Sage Journals: Discover world-class research

Abstract

The establishment of new school districts in predominantly White municipalities in the South is restructuring school and housing segregation in impacted countywide school systems. This article compares the contribution of school district boundaries to school and residential segregation in the Southern counties that experienced secession since 2000. Merging together several data sets, including Common Core of Data, census data, and shapefiles at multiple geographic scales, we measure segregation of public school students and the entire population over time. We show that school district secession is restructuring school segregation in the counties where secession is occurring, with segregation increasingly occurring because students attend different school districts. Additionally, in the most recent year of analysis, residents were increasingly stratified by race in different school districts. Segregation patterns differ substantially, however, depending on the history of secession in the county.

Keywords

race school districts secession residential segregation school segregation South

School district secession—a political process forming small, new school systems that tend to serve a higher-proportion White and more affluent residential population than the large districts from which they splinter—has received growing attention in media, legal, and policy circles (Brown, 2016; Eaton, 2014; EdBuild, 2019a; Hannah-Jones, 2017; Spencer, 2014; Wilson, 2016). Although district secessions occur around the country, they are concentrated in the South.¹ Recent Southern secessions reflect a narrowing conception of what is “public” about public education as newly created districts seek to preserve relative racial and economic advantages for more homogeneous White areas (Siegel-Hawley, Diem, & Frankenberg, 2018). Often couched in the race-neutral rationale of “local control”—language long used by White families to resist school desegregation—Southern secessions raise particular concerns about educational inequity, occurring as they do against the region’s historical backdrop of de jure segregation and subsequent resistance to the Brown v. Board of Education decision. Research has documented the resegregation of students in the South (Boger & Orfield, 2005; Reardon, Grewal, Kalogrides, & Greenberg, 2012), a trend that follows a post-Brown period when Southern schools became the most integrated (Frankenberg, Lee, & Orfield, 2003). In a region undergoing rapid demographic change, is contemporary school district secession poised to become yet another tactic in the South’s “sordid history” of White efforts to thwart equal educational opportunity (Wilson, 2016)? The following study is the first to systematically explore whether and to what extent new school district boundaries increasingly segregate students and residents in the South.

Fragmented school districts have been a phenomenon primarily associated with the Northeast and Midwest (Bischoff, 2008), but district secession in Southern countywide systems is upending, to some degree, the one-county, one-school-system archetype in parts of the South. The end of court-ordered school desegregation has meant that schools in the metropolitan South, once more integrated than neighborhoods, have lost that distinction (Reardon et al., 2012; Reardon & Yun, 2005). Thus, the erosion of the South’s countywide advantage also comes amid the erosion of the region’s schooling advantage—that is, a pattern in which schools are more integrated than neighborhoods.

Despite long-standing White resistance to desegregation, judicial pressure meant that the South’s countywide school systems (or ones that include both cities and suburbs) historically have been some of the most integrated for Black and White students (Siegel-Hawley, 2016). Countywide systems with comprehensive school desegregation plans also report lower residential segregation because housing choices become—to some degree—decoupled from school choices once all households in a county are part of the same school system (Pearce, 1980). Families come to understand they can move across the metropolitan area (also referred to herein as “metro”) community and still be connected to a diverse school. By contrast, school and housing segregation tend to be higher in metropolitan areas outside the South where districts are highly fragmented, with single metros containing multiple suburban school systems alongside one or more central-city school districts (Bischoff, 2008; Frankenberg, 2009). It stands to reason, then, that the secession of White, suburban municipalities from larger, more diverse countywide systems is therefore likely to impact not only school segregation but also residential segregation.

We found that school district secession over the period we studied occurred in counties that were more racially diverse than the region or country. In the seven Southern counties that experienced a secession, between-district segregation for Black and White students has increased since 2000, as it has for White and Hispanic and White and Asian students. Additionally, although overall residential segregation declined during the period studied, school district boundaries accounted for an increasing share of residential segregation from 2010 to 2014. Thus, while absolute levels of residential segregation declined, existing residential segregation in impacted counties was increasingly due to residents living in different school districts. Our analysis contributes to our understanding of the erosion of mechanisms (like countywide school districts) that historically have permitted integration. More broadly, we illuminate the evolving relationship between school and residential segregation, at least in the short term, as new school district boundaries are erected and begin to acquire racialized meaning for both parents and home seekers (Weiher, 1991).

Review of the Literature

We explore several multidisciplinary strands of literature as context for the contemporary phenomenon of Southern school district secession. History and law illuminate White Southern resistance to school integration, of which district secession is a past and present part. We then delve into the issue of school district boundary lines, seeking to understand how they contribute to segregation, especially as court-ordered desegregation is waning, and what is known about contemporary secession efforts in the South creating new district boundary lines.

Southern Resistance to School Integration

The term secession has particular resonance in the American South after 11 states (including the three studied here) seceded from the Union, rather than abandoning the enslavement of African Americans. Despite the defeat of the Confederacy in the Civil War, White Southern political leaders passed laws sharply restricting access to and resources for education for Black children (Anderson, 1988). African American teachers across the South fought against those laws, relying on their professional training to prepare African American students in the face of deeply discriminatory circumstances (Walker, 2001). Still, political leaders across the region—virtually all White at that time—did not come close to providing equal resources of the Plessy era’s “separate but equal” mandate (Ashmore, 1954; Carter, 1953). These ongoing inequalities became the basis for the 1954 U.S. Supreme Court ruling in Brown that invalidated racially segregated schools in the South and border regions. Research since the time of Brown continues to show improved social and academic outcomes for students in desegregated schools, including long-term benefits once they leave schools that are important for communities, as well (Braddock & Gonzales, 2010; R. Johnson, 2019; Mickelson & Nkomo, 2012; Trent, 1997).

Despite Thurgood Marshall’s prediction that Southern schools would be desegregated within 5 years of the Brown decision, just 2% of Black students in the region attended majority-White schools 10 years later (G. Orfield, 1969). State and local officials implemented a range of mechanisms to delay and then limit the extent of desegregation, including state pupil placement laws and the founding of private school academies; White suburban resistance to urban annexation also prevented government efforts that may have had a by-product of creating integration (Connerly, 2005; Patterson, 2001; Pratt, 1992). Importantly, early post-Brown efforts to avoid desegregation included White attempts to secede from countywide districts in the South (Wilson, 2016). A judicial stance prioritizing equal protection under the law rather than local control prevailed, however. In 1972, the Supreme Court issued a decision in Wright v. Council of City of Emporia that prevented a district secession in Virginia. The secession would have exacerbated school segregation by creating a disproportionately White city district exempt from desegregation efforts in the majority-Black county district and was announced 2 weeks after the county district’s desegregation order was issued. The Emporia decision still stands (and was recently cited in a decision forbidding a secession attempt in Alabama), but broader legal and political forces have made it much easier for districts to secede today than in the past (Frankenberg, Siegel-Hawley, & Diem, 2017; Hannah-Jones, 2017; Siegel-Hawley et al., 2018). These structures and decisions use rational discourse citing race-neutral reasons or structures to further racial inequality without invoking intentional racist beliefs (Bobo, Kluegel, & Smith, 1997). This occurs in a larger context of a judiciary increasingly hostile to race-conscious remedies in a society predicated on racial discrimination as well as many more conservative state legislative bodies willing to remove existing backstops for secession or unwilling to pass new laws that would contain or oversee it (Siegel-Hawley et al., 2018).

