By Daniel Dockterman, PhD. Chief Methodologist of The School Review
This reports analyses student racial/ethnic diversity in U.S. pre-kindergarten through twelfth grade (PK-12) Public Schools using 2016-17 Common Core data from the National Center for Education Statistics (NCES). We measure how evenly distributed a school’s student body population is across racial/ethnic categories by calculating a Diversity Index, ranging from 0 (no diversity) to 100 (full diversity). Using this metric, we examine school diversity nationally as well as by state, metropolitan area, and county. In addition, we compare diversity across school type (public, charter, magnet), urbanicity (urban vs. rural), and level (elementary, middle, high). We find Asian, Hawaiian Native or Other Pacific Islander, and students of two or more races are substantially more likely to attend the most racially/ethnically heterogenous schools while Hispanic and American Indian or Alaska Native students disproportionately attend the most homogeneous schools. We also find greater racial/ethnic diversity in magnet and urban schools. The East and West Coasts contain schools with greater diversity than those outside the coastal areas.
Research suggests students attending more racially/ethnically diverse schools enjoy several academic, psychosocial, civic, and economic benefits. First, school diversity can enhance student learning experiences by promoting self-reflection and academic growth. Spending time among peers with divergent backgrounds exposes students to new perspectives. Different viewpoints expand students’ understanding of the world and can help them combat stereotypical thinking by challenging their preconceived beliefs. By recognizing and acknowledging cultural differences, students can reduce their own prejudices and become more tolerant. Moreover, greater diversity in schools can help minority students close racial achievement gaps and reduce their likelihood of dropping out.
Second, research indicates students feel safer when they are educated in a diverse setting. Learning about other cultures and backgrounds permits students to feel more comfortable in their own skin. One study found middle school students attending more racially diverse schools felt less lonely, less bullied, and safer. Greater racial diversity also increases the likelihood of cross-ethnic friendships, which can make students feel more socially competent and less vulnerable to peer harassment.
Finally, attending a racial and ethnically diverse school helps students learn to engage in civic discourse, thereby preparing them for citizenship in a multicultural democracy and a career in the global job market. In diverse classrooms, students learn to converse respectfully and dispassionately with students of different backgrounds. They gain an appreciation for their peers’ divergent life experiences. Such discourse and collaboration can improve cognitive skills, including problem solving, creativity, and critical thinking, and ready students to navigate adulthood in an increasingly international society—a skill that employers value.
To learn more about the benefits of greater student racial/ethnic diversity in schools, we recommend a 2016 report from The Century Foundation and a 2012 report from The Civil Rights Project at the University of California, Los Angeles.
Student diversity within a school and classroom setting can take several forms (e.g., religious, gender, socioeconomic background, English language skills). At The School Review we measure a school’s student body racial/ethnic composition by calculating a Diversity Index. While no measure of diversity can be perfect given available data, our Diversity Index is a measurement tool for capturing the degree of racial/ethnic heterogeneity in a single metric. Ranging from 0 (no diversity) to 100 (full diversity), the Diversity Index indicates how evenly distributed students are among the seven racial/ethnic categories collected by the U.S. Department of Education (White, Black or African American, Hispanic/Latino, Asian, Hawaiian Native or Other Pacific Islander, American Indian or Alaska Native, Two or more races). Student racial/ethnic information is from the 2016-17 school year, the most recent year for which Common Core data of school membership from the National Center for Education Statistics (NCES) is publicly available.
For reference, Table 1 provides the Diversity Index for various hypothetical racial/ethnic distributional scenarios. As shown in the table, a school with a student population comprised of only one racial/ethnic group will have a Diversity Index of 0, a school with a student population evenly divided between two racial/ethnic groups will have a Diversity Index of 35, a school with a student population evenly split between four racial/ethnic groups will have a Diversity Index of 65, and a school with a student population nearly evenly divided among all seven racial/ethnic groups will have a Diversity Index approaching 100. It is important to note that our Diversity Index does not discriminate between types of racial/ethnic groups. That is, a school with a student body that is half white and half black, and a school that is half Hispanic and half Asian will have the same Diversity Index. Similarly, a school that is 90% white and 10% black will have an equivalent Diversity Index to a school that is 90% black and 10% white (see Appendix A).
|Table 1: Diversity Index for Various Hypothetical Racial/Ethnic Distributions|
|White %||Hispanic %||Black %||Asian %||American Indian or
Alaska Native %
|Hawaiian Native or
Other Pacific Islander %
|Two or More Races %||Diversity Index|
Figure 1 displays the distribution of diversity across all PK-12 public schools in the U.S. The average Diversity Index for all 92,3081 schools in the United States is 29. This is the median school Diversity Index as well. We can also compute an average Diversity Index for all schools weighted by the number of students enrolled at each school. For example, Morningside Middle School in Tarrant County, Texas is twice as large as Northview Elementary School in Champaign County, Illinois (644 students vs. 322 students). Therefore, Morningside can be assigned twice as much weight as Northview in computing the average Diversity Index. When weighting by the total number of students, the average diversity for all schools—henceforth known as the “Weighted Average Diversity Index”—is 32. The difference between an unweighted and the Weighted Average Diversity Index indicates that schools with more students tend to be more diverse. Note that this association is not a function of the Diversity Index measure itself but reflects that fact that schools with fewer students are more often located in rural areas and tend to be more racially/ethnically homogeneous. Due to the wide disparity in the size of U.S. public schools2, the Weighted Average Diversity Index is used exclusively throughout this report.
Figure 1 also shows that diversity in PK-12 public schools exhibits a bimodal distributional shape, meaning there are two discernible peaks. The first (and highest) peak occurs at a Diversity Index of 7. This is the mode of the distribution and 2,131 schools have this Diversity Index. A Diversity Index of 7 means at least 94% of the students belong to one racial/ethnic group. Thus, the typical U.S. PK-12 public school has a student body population that is quite racially/ethnically homogeneous. Indeed, 19.5% of schools have a Diversity Index less than 10 and 1.6% of schools have a Diversity Index less than 1.
