School Desegregation, School Re-zoning, and Growth Management in Two Maryland Counties

This article examines how school and non-school policies interact to reinforce or disrupt school segregation in the context of suburban communities and how these systems are maintained by structural and institutional mechanisms. Methodologically, we use a case study approach to delve deeply into the interpretation and implementation of school attendance zone redesign and nonschool policies, specifically land use policies and tools. We draw on neo-institutionalist theory and Ray’s (2019) framework of racialized organizations to make sense of school districts, planning agencies, and their policies. We find that school district rezoning policies provide a weak regulatory framework for desegregating schools because school zoning decisions are not made in a vacuum but rather are shaped by policies and actions taken by other actors in a multi-level governance structure. School zoning policies themselves prioritized capacity over desegregation, and regulations and norms governing the public engagement processes privilege opposition to desegregation. Our study points to the importance of greater coordination across governmental levels and policy arenas, and Education Policy Analysis Archives, Vol. 29 No. 165 2 underscores how desegregation policy is part of a political and relational process between advocates, elected leaders, families, and youth across policy sectors.


School Desegregation, School Re-zoning, and Growth Management in Two Maryland Counties
The increasing diversity of America's suburbs is changing the dynamics of how we think about access to educational opportunity across and within metropolitan areas. As large numbers of low-income families and families of color migrate to the suburbs (Frey, 2018;Howell & Timberlake, 2014), how policy makers in these communities respond to growth and increasing racial and socioeconomic diversity has implications for the educational opportunities available to students. Unfortunately, we still know too little about the complex ways that school and non-school policies interact with each other to shape access to well-resourced schools in these suburban communities.
This study examines the interpretation and implementation of policy in response to increasing suburban diversity and school segregation in two Maryland jurisdictions. We focus on school attendance zones and non-school policies, and the ways these two arenas interact to foster or disrupt school segregation. School attendance zones play a central role in determining school composition and can be used to reinforce or disrupt segregation at both the school and neighborhood level. We use the term "non-school policies" to refer to land use policies and regulations that manage the "pace, location, and extent of development" (Pendall et al., 2006), residential and other wise, through tools like zoning, growth boundaries, growth caps, affordable housing, and impact fees. Specifically, we ask the following research question: How do school and non-school policy levers influence school rezoning within the context of diverse, suburban communities?
To address this question, we delve into technical aspects of agency-level policy implementation and examine how structural and institutional mechanisms constrain or facilitate efforts to foster school desegregation through school rezoning. We do not discount the salience of public mobilization and political actions to support or oppose desegregation, but in this study, we limit our inquiry to the roles of state and local public agencies as agents with the potential to disrupt school segregation. This approach recognizes that the key drivers of school segregation reach across state and local policy arenas and broadens our understanding of how non-school decisions interact with the informal and mutable boundaries of school attendance zones.
We draw on neo-institutionalist theories and Ray's (2019) framework of racialized organizations to make sense of school districts, planning agencies, and their policies. We find that school district rezoning policies provide a weak regulatory framework for desegregating schools because school zoning decisions are not made in a vacuum but are shaped by policies in a multi-level governance structure. School zoning policies themselves prioritized capacity over desegregation, and rules and norms defining the public engagement processes privilege opposition to desegregation. Our study points to the importance of greater coordination across governmental levels and policy arenas, and underscores how desegregation policy implementation requires attention to the political dimensions that are part and parcel of the relational process between advocates, elected leaders, families, and youth that shape policy development, implementation, and attendant outcomes.

