School segregation statistics show that racial and ethnic concentration remains common in U.S. public schools. In 2020–21, about 38% of public K–12 students—approximately 18.5 million children—attended a school where at least 75% of students belonged to one racial or ethnic group. About 14% of students, or roughly 7 million, attended a school with at least 90% from one group. These measures describe enrollment patterns; they do not by themselves establish unlawful discrimination or illegal segregation.
Contents
- How common is racial and ethnic concentration?
- How the pattern changed over time
- Which student groups experience concentrated enrollment?
- How concentration varies by school type and region
- Nearby schools, boundaries, and racial isolation
- District secession and demographic change
- Economic concentration alongside racial patterns
How common is racial and ethnic concentration?
The Government Accountability Office (GAO) defines a predominantly same-race/ethnicity school as one in which at least 75% of students are from one racial or ethnic group. An almost-exclusively same-race/ethnicity school has at least 90% from one group. Using these definitions, about 14% of all public K–12 schools were predominantly same-race/ethnicity schools in 2020–21. The student-level share was much larger: 38% of public K–12 students attended a predominantly same-race/ethnicity school, while 14% attended an almost-exclusively same-race/ethnicity school. GAO, K-12 EDUCATION: Student Population Has Significantly Diversified, But Many Schools Remain Divided Along Racial, Ethnic, and Economic Lines, 2020–21.
An earlier NCES measure uses a different definition. It counts schools with at least 75% minority enrollment, combining minority racial and ethnic groups rather than measuring whether one group alone makes up at least 75% of enrollment. In fall 2015, approximately 30% of public elementary and secondary students attended schools with at least 75% minority enrollment. The same NCES measure found that 60% of Hispanic students, 58% of Black students, and 53% of Pacific Islander students attended such schools. NCES, Indicator 7: Racial/Ethnic Concentration in Public Schools, fall 2015.
The NCES measure also found that 51% of White public-school students attended schools where at least 75% of enrollment was White in fall 2015. Because the NCES threshold combines minority groups while the GAO threshold focuses on a single group, the two measures should not be treated as interchangeable.
How the pattern changed over time
The GAO data show a decline in the overall share of students and schools classified as predominantly same-race/ethnicity between 2014–15 and 2020–21. The student share fell from 42% to about 38%, while the school share declined from 47% to 43%.
The composition of predominantly same-race/ethnicity schools also changed. The share of predominantly Hispanic public schools rose from 7.4% in 2014–15 to 8.4% in 2020–21. The share of predominantly White public schools declined during the same period. The share of predominantly same-race/ethnicity schools other than predominantly White or Hispanic schools remained unchanged. GAO, 2014–15 to 2020–21.
The longer NCES series uses the 75% minority-enrollment measure. The share of all public students attending schools with at least 75% minority enrollment increased from 21% in fall 2000 to 30% in fall 2015. For American Indian/Alaska Native students, the share rose from 29% in 2000 to 37% in 2015. For Black students, it rose from 51% to 58% over the same period. NCES, 2000 to 2015.
These trends are not contradictory: they use different thresholds and group definitions. The GAO series measures concentration around one racial or ethnic group, while the NCES series measures schools where minority students collectively represent at least three-quarters of enrollment.
Which student groups experience concentrated enrollment?
In 2020–21, nearly half of White students attended predominantly White schools. Nearly one-third of Hispanic students attended predominantly Hispanic schools, and nearly one-quarter of Black students attended predominantly Black schools. More than one-quarter of Hispanic and Black students attended almost-exclusively Hispanic, Black, or American Indian/Alaska Native schools. More than one-fifth of American Indian/Alaska Native students attended almost-exclusively American Indian/Alaska Native schools.
The GAO also reported that nearly one-third of White students attended schools that were almost-exclusively White, Asian, or multiracial. These categories describe the composition of the schools students attended, not the demographic makeup of every student’s surrounding community. GAO, 2020–21.
The NCES figures provide additional group-specific thresholds for fall 2015:
| Student group | Share attending schools at least 75% that group |
|---|---|
| American Indian/Alaska Native | 17% |
| Asian | 3% |
| Pacific Islander | 3% |
| Hispanic | 33% |
The table uses the NCES group-specific measure and therefore cannot be directly compared with the GAO percentages for predominantly same-race/ethnicity schools. NCES, fall 2015.