The second (and smaller) peak of the distribution occurs at a Diversity Index of 38. Nearly 1,900 schools have a Diversity Index of 38 and roughly 18.7% of schools have a Diversity Index between 34 and 44. Thus, the second most common U.S. PK-12 public school is significantly more racially/ethnically heterogeneous. After the second peak, there are fewer and fewer schools with greater Diversity Index values. Indeed, only 2.8% of schools have a Diversity Index greater than 60.
Homogeneous Schools vs. Heterogeneous Schools. Also highlighted in Figure 1 are two school diversity classification groups. Shaded in red are the least racially/ethnically diverse 10% of U.S. public schools. These schools have Diversity Indices roughly less than 6 and are henceforth referred to as “homogeneous schools”. Homogeneous schools tend to have smaller student populations and are more often in rural geographic locales. In total, there are 9,231 homogeneous schools.
Shaded in blue are the most racially/ethnically diverse 10% of U.S. public schools. These schools have Diversity Indices roughly greater than 53 and are henceforth referred to as “heterogeneous schools”. Heterogenous schools have larger student body populations (678 vs. 381) and are more often situated in urban communities. In total, there are 9,231 heterogeneous schools. While the parameters of homogeneous and heterogeneous schools are somewhat arbitrary, creating these groupings is useful for differentiation, and so these two school classifications will be compared across various characteristics and outcomes throughout this report.
As shown in Figure 2, student racial/ethnic groups are distributed very differently between these two school classification groupings. Whites comprise half (48.2%) of the U.S. PK-12 student population, yet only a third (32%) of the student body in heterogeneous schools. Hispanics, on the other hand, disproportionately attend more racially/ethnically isolated schools. While Hispanics comprise a quarter (26.5%) of the U.S. PK-12 student population they make up 34.4% of the student body in homogeneous schools. Interestingly, black students are disproportionately concentrated in both the most and least racially/ethnically segregated schools. Black students are more likely to attend homogeneous (18.4%) and heterogeneous (20.5%) schools than their overall proportion of the school population would suggest (15.2%).
Asian students are the most strikingly discrepant across school classification type. They comprise only 0.4% of the student body of homogeneous schools, yet 12.1% of the population in heterogenous schools. Hawaiian Native or Other Pacific Islander students and students that are two or more races exhibit similar attendance patterns as Asian students. Hawaiian Native or Other Pacific Islander students comprise less than 0.1% of students in homogeneous schools, yet are 1.1% of the heterogeneous school population. Likewise, students that are two or more races comprise 0.6% of the homogeneous population, but 6.4% of the heterogenous population.
Finally, the school enrollment patterns of American Indian or Alaska Native students are unlike any of the other racial/ethnic groups. In contrast to Asian, Hawaiian Native or Other Pacific Islander, and students that are two or more races, this racial/ethnic group makes up a greater percentage of the student population in homogeneous (1.6%) than heterogeneous schools (1.2%). Moreover, American Indian or Alaska Native students constitute 3.1% of the homogeneous school population (see Appendix Table B1).3 Of the 9,231 total homogeneous schools, 282 are predominately American Indian or Alaska Native.4
The fact that American Indian or Alaska Native students constitute 3.1% of the homogeneous school population is remarkable given that this racial/ethnic group comprises only 1% of the student population. By comparison, Asian, Hawaiian Native or Other Pacific Islander, and two or more races total 8.7% of U.S. student population, but are the predominate racial/ethnic group in only 7 of the 9,381 homogeneous schools (0.1%). This difference is staggering and reveals the degree of racial/ethnic school segregation that exists for a sizeable portion of American Indian or Alaska Native students.
Another way to examine racial/ethnic differences between heterogenous and homogeneous schools is to compare the total number of students in each of these school types. Table 2 reveals that there are 2.7 million more students in heterogenous schools (6,256,974) than in homogeneous schools (3,520,867)—even though the total number of each of these school classifications is the same (9,231). Indeed, the average size of heterogenous schools is 678 students and the average size for homogeneous schools is 381 students. This difference results in a greater number of students from each of seven racial/ethnic groups attending heterogenous schools than homogeneous schools.
|Table 2: Number of Students, by Race/Ethnicity, Homogeneous Schools vs. Heterogeneous Schools|
|Number of Students||Ratio|
|Race/Ethnicity||Heterogeneous Schools||Homogeneous Schools||All Schools||Heterogenous to Homogeneous|
|White||2,000,067||1,568,434||23,936,385||1.3 to 1|
|Hispanic||1,676,331||1,212,124||13,144,830||1.4 to 1|
|Black||1,279,950||647,734||7,572,774||2.0 to 1|
|Asian||757,543||14,199||2,548,842||53.4 to 1|
|Two or more races||401,971||21,252||1,803,387||18.9 to 1|
|American Indian or Alaska Native||72,340||55,824||497,499||1.3 to 1|
|Hawaiian Native or Other Pacific Islander||68,772||1,300||181,906||52.9 to 1|
|Total||6,256,974||3,520,867||49,685,623||1.8 to 1|
Table 2 also shows that the ratio of the student population in heterogenous versus homogeneous schools follows two distinct trends, which align with the findings of Figure 2. White, Hispanic, Black, and American Indian or Alaska Native students all have a heterogenous to homogeneous ratio ranging from 1.3-2.0 to 1. Thus, for these four racial/ethnic groups, the greater number of students in heterogenous schools reflects the larger student body size in heterogenous schools on average. By contrast, Asians students, Hawaiian Native or Other Pacific Islander students, and students of two are more races are far-and-away disproportionately concentrated in heterogenous schools. Incredibly, there are 53.4 times as many Asian students, 52.9 times as many Hawaiian Native or Other Pacific Islander students, and 18.9 times as many students of two or more races in heterogenous schools than in homogeneous schools.
