Residential Segregation is the Linchpin of Racial Stratification

Three decades of research have amply confirmed Pettigrew’s (1979) prescient observation that residential segregation constitutes the “structural linchpin” of racial stratification in the United States. Although the centrality of segregation as a stratifying force in American society remains, however, patterns of segregation have changed substantially since the 1970s. At that time, African Americans were highly segregated almost everywhere and socioeconomic attainments had no effect on the degree of segregation experienced by African Americans. Race was very much a master status and most whites subscribed to an ideology of segregation, either de jure or de facto. In the early 1960s, for example, absolute majorities of white Americans still supported segregation as a matter of principle, agreeing on surveys that schools, transportation, occupations, and neighborhoods should be racially segregated and that intermarriage should be prohibited (Schuman et al. 1998).

White racial attitudes towards black Americans shifted during the Civil Rights Era, however, with important consequences for patterns of racial segregation. During the 1980s, principled support for segregation all but disappeared; but despite this retreat from segregationist ideology, whites nonetheless continued to harbor strong anti-black sentiments rooted in negative stereotypes about the low intelligence, lack of motivation, propensity toward criminality, and predatory sexuality of African Americans (Bobo et al. 2012). Even though whites had come to reject segregation in principle, they continued to feel uncomfortable in the presence of many African Americans in practice; and they grew progressively more uncomfortable as black numbers in the social setting rose (Charles 2003).

With respect to neighborhoods, surveys have consistently shown that white avoidance increases rapidly as the percentage of potential black neighbors rise; that such avoidance is rooted in anti-black stereotypes; and that it persists even when objective characteristics of the neighborhood are experimentally controlled (see Emerson, Yancy, and Chai 2001; Charles 2003; Krysan et al. 2009; Swaroop and Krysan 2011). African Americans seem to be tolerated as potential neighbors for whites mainly when they are small in number and are not perceived in stereotypical terms. Indeed, in American social cognition today, black professionals are perceived to be on a par with others in the middle class and generally accepted as mainstream “Americans” (Cuddy, Fiske, and Glick 2007), in contrast to poor blacks who continue to be seen in very negative terms and are largely blamed for their own poverty and problems (Bobo et al. 2012). In keeping with their growing acceptability, in 2000 affluent African Americans for the first time were able to achieve declines in segregation as their socioeconomic status rose (Iceland and Wilkes 2006).

Racial attitudes have thus evolved such that whites no longer insist on segregation in all circumstances and are willing to share social space with black Americans under certain limited conditions. Specifically, in metropolitan areas where blacks constitute a small share of the population and are relatively affluent they should observe a shift toward integration over time, whereas in areas where blacks comprise a large share of the population and display high rates of poverty they should continue to experience high levels of segregation. Both predictions are borne out by recent trends in the United States.

According to Rugh and Massey (2014), the most rapidly integrating metro areas from 1970 to 2010 were Provo, Utah, Missoula, Montana, and three small metropolitan areas in Colorado: Boulder, Grand Junction, and Fort Collins, all places with small and more affluent black populations. In contrast, movement toward integration was least and segregation remained highest in Milwaukee, Gary, Detroit, Newark, and New York, all places with large, poor black populations. Rugh and Massey (2014) also determined that desegregation was also predicted by small metropolitan population size, low levels of anti-black prejudice, permissive density zoning regulations in suburbs, a newer housing stock, the presence of a college or university, and a sizeable military population, conditions that more often prevail in the West than the Northeast, Midwest, or South.

Many of the metropolitan areas displaying high levels of black segregation in 2010 also satisfied the criteria for hypersegregation, an extreme form of segregation and isolation across multiple geographic dimensions defined originally by Massey and Denton (1989). More recently Massey and Tannen (2015) found that the number of hypersegregated black areas fell from 40 to 21 between 1970 and 2010 while the share of black metropolitan residents experiencing hypersegregation dropped from 61% to 32%. As of 2010, the list of hypersegregated areas included, among others, Baltimore, Boston, Chicago, Cleveland, Detroit, Kansas City, Milwaukee, New York, Philadelphia, and St. Louis. In contrast, areas that had moved away from hypersegregation included Amarillo, Texas, Muncie, Indiana, Omaha, Nebraska, and Witchita, Kansas, again places with small black populations.

By 2010, of course, African Americans are no longer the nation’s principal minority group a distinction that increasingly fell to Hispanics after the turn of the century. Whereas in 1970 Hispanics constituted just 4.7% of the population and were regionally isolated and ethnically segmented, with Puerto Ricans in the Northeast, Cubans in South Florida, and Mexicans in the Southwest. Outside of these regions, only Chicago displayed significant numbers of Mexicans, Puerto Ricans, and Cubans living in the same place. As of2010, however, Hispanics had risen to comprise 17.3% of the U.S. population and Mexicans had dispersed throughout the nation along with new arrivals from Central Americans to create multi-ethnic Hispanic populations in all 50 states.

As the Hispanic population grew rapidly after 1980, levels of Hispanic segregation trended upward and levels of spatial isolation within neighborhoods sharply increased (Rugh and Massey 2014). A large share of this population growth occurred through unauthorized migration and in 2010 around 60% of Mexican immigrants and two-thirds of those from Central America were undocumented (Massey and Pren 2012). Research indicates that the degree of Hispanic segregation is positively associated with a larger share of undocumented migrants in the metropolitan population (Hall and Stringfield 2014), higher levels of prejudice toward “illegal aliens” (Rugh and Massey 2014), and with larger shares of Hispanics of Afro-Caribbean origin (Denton and Massey 1989), in addition to the aforementioned metropolitan characteristics that are also conducive to high black segregation. Whereas Hispanics were not hypersegregated in any metropolitan area in 1980 or 1990, by 2000 Wilkes and Iceland (2004) found they satisfied the criteria for hypersegregation in both New York and Los Angeles, the nation’s two largest Hispanic communities.

In sum, although racial residential segregation is no longer universal in urban America, it continues actively to be produced and perpetuated within an important subset of metropolitan areas that together contain a disproportionate share of the nation’s black and Hispanic residents. The active, ongoing production of residential segregation today occurs within a context of sharply rising inequality and growing segregation on the basis of wealth and income (Massey, Rothwell, and Domina 2009; Bishoff and Reardon 2014), thereby creating a new and more complex urban ecology in which race and class interact powerfully to determine individual and family wellbeing (Massey and Brodmann 2014).

As income distributions polarize and poverty intensifies within metropolitan areas characterized by high levels of black and Hispanic segregation, the inevitable result is the spatial concentration of poverty within black and Hispanic neighborhoods (Quillian 2012); and as poverty is concentrated spatially, so are its correlates such as crime, violence, family disruption, dependency, and substance abuse, in the process creating a uniquely disadvantaged residential environment that is rarely, if ever, experienced by white Americans (Massey and Denton 1993; Peterson and Krivo 2010).

As Sampson’s (2012) comprehensive analysis of Chicago has shown, under conditions of hypersegregation things go together ecologically such that low neighborhood income is strongly correlated with a host of other deleterious conditions that collectively undermine human well-being across a variety of dimensions. For this reason, exposure to concentrated neighborhood disadvantage has unsurprisingly emerged as the critical nexus for reproduction of socioeconomic disadvantage over the life course and across the generations, especially for African Americans but increasingly also for Hispanics (Massey 2007; Sharkey 2013). In a very real, the perpetuation of poverty among blacks and Latinos today prevails because segregation is not a thing of the past, but a condition that continues to be generated and reinforced by ongoing social and economic processes that continue to operate within distinct segments of American society.