Persistent Segregation, Uneven Geographies of Opportunity, and School Attendance Boundaries
The persistence of segregated schools is a long-standing and enduring problem facing American society (Rosiek, 2019;Rothstein, 2017). Nationally, high-poverty segregated schools with majority students of color have fewer resources, higher suspension rates, and employ less experienced teachers (Carter & Welner, 2013;Orfield et al., 2012;Orfield & Lee, 2005; U.S. Government Accountability Office, 2016). Segregation negatively impacts students, both academically and socially (Fantuzzo et al., 2014;Konstantopoulos & Borman, 2011). This holds true in Maryland, where a school's demographic composition, especially the percentage of low-income students in a school, is a strong predictor of the school's performance on standardized achievement tests (Waldron et al., 2017b).
Segregated schools stem from not only educational policies, but also from larger segregated patterns across metropolitan areas. These patterns-uneven "geographies of opportunity"-are the spatial manifestations of systems of racial and socioeconomic inequality in which risks and resources are unevenly distributed across metropolitan areas (Briggs, 2005;Galster & Killen, 1995;Squires & Kubrin, 2005). Some people-often White and with higher incomes-have access to quality housing they can afford, living-wage jobs, reliable transportation, high quality health care, well-resourced schools, open space, and other amenities. Othes-people of color, particularly African American/Black and Latinx peoples, new immigrants, and non-native English speakers and those with lower incomes-however, often live in places without these assets. Metropolitan fragmentation of housing markets accompanies that of school districts, resulting in both racially and economically segregated communities that are home to similarly segregated public schools (Ayscue & Orfield, 2016;Owens et al., 2016).
These geographies are no accident. For generations, politics and economics have shaped metropolitan change through the actions of city leaders and planners, real estate developers, and lenders. As Feagin and Parker (1990)  The processes that foster segregation cut across federal, state, and local levels. Since at least the 1930s, the federal government has underwritten residential segregation through highway infrastructure and racially restrictive home mortgages (Jackson, 1985;Massey, 2008;Rothstein, 2017). At the state and local levels, zoning in particular became a tool to enforce racial segregation and eliminate blighted conditions associated with certain classes of "undesirables" (Silver, 1997). The City of Baltimore instituted the first racial zoning ordinance in 1910; in 1917 the Supreme Court deemed a similar law in Louisville, Kentucky unconstitutional. Subsequently, jurisdictions worked to find legally defensible ways to perpetuate exclusionary spatial patterns that cemented neighborhood and school segregation, including shifting the burden to school district facilities and enrollment plans (Benjamin, 2012;Erickson, 2016;Highsmith & Erickson, 2015). Today the link between residential and school segregation at the city and school district levels persists (Bischoff, 2008;Frankenberg, 2013;Reardon et al., 2012;Rothstein, 2017;Wells et al., 2012). Historically, a divide between "good" and "bad" neighborhoods and schools fell across urban-suburban lines; wealthier, whiter suburbs had better resourced school districts, while lowerincome, communities of color in urban areas suffered from a diminished tax base and neglect of public spending. This pattern is shifting, however. Over the past 20 years, large numbers of lowincome families, families of color, and new immigrants have been moving outside of central cities and into suburban communities (Frey, 2018;Howell & Timberlake, 2014, Lewis-McCoy, 2014. These migration patterns sometimes mean that families find themselves in a new suburban school district or, in the case of county-wide school districts, with access to different schools within the same district. These shifting demographic patterns pose new opportunities and challenges for disrupting persistent segregation and ensuring equitable access to opportunity, particularly to well-resourced schools. New patterns also require a reconsideration of tools available to school districts to manage school enrollment and composition, including school attendance zones. School attendance zones are intermediary spatial units that determine which school students attend and thus also determine school composition. School districts design these zones to manage enrollment, building capacity, and sometimes alter school composition at particular school sites. They are not a formal jurisdictional boundary (like city or school district lines) and are dynamic. Because attendance zones play a central role in determining school composition, they can be used to reinforce or disrupt segregation.
Only a small number of studies have looked specifically at the segregative effects of school attendance zones (Monarrez et al., 2021;Richards, 2014;Saporito, 2017aSaporito, , 2017bSaporito & Sohoni, 2006, 2007Saporito & Van Riper, 2016;Siegel-Hawley, 2013;Sohoni & Saporito, 2009). These studies rely predominantly on quantitative measures to characterize the relationship between the shape of attendance zones and levels of segregation. They found, for example, that racial segregation is greater in traditional neighborhood-based public schools than it is across school catchment areas and that economic segregation in public schools is higher than expected given the distribution of poverty across neighborhoods. Other researchers found that school districts with compact school zones have higher income segregation than districts with irregularly shaped attendance zones, suggesting that residential segregation is replicated in school attendance zones (Saporito & Van Riper, 2016).
Other studies examined the relationship between the residential segregation and school segregation. One study in a small Wisconsin suburb found that while stated rezoning goals elevated diversity, the results were nil; students remained socioeconomically and racially segregated in large part because of persistent residential segregation that the school rezoning could not mitigate (Mawene & Bal, 2020). Another study looked specifically at the intersection of school attendance zones and the location of federally subsidized housing in four Texas counties. It found that these housing developments "are zoned to racially and economically isolated schools, and that developments are associated with especially high levels of economic and racial isolation for Black and Latinx students" (Holme et al., 2020, p. 2). In addition to school attendance zones, school choice policies, which facilitate the movement of students from traditional public schools to private, magnet, and charter schools, also have been shown to contribute to segregation (Kotok et al., 2017;Saporito, 2017a;Saporito & Sohoni, 2006;Sohoni & Saporito, 2009).
These studies (and other more policy-oriented reports and press (Carey, 2019;EdBuild, 2020;Hannah-Jones, n.d.;Illing, 2018) are largely descriptive, motivated by the need to document the presence and extent of segregation. While informative and important, they tell us little about local policy implementation. They also focus primarily on school district policy levers, neglecting how non-school policies and actions interact with the informal and mutable boundaries of school attendance zones.
Scholarship that focuses on land use planning and its links to schools focuses narrowly in three areas: environmental and transportation impacts of school siting in outer lying suburban areas (Ewing & Greene, 2004;Hoskens et al., 2004;Vincent et al., 2017); school siting policy and decision-making (McDonald, 2010;McDonald et al., 2014); and travel to school and the declining prevalence of walking and bicycling (McDonald, 2007(McDonald, , 2008. These studies often disentangle land use policies and practices from those of the public schools, thus obscuring the historical links between racial discrimination in land use, property ownership, and public education (Ladson-Billings & Tate, 1995). Analyzing these policy domains separately also limits attention to interactions across sectors that have material consequence.
Our study builds on the foundation of this prior research and extends it by specifically examining how the intersection of education with land use policies contributes to the persistence of segregated schools and communities in two Maryland county-wide school districts. We contribute to what has been a predominantly a-theoretical conversation, driven more by the need to document and describe the phenomenon through quantitative measures. This earlier research has been critical in motivating interventions. Our project offers a next step-to examine the local policies in place and their intended and unintended outcomes, and to delve deeply into the nuance of implementation of land use decisions, policies and tools, and a key mechanism driving school composition-school attendance zones.

Theoretical Framework
We start from the premise that how education and land use policies interact to reinforce or disrupt school segregation is related to how these systems are maintained by structural and institutional mechanisms (Delmestri, 2008;Healey, 1999;Powell, 2008;Scott, 2008). Structural factors refer to how systems are organized though multilevel governance and highlight the complexity of policy and its implementation (Diem et al., 2014). Within this broader environment, policies enacted and implemented in one policy arena interact with policies in another arena to shape agency behavior by rendering some choices unavailable, precluding or advancing particular courses of action, and modifying how resources are allocated. Specifically, efforts to regulate school boundaries are constrained not just by the rules guiding the school rezoning process but by the rules governing where growth and development are targeted. Land use decisions influence the need for schools by directing population growth in certain areas. Where schools are built to accommodate these growing populations subsequently may constrain a community's ability to invest in maintaining or improving existing school buildings. Likewise, school facilities decisions and capacity constraints influence if and when new housing can be built and may constrain a community's ability to accommodate new populations or to provide affordable housing. Thus, district rezoning policies may be insufficient to address school segregation if the structures that support or produce segregated schools are not considered.
Institutional factors refer to the regulatory processes and normative mechanisms that operate within public systems in interdependent and mutually reinforcing ways to shape behavior, roles, and responsibilities by determining who can act, when, and how (Scott, 2008). Regulatory processes highlight the importance of rules, monitoring for conformity, and manipulating sanctions or incentives to influence behavior. For example, how school rezoning policy defines the rules around who participates in the decision-making process influences which groups may mobilize and shapes how they legitimately participate (Bachrach & Baratz, 1963;Bonal, 2012;McDonnell, 2009).
Normative mechanisms include conceptions of preferred behavior and confer legitimacy on policies by specifying appropriate actions. They may reinforce or perpetuate school segregation by constraining the options districts are willing to consider when rezoning schools. School boundary decisions are likely to trigger conflict between school officials and residents (Lareau et al., 2018). Districts may manage this conflict by avoiding explicit desegregation policies and instead define school rezoning policies as issues of school capacity (or something else) that help to maintain the segregated status quo. Likewise, in land use planning, seemingly race-neutral policies aiming to restrict growth are more likely to occur in places with larger White populations and result in persistent segregation (Trounstine, 2020). These race-neutral approaches are manifestations of a post-Civil Rights racial structure in which the explicit racist structures and overt discrimination of the antebellum and Jim Crow eras have been replaced by covert racial discourse, practices and terminology, and a "color-blind racism" (Bonilla-Silva, 2001, 2015. In both school and nonschool arenas, regulatory processes and normative mechanisms collide when public policies reify the deeply held albeit discriminatory preferences and values of the community through processes that residents recognize and respond to as legitimate. Institutional factors also contribute to how race is embedded in organizational processes and structures, often in seemingly race-neutral ways (Bonilla-Silva, 2015; Ray, 2019). Ray's (2019) framework on "racialized organizations" connects organizational rules and normative mechanisms to social and material resources in ways that legitimatize the unequal distribution of resources in racially disparate ways "in the absence of conscious discriminatory intent" (p. 34). Notably, while school districts and planning agencies may publicly commit to values of equity and diversity in vision statements or policy goals, they may divorce this rhetoric from the actual policies or implementation that would disrupt racial disparities and hierarchy. Thus, "decoupling allows organizations to maintain legitimacy and appear neutral or even progressive while doing little to intervene in pervasive patterns of racial inequality" (p. 42). Critical policy approaches also address how race is embedded within the structure of organizations and the space between policy rhetoric, development, and implementation and "practiced reality" (Diem et al., 2014). Specifically, we grapple with the ways that the technical specifications in policy reflect and reinforce particular configurations of power, knowledge and privilege, in particular in our discussion of the structure of public participation in these rezoning efforts.