Diversity by School Type. Of the more than 92,000 PK-12 public schools included in this report, 82,898 are classified as traditional public schools, 6,306 are classified as charter schools, and 3,012 are classified as magnet schools.5 As shown in Figure 3, patterns in school diversity vary noticeably between these school types. The distribution in the Diversity Index of traditional public schools and charter schools are the most similar. The Weighted Average Diversity Indices for these two school types are nearly equivalent—30 for charter schools and 31 for public schools—and both school types exhibit distinct bimodal distributions. By contrast, the spread in diversity across magnet schools is quite different, particularly when compared to public schools. The Weighted Average Diversity Index for magnet schools, at 38, is noticeably higher. Moreover, the most common index value for magnet schools is 43 versus only 3 for charter schools and 7 for traditional public schools. Finally, more than four times as many magnet schools are racially/ethnically heterogenous (559) than homogeneous (122) (see Table 3).
Diversity by School Urbanicity. Even larger racial/ethnic diversity differences are apparent depending on the degree of school urbanicity. Simply put, urban and suburban students are significantly more likely to attend racial/ethnically diverse schools than students living in towns or rural areas. The Weighted Average Diversity Index for the 54,707 urban and suburban schools is 35 and the most common Diversity value is 43. Moreover, there are 2.6 times as many heterogenous urban/suburban schools (8,063) than homogeneous (3,147). By contrast, the Weighted Average Diversity Index for the 37,601 schools in towns or rural areas is 24 and the modal Diversity Index is 7. This means the typical rural school is extremely racially/ethnically homogeneous. Indeed, there are 80% fewer heterogenous schools (1,167) in towns or rural areas than homogeneous schools (6,084) (see Table 3).
Diversity by School Level. Elementary, middle, and high schools are all similarly diverse (DI=32). This equivalency is a function of two completing forces. On the one hand, a greater proportion of students in younger grades are non-white. On the other hand, school size is correlated with greater diversity and elementary schools have fewer students (Mean=463 students) than middle schools (Mean=598 students), which tend to have fewer students than high schools (Mean=837 students). Interestingly, schools that combine grades 6-12 (i.e., middle/high schools) tend to be less diverse, with an average Diversity Index of 23. This is driven by the fact that middle/high schools are smaller (Mean=332 students), disproportionately rural (62.1%), and disproportionately non-magnet (97.5%). As a result, there are one-third as many heterogeneous middle/high schools (267) as homogeneous middle/high schools (826) (see Table 3).
|Table 3: Homogeneous vs. Heterogeneous School Counts, by School Level, School Type, and Urbanicity|
| Weighted Average
|All Schools||Heterogenous to
|Elementary||32||5,227||3,912||46,166||1.3 to 1|
|Middle||32||1,453||1,254||14,600||1.2 to 1|
|High||32||1,327||1,667||16,051||0.8 to 1|
|Elementary-Middle||30||678||1,225||7,723||0.6 to 1|
|Middle-High||25||267||826||4,692||0.3 to 1|
|PK-12||30||270||299||2,828||0.9 to 1|
|Regular||31||8,063||8,326||82,898||1.0 to 1|
|Charter||30||596||771||6,306||0.8 to 1|
|Magnet||38||559||122||3,012||4.6 to 1|
|Urban/Suburban||35||8,063||3,147||54,707||2.6 to 1|
|Town/Rural||24||1,167||6,084||37,601||0.2 to 1|
|Total||32||9,231||9,231||92,308||1 to 1|
There is significant variation in school racial/ethnic diversity across states. Due in part to their large Native Hawaiian or Other Pacific Islander and Asian student populations, schools in Hawaii are the most racially/ethnically diverse in the U.S. (see Figure 5). The Weighted Average Diversity Index for a school in Hawaii is 56. Hawaii’s high diversity is evenly spread across schools in the state as the average Diversity Index in all four counties—Hawaii, Honolulu, Kauai, and Maui—is above 50. In the contiguous U.S., schools in Nevada, Delaware, and Oklahoma are on average the most racially/ethnically diverse. These states have Weighted Average Diversity Indices of 46, 45, and 44, respectively.
West Virginia, Maine, and Vermont have the least diverse schools. The Weighted Average Diversity Indices for schools in each of these states is 11. In fact, there are no schools in these three states as diverse as the average school in Hawaii (i.e., with a Diversity Index greater than or equal to 56). West Virginia, Maine, and Vermont also have the smallest proportion of non-white students of any of the states, at 9.6%, 9.5%, and 10.3%, respectively.
Figure 6 plots the Diversity Index distributions for all 50 states and the District of Columbia with states ordered from lowest Weighted Average Diversity Index (West Virginia) to highest (Hawaii). In addition to differences in average school diversity, Figure 6 reveals dissimilarities between the states in their degree of variation in school diversity across schools. Racial/Ethnic diversity in West Virginia, Maine, and Vermont varies little from school to school, as nearly all schools in these states have very low Diversity Indices. Conversely, with a standard deviation of 24, Alaska has the greatest spread in Diversity Index values across schools. Whereas roughly a quarter of Alaska schools have a Diversity Index in the single digits, another quarter of schools have an Index value greater than 50. This is due to the fact that most rural schools in Alaska are very racially/ethnically homogeneous while schools in Anchorage, Alaska are especially racially/ethnically heterogeneous. Indeed, many of the most diverse schools in the country are located in Anchorage.
Of all the states, student racial/ethnic diversity in New York public schools are most similar to the U.S. PK-12 public schools as a whole. Like in the U.S., schools in New York have a Weighted Average Diversity Index of 32. Moreover, the distribution of Diversity Indices across New York schools is bimodally shaped. The smaller peak of the distribution occurs at a Diversity Index of 8 while the higher peak occurs at an Index of 40. Two factors contributing to these distributional similarities in racial/ethnic diversity are the comparable share of students by racial/ethnic group and the fact New York—like the U.S.—has a sizeable proportion of both urban and rural schools.
Next, Figure 7 shows the nuanced relationship between the proportion of a state’s PK-12 student population that is non-white and their Weighted Average Diversity Index. The relationship between these variables is essentially linear when a small percentage of a school’s student body population is non-white. For example, West Virginia, Maine, and Vermont have few non-white students and low school diversity. However, as the proportion of non-white students in a state increases, the variation in diversity between states widens.
