Racial segregation in the United States has changed unevenly across metropolitan neighborhoods and public schools. From 1980 to 2020, Black-white dissimilarity declined nationally, while some metropolitan areas remained highly separated. School data for 2020-21 also show that many students attended schools dominated by one racial or ethnic group.
Contents
- How residential segregation is measured
- National metropolitan trends since 1980
- Black-white segregation across metropolitan areas
- Black isolation and persistent high-segregation metros
- Hispanic-white segregation and isolation
- Asian-white segregation and isolation
- Racial separation in K-12 public schools
How residential segregation is measured
The Census-hosted working paper Metropolitan Segregation: No Breakthrough in Sight uses the Index of Dissimilarity, or D. The index ranges from 0 to 100 and measures the percentage of either group that would have to move for the two groups to have equal distributions across census tracts. The paper treats a score of 60 or more as very high.
The residential figures below compare non-Hispanic White, non-Hispanic Black, non-Hispanic Asian/Pacific Islander, and Hispanic populations in metropolitan areas. The authors note estimation and classification changes affecting some 1980 categories, so comparisons across the full period should be read with that limitation in mind. A dissimilarity score describes distribution between groups; an isolation score describes the racial or ethnic composition experienced by the average member of a group.
National metropolitan trends since 1980
National averages show the largest decline in Black-white dissimilarity. The score fell from 72.7 in 1980 to 55.2 in 2020. Hispanic-white dissimilarity declined from 50.2 to 45.3 over the same years, while Asian-white dissimilarity changed only slightly, from 40.4 to 40.0. The working paper’s national Hispanic-white average is reported as 45.5 in 2020 in another table, a small table-level difference that should not be treated as a new estimate.
| Metropolitan measure | 1980 | 2020 |
|---|---|---|
| Black-white dissimilarity | 72.7 | 55.2 |
| Hispanic-white dissimilarity | 50.2 | 45.3 |
| Asian-white dissimilarity | 40.4 | 40.0 |
| White share of average White resident’s neighborhood | 88.3% | 69.4% |
| White share of average Black resident’s neighborhood | 31.3% | 34.2% |
The average White metropolitan resident’s neighborhood was 88.3% White in 1980 and 69.4% White in 2020. For the average Black metropolitan resident, the White share of the neighborhood was 31.3% in 1980 and 34.2% in 2020. These neighborhood-composition measures do not mean that every resident lived in a neighborhood with the average composition.
Other group-exposure measures moved in different directions. The average Hispanic resident’s neighborhood was 47.5% White in 1980 and 33.6% White in 2020. The average Asian resident’s neighborhood was 61.3% White in 1980 and 44.3% White in 2020. At the same time, the average Black resident’s neighborhood was 60.8% Black in 1980 and 40.8% Black in 2020; the average Hispanic resident’s neighborhood was 38.2% Hispanic in 1980 and 44.6% Hispanic in 2020; and the average Asian resident’s neighborhood was 18.0% Asian in 1980 and 24.5% Asian in 2020. Source: Metropolitan Segregation: No Breakthrough in Sight
Black-white segregation across metropolitan areas
Metropolitan population composition is associated with different Black-white segregation patterns. In metropolitan areas with less than 5% Black residents, average Black-white dissimilarity declined from 56 in 1980 to 34 in 2020. In areas with more than 20% Black residents, the average was 70 in 1980. It fell 3 points during the 1990s, 4 points from 2000 to 2010, and 2 points from 2010 to 2020.
The overall decline therefore did not eliminate very different local experiences. Selected metropolitan comparisons from 1980 and 2020 illustrate the range:
| Metropolitan area | 1980 Black-white dissimilarity | 2020 Black-white dissimilarity |
|---|---|---|
| Nassau County-Suffolk County, New York | 76.9 | 65.5 |
| Kansas City | 77.6 | 51.6 |
| Chicago | 89.1 | 73.8 |
| Detroit | 83.0 | 74.5 |
| Las Vegas | 62.9 | 37.4 |
Chicago’s score was 89.1 in 1980 and 73.8 in 2020, while Detroit’s was 83.0 and 74.5. Both remained above the working paper’s very-high threshold in 2020. Las Vegas moved from 62.9 to 37.4, and Kansas City moved from 77.6 to 51.6. Nassau County-Suffolk County, New York, declined from 76.9 to 65.5. Source: Metropolitan Segregation: No Breakthrough in Sight
Black isolation and persistent high-segregation metros
Isolation provides a different view of residential separation. In 2020, Black isolation was 75.7 in Detroit, meaning the average Black resident lived in a tract that was about three-quarters Black. The measure was 67.2 in Jackson, Mississippi; 67.1 in Memphis; 65.2 in Philadelphia; 62.8 in Chicago; and 60.7 in Cleveland.