Data and Methods
This interdisciplinary research used qualitative case study methods and spatial analytic strategies to advance our understanding of the relationships between education and place-based policies. An in-depth qualitative case study approach allows us to capture a rich and holistic account of policy processes as they are embedded in historical and socio-spatial contexts (Yin, 2009). We examine specific evidence on policies and plans, engagement and decision-making processes, and arguments that imply particular values and/or action around school attendance zone policy implementation and their links to land use policies.
We situated our study in Maryland where school districts are county-wide. County-wide districts mean policymakers have greater flexibility in drawing school boundaries than smaller districts constrained by jurisdictional boundary lines. We identified two counties-Baltimore and Howard counties-to study. We selected these two counties for several reasons. First, when we launched this study in early 2018, Howard County had recently approved redistricting elementary and middle school boundaries for the 2018/19 school year (AAC 2017 Meetings -HCPSS, n.d.) and Baltimore County had recently completed two elementary rezoning processes and was in the midst of a comprehensive high school rezoning. Second, both counties are in the Baltimore metropolitan statistical area (MSAs), share a regional housing market and have growing and increasingly diverse populations. Third, as we describe in greater detail below, the two counties vary in terms of their regulatory and policy approaches to land use, housing, and neighborhood segregation. These differences are proxies for other institutional and political differences that manifest in local governance arrangements, that may either facilitate or constrain policy options.
We conducted our study between February 2018 and March 2019. We relied on four sets of data: geospatial data, semi-structured interviews, official policy and planning documents, and newspaper media coverage. Each is described below. First, using 1990First, using , 2000First, using , and 2010 Census data and 2015 American Community Survey data, we mapped the spatial relationships between the demographic and socioeconomic compositions of the county, school district, neighborhood, and school populations. Demographic variables included race/ethnicity; socioeconomic variables included household median income and student eligibility for free and reduced-price meals.
Second, we conducted in-depth, semi-structured interviews with regional, county, and school officials, advocates, and housing professionals. Each interview lasted 60 to 90 minutes and included questions tailored to the specific responsibilities of each agency. Our interview protocol did not include questions about how individual respondents identify by race, socioeconomic status, gender, etc. We used purposive sampling to identify respondents with significant policy or decision-making authority and knowledge about the programs and policies in their organization. We also used a snowball sampling procedure to identify additional individuals that could address issues particular to each county.
We conducted a total of 23 interviews, including nine in Baltimore County, eight in Howard County, and six at regional organizations (Table 1). In Baltimore County, where there was an active rezoning process, we conducted participant observation and informal interviews at public meetings hosted by the school district. We audio-recorded interviews, except in one case in which an interviewee requested that we not and in five instances when we conducted interviews by telephone and recording was not possible. We transcribed all recorded interviews and took copious real-time notes during non-recorded interviews. We analyzed all transcriptions and interview notes using Dedoose qualitative analysis software. We began with a set of thematic codes drawn from prior studies and literature. We refined these themes based on en vivo coding of our interview transcripts. These codes included state and regional planning context (e.g., Adequate Public Facilities Ordinances, housing affordability, growth mitigation, transit); policy processes and institutional constraints (e.g., constraints imposed on school redistricting by Adequate Public Facilities Ordiances, growth, transit, etc.); race, diversity and othering (e.g., segregation, discrimination, demographic change); and transportation (e.g., how transportation intersects with housing and opportunity). Third, we analyzed relevant policy documents, plans, and public records to understand the formal policy and regulatory processes. School district documents included school board attendance area and zoning policies, construction and boundary changes studies, capital planning and attendance area studies, school rezoning recommendations, and attendance area maps. To track land use decisions, we collected county master plans; housing, development, land use and zoning polices; Adequate Public Facilities Ordinances (APFO); and county operating and capital improvement budgets. We also reviewed state agency planning and growth management policies and guidance.
Fourth, we analyzed newspaper coverage of key education and land use planning efforts for evidence of local priorities. We used the ProQuest newspaper database to search for education and school rezoning related stories in Baltimore and Howard counties between January 1, 2014, and September 30, 2018. Our final sample included 112 articles about Howard County and 84 articles about Baltimore County. As with our interviews, we analyzed articles using Dedoose qualitative analysis software. We began with a set of thematic codes drawn from prior studies, literature, and our interviews (which started prior to our news analysis) and refined these themes based on en vivo coding of the articles. The final media codes included growth and change (e.g., enrollment growth, school capacity, changes in school diversity); interventions (e.g., new school construction versus school renovations, redistricting); ways to engage (e.g., via public hearings, petitions, protests); race/ethnicity (e.g., how race/ethnicity is talked about, frequency); funding and the capital budget (e.g., money for new school construction, funding formulas, source of funding); values (e.g., safety, equity, transparency, stability, community, school performance); voice (e.g., which voices appear most often); issues of privilege/entitlement (e.g., which schools are prioritized for renovations, how resources are distributed across affluent/less affluent communities). We triangulated the codes and analysis across the qualitative data sources (interviews, policy documents, and newspaper coverage) to identify the themes and findings presented in this article.