The Contribution of School District Boundaries to Segregation

The formation of new district boundary lines through secession matters because those lines help structure school and residential segregation (Bischoff, 2008; Owens, 2017). As Erickson (2015) wrote, “The causal roots of educational inequality . . . can be found in the interactions between schools and the basic political and economic structures of the city and the metropolis” (p. 4). Whether segregation or desegregation would occur, she added, was the result of many decisions made by political leaders at all levels. In U.S. metropolitan areas, approximately two-thirds of school segregation is due to segregation between districts, and one third is due to segregation within districts (Stroub & Richards, 2013). And for residents, school district boundaries account for under half of slowly declining residential segregation by race in metropolitan areas and for just over half of the residential segregation of school-age children (Frey, 2014; Owens, 2017). School district boundaries are therefore much more segregative for students in schools, on average, than for residents of all ages as well as for school-age children residing in neighborhoods.

Despite divergent impacts on school and residential segregation, school districts remain critical organizational units for households with and without children. More than 40 years ago, desegregation scholar Gary Orfield (1978) noted, “United States society [has] . . . an increasing tendency for political boundaries to become racial boundaries” (p. 381). School district boundary lines are one such critical political boundary. Home seekers make residential choices with school considerations in mind (Kane, Staiger, & Riegg, 2005; Lareau & Goyette, 2014). In an analysis of non-Southern metropolitan areas, suburban house prices increased in metros where central city districts came under a desegregation order (Boustan, 2012; see also O. Johnson, 2017). This means households without children are also considering the housing–school link, because increased property values are linked to neighborhoods offering access to highly-sought-after schools (Kane et al., 2005). We might expect housing–school trends to diverge, however, in places with higher concentrations of charter schools because of the nongeographical nature of such schools (e.g., Pearman & Swain, 2017).

In areas with fewer political boundaries encompassing city and suburban areas, however, when coupled with the implementation of a desegregation plan, prior research shows sharper housing segregation declines than in areas with boundaries dividing suburbia from central cities (Frankenberg, 2005; Pearce, 1980; G. Orfield, 2001; M. Orfield, 2015; Siegel-Hawley, 2016). Fewer political boundaries interrupt the school–housing calculus for advantaged families, opening the real estate market across a wide swath of the metropolitan community. It follows that housing segregation might decline less slowly—or not at all—in counties marked by the fragmented housing market associated with school district secessions.

Most prior research has examined overall residential racial segregation, not ascertaining whether there is a changing contribution to segregation within or between districts as we do here. Based on past studies, though, we expect to see increasing between-district segregation in counties experiencing secession, although such effects might be modest at first due to factors such as the time it takes for districts to develop reputations or for families to make residential moves.

Southern Demographic Patterns and Recent Southern Secessions

A study of 350 U.S. metros over a 30-year period found that school segregation was highest in areas where families could racially sort themselves across numerous district boundaries (Fiel, 2015). Competition for access to certain schools and school districts intensified when resources were distributed in markedly unequal ways across organizational boundaries and through school choice in the form of private or charter schools. Drawing on theories surrounding racial threat, the study also found that school segregation was higher when two major racial groups approached numerical parity; this meant the dominant group was likely to employ exclusionary tactics, like expanding school choice or policing school district boundaries, to maintain its relative advantage in regard to opportunities. More hopefully, growing multiracial diversity may alleviate feelings of intense competition activated by the presence of two focal groups. Fiel (2015) offers growth of Latino and Asian enrollment as one explanation for gradual overall declines in segregation between most student groups. These findings seem particularly relevant for the South, a region historically characterized by a sizeable Black enrollment and highly unequal resource allocation between Black and White students because of de jure segregation. Demographic change has swiftly altered the region, however. In 2016, White students made up a minority share of the enrollment (41%), and Hispanic students (28%) surpassed Black students (24%) as the largest racial minority (Frankenberg, Ee, Ayscue, & Orfield, 2019).

Accompanying broader research related to segregation is a small set of prior studies specifically dealing with secession. In 2017, for instance, EdBuild released a national analysis of secession attempts with an update in 2019 (EdBuild, 2019a). The group found in 2017 that 47 U.S. school districts had successfully seceded since 2000; 10 more had seceded by 2019, with more attempted and some still in process (EdBuild, 2019a). Secessions have deleterious fiscal and racial effects for school districts left behind. Further, sharp variations in state law and policy exist governing the ease of district secession (see also Wilson, 2016).

A series of single-case studies (Frankenberg et al., 2017; Siegel-Hawley et al., 2018) examined suburban secession in several Southern metros. In the postsecession Memphis, Tennessee, area, in the first year after secession occurred, more school segregation occurred because students of different races enrolled in separate school districts rather than separate schools within the same district (Frankenberg et al., 2017). Additional research focusing solely on Jefferson County, Alabama, found that White residents have increasingly populated seceding suburban school districts. Some municipal districts that seceded decades ago, however, have experienced racial transition and had few White students (Frankenberg, 2009; Frankenberg & Taylor, 2017). The secessions had residential implications—a sorting of households particularly in the aftermath of secession by district boundary lines—as well as educational implications. Moreover, had the secessions not occurred, the Jefferson County School District would today enroll a majority-White student population and an additional 20,000 more students, the latter affecting funding support. A comparative look at these two metros alongside Jefferson County, Kentucky—still a merged city–suburban countywide district—found that Jefferson County, Kentucky, reported minimal school segregation, whereas between 1960 and 2012, both Jefferson County, Alabama, and Memphis-Shelby County, Tennessee, experienced high and increasing school segregation unevenly concentrated in each area’s urban school systems (Diem, Frankenberg, Siegel-Hawley, & Cleary, 2015).

Although these case studies illuminate important aspects of school district secession, we know little about the systemic, longitudinal impacts of secession on school and housing segregation in the South. The present study will explore the role of school district boundaries over time in all Southern counties experiencing secession. Within each county, we examine school-level segregation using schools as the unit of analysis and residential segregation using block groups as the unit of analysis. This approach allows us to understand whether and, if so, how school district fragmentation is related to increased sorting of students and residents into different school districts. We examine these interrelated issues from the year 2000 to 2015, assessing segregation during the period most associated with secession. We would expect that as school district secession occurs, more school and residential segregation in the counties studied would be related to school district boundaries. And, as expected, we find that the proportion of school segregation due to school district boundaries increased from 2000 to 2015. For example, whereas school district boundaries accounted for 60% of Black and White school segregation in 2000 in our sample of counties, by 2015 that proportion had increased to 70%. At the same time, school district boundaries may have had a delayed effect in impacted counties on residential segregation, whereby from 2010 to 2014 there was an increase in the proportion of residential segregation due to school district boundaries. Such trends will be important to monitor as district boundary lines in these counties continue to accrue meaning for home seekers, particularly households with school-age children.