At the extreme, Hawaii and the District of Columbia have similar non-white student proportions, 87.6% and 89.3%, respectively. However, Hawaii schools are far more diverse on average. This is partly because Hawaiian students are predominantly comprised of five racial/ethnic groups (29.6% Native Hawaiian or Other Pacific Islander, 29.5% Asian, 12.4% white, 13.2% Hispanic, and 13.2% two or more races) whereas students in D.C. are mostly comprised of only three groups: black (69.5%), Hispanic (16.1%) and white (10.7%). Yet, it is also a result of greater racial/ethnic segregation in D.C. schools. Of the 222 public schools in D.C., 88 (39.6%) have a Diversity Index less than 6—with black students the homogeneous racial/ethnic group in all 88 of these schools. By comparison, only 2 of the 289 total public schools in Hawaii (0.7%) have Diversity Indices less than 6 (see Figure 8).
Another noteworthy diversity comparison is between the schools in Delaware and Mississippi. A very similar proportion of these states PK-12 public school populations are non-white: 54.9% of students are non-white in Delaware and 55.5% of students are non-white in Mississippi. Yet, the Weighted Average Diversity Index is 19 points higher in Delaware (45 versus 26). Again, this is partly due to the fact that students in Delaware are more racial/ethnically diverse. However, it is also because schools in Mississippi are far more racially/ethnically segregated. Only one of the 218 total schools in Delaware is homogeneous (0.5%) versus 192 out of the 892 total schools in Mississippi (21.5%) (see Figure 8). In these 192 homogeneous schools, blacks are the predominate racial/group in 170 schools and whites are the prevailing group in 22 schools.
Core based statistical area (CBSA) is a U.S. Census term that defines a geographic area with an urban core (at least 10,000 people) with any adjacent counties socially and economically integrated. CBSAs include both metropolitan and micropolitan statistical areas. As of 2015, there were 933 CBSA’s in the U.S. This section focuses on school racial/ethnic diversity in Large, Midsize, and Small metro areas.6 School Diversity in Large Metro Areas. Figure 9 ranks the Weighted Average Diversity Index of schools in Large metro areas (defined as having at least 250,000 PK-12 public school students). There are 37 Large metro areas with New York-Newark-Jersey City, NY-NJ-PA—at 4,758 public schools and 2.8 million students—the most populous. As shown in the figure, three Large metro areas have average School Diversity Indices of 48: Sacramento—Roseville—Arden-Arcade in California, Seattle-Tacoma-Bellevue in Washington, and Las Vegas-Henderson-Paradise in Nevada. Conversely, with a Weighted Average Diversity Index of 17, Pittsburgh, Pennsylvania is the only metro area for which the average school’s diversity is below 20.
Figure 9 also shows that 26 of the 37 largest metro areas have indices greater than or equal to the U.S. Weighted Average Diversity Index of 32. This finding is not surprising given the greater racial/ethnic heterogeneity in urban/suburban locales (see Figure 4). Figure 9 also reveals the clear geographic patterns in school diversity amongst Large metro areas. The eight metro areas with the highest average school diversity are all situated on the East or West Coasts.7 By contrast, of these Large metro areas, seven of the eight with the least diverse schools are located in the middle region of the U.S.8 The lone exception is Boston-Cambridge-Newton, MA-NH, which has an average school diversity of 29. These metro areas with less diverse schools have smaller proportions of non-white students and/or tend to be more racially/ethnically segregated.
School Diversity in Midsize Metro Areas. Figure 10 displays the Midsize metro areas (defined as having between 50,000 and 250,000 public school students) with the most and least racially/ethnically diverse PK-12 public schools. Of the 117 metro areas that fit this size criterium, four have schools for which the Weighted Average Diversity Index is greater than 50: Vallejo-Fairfield in California, Urban Honolulu in Hawaii, Anchorage in Alaska, and Killeen-Temple in Texas. Three of the top 15 Midsize metro areas with highly diverse schools are also in Florida—Lakeland-Winter Haven, Port St. Lucie, and Cape Coral-Fort Myers—and three are in North Carolina—Fayetteville, Raleigh, and Durham-Chapel Hill.
Except for Portland-South Portland in Maine, Midsize metro areas with the least diverse schools are not situated on the East or West Coasts. Rather, several of these Midsize metro areas are in Ohio, Utah, and Texas. Indeed, all three Midsize metro areas that have an average school diversity less than 5 are in Texas. Laredo, McAllen-Edinburg-Mission, and Brownsville-Harlingen metro areas each abut the U.S.-Mexico border and have PK-12 public school students that are at least 96% Hispanic. Thus, the low school diversity for schools in these metro areas is entirely driven by the homogeneous student populations in these metro areas as a whole.
School Diversity in Small Metro Areas. Figure 11 displays the Small metro areas (defined as having between 25,000 and 50,000 public school students) with the most and least racially/ethnically diverse PK-12 public schools. Of the 97 metro areas that fit this size criterium, four have schools for which the Weighted Average Diversity Index is greater than or equal to 45: Yuba City in California, Ocala in Florida, Atlantic City-Hammonton in New Jersey, and Warner Robins in Georgia. Additionally, three of the most diverse Small metro areas are in Texas: College Station-Bryan, Longview, and Tyler.
Small metro areas with the least diverse schools are spread across the U.S. Indeed, no state has more than two Small metro areas for which the schools have a Weighted Averages Diversity Index of 20 or lower. The four Small metro areas with the least diverse schools are El Centro, California; Kingsport-Bristol-Bristol, Tennessee-Virginia; Charleston, West Virginia; and Johnson City, Tennessee.
