The population of the United States is increasingly diverse both racially and ethnically. In the 1990 census 12 percent of Americans identified themselves as black, 9 percent as Hispanic, 3 percent as Asian, and 1 percent as American Indian, Eskimo, or Aleut. Today approximately one-quarter of Americans fit into these socially defined ethnic-minority categories; by the mid-twenty-first century, nearly half will be. As part of this changing demographic structure, blacks, who once constituted the vast majority of the minority population but are now less than one-half of it, will continue to decline numerically relative to Asians and Hispanics.
Along with a more multiethnic United States has come a reevaluation of the melting-pot concept. A heightened appreciation of ethnic experience and identity underlies the discussions of diversity that are prominent today in many educational and political settings. Yet, even in a climate of appreciation for ethnic identity and solidarity, many commentators are concerned about dramatic and persistent patterns of segregation, especially between whites, a term that throughout this article refers to non-Hispanic whites, and blacks.
Many impressions of ethnic segregation and related problems derive from the largest metropolitan areas, including New York City, Los Angeles, Chicago, San Francisco, Philadelphia, Detroit, Washington, D.C., and Miami. As a group these areas are characterized by a large number of minorities and high levels of residential segregation. They are the places where terms such as hypersegregation, American apartheid, and chocolate city-vanilla suburbs are appropriate (Massey and Denton 1993; Farley and others 1994). To broaden perspective on the issue we focus on multiethnicity in medium-sized cities and examine Sacramento, California, as the case study. This article both provides data on a specific city and offers an analytical format that can be applied to other cities of similar size.
In this context, size matters: large metropolitan areas generally have high levels of segregation. For example, in California's ten largest metropolitan areas, segregation indexes for 1990 were generally lower for Sacramento and the six smaller places than they were for the Los Angeles, San Francisco, and San Diego regions. The Sacramento experience with segregation may be usefully compared with that of Cincinnati, San Antonio, Memphis, and other places with populations of 1 million to 2 million. Functional specialization is crucial, because low levels of segregation often are associated with strong governmental, higher-educational, or military functions. The segregation experiences of other state capitals, university towns, and communities with large military bases are likely to have something in common with Sacramento.
Region is another important factor, in large part because it summarizes many of the characteristics that bear on segregation. Because of their functional specialization, ethnic mixture, and age, urban areas in the western United States have long had lower levels of segregation than have cities in other parts of the country. In many western urban centers a large portion of the housing stock was built during the last quarter-century, a pattern with positive implications for reduced segregation. Only a small share of the housing supply is obsolescent and can be tagged as minority housing, and a sizable amount has been built since the implementation of open-housing legislation. Minorities are less likely to have been crowded into public-housing projects in a central city; likewise, a belt of politically independent, exclusionary suburbs is often not present (Farley and Frey 1994, 25-30). Sacramento thus provides a reasonable model for understanding contemporary patterns of ethnic segregation in medium - sized metropolitan areas, especially those in the American West.
SEGREGATION IN SACRAMENTO
When quantitative measures of segregation were first applied systematically to census data, Sacramento was one of the least-segregated metropolitan areas in the United States. …