County and School District Context
Our two case study sites-Howard County and Baltimore County-share similarities in terms of demographic composition and growth patterns. Both are located in central Maryland within the Baltimore metropolitan area; have a mix of urban, rural, and suburban areas; and have seen high rates of population growth over the past 25 years. Although both counties have become more diverse, changes in the racial composition of the population were not evenly distributed across the counties. Population growth led to greater diversity in the enrollment of students in each county's public school system.
The jurisdictions differ in their historical development and collective mythologies of place. Howard County's politics and governance are shaped largely by the origin story of Columbia. Built by developer James Rouse in the 1960s, Columbia is one of the country's most famous planned communities designed around principles of racial, economic, and religious integration (Columbia Archives, n.d.). In Howard County, the legacy of Columbia informs a public rhetoric and collective imagination around integration, progressive politics, and its high-quality public schools, which residents consistently vote to support through bond measures and local taxation. In contrast, Baltimore County has a long history of resisting the development of affordable housing and adopting land management policies that isolated Black residents from predominantly White neighborhoods and sought to reduce the number of Black people residing in the county (Maryland State Advisory Committee, 1971; Pietila, 2010; U.S. Department of Housing and Urban Development, 2016). Unlike Howard County's willingness to self-tax for public education, Baltimore County relied on population growth to maintain county revenues, which limited the school district's ability to address ongoing facility maintenance and construction.

Demographic Change and Population Distribution
Over the course of 25 years, Howard County experienced population growth, increasing 62.3% since 1990, from 108,328 to 304,115 in 2015 (Mullin & Lonergan Associates, 2011). The population growth is concentrated in the eastern part of the county, while the western part of the county shows not only a smaller population, but a concentrated White population. Howard County does not have any incorporated municipalities, but is home to several more urbanized, censusdesignated places (CDP), including Columbia, Ellicott City, Elkridge, North Laurel, and Savage-Guilford. Columbia is perhaps the most notable, as it is one of the country's first planned communities and was designed as an intentionally racially and socioeconomically integrated community. Net migration into Howard County occurred mostly among households earning $100,000 or more, resulting in a decrease in the proportion of lower ($25,000 -$44,999) and middle ($45,000 -$99,999) income families. In 1990, the highest income group made up 14.3% of the population; by 2000, that income group comprised 62.6% and in 2015, 64.5% of the population. As a result, new residents tend to be well-educated and employed in higher wage jobs, resulting in less economic diversity. While the county continues to be predominantly White (56.1% in 2015 compared to 81.9% in 1990), racial diversity among the minority population is increasing, specifically among Asian and Hispanic residents (Figure 1).

Figure 1
Howard County Race/Ethnicity  Source: U.S. Census, American Community Survey To accommodate the population influx, planning efforts directed development into urban areas, such as Columbia, thereby shifting development patterns to the central, southern, and eastern portions of the county. The county's planning efforts since the 1960s consistently insulated the rural western portion of the county from development pressure. These efforts were increasingly guided and reinforced by the state's planning vision, initially laid out in legislation in 1992. Today, higher concentrations of Black residents are found in the eastern part of the county. Asian residents are concentrated in the southwest, while many Hispanic residents live in or around Columbia, the most "urbanized" part of the County (Figure 2).

Figure 2
Howard County Race/Ethnicity (2015) Source: U.S. Census, American Community Survey Although the number of housing units increased by 56.4% from 1990 to 2015, there were only minor fluctuations between the proportion of owner-occupied and renter-occupied units, and the vacancy rate remained under 5%. The proportion of single family versus multi-family units also remained consistent, with 74.3% of units being single family in 2015.
Baltimore County is the third most populous county in Maryland, with a population of 822,959 in 2015 for an increase of 18.9% since 1990 when the population was 692,134. Much of Baltimore County is suburban, while the northern portion of the county is rural and areas adjacent to the border with Baltimore City are more urban. As in Howard County, Baltimore County has no incorporated municipalities, however, there are several census-designated places including Catonsville, Dundalk, Middle River, Owings Mill, and Towson. In Baltimore County, what at first looks like minor fluctuation in the county's overall family income composition represents a widening wealth gap. Like Howard County, the proportion of residents at the highest income group increased (from 36.5% to 39.8%). But those in the low and very low-income groups also increased (from 6.6% to 9.5%), while the middle-income group ($45,000-$99,999) saw the largest percentage point change, decreasing from 44.9% of residents in 1990 to 36.9% in 2015.
Since 1990, Baltimore County has diversified. While all minority groups have grown, the largest increase was in the Black population which went from 12.2% of county residents in 1990 to 26.7% in 2015 ( Figure 3).

Figure 3
Baltimore County Race/Ethnicity  Source: U.S. Census, American Community Survey During the same time, the percentage of White residents decreased from 84.1% in 1990 to 60.3% in 2015. Black residents are concentrated in the western and southeastern portions of the county that border Baltimore City, while White residents live in the northern and eastern parts of the county (Figure 4). Asian and Hispanic residents are more broadly dispersed in a U-shape surrounding Baltimore City.

Figure 4
Baltimore County Race/Ethnicity (2015) Source: U.S. Census, American Community Survey To keep pace with population growth, the number of housing units in Baltimore County increased, although there was no significant change in the share of owner occupied versus renters in the county over the 25-year time period. However, the vacancy rate increased since 1990, almost doubling its percentage share. While growth in multifamily units outpaced growth of single family with 19.1% growth in single family and 24.0% in multifamily, overall, the share of housing stock by housing type remained consistent. Like Howard County, Baltimore insulated the northern, rural area of the county from development, designating the remaining one/third of the county for development and revitalization of aging areas.

School District Enrollment and Composition
Howard County Public School System (HCPSS), the sixth largest school district in Maryland, ranks among the top school districts based on student performance on state assessments and local share of per pupil revenues for public schools (Department of Legislative Services, Office of Policy Analysis, 2018). School district enrollment and composition reflects the population growth and increasing diversity of the respective county. Enrollment in HCPSS increased 179% between 1990 and 2015, from 29,949 to 53,685. To accommodate such growth, the number of schools in Howard County increased from 48 in 1990 to 73 in 2015. Although White students still comprised the largest racial group in the school system, public school enrollment in Howard County became more diverse. Between 1990 and 2015, White student enrollment fell from 79.4% of total enrollment in 1990 compared to 42.7% in 2015. Black students made up the second largest racial group, accounting for nearly 21.9% of students in 2015 compared to 13.9% in 1990. Enrollment among Asian/Pacific Islander students represented 19.3% of Howard County students 2015 compared to 5.6% in 1990. Additionally, Hispanic student enrollment also grew significantly from 1.0% in 1990 to approximately 9.5% in 2015 ( Figure 5).