The working paper identifies six persistent high-segregation metropolitan areas. Together, they contained 4,846,085 Black residents in 1980 and 5,284,094 in 2020. The Black population in all other metropolitan areas combined grew by nearly two-fold between 1980 and 2020. Across the report’s metropolitan table, the national Black metropolitan population rose from 21,744,904 in 1980 to 40,472,563 in 2020.
These population counts and isolation scores answer different questions. The counts describe how many Black residents were represented in the table, while isolation describes the average racial composition of the tract where Black residents lived. Neither measure alone explains why a metropolitan pattern exists or how housing, income, school boundaries, or other conditions contributed to it. Source: Metropolitan Segregation: No Breakthrough in Sight
Hispanic-white segregation and isolation
The national Hispanic-white dissimilarity average was 50.2 in 1980 and 45.5 in 2020 in the paper’s reported average. The share of Hispanics living in metropolitan areas classified at 50 or higher fell from 55.3% in 1980 to 37.4% in 2020. Meanwhile, the share living in the least-segregated 1980 metropolitan category rose from 20.2% to 35.1%.
The averages differed by metropolitan category. Areas with Hispanic-white dissimilarity below 40 averaged 32.4 in 1980 and 38.1 in 2020. Areas with scores of 50 or higher averaged 59.0 in 1980 and 53.0 in 2020. Thus, the lower-segregation category’s average rose even as the high-segregation category’s average fell.
Several metropolitan areas had high 2020 scores: Salinas, California, had Hispanic-white dissimilarity of 61.3; Newark, New Jersey-Pennsylvania, had 61.1; and Los Angeles had 61.0. Hispanic isolation was 95.3 in Laredo, 92.4 in McAllen-Edinburg-Mission, and 90.6 in Brownsville-Harlingen in 2020. Source: Metropolitan Segregation: No Breakthrough in Sight
Asian-white segregation and isolation
The national Asian-white dissimilarity average was 40.4 in 1980 and 40.1 in 2020 in the paper’s reported average. The share of Asians living in metropolitan areas with a score of 40 or higher fell from 59.7% to 41.0%. The share in areas below 30 rose from 11.5% to 18.7%.
The Asian population represented in the metropolitan table rose from 3,100,635 in 1980 to 22,687,369 in 2020. At the same time, Asian isolation increased in several metropolitan areas: San Jose rose from 10% to 51%, Oakland from 12% to 43%, and Anaheim-Santa Ana-Irvine from 6% to 36%.
Category averages show why a national average can conceal local changes. Metropolitan areas with Asian-white dissimilarity of 40 or higher averaged 45.4 in 1980 and 44.2 in 2020. Areas from 30 to 34.9 averaged 31.9 and 37.9, while areas from 35 to 39.9 averaged 37.7 and 39.9. Source: Metropolitan Segregation: No Breakthrough in Sight
Racial separation in K-12 public schools
The Government Accountability Office’s K-12 Education: Student Population Has Significantly Diversified, but Many Schools Remain Divided Along Racial, Ethnic, and Economic Lines analyzed Department of Education Common Core of Data information for public K-12 schools and students in the 2020-21 school year. In that analysis, a predominantly same-race/ethnicity school had at least 75% of its students in one racial or ethnic group.
About 18.5 million public K-12 students attended schools where at least 75% of students were one race or ethnicity. Fourteen percent attended schools where at least 90% of students were one race or ethnicity.
The GAO also identified about 13,500 predominantly same-race/ethnicity public schools within 10 miles of a predominantly same-race/ethnicity school of a different race or ethnicity. Those schools represented about 14% of all public K-12 schools. Ninety percent of the nearby predominantly same-race/ethnicity school pairs were in different school districts.
The share of students attending schools that were at least 75% their own race or ethnicity varied by group:
| Student group | Share attending a school at least 75% their own race or ethnicity |
|---|---|
| White | 45% |
| Hispanic | 31% |
| Black | 23% |
| American Indian/Alaska Native | 19% |
| Asian | 4% |
These school figures describe concentration within schools during 2020-21, not the full residential conditions surrounding each school. The nearby-school analysis also indicates that school boundaries can separate predominantly same-race/ethnicity schools even when another such school of a different racial or ethnic composition is within 10 miles. Source: K-12 Education: Student Population Has Significantly Diversified, but Many Schools Remain Divided Along Racial, Ethnic, and Economic Lines