Method

In this study, we use a research design that allows us to determine the unique contribution of school district boundaries to racial school and residential segregation in Southern counties experiencing secession. Decomposing Theil’s H (Theil, 1972), an evenness measure of segregation, means that we can identify the proportion of school and residential segregation due to school district boundaries at different time points as school districts secede from county districts.² We compare resident trends to public school student trends to help us understand to what extent district secessions are related to both residential and school segregation. We also explore how school district boundaries segregate populations and students depending on race (i.e., if school district boundaries are more segregative for Black and White residents than for Hispanic and White residents).

Sample

From 2000 to 2017, 47 school districts have successfully seceded from a larger school district in the United States (EdBuild, 2019a). These secessions have affected 13 counties across the United States, seven of which are in the South. In this study, we focus on the school district secessions that have occurred in the South, including the 18 new districts identified by EdBuild that have formed during this time period.³ We focus on the South because of the unique role countywide school systems have played in the region in terms of school desegregation and because secession can be seen as a continuation of Southern strategies used to avoid integration.⁴ Although many of these districts were once under court-ordered desegregation plans, today most have been declared unitary.⁵ We examine racial segregation, both school level and residential level, in the seven Southern counties where school district secessions have occurred since 2000: Jefferson, Marshall, Mobile, Montgomery, and Shelby Counties in Alabama; East Baton Rouge Parish, Louisiana; and Shelby County, Tennessee. In three counties, secessions did not occur until after 2010. Thus, for the first decade of our analysis, we expect relatively lower changes in between-district segregation.

To conduct our analysis, we measure the residential segregation that occurs within school districts, as compared to the residential segregation that occurs between districts, in the seven counties. Similarly, we measure the segregation of students within school districts, and between school districts, to determine changes over time as school districts secede. We select school districts that are geographically within each county that experienced a secession (see Table 1 for frequency of school districts at each time point).⁶

Table 1

Demographic Characteristics of Seven Counties Where School Districts and LEAs Have Seceded Since 2000

Variable	2000	2010	2014-2015
Number of school districts/LEAs	20	29	54
Number of schools	669	709	743
Total population	2,783,073	2,912,265	2,913,761
White	56.75%	50.60%	49.22%
Black	39.86%	42.76%	43.80%
Hispanic	2.01%	4.54%	4.78%
Asian	1.38%	2.10%	2.21%
Student enrollment	446,978	450,976	454,872
White	41.6%	34.4%	32.2%
Black	56.2%	58.2%	55.3%
Hispanic	1.0%	4.8%	8.2%
Asian	1.3%	2.1%	2.3%

Source. Common Core of Data 1999-2000, 2009-10, & 2014-15, Decennial Census 2000 and 2010, and American Community Survey 2012-2016.

Note. Between 2000 and 2010, these schools changed racial-ethnic categories and in 2015 reported 1.6% multiracial students and 0.1% Native Hawaiian and Pacific Islander students who were previously counted in other categories. There were 0.1% American Indian and Alaska Native students during these years. LEA = local education agency.

In the seven Southern counties impacted by secession, the aggregate residential population became majority non-White between 2000 and 2014. In the same counties, the public student enrollment was already majority non-White in 2000 and became increasingly so over time. White students made up only 32% of enrolled students in 2015, down from 42% in 2000. These counties had a disproportionately more diverse enrollment than the region: The entire South region had 54% White students in 2000 and 43% in 2014. A sharp increase in the share of Hispanic students occurred in several of the counties. For the seven counties experiencing secession, the enrolled public school student and resident population were disproportionately non-White in comparison to the entire region.

By 2015, non-geographically-based schooling options arose in the two non-Alabama counties. In Tennessee, the state-run Achievement School District (ASD) operated 27 schools comprising 6.3% of the enrollment (Table A1 in the appendix). In 2010 there were two charter local education agencies (LEAs) in East Baton Rouge Parish, which had grown to 16 by 2015. We treat each of these charter LEAs and ASD as separate districts. Finally, in 2015, the two central-city districts, Shelby County and East Baton Rouge, contained a handful of charter schools along with traditional schools, which we consider to be part of the district like other noncharter schools. The first charter school in Alabama opened after 2015 and was not included in this study.

Data

The data in this study are drawn from several sources, including census demographic data at the block group level, Common Core of Data (CCD) demographic data at the school level, and TIGER/Line geographic data at the block group, school district, and county levels. Bringing together these data sources allows us to examine population and student patterns simultaneously and to study populations in relevant educational geographies.

Census

We measure racial residential segregation using the four largest racial groups in the United States: White, Black, Hispanic, and Asian residents. To conduct the residential portion of this analysis, measuring the extent to which residents are sorted unevenly between school districts and unevenly among neighborhoods within school districts, it is necessary to utilize census units within school districts. We use block groups for this purpose because they are smaller than tracts and yield a more precise estimate of segregation within educational geographies. We use areal interpolation with the TIGER/Line block group and school district shapefiles to estimate block group populations within school districts (Logan, Xu, & Stults, 2014; Saporito, Chavers, Nixon, & McQuiddy, 2007) because block groups (and other census geographies) do not nest perfectly within school districts. Areal interpolation is a useful technique for education researchers that utilize census data because it makes it possible to estimate populations within school boundaries and, in this study, allows for the measurement of residential segregation within school district boundaries. We created block group estimates from 2000, 2010, and 2014 for each school district. Often, racial residential segregation is measured using tracts, which are larger administrative units than block groups. When measuring segregation with smaller administrative units, like block groups, the result will usually be higher segregation because of the nature of the unit of measurement. It is important to keep this in mind when interpreting the findings on racial residential segregation. In our residential analysis, the overall county segregation and within-district component will be relatively higher when using block groups as compared to using tracts. Therefore, considering changes over time in segregation at the residential level will be most useful, rather than considering absolute levels of segregation. Additionally, comparisons between school and residential segregation must also be made in the context of the differences in composition between the student population and the residential population. Consistently, White residents are a higher percentage of the county population than are White students of the county’s public school student enrollment.

CCD

To study school-level segregation, we measure segregation in schools among the four largest racial groups during the 1999–2000, 2009–2010, and 2014–2015 school years. We include the same school districts in the school-level analysis as are in the residential analysis to allow for comparison of patterns at the school level and residential level. Only regular public schools (e.g., not including vocational, special education, or alternative schools) are included as we explore the extent to which students are segregated between and within school districts in the counties where school districts have seceded since 2000. We also include charter schools in Tennessee and Louisiana if they are not vocational, special education, or alternative schools.

Analysis

Using Theil’s H (Theil, 1972), we report on the magnitude of segregation at different geographic scales from 2000 to 2014–2015. These scales include segregation in the entire county, segregation between school districts, and segregation within school districts. For each, segregation is measured for both the residential population and public school enrollment. Theil’s H is an evenness measure of segregation that reports on how much less diverse a subunit is as compared to the area of study (for example, how much less diverse, on average, schools are as compared to the entire school district). Ranging in value from 0 to 1, a low H value indicates students/residents are evenly sorted among subunits, whereas a high H value indicates that students/residents are unevenly sorted among subunits. We examine multigroup segregation, which allows us to evaluate the segregation of all groups simultaneously, as well as racial dyads. H has the advantage of being decomposable across geographic units (Reardon & Firebaugh, 2002), which makes it possible to determine the unique contribution of a subunit, in this case, school districts, to the overall segregation of counties. For example, an H value of .25 at the school district level means that the school district is 25% less diverse than the county. At the population level, we decompose Theil’s H to determine the contribution of school districts to county residential segregation. Similarly, we decompose H at the school level to understand the contribution of school districts to county school segregation (relative to within-district segregation). We are particularly interested in how county H values and their components (e.g., school districts) have changed over time as school districts have seceded.