Figure 5
Howard County Public Schools Race/Ethnicity  Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data Baltimore County Public Schools (BCPS) is the third largest school district in Maryland and 25 th largest in the United States, with an enrollment of 109,830 students in 2015. Enrollment in the county schools has fluctuated between 1990 and 2015. Since 1990, total enrollment increased by 26.1%, from 86,737 to 109,830, however, like the growth in the county population, this occurred during the 1990-2000 period. Total enrollment decreased slightly between 2000 and 2010 and increased between 2010 and 2015, which brought the county slightly above the total enrollment of 2000. As enrollment increased, Baltimore County ignored building facility maintenance and construction until 2011 when the county adopted the Schools for Our Future school renovation and construction program aimed at addressing school overcrowding and deferred maintenance (Baltimore County Public Schools, 2011). The schools also became more racially diverse (Figure 6).

Figure 6
Baltimore County Public Schools Race/Ethnicity (1990-2015 Source: U.S. Department of Education, National Center for Education Statistics, Common Core of Data In 1990, White students represented 77.5% of students in public schools and 42.3% in 2015. Similarly, Black students represented 18.5% of students in 1990 and 38.6% in 2015. Additionally, the proportion of students eligible for FARMS has increased from 26.5% of students in 2000 to 46.9% in 2015. As schools became more diverse, they also became more segregated by race and income in both counties. Segregation by race is most pronounced between Black and White students where the dissimilarity index was 0.511 in Baltimore and 0.380 in Howard in 2010 (Dayhoff & Sunderman, 2014). Comparatively, Baltimore County was the third most segregated district by race (i.e., between Black and White) in the state while Howard County was the ninth most segregated on this measure. An analysis of school segregation by income in Baltimore and Howard County schools showed that the dissimilarity index was 0.413 for Howard and 0.391 for Baltimore, suggesting moderate segregation by income in both school districts. 1 However, put in context with other districts in the state, low-income students in the HCPSS were among the most segregated from non-poor students in the state. There were just two other districts in Maryland with higher levels of segregation by income.

School Rezoning Context
Due to rapid growth, Howard County opened 31 new schools since 1990. Many of these required individual school boundary changes, but the district did not initiate a comprehensive boundary redesign until 2019 (after we finished our data collection). In contrast to Howard County, boundary changes were infrequent in Baltimore County, but increased following the 2011 adoption of the Schools for Our Future school renovation and construction program (Baltimore County Public Schools, 2011). This $1.3 billion school renovation and construction program was aimed at addressing overcrowding in elementary schools, modernizing schools, and installing central air conditioning in all non-air-conditioned schools. In response to the new construction and building renovations, BCPS conducted 10 boundary change studies between 2014 and 2018. We describe more about the specific policies and mechanisms governing school rezoning processes in each school district in our findings below.

Different Approaches, Similar Material Outcome
We situate school zoning decisions within a broader institutional environment to explain how a multi-level governance structure shapes the policy choices available to school district actors. Within this multi-level structure, our research uncovered many more layers of complexity in the policy landscape than we anticipated on both the school and non-school side. Our initial goal was to understand the relationship between affordable housing and school rezoning implementation, as a response to desegregation advocates and scholarship that argue for better connections between specifically housing and school policies. But we heard much less about housing policy per say, and much more about land use and other growth management tools.
We found that the policies and procedures governing school rezoning are not designed to facilitate desegregation, but rather were a response to capacity constraints tied to land use policy. The formal policy and implementation mechanisms across school and non-school sectors foster the segregated status quo in three specific ways. First, state level growth management vision and framework fosters segregation at the county level by directing where growth and development take place. Second, school boundary change policies and county level non-school policies are designed to manage capacity, not composition. Third, regulations and norms governing the school rezoning public engagement processes privilege opposition to desegregation. The technical and public engagement specificities of both school rezoning and land use tools for growth management efforts (from state and local levels) converge in ways that hinder school desegregation and shield district administrators, school board members, and other county leaders from articulating and operationalizing a desegregation agenda.