Finally, the low school race/ethnicity diversity in Monroe, Louisiana merits further examination. Recall that for the Midsize metro areas of Laredo, McAllen-Edinburg-Mission, and Brownsville-Harlingen in Texas, low school diversity was driven entirely by the extremely homogeneous student populations in these metro areas. Again, we find a similar trend for most of the Small metro areas with the least diverse schools. For instance, students in El Centro, California are 92.0% Hispanic and students in Kingsport-Bristol-Bristol, Virginia are 91.0% White. However, the largest racial/ethnic group in Monroe, Louisiana (Black) only comprise 48.5% of the student population while the second largest racial/ethnic group (White) comprise 46.1%. The fact that there is no majority racial/ethnic group in Monroe, yet the typical school is quite homogeneous indicates substantial racial/ethnic school segregation in this metro area. Indeed, greater racial/ethnic school segregation is more common in southern U.S. states. Among majority-minority metro areas with at least 25,000 PK-12 public school students, only five have schools for which the Weighted Average Diversity Index is less than 30: Mobile and Tuscaloosa in Alabama and Shreveport-Bossier City, Baton Rouge, and Monroe in Louisiana.
Figure 12 presents the 15 metro areas (Small, Midsize, and Large) with the greatest proportion of racially/ethnically homogeneous schools. In 6 of these 15 metro areas all the homogeneous schools are majority Hispanic students. These metro areas either abut or are within 50 miles of the U.S.-Mexico border. This includes four metro areas in Texas (Laredo, McAllen-Edinburg-Mission, Brownsville-Harlingen, El Paso), El Centro in California, and Las Cruces in New Mexico. There are also six metro areas in which all the homogeneous schools have predominately white students. Five of these six metro areas are located in the Appalachian Region of the U.S., while Utica-Rome, New York is located within 100 miles of this region.
The remaining three metro areas with the greatest proportion of homogeneous schools are comprised predominately of black students and are in the southern region of the U.S. In Jackson, Mississippi, all 60 of the homogeneous schools have predominately black students. Lastly, in Monroe, Louisiana there are 18 homogeneous schools that are majority black and 3 homogeneous schools that are majority white and in Alexandria, Louisiana there are 15 homogeneous schools (ten majority black and five majority white). The fact that there are multiple homogeneous schools with different majority racial/ethnic groups indicates there is significant racial/ethnic student segregation across the schools in these two Louisiana metro areas.
Counties (and equivalents) are the primary legal or administrative subdivision of a state. Currently, there are 3,141 counties and county equivalents in the 50 states and District of Columbia9. This section focuses on school racial/ethnic diversity in Large, Midsize, Small, and Tiny counties.
School Diversity in Large Counties. Table 4 lists the Large counties (defined as counties having at least 100,000 public school students) with the most racially/ethnically diverse PK-12 public schools in the U.S. Counties are ranked according to their Average Diversity Index weighted by total number of students at each school. Honolulu County in Hawaii, Sacramento County in California, and Montgomery County in Maryland are the Large counties with the most diverse public schools. The average school in each of these counties has a Diversity Index of 53 or greater. Of the 20 Large counties with the most racially/ethnically heterogenous schools, four are in California and four are in Texas. Sacramento County in California, King County in Washington, and Clark County in Nevada are the most populous counties of these 20 most diverse.
Table 4: Large Counties with Most Racially/Ethnically Heterogeneous Schools,
Ranked by Average School Diversity Index
|Most Diverse School
in the County
|Honolulu County, HI||55||Pearl Harbor Elementary School||77|
|Sacramento County, CA||54||Harriet G. Eddy Middle School||74|
|Montgomery County, MD||53||Cloverly Elementary School||69|
|Fairfax County, VA||52||Saratoga Elementary School||68|
|Fort Bend County, TX||51||Irma Dru Hutchison Elementary School||67|
|King County, WA||51||Panther Lake Elementary School||79|
|Gwinnett County, GA||51||Trickum Middle School||68|
|Tulsa County, OK||49||Jarman Elementary School||74|
|Collin County, TX||49||Rasor Elementary School||67|
|Polk County, FL||49||Dixieland Elementary School||59|
|Clark County, NV||48||Tony Alamo Elementary School||74|
|Pierce County, WA||48||Helen B. Stafford Elementary School||75|
|Wake County, NC||46||Wake Early College of Health and Science||66|
|Hillsborough County, FL||46||Tampa Palms Elementary School||69|
|San Joaquin County, CA||46||Podesta Ranch Elementary School||67|
|Contra Costa County, CA||45||Rodeo Hills Elementary School||71|
|Alameda County, CA||45||Sequoia Elementary School||73|
|Williamson County, TX||45||Harmony School of Political Science & Communication-Austin||62|
|Denton County, TX||44||Kent Elementary School||66|
|Orange County, FL||44||Orlando Science Elementary Charter School||66|
School Diversity in Midsize Counties. Tables 5 lists the Midsize counties (defined as counties having between 25,000 and 99,999 public school students) with the most racially/ethnically diverse PK-12 public schools in the U.S. Counties are ranked according to their Average Diversity Index weighted by total number of students at each school. Anchorage Municipality in Alaska, Hawaii County in Hawaii, and Solano County in California are the Midsize counties with the most diverse public schools. The Weighted Average Diversity Index for schools in each of these counties is 56 or greater. Indeed, with a Weighted Average Diversity Index of 61, public schools in Anchorage are the most diverse in the U.S. while North Star Elementary School is the most diverse school in the U.S. Of the 20 Midsize counties with the most racially/ethnically heterogenous schools, five are in Virginia and three are in Florida.