Other segregation measures, like exposure, indicate the interracial contact that students or residents might have in a unit of analysis. One weakness of using an evenness measure of segregation as we do here is that it can fail to account for the exposure dimension of segregation (G. Orfield, Siegel-Hawley & Kucsera, 2014). If a school district is 85% one race, and if all the schools in that district are close to 85% that race, then the school district can be described as experiencing low segregation. However, in this scenario, there is little exposure to students of other races. When measuring school segregation using an evenness measure, school segregation is often declining, but when measured using an exposure measure, school segregation is often increasing (G. Orfield et al., 2014). Segregation often declines when using an evenness measure because the diversity of schools is, on average, increasing. However, at the same time, White, Black, and Hispanic students tend to be increasingly isolated in schools. The changing racial composition of the counties under study (see Table 1) demonstrates that even if students are distributed more evenly among schools, students themselves may increasingly be racially isolated. This is related to the increasingly non-White composition of the student population. Although not described below, we find some evidence of such trends in exposure in some of the counties studied here.⁷

Findings

In the seven Southern counties experiencing school district secession between 2000 and 2015, the proportion of school segregation due to school district boundaries has increased. That has been particularly true since 2010, the time period in which all seven counties had at least one district that seceded. In other words, as school districts seceded, school district boundaries played a larger role in school segregation measured at the county level. Although the recent wave of secession had not yet started in 2000, by 2015 the number of school districts in the seven counties had nearly doubled (and more than doubled when including new charter districts). Additionally, for most racial-ethnic groups, the between-district share of residential segregation in 2014 was higher than in 2010 (Table 2). Finally, patterns of segregation varied among the seven counties according to the history and frequency of secession in the county. We explore these findings in more detail next.

Table 2

School and Residential Segregation Due to Segregation Between School Districts in Counties Where School Districts Seceded, 2000 to 2014–2015 (in percentages)

Variable	2000	2010	2014–2015
School segregation
Multiracial	57.7	58.2	63.8
Black/White	59.9	62.8	70.3
Hispanic/White	37.1	53.3	65.1
Asian/White	31.8	52.8	63.0
Residential segregation
Multiracial	17.8	15.2	16.1
Black/White	18.6	17.8	19.1
Hispanic/White	16.5	9.1	11.1
Asian/White	5.8	6.8	6.0

Source. Common Core of Data SY 1999-2000, SY 2009-2010, SY 2014-15, Decennial Census 2000 and 2010, American Community Survey 2012-16 estimates.

School Segregation and School District Boundaries

Across the seven counties, the contribution of school district boundary lines to overall school segregation was higher in 2015 than in 2000. In 2000, school district boundaries accounted for, on average, 57.7% of multiracial school segregation, a figure that grew to 63.8% by 2015. School district boundaries contributed the most to the segregation of Black and White students, the largest two racial groups in these counties. In 2000, school districts contributed, on average, to 59.9% of the school segregation for Black and White students. The contribution of school district boundaries to overall county school segregation grew to 70.3% by 2015. As the Hispanic student enrollment increased (see Table 1), Hispanic and White student segregation due to school district boundaries also increased. In 2000, on average, 37.1% of school segregation between Hispanic and White students was because of school district boundaries; by 2015, that proportion had increased sharply, to 65.1%. Both the relative and absolute levels of within-district segregation for Hispanic and White students declined, particularly after 2010 (see Table 3).

Table 3

Multiracial School Segregation Between and Within School Districts in the Counties Where School Districts Have Seceded, 2000 to 2015

School segregation	2000	2010	2015
Multiracial	.492	.431	.398
Between school districts	.285	.251	.254
Within school districts	.207	.180	.144
Black/White
Total segregation	.549	.548	.535
Between school districts	.329	.345	.377
Within school districts	.220	.204	.159
Hispanic/White
Total segregation	.248	.331	.366
Between school districts	.092	.176	.239
Within school districts	.156	.154	.128
Asian/White
Total segregation	.211	.191	.200
Between school districts	.067	.101	.126
Within school districts	.144	.090	.074

Source. Common Core of Data SY 1999-2000, SY 2009-2010, SY 2014-15.

Note. Includes charter local education agencies in 2010 and 2015 and Achievement School District in 2015.

Black and White students were increasingly sorted between school district boundaries within counties in 2015 (see Table 3). School district secession in the counties under study resulted in splinter districts that reported higher percentages of White students enrolled in them than was the case in most of the “left-behind” county districts. Most county districts, in turn, had a higher percentage of Black and Hispanic students. This meant that within each school district, there was less racial diversity, and therefore racial sorting between schools within one district became relatively less important to overall county segregation, whereas racial sorting between school districts became relatively more important. Put differently, the results bear out our expectation that these counties would experience declines in within-district segregation while at the same time experiencing increases in between-district segregation. In 2000, school districts were on average 32.9% less diverse than the county they were in, but by 2015, this figure had increased to 37.7% for Black and White students. Much larger increases happened for Hispanic and White students as well as for Asian and White students. For example, in 2000, school districts were, on average, 9.2% less diverse for White and Hispanic students, and by 2015, this figure had increased to 23.9%.

In absolute terms, however, from 2000 to 2015, multiracial school segregation overall and within school districts declined in the Southern counties that experienced school district secessions. There was a slight increase in multiracial segregation between 2010 and 2015, the period of time in which all seven counties had experienced secession. Despite declines in absolute segregation values over this period, school segregation remained persistently high, and school district boundaries accounted for an increasing share of the existing segregation. School districts themselves were, on average, 25.4% less diverse than the counties in 2015, indicating that school district boundary lines were highly segregative for students. Declines in the absolute level of multiracial school segregation are consistent with previous research using evenness measures of segregation (Reardon & Owens, 2014; Stroub & Richards, 2013). Although the absolute magnitude of multiracial school segregation declined,⁸ multigroup school segregation overall in the counties, and school segregation between school districts, remained quite high. In 2015, schools in the counties were, on average, 39.8% less diverse than the county student population.

Residential Segregation and School District Boundaries

In contrast to segregation of public school students, school district boundaries contributed relatively less to residential segregation of the entire population. District boundaries accounted for, on average, 16.1% of the multiracial residential segregation in the counties in 2014 (see Table 2). This was up slightly from 15.2% in 2010, indicating that as school district secessions picked up speed, so too did the percentage of residential segregation due to school district boundaries. In contrast, between 2000 and 2010, the relative contribution of school district boundaries to racial residential segregation decreased. In three of the seven counties studied, no secessions took place prior to 2010, and therefore it is not surprising that we would not find significant increases in residential segregation between 2000 and 2010.