How concentration varies by school type and region
School type is associated with different rates of predominantly same-race/ethnicity enrollment. In 2020–21, 44% of traditional public schools were predominantly same-race/ethnicity, compared with 41% of charter schools and 26% of magnet schools. Nearly 90% of public K–12 students attended traditional public schools in that year, so the traditional-public-school figure describes the largest part of the public-school system.
Regional patterns also differed. The Midwest and Northeast had the highest regional percentages of predominantly same-race/ethnicity schools. The West had more predominantly Hispanic schools than any other U.S. region. The South had more predominantly Hispanic and predominantly Black schools than other U.S. regions.
Locale mattered as well. Predominantly White schools were more often located in rural areas, while predominantly Black and predominantly Hispanic schools were generally located in urban and suburban locales. These are descriptive patterns from the GAO analysis; they do not identify a single cause for the differences. GAO, 2020–21.
Nearby schools, boundaries, and racial isolation
Physical proximity does not necessarily mean students can attend the nearby school. In 2020–21, nearly 8,000 predominantly same-race/ethnicity schools—8% of all public K–12 schools—were within 5 miles of another predominantly same-race/ethnicity school representing a different race or ethnicity. More than 13,000 such schools—14% of all public K–12 schools—were within 10 miles of a different-race/ethnicity counterpart. GAO also reported approximately 13,500 nearby schools within 10 miles, representing about 14% of all public K–12 schools; the source rounds these figures.
Among predominantly same-race/ethnicity schools with a different-race/ethnicity counterpart within 5 miles, 86% had that counterpart in a different school district. At the 10-mile distance, 90% had the counterpart in a different district. These figures show how district lines can separate schools that are geographically close while serving different student populations. GAO, 2020–21.
About 70% of U.S. schoolchildren attended their local neighborhood schools, according to the national estimate cited in the GAO report. GAO’s 2020–21 analysis also found that school district boundaries typically determined which public school a student could attend. The neighborhood-school estimate is cited by the report and is not presented as a new GAO survey estimate.
District secession and demographic change
District secession can change the demographic composition of the districts on either side of a new boundary. GAO identified 36 new school districts created through secession from 2009–10 through 2019–20. One year after secession, the new districts had roughly three times the White-student share of the districts that remained and roughly twice the Asian-student share. Their Hispanic-student share was about two-thirds the rate in the remaining districts, while their Black-student share was about one-fifth the rate. The percentage of students eligible for free or reduced-price lunch in new districts was about half the rate in the districts that remained.
The Southern results were mixed but frequently reflected economic and racial differences. Across three states, 13 district secessions occurred. Nine of the 13 new districts were Whiter and wealthier, while four were more Hispanic or Black and poorer. Three of those four secessions involved districts under active federal desegregation orders or plans. These statistics describe demographic differences associated with the secessions; GAO cautions that its descriptive analysis should not be used to determine unlawful discrimination or illegal segregation. GAO, 2009–10 to 2019–20.
Economic concentration alongside racial patterns
Racial and ethnic concentration often appears alongside economic concentration, but the measures are distinct. In 2019–20, 59.42% of comprehensive-support-and-improvement (CSI) schools had 75%–100% of students eligible for free or reduced-price lunch. The comparable share was 39.27% for Title I schools and 27.15% for all public schools.
The racial composition of these school groups also differed:
| School group | Predominantly Black | Predominantly Hispanic | Predominantly White |
|---|---|---|---|
| CSI schools | 18.62% | 12.13% | 14.04% |
| Title I schools | 7.01% | 14.06% | 25.77% |
| All public schools | 4.63% | 8.98% | 29.81% |
CSI means comprehensive support and improvement. Title I and CSI comparisons are school-level composition statistics, not causal estimates of segregation. GAO, Education Could Enhance Oversight of School Improvement Activities, 2019–20.
Locale statistics add another dimension. In 2019–20, 47.01% of CSI schools were urban, compared with 30.83% of Title I schools and 27.68% of all public schools. CSI schools were 20.71% suburban, compared with 28.27% of Title I schools and 31.44% of all public schools. Rural schools represented 32.28% of CSI schools, 40.89% of Title I schools, and 40.88% of all public schools. These figures describe where the school categories were located and how their enrollment was distributed; they do not establish that race, poverty, school improvement status, or locale caused another measured outcome. GAO, 2019–20.