Table 5: Midsize Counties with Most Racially/Ethnically Heterogeneous Schools,
Ranked by Average School Diversity Index
|Most Diverse School
in the County
|Anchorage Municipality, AK||61||North Star Elementary School||92|
|Hawaii County, HI||58||Pa’auilo Elementary & Intermediate School||74|
|Solano County, CA||56||Angelo Rodriguez High School||72|
|Howard County, MD||54||Gorman Crossing Elementary School||66|
|St. Lucie County, FL||54||Bayshore Elementary School||60|
|Bell County, TX||52||Saegert Elementary School||64|
|Prince William County, VA||52||Sonnie Penn Elementary School||68|
|Stafford County, VA||50||Kate Waller Barrett Elementary School||64|
|Brazoria County, TX||49||Glenn York Elementary School||67|
|Cleveland County, OK||49||Sky Ranch Elementary School||64|
|Virginia Beach City, VA||49||Glenwood Elementary School||67|
|Cumberland County, NC||48||C. Wayne Collier Elementary School||62|
|Seminole County, FL||48||Goldsboro Elementary Magnet School||64|
|Cabarrus County, NC||48||Coltrane-Webb Elementary School||61|
|Multnomah County, OR||47||Margaret Scott Elementary School||76|
|Arlington County, VA||47||Hoffman-Boston Elementary School||67|
|Marion County, FL||47||Greenway Elementary School||60|
|Jefferson Parish, LA||47||Thomas Jefferson High School for Advanced Studies||61|
|San Francisco County, CA||47||Parks (Rosa) Elementary School||75|
|Loudoun County, VA||46||Frederick Douglass Elementary School||63|
School Diversity in Small Counties. Tables 6 lists the Small counties (defined as counties having between 10,000 and 24,999 public school students) with the most racially/ethnically diverse PK-12 public schools in the U.S. Counties are ranked according to their Average Diversity Index weighted by total number of students at each school. Comanche County and Muskogee County in Oklahoma and Maui County in Hawaii are the Small counties with the most diverse public schools. The average school diversity in each of these counties is 55 or greater. Of the 20 most diverse small counties, six are in North Carolina and three are in Oklahoma.
Table 6: Small Counties with Most Racially/Ethnically Heterogeneous Schools,
Ranked by Average School Diversity Index
|Most Diverse School
in the County
|Comanche County, OK||57||Washington Elementary School||72|
|Maui County, HI||57||King Kamehameha III Elementary School||73|
|Muskogee County, OK||55||Alice Robertson Junior High School||73|
|Lee County, NC||53||J. R. Ingram Junior Elementary School||57|
|Coryell County, TX||52||Hettie Halstead Elementary School||65|
|Alexandria City, VA||52||John Adams Elementary School||60|
|Harnett County, NC||51||Overhills Middle School||58|
|Highlands County, FL||51||Sun 'N Lake Elementary School||58|
|Sampson County, NC||50||Sampson Middle School||60|
|Yuba County, CA||48||Linda Elementary School||58|
|Liberty County, GA||48||Taylors Creek Elementary School||55|
|Robeson County, NC||47||East Robeson Primary School||70|
|Rusk County, TX||47||Chandler Elementary School||53|
|Craven County, NC||47||Trent Park Elementary School||66|
|Terrebonne Parish, LA||47||Grand Caillou Elementary School||73|
|Orange County, NC||47||Northside Elementary School||59|
|Rogers County, OK||47||J. W. Sam Elementary School||60|
|Sutter County, CA||47||River Valley High School||59|
|Clarke County, GA||46||Cleveland Road Elementary School||58|
|Wicomico County, MD||46||North Salisbury Elementary School||58|
School Diversity in Tiny Counties. Tables 7 lists the Tiny counties (defined as counties having between 5,000 and 9,999 public school students) with the most racially/ethnically diverse PK-12 public schools in the U.S. Counties are ranked according to their Average Diversity Index weighted by total number of students at each school. Kauai County in Hawaii, Hoke County in North Carolina, and Fairfax City in Virginia are the Tiny counties with the most diverse public schools. The average school diversity in each of these counties is 59 or greater. Of the 20 most diverse Tiny counties, seven are in Oklahoma, four are in North Carolina, and four are in Virginia.
Table 7: Tiny Counties with Most Racially/Ethnically Heterogeneous Schools,
Ranked by Average School Diversity Index
|Most Diverse School
in the County
|Kauai County, HI||61||Kalaheo Elementary School||71|
|Hoke County, NC||60||Sandy Grove Elementary School||66|
|Fairfax City, VA||59||Lanier Middle School||62|
|Geary County, KS||54||Washington Elementary School||65|
|Charlottesville City, VA||53||Greenbrier Elementary School||60|
|Walker County, TX||52||Huntsville Intermediate School||55|
|Washington County, TX||51||Alton Elementary School||56|
|Pontotoc County, OK||51||Washington Elementary School||63|
|Seminole County, OK||51||Wewoka Middle School||62|
|Richmond County, NC||51||Mineral Springs Elementary School||58|
|McCurtain County, OK||51||Idabel Middle School||74|
|Okmulgee County, OK||50||Preston High School||65|
|Ottawa County, OK||50||Commerce High School||60|
|Tift County, GA||50||G. O. Bailey Primary School||55|
|Prince George County, VA||50||David A. Harrison Elementary School||57|
|Sequoyah County, OK||50||Moffett Public School||61|
|Scotland County, NC||49||Scotland Early College High School||62|
|Cherokee County, OK||49||Greenwood Elementary School||64|
|Franklin County, NC||49||Louisburg Elementary School||54|
|Harrisonburg City, VA||48||Stone Spring Elementary School||53|
Counties with Least Diverse Schools. Tables 8 lists the Midsize and Large counties with the least diverse PK-12 public schools in the U.S. Webb County, Hidalgo County, and Cameron County in Texas have the most racially/ethnically homogeneous PK-12 public schools. Schools in each of these counties have a Weighted Average Diversity Index of less than 5. Eleven of the least diverse Midsize and Large counties are majority white (three of which are in Pennsylvania), six are majority Hispanic (four of which are in Texas), and three counties are majority Black. Recall that four of the 20 Large counties with the most racially/ethnically diverse schools are in Texas (see Table 4). Thus, Texas has some of the most heterogenous and homogeneous counties in the U.S. Finally, it worth noting that Hidalgo County in Texas is the largest county of the group and is the 23nd largest county in the U.S in terms of PK-12 public school students.