State Level Growth Management Vision and Framework Fosters Segregation at the County Level
The institutional literature suggests ways in which the broader intergovernmental structure influences the school rezoning processes. It posits that policies enacted in one policy arena interact with policies in another arena to shape agency behavior. In this case, the state's growth management policies provided local jurisdictions with a vision and framework, often accompanied with incentives. To implement this vision counties used a set of planning tools designed to disperse growth and distribute population across the county. Taken together, these policies implemented at the state and local levels affected school boundary geography and fostered segregation at the local level. 2 The state's growth management framework, enacted into legislation in 1992, puts forth a "smart growth" and "sustainable development" agenda that directs new development in places with existing infrastructure. These polices are designed to protect environmentally sensitive areas, preserve rural and agricultural lands, ensure stewardship of the Chesapeake Bay, conserve resources, encourage economic growth through streamlined regulations and permitting processes, and ensure the provision of adequate public facilities and infrastructure where new growth would occur. 3 In 2009, the Smart, Green, and Growing legislative package strengthened local government comprehensive plans by establishing a statewide land use goal and directing local jurisdictions and the State to collect smart growth measures. It also updated the planning process to include 12 new planning visions (Maryland Department of Planning, n.d.-a, 2009).
Notably, social equity issues were absent from the state vision. Specifically, it omits any reference to racial and socioeconomic segregation as a question of neighborhood and community composition. The vision for housing calls for a "range of housing densities, types, and sizes . . . for all ages and incomes" but ignores race/ethnicity and does not include attending to the persistence or reproduction of segregated communities as a growth management goal. Where the Department of Planning mentions schools, the focus remains on school construction and rehabilitation and official language promotes walkable communities and neighborhood schools.
…schools serve as community focal points by promoting shared uses, preservation of community identity and landmarks, and promoting healthy, walkable communities. Therefore, as much as possible, the Maryland Department of Planning aims to advance policies that encourage local jurisdictions to construct and rehabilitate neighborhood schools. (Maryland Department of Planning, n.d.-c) While this meets smart growth and sustainable development goals, it simultaneously may exacerbate or reproduce segregation at the school level. Counties use myriad planning tools to achieve these smart growth goals -comprehensive planning, zoning and land use, and housing policies for preservation and new construction to ensure adequate supply of housing affordable across income levels. Two specific tools-the Adequate Public Facilities Ordinance (APFO) and the Urban-Rural Demarcation Line (URDL)-are particularly impactful in the design and implementation of school boundaries. APFO dictates how much residential and other development can take place based on impacts it would have on public infrastructure, including schools. URDL directs growth to particular areas of a county (i.e., inside the URDL line) in order to maintain rural areas (Nortrup et al., 1995). Counties design specific parameters around each of these tools, which set the stage for school rezoning by dispersing growth and determining population distribution across the county. These tools are necessary for growth management, but they are insufficient to promote integration because they do not proactively specify the type of housing that might promote integration, a key component determining school composition (Maryland Department of Planning, n.d.-b). Rather than address neighborhood composition, APFOs are state-enabled and locally implemented growth management tools that attempt "to link the timing of a new development to the availability of facilities needed to serve it" (The National Center for Smart Growth, 2006, p. 6). If a jurisdiction does not have the resources to provide the facilities needed, development is delayed, or the developer may pay to cover the costs for additional infrastructural needs. APFOs cover a number of "tests" for different kinds of infrastructure like roads, water, sewers, fire/rescue, and schools. In the case of schools, APFOs measure the contribution of new residential development to school enrollments and sets thresholds for impacts on school building capacity.
Maryland school systems are aligned geographically with counties, and county councils have final budgetary control over school district budgets. This alignment does not come without tensions, however, as it "pits school boards against county councils and educators and parents against builders and developers" (The National Center for Smart Growth, 2006, p. 19). The decision as to whether to rezone school attendance areas in response to new residential development is particularly volatile. While rezoning attendance boundaries may be the most desirable option for builders-who may have a financial interest that could benefit from a particular school attendance zone-and perhaps planners because it allows development to proceed, it is perhaps the least favorable option for parents and families if it means the reassignment of their child to another school.
This growth management tool relies on a set of technocratic calculations around population generation and facilities' capacity, yet the competing interests of developers, public officials, planners, and parents render it a highly politicized instrument in the dynamics of land use, growth management, and school composition. In both Howard and Baltimore Counties, APFO impeded school rezoning efforts, albeit in different ways. In Baltimore County, a school is defined as overcrowded when enrollment is at 115% of full-time equivalent students of the state-rated capacity of the building. That could mean technically that a development could not move forward if it would result in school overcrowding within a particular school attendance area. However, these capacity tests are malleable. If the schools in one school attendance zone are overcrowded, spare capacity in an adjacent zone may be used to mitigate that overcrowding, thus allowing the development to move forward if approved by the county council and school board (Baltimore County, Maryland Office of Planning, 2006).
In Howard County the 2017 school attendance boundary rezoning process was conflated with an initiative to redefine the APFO capacity tests. This effort sought to increase school capacity tests as a means to slow development and succeeded in redirecting attention away from the school system adopting a major redistricting plan (Personal interview, Howard County Department of Planning and Zoning, August 13, 2018). Instead of rezoning school boundaries, the county council amended APFO to include a capacity test for high schools and reduce the percentage at which an elementary or middle school would be considered "closed" to development. Increasing school capacity tests makes it harder to develop, thereby limiting supply and driving up costs of housing. This complicates efforts to build housing for moderate-and low-income residents, a critical vehicle to integrate neighborhoods (Personal Interview, Howard County Housing Commission, April 26, 2018).
The second tool-the URDL (adopted in 1967 in Baltimore County, and 1972 in Howard County)-is designed to separate rural areas from other areas where growth should be concentrated. In both counties, about one-third of the county area lies within the URDL (i.e., urban growth area) and two thirds lies outside (i.e., rural). The areas inside the URDLs are where school attendance rezoning takes place to accommodate a growing and increasingly diverse population.
Baltimore County was the first Maryland county to adopt an URDL, which includes all the county's urban areas around Baltimore City and protects the primarily agricultural and rural northern portion of the county from development. Howard County designated the eastern part of the county, which includes Columbia and Ellicott City, as a growth area. Both counties adopted URDLs to substantially restrict development and mitigate suburban sprawl in rural areas that lacked infrastructure and to preserve open space. Howard County used its 20-year sewage service area to establish its boundary. Baltimore County did much the same, while ensuring that already built-up areas were inside its URDL.
Outside of the URDLs, both counties see limited population growth but the largest increases in the proportion of White residents (see Figures 1 and 3). White residents in Howard County are moving from the eastern growth areas to the more rural western part of the county as the county diversifies. In Baltimore County, there is a similar pattern, where most of the White population is in the central and northern part of the county, while there is a larger population of Black residents in the southwestern part of the county. The most pronounced geographic change in race has occurred in the southwest portion of the county -the area bordering west Baltimore City where census tracts have seen a 51-100% decrease in White residents from 1990 to 2015. While Howard County looks to have more a diverse population concentrated in the eastern part of the county, Baltimore County appears to be more racially segregated. This reflects Baltimore's historical segregation and Howard County's prior integration patterns.
These planning tools and the lack of policy coordination across policy sectors gives school district administrators little incentive to address the geographic distribution of students that is largely tied to these growth management policies. By linking school rezoning processes to the county's growth and development processes, as we discuss in the next section, boundary adjustments are initiated to manage enrollment growth and address school capacity issues, but not to disrupt entrenched patterns of school and neighborhood composition -the key dimension in school and neighborhood segregation.