Black and White residents were consistently more segregated by school district boundaries than other racial groups, despite a large increase in residential segregation for Hispanic and Asian residents. From 2010 to 2014, the percentage of residential segregation due to district boundaries increased for all groups except Asian/White, reversing declines from the preceding decade (a finding consistent with overall declines in residential segregation). That school district boundaries contributed relatively little across all years to the residential segregation in the counties can be attributed to several factors, including the use of block groups to measure within-district segregation (which, as noted, will drive up the contribution of the within-district portion of county segregation), the difference in diversity of the resident and student population, and the likely longer-term effect of residential responses to new school district boundaries. Additionally, previous research has shown that in the South, private school enrollment rates have increased for children from high-income families in recent decades and that White students enroll in private school at higher rates than Black and Hispanic students (Murnane & Reardon, 2018). Moreover, although the school district has some meaning for all home seekers because of the correlation with home values, it is likely more influential for home seekers also considering locations for their child’s school enrollment. However, that we see an increase in the role of school district boundaries in residential segregation from 2010 to 2014, a period when more secessions had occurred, suggests that there may be a delayed impact of school district secessions on residential trends and/or that it may be a result of all seven counties having experienced secession during this time period.

Across all seven counties, we find that residents are segregated more because they live in different neighborhoods within the same school district rather than because they live in different school districts. In absolute terms, residential segregation declined, driven largely by declines in within-district segregation during the last 4 years examined (Table 4). Or in other words, declines in residential segregation in the counties under study were mostly due to residents living in less segregated neighborhoods rather than due to residents being less segregated by school district boundaries. For example, in 2000, Black and White residents lived, on average, in a neighborhood that was 34.6% less diverse than their school district, but by 2014, this figure had declined to 30.9%. Residential segregation between school districts was not nearly as high as residential segregation within districts. In contrast to students, residential patterns did not display the same fragmenting results in the short term across all seven districts (see next paragraph, though, for variation in these trends among the seven counties). However, the residential population had a disproportionately higher percentage of Whites than the student population in these counties, and therefore we would not expect the same level of segregation at the residential level as we would at the school level. Moreover, it may require a longer period of time than examined here to fully see the residential effects.

Table 4

Multiracial Residential Segregation Between and Within School Districts in the Counties Where School Districts Have Seceded, 2000 to 2014

Residential segregation	2000	2010	2014
Multiracial	0.360	0.362	0.342
Between school districts	0.064	0.055	0.055
Within school districts	0.297	0.307	0.287
Black/White
Total segregation	0.425	0.405	0.382
Between school districts	0.079	0.072	0.073
Within school districts	0.346	0.332	0.309
Hispanic/White
Total segregation	0.158	0.339	0.326
Between school districts	0.026	0.031	0.036
Within school districts	0.132	0.308	0.289
Asian/White
Total segregation	0.139	0.280	0.251
Between school districts	0.008	0.019	0.015
Within school districts	0.130	0.261	0.236

Source. Decennial Census 2000 and 2010, American Community Survey 2012-16 estimates.

Different racial groups experienced varying levels of residential segregation in the Southern counties where school districts seceded. Consistent with previous research (Logan & Stults, 2011), Black and White residents experienced the highest magnitude of racial residential segregation at all levels, with school district boundaries contributing more to their segregation than for other dyads (e.g., .073 is 19.1% of Black/White residential segregation in 2014; see Table 4). The largest disparities in Black/White segregation and other racial pairs occurred in 2000, with declining, although persistently high, Black/White segregation accompanied by increasing Hispanic/White and Asian/White segregation through 2014. School district boundaries contributed a comparatively low percentage for Hispanic and White residents (11.1%) and Asian and White residents (6.0%). In sum, school district boundaries in absolute and relative terms mattered the most for Black and White residents in counties where school districts seceded, and they had the smallest impact for Asian and White residents.

Variation in Segregation by County’s History of Secession

When examining patterns for each of the seven counties separately, distinct patterns emerge for multigroup segregation as well as for Black/White segregation, which was typically higher. First, in three Alabama counties, Mobile, Montgomery, and Shelby, most secessions have taken place recently (only Saraland City Schools prior to 2010), and the county district still enrolls most students (see Table 5); recall also that Alabama did not permit any charter schools during this period. These three counties report low between-district segregation, likely because most students and residents still are in the one large district from which the small secessions are occurring (see, e.g., Figure 1, Mobile County, Alabama). However, even in these contexts, we saw substantial increases in between-district school segregation. For example, in Mobile County, between-district segregation increased from 2.2% of segregation in 2010 to 8.7% in 2015. Residential segregation in these Alabama counties was lower, as we would expect, given that secessions occurred recently and residents without school-age children in the public schools may be less likely to rapidly respond to such a change. As the seceded districts increase in size and more time passes since secession, it stands to reason that patterns may change.

Table 5

District Characteristics by County, 2015

District name	Year formed (if post-2000)	No. of schools	% County’s Enrollment	% County’s residential population
Jefferson County, Alabama
Bessemer		7	3.6	4.1
Birmingham		41	23.0	32.2
Fairfield		4	1.7	1.7
Homewood		5	3.9	3.9
Hoover		16	13.5	12.6
Jefferson County		55	34.9	34.9
Leeds	2003	3	1.8	1.8
Midfield		3	1.1	0.8
Mountain Brook		6	4.2	3.1
Tarrant		3	1.1	1.0
Trussville	2005	4	4.3	3.2
Vestavia Hills		8	6.8	5.2
Marshall County, Alabama
Albertville		6	29.5	22.7
Arab		4	14.4	8.7
Boaz	2004	5	13.1	10.2
Guntersville		4	11.3	8.8
Marshall County		14	31.8	50.7
Mobile County, Alabama
Chickasaw	2011	2	1.7	1.4
Mobile County		82	91.4	93.4
Saraland	2007	3	4.8	3.3
Satsuma	2012	2	2.1	1.5
Montgomery County, Alabama
Montgomery County		51	96.4	95.6
Pike Road	2015	1	3.6	3.4
Shelby County, Alabama
Alabaster	2012	6	20.9	15.7
Pelham	2014	4	10.4	11.0
Shelby County		28	68.7	60.6
East Baton Rouge, Louisiana
Central Community	2007	5	7.8	6.6
Charter schools (sum)^a		16	11.7
City of Baker	2003	5	2.4	3.1
East Baton Rouge Parish^b		81	69.1	85.2
Zachary Community	2003	7	8.9	5.2
Shelby County, Tennessee^c
Achievement School District (state-run)	2012	27	6.3	n/a
Arlington	2014	4	3.3	1.4
Bartlett	2014	11	5.7	6.2
Collierville	2014	8	5.3	5.1
Germantown	2014	5	3.8	4.2
Lakeland	2014	1	0.6	1.1
Millington	2014	4	1.8	1.8
Shelby County^b		202	73.1	80.1

Source: Common Core of Data 2014-15, American Community Survey 2012-16.

In 14 of the 16 charter schools in East Baton Rouge Parish in 2015, White students were 5% or less of their enrollment. The largest charter school in the parish, however, was 70% White. In the entire county in 2015, 20% of students were White. The two charter local education agencies that were operating in 2010 were among the 14 with almost no White students in either 2010 or 2015.

Indicates district also included district-run charter schools.

Shelby County, Tennessee, absorbed Memphis Schools in 2013.