Table 8: Midsize and Large Counties with Most Racially/Ethnically Homogeneous Schools,
Ranked by Average School Diversity Index
|Webb County, TX||98.8% Hispanic||70,059||1|
|Hidalgo County, TX||97.5% Hispanic||224,753||3|
|Cameron County, TX||96.7% Hispanic||105,969||4|
|Butler County, PA||94.3% White||25,491||7|
|Livingston County, MI||93.6% White||26,530||7|
|Jefferson County, MO||93.2% White||34,797||8|
|Imperial County, CA||92.0% Hispanic||38,484||8|
|York County, ME||92.1% White||25,022||9|
|Medina County, OH||91.9% White||26,518||9|
|Clermont County, OH||91.1% White||25,542||10|
|El Paso County, TX||90.4% Hispanic||177,797||11|
|Westmoreland County, PA||90.1% White||45,119||11|
|Rockingham County, NH||90.4% White||43,886||11|
|Hinds County, MS||88.4% Black||38,776||11|
|Wright County, MN||90.1% White||26,903||11|
|Washington County, PA||88.7% White||26,170||13|
|Saratoga County, NY||87.9% White||32,238||14|
|Baltimore City, MD||80.6% Black||81,704||16|
|Orleans Parish, LA||81.4% Black||47,709||17|
|Yuma County, AZ||83.3% Hispanic||38,103||17|
School Segregation by County. Another way to measure PK-12 racial/ethnic diversity in a county is by computing an aggregated diversity index of all students. That is, the total number of students in each racial/ethnic group can be summed across all the schools in a county and then a diversity index can be computed on these aggregate counts. This method ignores the racial/ethnic composition of students within schools, and instead measures the diversity of the PK-12 student population in the county. This diversity metric is henceforth referred to as the “County Diversity Index”.
Figure 13 plots the County Diversity Index and Weighted Average School Diversity Index for all counties with at least 10,000 public school students. Note that all points are below the line y=x. This is because it is not mathematically possible for the County Diversity Index to be greater than the Weighted Average School Diversity Index for a given county as the line represents the highest achievable Diversity Index value if students from each racial/ethnic group are optimally distributed across schools. Nonetheless, the County Diversity Index can be useful as an upper bound of racial/ethnically heterogeneity.
One way to measure the degree of racial/ethnic school segregation of students within a county is to regress the Average School Diversity Index on the County Diversity Index, as shown by the green line of best fit in Figure 13. Counties with the greatest positive difference between their Actual Average School Diversity Index and their Predicted Average School Diversity Index (based on their County Diversity Index) all lie above this green line and are highlighted in blue in Figure 13. They are also listed in Table 9 (below). Essentially, these counties have schools that are extremely racially/ethnically integrated (i.e., students of different racial/ethnic groups are spread very evenly across their schools). Using this methodology, Anchorage Municipality in Alaska, Lee County in North Carolina, and Comanche County in Oklahoma have the most racially/ethnically integrated schools in the U.S.
Table 9: Counties with the Most Racially/Ethnically Integrated Schools in the U.S.,
Average School Diversity Index vs. County Diversity Index
|County||Number of Students||County
|Anchorage Municipality, AK||48,789||66||61||51||10|
|Lee County, NC||10,210||61||57||48||9|
|Comanche County, OK||21,265||61||57||48||9|
|Coryell County, TX||14,704||54||52||43||9|
|Stafford County, VA||28,386||52||50||41||9|
|Cleveland County, OK||45,556||50||49||40||9|
|Highlands County, FL||12,289||53||51||42||9|
|Harnett County, NC||21,015||53||51||43||8|
|Liberty County, GA||10,020||49||48||40||8|
|St. Lucie County, FL||41,631||58||54||46||8|
Counties with the greatest negative difference between their Actual Average School Diversity Index and their Predicted Average School Diversity Index, based on their County Diversity Index, lie below the green line and are highlighted in red in Figure 13. They are listed in Table 10 (below). Essentially, these counties have schools that are especially racially/ethnically segregated. Although the overall PK-12 student population may be diverse within these counties, students of the same racial/ethnic group are more often clustered within the same individual schools. Using this methodology, Wayne County in Michigan, Cook County in Illinois, and Milwaukee County in Wisconsin have the most racially/ethnically segregated schools in the U.S.
Table 10: Counties with the Most Racially/Ethnically Segregated Schools in the U.S.,
Average School Diversity Index vs. County Diversity Index
|County||Number of Students||County
|Wayne County, MI||271,656||48||21||39||-18|
|Cook County, IL||729,779||59||30||47||-17|
|Milwaukee County, WI||133,816||59||31||47||-16|
|Ouachita Parish, LA||28,255||39||18||33||-15|
|Cuyahoga County, OH||164,539||48||25||33||-14|
|Essex County, NJ||132,420||56||30||44||-14|
|Apache County, AZ||10,576||25||9||23||-14|
|Navajo County, AZ||17,654||49||26||39||-13|
|Forrest County, MS||11,278||41||21||34||-13|
|Jefferson County, AL||99,112||44||24||36||-12|
Table 11: Student Racial/Ethnic Diversity in PK-12 Public Schools,
Cook County in Illinois vs. King County in Washington
|Cook County, Illinois (Chicago)||King County, Washington (Seattle)|
|County Diversity Index||59||59|
|White students (%)||26.6||45.5|
|Hispanic students (%)||38.2||17.2|
|Black students (%)||26.1||8.5|
|Asian students (%)||6.0||17.9|
|American Indian or Alaska Native students (%)||0.7||0.5|
|Hawaiian Native or Other Pacific Islander students (%)||0.1||1.4|
|Two or more race students (%)||2.2||8.9|
|Average School Diversity Index||30||51|
|Majority racial/ethnic group: White||--||1|
|Majority racial/ethnic group: Hispanic||51||--|
|Majority racial/ethnic group: Black||194||--|
However, as is evident from Figure 14 the racial/ethnic diversity within schools for these counties is starkly different. Nineteen percent of Cook County’s schools (245 out of 1,267) are racially/ethnically homogeneous, meaning they are in the bottom 10% of U.S. schools in terms of diversity. By contrast, only one public school out of 473 in King County is racially/ethnically homogeneous. Likewise, Cook County has 82 fewer racially/ethnically heterogeneous schools (144 vs. 226) even though King County has 794 fewer schools overall. This difference in school racial/ethnic segregation results in Cook County having a Weighted Average School Diversity Index that is 21 points lower than King County (30 vs. 51).
