Data recently released from Census 2000 provide an opportunity to examine the extent of changes in racial and ethnic residential segregation in the last 2 decades of the 20th century. Segregation can result from, among other factors, voluntary choices people make about where they want to live or from the involuntary restriction of choices, such as through discrimination in the housing market, or from a lack of information about residential opportunities. This study does not attempt to identify the causes of racial and ethnic residential segregation (or simply "segregation"), nor do we argue that segregation is a more serious problem in one area than another. This report simply describes the extent of, and changes in, segregation over the 1980 to 2000 period. Because segregation is much more of an issue in urban environments, we focused on segregation patterns in metropolitan areas across the United States.
B. RACE AND ETHNICITY
Residential segregation measures are influenced by how race and ethnicity are defined. In 1977, the Office of Management and Budget (OMB) issued Statistical Policy Directive 15, which provided the framework for data collection on race and ethnicity to federal agencies, including the Census Bureau, for the 1980 decennial census. That directive identified four racial groups:
-- and one ethnicity --
The questions on the 1980 and 1990 censuses asked individuals to self-identify with one of these four racial groups and indicate whether they were Hispanic.(1)
In the 1990s, after much research and public comment, OMB revised the racial classification to include five groups:
An additional major change was to permit the individuals to report more than one race. Census 2000 figures indicate that 6.8 million people, or 2.4 percent of the population, reported more than one race.(2)
C. SEGREGATION MEASURES
This report examines five dimensions of segregation proposed by Massey and Denton (1988). Within each of these dimensions, several segregation measures are possible. In this report we focus on only one segregation measure from each dimension as follows:
|evenness dimension:||dissimilarity index|
|exposure dimension:||isolation index|
|concentration dimension:||delta index|
|centralization dimension:||absolute centralization index|
|clustering dimension:||spatial proximity index|
These dimensions and indexes are discussed in more detail in Chapter 2 and in Appendix B.
E. PLAN OF THE REPORT
Chapter 2 discusses the data and methods employed in this report. It covers race and ethnicity measurement, geographic areas and units of analysis, residential segregation measurement, the data used, statistical testing, and guidance on how to interpret the findings. Chapters 3 through 6 focus on the 1980 to 2000 residential segregation patterns of four major racial and ethnic groups:
In each case, non-Hispanic Whites serve as the reference (majority) group, even though in some metropolitan areas they are actually in the minority.
Each chapter presents information in the same way using the same table structure. First, descriptive statistics about the five indexes are presented and discussed. Second, changes over time are discussed using the characteristics of the metropolitan areas to understand differences. Third, the magnitudes of changes are examined and any differing patterns are discussed. Fourth, statistics are presented for all large metropolitan areas (1 million people or more) that have at least 20,000 people or three percent of their population in the minority group. Then, the metropolitan areas with the biggest increases and decreases in segregation are discussed. Each chapter includes graphical representations of residential segregation in the form of scatter plots, histograms, and maps. The chapters close with a summary of findings. Finally, Chapter 7 presents some cross-group comparisons and analyses.
1. The population censuses have a special dispensation from OMB to allow individuals to designate "Some Other Race" rather than one of those specifically listed. The vast majority of individuals choosing that option are Hispanic (Grieco and Cassidy 2001). The decennial census questions also ask about specific Asian and Pacific Islander races (e.g., Chinese).
2. Many of those who report more than one race list "other" as one of the races. About 1.2 percent of the population selected two races which did not include the "other" race. Another 0.2 percent of the population selected three or more races (indicating that they selected at least two races which were not "other").