School and Non-School Policies Manage Capacity, Not Composition
Baltimore and Howard County school district boundary change policies provide a weak regulatory framework for desegregating schools. These formal rezoning policies encourage values such as equity, quality, and efficiency, but avoid identifying specific goals, objectives, or action steps that operationalize these values. They also omit any explicit attention to reducing school segregation. In Howard County, the Board of Education "establishes school attendance areas to provide quality, equitable educational opportunities to all students and to balance the capacity utilization of all schools" (Howard County Board of Education Policy 6010: School Attendance Zones, 2017). The intent is similar in Baltimore County where the Board of Education "establishes attendance areas in order to provide quality educational opportunities for all students and to promote the efficient use of school facilities and resources" (Baltimore County Board of Education Rule 1280: Community Relations: Community Involvement, 2016).
Absent an explicit regulatory framework for reducing segregation, the technical components of these boundary change policies support the status quo by constraining policy options to those that elevate building capacity and utilization over other concerns, like student composition and segregated outcomes. The criteria that trigger a boundary change process and the technical specifications for adjusting boundaries (once a process is initiated) are oriented around capacity issues, but not around school composition.
In both school districts, the superintendent may initiate a boundary area study when a new school is opening or a school is closing, a school is over or under capacity, to align school feeder pattern areas, respond to changes in an academic program, or when school attendance area projections are outside the target utilization (Baltimore County Board of Education Rule 1280: Community Relations: Community Involvement, 2016; Howard County Board of Education Policy 6010: School Attendance Zones, 2017). Assessments of student composition and/or segregation indices cannot trigger a boundary study, however. County growth management policies reinforce the school district policies' emphasis on capacity over composition. APFO and URDL limit where and how much growth can occur, but not the kinds of development that may promote integration (e.g., mixed income or multifamily housing). As county staff commented, "People think APFO does a lot more than it really does. APFO is not about redesigning communities" (Personal interview, Howard County Executive's Office, September 13, 2018).
The Howard County comprehensive plan suggests that drawing "boundaries to optimize any available capacity in adjoining schools is the most cost-effective means of addressing expanding enrollments," rather than reconsidering the land use and development patterns as drivers of that enrollment (Howard County Department of Planning and Zoning, 2018). But residents-some parents and other anti-growth advocates-approached the process with different priorities. In the 2017-18 school attendance rezoning process, many parents mobilized not only against the rezoning, but also around the APFO rewrite that the county council was considering. APFO presented an unexpected opportunity; the rewrite was delayed a few years, which meant that through an accident of bureaucratic timing, the look at this growth management tool would now happen simultaneously with HCPSS's look at their school rezoning. As these stakeholders saw it, if they could change the thresholds in APFO, they could limit development, mitigate school overcrowding, and avoid school attendance zone redesign. Despite the centrality of APFO in the debates over school rezoning, the planning tool was never going to support goals of desegregation because it is a tool to manage growth and therefore capacity, not student composition of schools.
Baltimore County's approach was similar in how it positioned school boundary rezoning processes as a response to the conditions of growth and development, rather than a tool to affect the distribution of students (Bonal, 2012) As noted by a Baltimore County Council member: "School redistricting is always at the end of the line. In terms of development and housing policy, those things are at the front of the line" (Personal interview, Baltimore County Council member, October 9, 2018). Further, the County's 2020 master plan focuses on school facility management in the context only of enrollment capacity, not demographic distribution: The County must balance the needs between new school construction and the renovation of existing schools when providing adequate facilities to mitigate overcrowding . . . New school construction, to the extent possible, should occur within the Urban Rural Demarcation Line (URDL). . . including growth areas featuring higher density, mixed use development. (Baltimore County Executive & the Baltimore County Council, 2010) As outlined, school facilities capital investments are assessed by standards of "overcrowding" and tied to the generic category of "growth," with no attention to the specificities of demographic composition of these growing populations. Once a boundary adjustment process is initiated, the technical components of school district policies provide a framework for what must be considered when adjusting the school boundaries. These technical components prioritize facility utilization, maintaining the continuity and stability of the neighborhoods, and consideration of the demographic characteristics of the students and the region. They also seek to ensure that students generally stay together through feeder patterns and are not reassigned multiple times throughout their school career, there is an equitable distribution of programs across the district, and that school areas are made up of contiguous communities or neighborhoods. Both districts recognized that the geographic place or neighborhood where students live is important in promoting a sense of community (Baltimore County Board of Education Rule 1280: Community Relations: Community Involvement, 2016; Howard County Board of Education Policy 6010: School Attendance Zones, 2017).
The inherent tensions in these technical components of the school rezoning policies operate to ensure the status quo by structuring calculations and baseline metrics in ways that reproduce segregated outcomes. The primary mechanisms that trigger rezoning are school capacity needs and enrollment projections. Every year, both Baltimore and Howard County school districts conduct school impact analyses based on school capacity and enrollment projections and use those to identify whether an attendance area adjustment is needed. Once a boundary adjustment process is initiated, the district conducts a feasibility study to identify possible boundary change scenarios.
The policies also specify priorities around student demographics, although Howard County's guidelines are more specific than Baltimore County. In Baltimore County, considerations are limited to "maintaining or increasing the diversity among schools to reflect the diversity of the region and school system" (Baltimore County Board of Education Rule 1280: Community Relations: Community Involvement, 2016). The Howard County policy seeks to promote diversity, where reasonable, within and across schools by racial/ethnic, socioeconomic, English language proficiency, and academic performance measures (Howard County Board of Education Policy 6010: School Attendance Zones, 2017).
Yet, the benchmark for composition is not about some envisioned target, but rather, pegged to the status quo (regardless of how segregated that may be). As one Howard County district staff described, "when we look at a plan, the changes we make, we aim to make sure we are not making the balance [of student composition] worse. We try to make it look better. Say before it looks like X, now it looks like Y -is it better or worse?" When asked about the standard or metric to which they referenced "better or worse," the staff member explained that it was the current configuration, rather than some envisioned target goal for integration. Most important, this staff member explained, "We want to minimize the instability of student moves while achieving some of these other goals." (Personal interview, Howard County Public School System, August 28, 2018) In Baltimore County redistricting, the diversity priority was operationalized as maintaining to the extent possible the demographic distribution of students that existed prior to rezoning, thus ensuring that desegregation would not occur. In Howard County, the equity priority could be used for reducing school segregation, but it wasn't. Instead, district feasibility studies, which form the foundation for boundary changes, were based on enrollment projections and prioritized capacity utilization and the alignment of feeder patterns (Howard County Public School System, 2017.

