-- nationals at both origin and destination. It is also caused by the establishment in these areas of real ethnic communities replete with their own institutions, small businesses, clubs, churches, and social networks that are not easily replicated in other regions of the country. For new ethnic minorities from Latin America, Asia, or elsewhere, a move to the suburbs or another community within the Melting Pot region is likely to be more comfortable than becoming a "pioneer" in another part of the country. It is for this reason that the suburbs in the Melting Pot region are becoming almost as multi-ethnic as the cities (Frey, 2001). And it is for this reason that cities and suburbs in the Melting Pot region will increasingly have more in common with each other than with cities and suburbs in the New Sunbelt.
The remaining reason the Melting Pot region is becoming demographically more distinct lies with the out-migration of its middle-class Whites. During the 1990s, the greater Los Angeles area lost over 800,000 Whites, the greater New York City area lost over 600,000, and other immigrant gateway metropolitan areas, such as Miami, Chicago, and San Diego, experienced somewhat smaller losses. These losses are occurring in both the central cities and suburban communities and reflect a "flight from urbanism" more than a flight from diversity. However, these White migrants are from the same population groups that are moving to the New Sunbelt: young people, married couples, parents, and new retirees. This shift represents an ongoing displacement of the White, middle-class core populations of suburbs surrounding the nation's largest urban areas which, for the most part, are located in the Melting Pot states.
The good news is that cities and suburbs in the Melting Pot region are being infused with new immigrant minorities that, by virtue of their younger ages and proclivity for more traditional families, will be contributing to a new sense of community in these areas. The 2000 Census (U.S. Census Bureau, 2001) shows that the large city with the highest percentage of Ozzie-and-Harriet-- style families is Santa Ana, California, where such families comprise 42% of all households (Frey, 2002). Close behind are Anaheim and San Jose, California; and El Paso, Texas, where at least 3 out of 10 households are traditional families.
The Heartland region consists of the remaining 28 states and the District of Columbia that have in common relatively modest growth levels and populations that are largely White or White and African American (see Figure 1). In 2000, Heartland states contained 39% of the U.S. population. They include all northeastern and midwestern states that are not classed as Melting Pot and selected southern and western states that are lagging in population growth. The least racially diverse of the three regions, the Heartland is 81% White and 12% Black, with Blacks primarily located in the region's industrial cities. In the 1990s, only about 14% of the nation's Asian and Hispanic gains came to the Heartland, but this small infusion of minorities helped to stem losses in several of its declining cities.
A large part of the Heartland has not attracted many migrants for decades. This is reflected in its older age structure and the fact that a high percentage of its population was born in-state (e.g., 78% in Pennsylvania, compared with only 24% in Nevada). Its suburbs are more middle aged and poised toward rapid "graying," in contrast to their counterparts in the New Sunbelt or Melting Pot regions, which have been attracting more Gen-Xers and immigrants. Therefore, Heartland states will have larger shares of "baby boomers," now spanning their mid 30s to mid 50s, who will have considerable influence on the Heartland's government decisions, consumer spending patterns, and politics. As seen in the November 2000 presidential election, several important "swing states" are located in the Heartland, and this will serve to magnify the national visibility of issues espoused by its relatively older, Whiter, and more-blue-collar population.
How Sharp the Divide?
The fact that new regional distinctions are taking precedence over the older, local ones raise the question: Can divides across these regions be bridged as easily as those across local areas? After all, the picture being painted here is of one region (the Melting Pot) possessing the youngest age structure and the most multi-ethnic population, and likely to be the most economically vibrant in the global economy; a second region (the New Sunbelt) becoming more suburban and middle class, with its residents choosing to live in safe, dispersed communities; and a third (the Heartland) having the least exposure to new immigrant minorities as it becomes older, Whiter, and demographically more stagnant.
In some respects, these distinctions overlie the distribution of votes by state in the 2000 presidential election, which has been subject to much discussion by pundits and the media (Barnes, 2002; Barone, 2001; Brooks, 2001). One prevalent thesis holds that the states that favored Gore ("Blue America") represent a more individualistic, secular, and liberal lifestyle; whereas those that favored Bush ("Red America") adhere to a more community and family-centered religious and conservative way of life. It is tempting to apply these interpretations to our three regions. In fact, the Melting Pot region coincides closely with "Blue America," since all those states except Florida, Texas, and Alaska favored Gore. As a group, the residents of these states are culturally diverse, economically heterogeneous, and would likely support a larger role for government, especially for education and programs directed to the less well off. The Melting Pot region is also more cosmopolitan and tends to attract educated-some might say culturally elite-people who tend to be more agnostic with regard to religious belief.
Yet it would be difficult to square "Red America" with our other two regions, because to an increasing degree the New Sunbelt is composed of refugees from the more urbane Melting Pot region. While they may be in quest of family-friendly neighborhoods and hold conservative views on economic issues, their roots will make them take a more moderate stance on social issues such as abortion, gun control, and affirmative action. In this respect, they will pull more traditional "Old South" and "Frontier West" attitudes toward the center of the ideological spectrum. In fact, it is the Heartland that most closely fits the stereotype of "Red America," given its Whiter, older, and more socially conservative population. Still, suburbanites in several Heartland states surprised the pundits by voting Democratic in the November 2000 presidential election, even though they were seen as belonging culturally to "Red America."