Figure 1.

School district boundaries and the percentage of White residents in Mobile County, Alabama, 2014.

In the second category, which includes Jefferson County and Marshall County in Alabama, secessions largely took place before 2000, although some additional secessions occurred in each county between 2000 and 2005. As with the first group of counties described earlier, there were no charter schools. For these two counties, however, no one system enrolls the majority of students, although the Marshall County school district has a slim majority of residents in the county (Table 5). These two counties have high between-district school and residential segregation, although it has declined somewhat over time (see Figures 2 and 3). Still, by 2015, nearly two thirds or more of the school segregation was due to between-district segregation (and between-district segregation is an even higher share of Black/White segregation in these counties; see Table A2 in the appendix). Jefferson County, Alabama, consistently had the highest segregation, in schools and neighborhoods, among all seven counties. Jefferson County’s residential segregation declined more slowly than school segregation did during this time—and by 2015, neighborhoods were more segregated than schools were, the inverse of 2000. By contrast, Marshall County had declining school segregation and rising residential segregation. Both counties are indicators that long-term residential patterns may be important to monitor in counties where secession is occurring.

Figure 2.

School segregation between and within school districts in counties experiencing secession, 2015.

Figure 3.

Residential segregation between and within school districts in counties experiencing secession, 2014.

East Baton Rouge and Shelby Counties (Tennessee) are a third category of counties where the presence of nonresidential LEAs (e.g., ASD in Tennessee, charter schools in Louisiana) decouple the link between the school enrollments and residential patterns. Thus, East Baton Rouge Parish enrolls a much smaller percentage of students than the share of the population that resides in the district. In East Baton Rouge, three secessions occurred between 2000 and 2010, and by 2010 most of the segregation in the county was due to between-district segregation. Between 2010 and 2015, 14 charter school LEAs formed (in addition to two that existed in 2010). As between-district school segregation continued to rise (see Table A1 in the appendix), there was also a very small increase in between-district residential segregation during this time period, as the charter schools permitted families to choose a different education provider without moving or creating new boundary lines. Likewise, the ASD, as a non-geographically-based system of state-run schools, plays a similar role in Shelby County, Tennessee. The school–residential segregation relationship in this county was further complicated when there was a consolidation of the city and county districts and then subsequent secession of six separate suburban districts. Both school and residential segregation increased during this time, and the residential patterns likely do not fully reflect the new suburban secessions. Between-district segregation was especially high for Black and White students.

Discussion

In a region known for the desegregative influence of its countywide school districts, particularly for Black and White students, school district secession in seven counties in the South has erected new boundaries that largely prohibit student transfer across them, and in many places, the newly formed districts, by and large, were more racially homogeneous than the larger county districts that housed them in 2000. Such new boundaries contribute to an increasing share of overall county-level segregation by 2015. Furthermore, secession continues apace beyond the time period we examined. With the passage of time, initial effects, particularly for residential segregation, may grow stronger.⁹ In these seven counties, where a combined 18 new school districts—as well as charter schools in states that authorized them—formed from 2000 to 2015, increased residential integration and racial diversity in the counties may activate perceptions of racial threat as Black student enrollment approached parity or surpassed that of White student enrollment (Fiel, 2015). This, in turn, can fuel the perceived stakes around securing relative schooling advantage through secession efforts (Stout v. Jefferson County Board of Education, 2018). Although residential segregation patterns suggest there may have been a delayed effect, in the counties with the most extensive patterns of secession, we saw that boundaries were beginning to be racialized boundaries for residents as well as for students.

Indeed, the public school enrollment in the counties experiencing secession was increasingly non-White, whereas the residential population remained about 50% White. This demographic divergence from the rest of the region and country may be driving some of the motivation for secession. Our county-level analysis shows that secession has occurred in large Southern school systems with substantially lower shares of White students (roughly 33%) than the South as a whole (43%), suggesting that racial threat and competition may be at work (Fiel, 2015). Through the creation of new boundary lines, secession becomes a political mechanism for disproportionately White communities to maintain relative advantage in terms of student composition and, likely, financial resources, given the funding gaps between predominantly minority and predominantly White districts (EdBuild, 2019b). In the recent Gardendale, Alabama, secession attempt, a member of the advisory board for the secession wrote, “A look around at our community sporting events, our churches are great snapshots of our community. A look into our schools, and you’ll see something totally different” (Stout v. Jefferson County Board of Education, 2018, p. 11). The something “totally different” was the racial diversity of public schools, a result of court-ordered school desegregation efforts in what remains of the countywide district Jefferson County that was 43% White and 47% Black in 2015. The differences between the community and the schools was also cited by residents as a reason to form their own separate district to keep students of color out of their community’s schools. Although the federal courts prohibited the Gardendale secession attempt, similar forces have successfully created new boundary lines in other places without ongoing court orders.

Our analysis shows that a relatively higher percentage of school segregation in these seven Southern counties now occurs because students are enrolled in separate school districts, not just separate schools. School district secession has meant that segregation between Black and White students, and between Hispanic and White students, deepens between school districts. By 2015, the proportion of school segregation due to school district boundaries was approaching three quarters for Black and White students. And for Hispanic and White students, whereas in 2000 school district boundaries accounted for about a third of school segregation, by 2015 that proportion had increased to two thirds.

This systematic examination of 15 years of secession in the South also provides preliminary evidence that school district secession is eroding what has historically been one of the cornerstones of school desegregation in the South: the one-county, one-school-system jurisdiction. In counties where secession has eroded the countywide advantage for school integration (i.e., Jefferson County and Marshall County), it has done so in places where multiple significant secessions have occurred over time, meaning that there is no longer one school district that enrolls most students in the county. District secession is associated with increasing segregation of students across district lines, particularly for Black and White students, and Hispanic and White students.

County-level results point to a nuanced relationship between school and housing segregation in communities experiencing secession. Counties experiencing secession over a longer period and containing several sizeable school districts experienced higher segregation between school districts for residents than counties undergoing more recent secessions impacting smaller numbers of residents. For example, in Jefferson County, Alabama, school district boundaries in 2014 contributed to approximately 40% of the racial residential segregation in the county—which was quite high. This underscores the fact that in a county with a long history of school district secession, school district boundaries contribute substantially to the residential segregation of the total population, including those without children in the public schools. Neighborhood segregation is often cited as the root cause of school segregation (see Rothstein, 2017), but the divergence here in school and residential segregation patterns demonstrates that the relationship is more complex and dynamic. In the counties impacted by secession, on average, the construction of new school district boundaries was associated with a rise in their segregative role for students, but not for residents, at least in the short term. Our findings are an important foundation for future research with a longer time horizon to assess changes in residential sorting associated with more recent district secessions. It remains important to monitor the residential as well as school trends in the long term as school composition becomes increasingly shaped by the new boundaries, which then may become an important amenity in marketing homes in the counties with more recently formed districts. It is also unclear how the growth of charter schools in the region might emerge as a complementary segregating mechanism from countywide districts and whether that has a different longer-term effect on residential segregation.