Student academic learning and social-emotional development is shaped in schools not only by teachers and administrators, but by the effects of peers, and the degree of racial/ethnic peer diversity that a student encounters varies widely across U.S. public schools. For many students, attending schools with very diverse peer populations is the norm.10 These heterogenous schools are disproportionately magnet, located in urban or suburban areas, and are often situated on the East and West Coasts. Other students attend schools with peers who are very often racially/ethnically like themselves. These homogeneous schools tend to have smaller student populations and are frequently located in rural locales and in the middle regions of the country. The U.S. is a multicultural society, so it is not unexpected that the racial/ethnic composition of students varies considerably across schools. However, the magnitude of this variation—both within states (e.g., Alaska, Texas) and at a local level within individual counties—is remarkable and leads to disparate schooling experiences for many students.
We also examine the propensity of White, Black, Hispanic, Asian, Hawaiian Native or Other Pacific Islander, American Indian or Alaska Native, and students of two or more races to attend PK-12 public school with diverse populations. There are several notable differences in school diversity, including that American Indian or Alaska Native students disproportionately attend homogeneous schools. However, the divergent school attendance patterns between Asian and Hispanic students, the two fastest growing racial/ethnic groups in the U.S., is particularly noteworthy. Incredibly, less than one percent of Asian students attend racial/ethnically homogeneous schools and there are only five homogeneous schools that are comprised predominately of Asian students. By contrast, Hispanic students are 17 times more likely to attend racial/ethnically homogeneous schools and there are more than 1,900 homogeneous schools comprised predominately of Hispanic students (and more than 900 in Texas alone). Simply put, Asian students are far more likely than Hispanic students to be exposed to racial/ethnic diversity during through their public schooling experience.
Finally, this report highlights that students of the same race/ethnicity often disproportionately attend the same school, even when the surrounding county or metro area is significantly more diverse. Detroit, Michigan; Chicago, Illinois; Milwaukee, Wisconsin; and Monroe, Louisiana are just a few of the places with public schools that are substantially more segregated than the PK-12 public student population in these areas would suggest. While perfectly even distributions of students across schools by race/ethnicity is of course only aspirational, many districts should strive to better integrate their schools. Purposeful enrollment changes, even if only small in scale, can lead to substantial gains in the degree of student racial/ethnic diversity for the most homogeneous schools.
The Census Bureau projects the U.S. population will become majority-minority in 2044. However, in PK-12 public schools this future has already arrived, as 52% of students in 2016-17 were non-white. Yet even while the overall population of PK-12 children becomes more diverse, there is no guarantee that schools will become more racially/ethnically heterogeneous. Indeed, there is evidence schools became more racially/ethnically and economically isolated from 2000 to 2014.
This apparent contradiction—that greater overall diversity in the U.S. population does not necessary lead to greater diversity in schools—is illustrated by the fact that although the average school has a Diversity Index of 29, the most common U.S. school has a Diversity Index of only 7, making it extremely racially/ethnically homogeneous (see Figure 1). The bimodal distributional shape of U.S. PK-12 public schools is driven by many factors, but none more influential than the greater racial/ethnic homogeneity of schools in rural locales and segregated metro areas. As the demographics of the U.S. population inexorability shift, embracing and facilitating school heterogeneity through integration efforts is both pragmatic and prudent policy given the social, civic, and academic benefits that greater racial/ethnic diversity affords.
Adapted from Ed.Data.org’s Ethnic Diversity Index, we give every public school a racial/ethnic Diversity Index score, ranging from 0 (no diversity) to 100 (full diversity) based on the seven racial/ethnic categories collected by the U.S. Department of Education. Our school Diversity Index is calculated as follows:
|Appendix Table B1: Majority Racial/Ethnic Group Counts in Homogeneous Schools|
|Majority Racial/Ethnic Group|
|State||Total Schools||Homogenous Schools||White||Hispanic||Black||Asian||American Indian
or Alaska Native
|Appendix Table B2: Majority Racial/Ethnic Group Counts in Heterogenous Schools|
|Majority Racial/Ethnic Group|
|State||Total Schools||Heterogenous Schools||White||Hispanic||Black||Asian||American Indian
or Alaska Native
While there are roughly 100,000 schools in the 2016-17 NCES School Characteristics database, all analysis in this report uses a subsample of 92,308 PK-12 public schools. Schools that are not governed by the 50 states and the District of Columbia (i.e., American Samoa, Bureau of Indian Education, Guam, Puerto Rico, U.S. Virgin Islands), schools that are face or full virtual, schools that are listed as closed, and schools with less than 10 students were omitted from this analysis. This subsample of 92,308 schools educates 49.7 million students, which works out to an average size of 538 students per school. ↩
On the low end, there are more than 700 U.S. PK-12 public schools that have between 10 and 15 students. On the high end, there are five schools with at least 5,000 students: Somerset Academy Sky Pointe, River Springs Charter School, Brooklyn Technical High School, Visions in Education, and Carmel High School. ↩
Among states with at least 10 homogeneous schools that are majority American Indian or Alaska Native, there are 101 schools in Alaska, 53 in Arizona, 34 in Montana, 25 in New Mexico, 20 in South Dakota, 15 in North Dakota, and 10 in Minnesota. ↩
There are 92 schools that identify as both magnet and charter schools. For simplicity, these schools have been omitted from the analysis of diversity by school type. ↩
Schools in metropolitan areas with less than 25,000 PK-12 public school students or students in micropolitan areas are not included in this analysis. ↩
This claim assumes Las Vegas, Nevada is located on the West Coast. ↩
This claim assumes Pittsburgh, Pennsylvania is located in the Midwest. ↩
County equivalents vary from state to state and include the designations borough, city, municipality, census area, and parish. ↩
Of course just because a school’s student population is diverse does not necessary mean classes are racially/ethnically heterogeneous. Undoubtedly there exists significant racial/ethnic sorting across classrooms and academic subjects—as well as by gender, socioeconomic background, and English language skills. Unfortunately, there is no student enrollment data at the classroom level. ↩