Rather than reflecting these two "Americas," I would prefer to cast the three regions as reflecting the older local distinctions of urban, suburban, and rural. The new regional "White flight" from the Melting Pot to the New Sunbelt region is analogous to the older local "White flight" from the central city to its suburban ring. The difference today is that both residents and jobs are much more mobile; and for middle-class Americans, lifestyle as well as economics is important in selecting a destination. Hence, while the Melting Pot region provides the intensity, ethnic diversity, and close contact that used to be associated only with cities, the New Sunbelt offers a quieter setting, large lot sizes, and local control that have always attracted people to suburbs. Finally, large swaths of the Heartland region now replicate the older, more conservative rural areas of the past.
What is missing in this new scenario is the opportunity that used to exist for daily, face-to-face interactions among people from these different social worlds. Shoppers and theater-goers from the suburbs would have to interact with urbanites on a regular basis; children in growing young families would still be in close proximity to their grandparents who lived in rural areas or the city. In Bowling Alone, Robert Putnam (2000) refers to a "sprawl civic penalty" (p. 215) that contributes to our overall civic disengagement. The fact that young couples, empty-nest boomers, and retirees will increasingly populate a region where sprawl is expanding rapidly suggests that greater social isolation will result from this trend. Moreover, the census trends discussed here reflect patterns that occurred prior to the September 11 terrorist attacks. On that day, White residents living in large cities or their suburbs were substantially more concerned about an attack in their community than Whites living in small towns or rural areas, according to my analysis of a national survey (see Table 1). The results of this analysis show that 65.1% of White residents in large cities were concerned about such an attack, compared with only 18.8% of White residents in rural areas. These new concerns about security may further reinforce an already strongly held White, middle-class preference for dispersed settlements.
Within the Melting Pot states, however, there is already evidence of greater interracial and intercultural dating and marriage, residential coexistence, and the propensity for second-generation children to become proficient in English as well as in the language of their parents. "Melting" is indeed occuring within the Melting Pot regions, if not across the broader national landscape. These trends imply that an important national challenge for the present century will be to find ways to bridge these new regional divisions among communities with different demographics, lifestyles, and values but probably similar aspirations. National political parties, big corporations, and religious and civic institutions, as well as local governments and planners, will all be affected by this increasing social and geographic divide, separating us into "urban," "suburban," and "rural" regions that divide the nation into three Americas.
Frey is a demographer on the faculty of the University of Michigan Population Studies Center and a senior fellow at the Milken Institute in Santa Monica, CA. He has been monitoring 2000 Census trends in American Demographics articles, in Brookings Institution Census Surveys, and on his Web site http://www.frey-demographer.org.
Sarah McKenzie's image Aerial #21 (Niwot) shows new housing developments encroaching on farmland in Colorado. It serves as a visual example of the residential redistribution discussed by William Frey in this Longer View, and one of its consequences. It also highlights the rapid growth of many mountain states in the region Frey labels the New Sunbelt.
The artist is a landscape painter whose work addresses issues of land use and development. She is particularly interested in the many ways our built environment has changed over time in response to changes in our culture. She is currently an assistant professor at the Cleveland Institute of Art. She holds a BA from Yale and an MFA in painting from the University of Michigan. To view more of her work, visit www.sarahmckenzie.com or contact the Carson-Masuoka Gallery in Denver, Colorado (303) 573-8585.
The results of the 2000 Census show the increasing importance of regional differences for understanding America's racial and demographic landscape. Well-worn local labels such as urban, suburban, and rural are becoming less descriptive of lifestyles, racial profiles, and age structures than distinctions that separate sets of states into three regions: the suburb-like "New Sun belt," the racially diverse "Melting Pot," and the slow-growing, aging "Heartland." These regional divisions are rooted in the somewhat distinct redistribution patterns of immigrant minorities, who have concentrated mostly in coastal areas, and streams of largely White domestic migrants, who have gravitated to newer, economically prosperous areas in the Southeast and West.
1. The New Sunbelt states are Arizona, Colorado, Delaware, Georgia, Idaho, North Carolina, Nevada, Oregon, South Carolina, Tennessee, Utah, Virginia, and Washington. The Melting Pot states are Alaska, California, Florida, Hawaii, Illinois, New Jersey, New York, New Mexico, and Texas. The Heartland states are Alabama, Arkansas, Connecticut, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi, Missouri, Montana, Nebraska, New Hampshire, North Dakota, Ohio, Oklahoma, Pennsylvania, Rhode Island, South Dakota, Vermont, West Virginia, and Wyoming.
2. These metropolitan areas are Chicago and New York (North); Dallas, Houston, and Miami (South); and Los Angeles, San Diego, and San Francisco (West).
ABC News/Washington Post. (2001, September 11). Terrorist Attack Poll No. 1. Washington, DC: Author.
Barnes, J. A. (2002, March 15). Democrats see red. National Journal, p. 758-765.
Barone, M. (2001, June 8). 49 percent nation. National Journal, p.1710-1716.
Brooks, D. (2001, December). Are we really one country?: A report from Red and Blue America. Atlantic Monthly, pp. 53-- 65.
Frey, W. H. (2001, June). Meltingpot suburbs: A Census 2000 study of suburban diversity (Census Survey). Washington DC: Brookings Institution Center on Urban & Metropolitan Policy.
Frey, W. H. (2002, Second Quarter). Ozzie & Harriet meet Will & Grace (Charticle). Milken Institute Review, pp. 5-7.
Putnam, R. D. (2000). Bowling alone. New York: Touchstone.
U.S. Census Bureau. (1991). Census of Population and Housing, 1990 [United States]: Public Law (P.L.) 94-171 Data. Washington, DC: Author.
U.S. Census Bureau. (2001). Census of Population and Housing, 2000 [United States]: Public Law (P.L.) 94-171 Data. Washington, DC: Author.