Ultimately, the trends documented here point to secession as a new form of resisting desegregation amid the growing diversity of the South’s public schools. Secession has weakened the potential for greater school integration across the South’s broadly defined communities, fracturing White and Black and White and Hispanic students into separate school systems. Given the obstacles to comprehensive cross-district integration policies, secession makes it ever more difficult to bring them back together. Although the link between school and housing segregation in Southern communities impacted by secession is less clear-cut, trends in places with long-standing secession experience suggest that their neighborhoods will become more residentially divided along with their schools. The array of negative consequences outlined here should give states and communities substantial pause as secession efforts accelerate. A framework for examining the damages of proposed secessions is sorely needed—and oversight related to the impact of existing secessions past due.

Footnotes

Appendix

Table A2

Black/White Segregation of Students and Residents in the Counties Where School Districts Have Seceded, 2000 to 2014–2015

Variable	East Baton Rouge, LA	Jefferson County, AL	Marshall County, AL	Mobile County, AL	Montgomery County, AL	Shelby County, AL	Shelby County, TN
School
2000	.321	.632	.252	.475	.282	.125	.551
Between	.000	.534	.229	.000	.000	.000	.324
Within	.321	.098	.023	.475	.282	.125	.228
2010	.355	.557	.147	.461	.364	.061	.537
Between	.223	.441	.131	.010	.000	.000	.267
Within	.132	.116	.016	.451	.364	.061	.270
2015	.396	.533	.126	.427	.330	.058	.543
Between	.284	.407	.112	.037	.053	.005	.380
Within	.112	.126	.014	.389	.277	.054	.163
Residential
2000	.504	.556	.297	.477	.411	.183	.538
Between	.000	.275	.131	.000	.000	.000	.148
Within	.504	.281	.166	.477	.411	.183	.391
2010	.467	.530	.362	.424	.370	.152	.530
Between	.038	.221	.138	.003	.000	.000	.107
Within	.428	.308	.224	.420	.370	.152	.423
2014	.434	.508	.362	.393	.348	.128	.498
Between	.037	.200	.122	.007	.003	.001	.143
Within	.397	.308	.241	.386	.346	.127	.355

Source. Common Core of Data SY 1999-2000, SY 2009-2010, SY 2014-15, Decennial Census 2000 and 2010, American Community Survey 2012-16 estimates.

Note. LA = Louisiana; AL = Alabama; TN = Tennessee.

Notes

Authors

KENDRA TAYLOR completed her PhD in the Department of Education Policy Studies at the Pennsylvania State University. Her research has focused on the role of education policies in structuring residential segregation and policies to promote school and residential integration.

ERICA FRANKENBERG is a professor of education and demography at the Pennsylvania State University. Her research interests focus on racial desegregation and inequality in K–12 schools and the connections between school segregation and other metropolitan policies.

GENEVIEVE SIEGEL-HAWLEY is an associate professor at Virginia Commonwealth University. Her research focuses on examining school segregation and resegregation in U.S. metropolitan areas along with strategies for promoting inclusive school communities and policy options for a truly integrated society.

References

Anderson

J. D.

(1988). The education of Blacks in the South, 1860–1935. Chapel Hill: University of North Carolina Press.

Ashmore

H. S.

(1954). The Negro and the schools. Chapel Hill: University of North Carolina Press.

Bischoff

(2008). School district fragmentation and racial residential segregation: How do boundaries matter? Urban Affairs Review, 44(2), 182–217.

Bobo

Kluegel

J. R.

Smith

R. A.

(1997). Laissez-faire racism: The crystallization of a kinder, gentler anti-Black ideology. In Tuch

S. A.

Martin

J. K.

(Eds.), Racial attitudes in the 1990s: Continuities and changes (pp. 15–41). Westport, CT: Praeger.

Boger

J. C.

Orfield

(2005). School resegregation: Must the South turn back? Chapel Hill: University of North Carolina Press.

Boustan

L. P.

(2012). School desegregation and urban change: Evidence from city boundaries. American Economic Journal: Applied Economics, 4(1), 85–108.

Braddock

J. H.

III Gonzales

A. D. C.

(2010). Social isolation and social cohesion: The effects of K–12 neighborhood and school segregation on intergroup orientations. Teachers College Record, 12, 1631–1653.

Brown

(2016, August 16). A Southern city wants to secede from its school district, raising concerns about segregation. Washington Post. Retrieved from https://www.washingtonpost.com/local/education/a-southern-city-wants-to-secede-from-its-school-district-raising-concerns-about-segregation/2016/08/25/13ce5398-694f-11e6-99bf-f0cf3a6449a6_story.html?utm_term=.9787ec24ffbc

Carter

R. L.

(1953). The effects of segregation and the consequences of desegregation: A social science statement. Journal of Negro Education, 22(1), 68–76.

10.

Census Bureau. (n.d) Decennial Census, 2000 & 2010. [data files]. Washington, DC: US Census Bureau. Retrieved from https://factfinder.census.gov/faces/nav/jsf/pages/download_center.xhtml.

11.

Census Bureau. (n.d.) American Community Survey, 2012–16. [data files]. Washington, DC: US Census Bureau. Retrieved from https://factfinder.census.gov/faces/nav/jsf/pages/download_center.xhtml.

12.

Connerly

(2005). “The most segregated city in America”: City planning and civil rights in Birmingham, 1920–1980. Charlottesville: University of Virginia Press.

13.

Diem

Siegel-Hawley

Frankenberg

Cleary

(2015). Consolidation vs. fragmentation: The relationship between school district boundaries and segregation in three southern metropolitan areas. Pennsylvania State Law Review, 19(3), 687–745.

14.

Eaton

(2014, May 15). How a new secessionist movement is threatening to worsen school segregation and widen inequalities. Nation. Retrieved from http://www.thenation.com/article/179870/how-new-secessionist-movement-threatening-worsen-school-segregation-and-widen-inequal

15.

EdBuild. (2019a). Fractured: The accelerating breakdown of America’s school districts. Jersey City, NJ: Author. Retrieved from https://edbuild.org/content/fractured/fractured-full-report.pdf

16.

EdBuild. (2019b). $23 billion. Jersey City, NJ: Author. Retrieved from https://edbuild.org/content/23-billion

17.

Erickson

A. T.

(2015). Making the unequal metropolis. Chicago, IL: University of Chicago Press.

18.

Fiel

(2015). Closing ranks: Closure, status competition, and school segregation. American Journal of Sociology, 121(1), 126–170.

19.

Frankenberg

(2005). The impact of school segregation on residential housing patterns: Mobile, AL and Charlotte, NC. In Boger

Orfield

(Eds.), School resegregation: Must the South turn back? (pp. 164–184). Chapel Hill: University of North Carolina Press.

20.

Frankenberg

(2009). Splintering school districts: Understanding the link between segregation and fragmentation. Law & Social Inquiry, 34(4), 869–909.

21.

Frankenberg

Ayscue

Orfield

(2019). Harming our common future: America’s segregated schools 65 years after Brown. Los Angeles, CA: Civil Rights Project/Proyecto Derechos Civiles and Center for Education and Civil Rights.

22.

Frankenberg

Lee

Orfield

(2003). A multiracial society with segregated schools: Are we losing the dream? Los Angeles, CA: Civil Rights Project/Proyecto Derechos Civiles.

23.

Frankenberg

Siegel-Hawley

Diem

(2017). Segregation by district boundary line: The fragmentation of Memphis area schools. Educational Researcher, 46(8), 449–463.

24.

Frankenberg

Taylor

(2017). School district secessions: How boundary lines stratify school and neighborhood populations in Jefferson County, Alabama, 1968–2014. University Park, PA: Center for Education and Civil Rights.

25.

Frey

(2014). Diversity explosion: How new racial demographics are remaking America Washington, DC: Brookings Institution.

26.

Hannah-Jones

(2017, September 10). Dividing lines. New York Times Magazine, pp. 40–49, 71–73.

27.

Johnson

Jr. (2017). The Changing influence of educational policy and race on metropolitan inequality, 1970–2010. Educational Forum, 81(2), 175–192.

28.

Johnson

(2019). Children of the dream: Why school integration works. New York, NY: Basic Books and Russell Sage Foundation Press.

29.

Kane

T. J.

Staiger

Riegg

S. K.

(2005). School quality, neighborhoods and housing prices: The impacts of school desegregation. Cambridge, MA: National Bureau of Economic Research.

30.

Lareau

Goyette

K. A.

(2014). Choosing homes, choosing schools. New York, NY: Russell Sage Foundation.

31.

Lassiter

M. D.

Crespino

(2010). The myth of Southern exceptionalism. New York, NY: Oxford University Press.

32.

Logan

J.R.

Stults

(2011). The persistence of segregation in the metropolis: New findings from the 2010 Census. Census Brief prepared for Project US2010.

33.

Logan

J. R.

Stults

B. J.

(2014). Interpolating U.S. decennial census tract data from as early as 1970 to 2010: A longitudinal tract database. Professional Geographer, 66(3), 412–420.

34.

Mickelson

Nkomo

(2012). Integrated schooling, life course outcomes, and social cohesion in multiethnic democratic societies. Review of Research in Education, 36, 197–238.

35.

Murnane

R. J.

Reardon

S. F.

(2018). Long-term trends in private school enrollments by family income. AERA Open, 4(1), 1–24.

36.

National Center for Education Statistics. (n.d.) Common Core of Data, Public School Universe, 1999–2000, 2009–10, & 2014–15. [data files]. Washington, DC: US Department of Education. Retrieved from https://nces.ed.gov/ccd/pubschuniv.asp

37.

Orfield

(1969). The reconstruction of Southern education: The schools and the 1964 Civil Rights Act. New York, NY: Wiley Interscience.

38.

Orfield

(1978). Must we bus? Segregated schools and national policy. Washington, DC: Brookings Institution.

39.

Orfield

(2001). Metropolitan school desegregation. In powell

j. a.

Kearney

Kay

(Eds.), In pursuit of a dream deferred (pp. 121–157). New York, NY: Peter Lang.

40.

Orfield

Siegel-Hawley

Kucsera

(2014). Sorting out deepening confusion on segregation trends. Los Angeles, CA: Civil Rights Project/Proyecto Derechos Civiles.

41.

Orfield

(2015). Milliken, Meredith and metropolitan segregation. UCLA Law Review, 62(2), 365–462.

42.

Owens

(2017). Racial residential segregation of school-age children and adults: The role of schooling as a segregating force. RSF: The Russell Sage Foundation Journal of the Social Sciences, 3(2), 63–80.

43.

Patterson

(2001). Brown v. Board of Education: A civil rights milestone and its troubled legacy. New York, NY: Oxford University Press.

44.

Pearce

(1980). Breaking down the barriers: New evidence on the impact of metropolitan school desegregation on housing patterns. Washington, DC: National Institute of Education.

45.

Pearman

F. A.

Swain

W. A.

(2017). School choice, gentrification, and the variable significance of racial stratification in urban neighborhoods. Sociology of Education, 90(3), 213–235.

46.

Pratt

(1992). The color of their skin. Charlottesville: University of Virginia Press.

47.

Reardon

S. F.

Firebaugh

(2002). Measures of multigroup segregation. Sociology Methodology, 32(1), 33–67.

48.

Reardon

S. F.

Grewal

E. T.

Kalogrides

Greenberg

(2012). Brown fades: The end of court-ordered school desegregation and the resegregation of American public schools. Journal of Policy Analysis and Management, 31(4), 876–904.

49.

Reardon

S. F.

Owens

(2014). 60 years after Brown: Trends and consequences of school segregation. Annual Review of Sociology, 40, 199–214.

50.

Reardon

Yun

(2005). Integrating neighborhoods, segregating schools: The retreat from school desegregation in the South, 1990–2000. In Boger

Orfield

(Eds.), School resegregation: Must the South turn back? (pp. 51–69). Chapel Hill: University of North Carolina Press.

51.

Rothstein

(2017). The color of law: A forgotten history of how our government segregated America. New York, NY: Liveright.

52.

Saporito

Chavers

J. M.

Nixon

L. C.

McQuiddy

M. R.

(2007). From here to there: Methods of allocating data between census geography and socially meaningful areas. Social Science Research, 36(3), 897–920.

53.

Siegel-Hawley

(2016). When the fences come down: Twenty-first century lessons from metropolitan school desegregation. Chapel Hill: University of North Carolina Press.

54.

Siegel-Hawley

Diem

Frankenberg

(2018). The disintegration of Memphis-Shelby County: School district secession and local control in the 21st century. American Educational Research Journal, 55(4), 651–692.

55.

Spencer

(2014, November 3). Should wealthy towns be able to secede from higher-poverty, higher-minority school districts? Hechinger Report. Retrieved from http://hechingerreport.org/wealthy-towns-able-secede-higher-poverty-higher-minority-school-districts/

56.

Stout v. Jefferson County Board of Education, 882 F.3d 988 (11th Cir. 2018).

57.

Stroub

K. J.

Richards

M. P.

(2013). From resegregation to reintegration: Trends in the racial/ethnic segregation of metropolitan public schools, 1993–2009. American Educational Research Journal, 50(3), 497–531.

58.

Theil

(1972). Statistical decomposition analysis: With applications in the social and administrative sciences. Amsterdam, Netherlands: North-Holland.

59.

Trent

(1997). Outcomes of school desegregation: Findings from longitudinal research. Journal of Negro Education, 66, 255–257.

60.

Walker

V. S.

(2001). African American teachers in segregated schools in the South, 1940–1969. American Educational Research Journal, 38(4), 751–780.

61.

Weiher

(1991). The fractured metropolis: Political fragmentation and metropolitan segregation. Albany: State University of New York Press.

62.

Wilson

E. K.

(2016). The new school segregation. Cornell Law Review, 102(1), 139–210.

63.

Wright v. Council of City of Emporia, 442 F. 2d 570 (1971).

Racial Segregation in the Southern Schools,School Districts,and Counties Where Districts Have Seceded

Abstract

Keywords

Review of the Literature

Southern Resistance to School Integration

The Contribution of School District Boundaries to Segregation

Southern Demographic Patterns and Recent Southern Secessions

Method

Sample

Data

Census

CCD

Analysis

Findings

School Segregation and School District Boundaries

Residential Segregation and School District Boundaries

Variation in Segregation by County’s History of Secession

Discussion

Footnotes

Appendix

Notes

Authors

References