Public Engagement Processes Privilege Better-Resourced Residents Who Oppose Desegregation
Public engagement efforts aim to confer legitimacy on school boundary change processes by making the many steps of the process transparent to the general public and linking parents and community members to the schools. However, in their implementation, these processes can also create constraints by reinforcing particular configurations of power, knowledge, and privilege. In both districts, school boundary changes were frequent occurrences, and most proceeded with little opposition. But when residents opposed proposed changes, they weaponized the public engagement tools to their advantage, albeit in different ways across these two places.
School board policies in both Baltimore and Howard Counties outline specific requirements designed to gain public feedback on school attendance boundary studies and changes. These requirements on their face promoted transparency, deliberation, and engagement. However, their implementation through advisory committees, public hearings and workshops, and online surveys favored some residents over others. Individuals and groups with resources to organize, ability to attend evening meetings, and access to technical and professional expertise could more easily navigate the public processes. In Baltimore and Howard Counties-like other communities in the U.S. (Lareau et al., 2018)-residents who opposed the boundary changes and supported the segregated status quo benefited, while voices supportive of boundary changes and a less segregated district remained marginalized.
Public engagement processes were similar in both counties. Both required residents to process a tremendous amount of highly technical information and commit time to read materials, navigate online surveys, and/or attend public meetings. In Baltimore County, the district appointed a boundary study committee comprised of parents, principals, teachers, and community representatives from the affected schools. The district also hired an outside consultant to conduct a feasibility study and develop alternative boundary realignment scenarios. The redistricting process in Howard County began with a feasibility study, conducted by district staff. The Board of Education vetted applicants and appointed selected members of the public to a technical advisory committee called the Attendance Area Committee. In both cases, feasibility studies took into consideration available capacity, student enrollment projections, and in some cases, facilities condition, feeder patterns, and transportation logistics to develop different boundary options and create maps showing the various boundary scenarios. The public processes included open meetings and other mechanisms, such as focus groups, online surveys, email comments, and public workshops to solicit community input on the boundary options. Using this input, the advisory committees developed final recommendations that they presented to the boards of education. Following a series of public hearings, the board voted on a course of action for realignment.
The formal boundary studies and use of consultants and advisory committees provided the process with a professional, value-free veneer. Maps visualized the various boundary options and other graphics and data tables showed how school capacity and the racial and socioeconomic composition of schools would change under each scenario. The visual presentations were impressive and appeared objective, although a close look at the data revealed that the scenarios rarely altered the existing demographic composition of schools, even if this meant that capacity imbalances persisted. For example, in Baltimore County, the 2018-19 high school redistricting study developed seven scenarios based on increasing capacity by building an addition to an existing school, constructing a new school, creating new magnet programs, or making boundary changes that better utilized new and/or existing capacity. The diversity of the schools was not considered, and the consultant readily admitted that students would not be redistricted into a particular under-capacity school serving predominately Black students because "no one would go there" (Public meeting, Baltimore County Public School System, October 2, 2018). A Black resident of that community was concerned that, as the process evolved, the district was responding to the loudest voices and were not necessarily doing what was best for the county as a whole. "It's already [happened]. The first scenarios that they came out with, [our school] was a part of the process. Redistricting, renovation or something. In all three of the new scenarios, [this school] gets nothing" (Solomon & Boteler, 2018).
Both districts used online surveys to solicit preferences on different boundary options. In Baltimore County, surveys were also available at public meetings and workshops, thereby capturing the voices of those most engaged in the process. In both counties, community members organized to submit identical or nearly identical responses and aimed to sway the process by turning out at meetings as a block and writing op-eds in local news outlets. For example, in one Baltimore County 2017 elementary school redistricting process, most of the responses to an online survey were residents of a particular neighborhood, all of whom gave similar reasons for opposing two of the three options developed by the consultant because they spilt the community (Baltimore County Public Schools and Cropper GIS Consulting, 2017). The parents explained that they wanted to keep their community together and that they did not want their students in schools with other kids they judged as "bad." The consultant developed new options to address these concerns, both of which left the school at overcapacity and other schools under capacity, and with little if any change to their demographic composition and segregative conditions.
The process in Howard County reinforced configurations of privilege and power: "Like all planning processes, the process shows the haves and have nots -[only] the people who have time come out and speak," commented on district staff member (Personal interview, Howard County Public School System, August 28, 2018). In Howard County, a place that prides itself on the integration aspirations of Columbia, "the haves" largely refrained from attacks on other families and focused on rezoning's potential negative impacts on property values. Some homeowners asserted that redistricting would compromise their property values because their current school catchment area raised their assessments. These arguments provided a color-blind veneer or "mask" over the deeper issues of segregation: "That's the way it gets masked. I can't imagine you would find anyone to say that a school has too many minorities or poor kids" (Personal interview, Howard County staff, September 13, 2018). An advocate also batted away the idea that opposition to redistricting was grounded in racism.
It Logistical concerns about splitting neighborhoods and travel distances to different schools also surfaced, often couched in terms of equity. A former Howard County council member noted: It's an issue that concerns me, but to me there's a balance between forcibly integrating those that are choosing not to integrate for good reasons. Like for immigrants to acculturate takes a generation or two. How do you make decisions of who's going to be stuck with a longer commute?" (Personal interview, former Howard County Council member, October 2, 2018) These arguments opposing boundary rezoning are situated in a segregated status quo that is historically and persistently tied to the relationship between race, land use, housing valuation, and school quality. They also challenge the collective mythology of Howard County as a place that embraces diversity and integration and sit unselfconsciously side-by-side with arguments that make assumptions about immigrant communities and reify property values. In both Howard and Baltimore Counties, opposition figured out ways to weaponize other kinds of investments and policies to avoid redistricting. In Baltimore County, high school survey results reflected a strong interest from parents in increasing school district spending on capital improvements in order to address overcrowding; parents would rather pay to build new schools than face widespread redistricting (Sage Policy Group, Inc., 2018). Capital investments also played out in some areas of Howard County. One community with an overcrowded high school in the more developed, eastern part of the county saw redistricting as a way for the school system to backtrack on an over-65-year promise to build a new school: You're letting people build [housing] like crazy, you finally got around to giving us a library and a fire station and now we want a high school. They're like, "Well you let us all build out here, and we all have houses out here, and now you want to tell us there's something wrong with the land and you can't build a school here?" You know. So they feel kind of looked down upon and kind of like second class citizens. (Personal interview, Affordable housing advocate, February 16, 2018) Other Howard County parents made impassioned arguments about transportation and the burdens of increased travel to and from school and extracurricular activities for not only themselves but also for low-income families that the policy was trying to help. One advocate explained: One of the downsides of creating [integration] in the schools by redistricting is do you really want to take the kids that are from lower income families and make them go further away to school, which makes it harder to participate in events?" (Personal interview, "Slow growth" advocate, October 18, 2018) A former council member described hearing this argument from his constituents, including those from less affluent families: "I heard from parents of kids who are less affluent -why should my kid have to travel an extra? Why not bus the rich kids in here?" (Personal interview, former Howard County Council member, October 2, 2018). As described previously, the coincidence of the APFO revision with the school rezoning process created an opportunity for opponents of development to align and leverage opposition to school rezoning and vice versa. This group of highly educated and politically savvy parent leaders turned to the nexus of development, school overcrowding, and redistricting: Everyone started asking why and how did this happen? By fighting for a reduction in the threshold that APFO "tests" for school capacity, these parents and anti-development advocates found common cause to stop growth in the county that stressed the capacity of individual schools under the current attendance zone configuration. However, their alignment exacerbated housing affordability and school and neighborhood segregation. The mechanisms for building alternative scenarios and gathering public input shielded the boards from controversy because the most contentious proposals, which would have considered desegregation of school composition, never even made it to the list of proposed alternatives. As the Baltimore County consultant commented, proposed high school scenarios did not include moving students to particular communities because "there is so much bitterness in this community [and] the [public meetings] have been hostile" (Personal conversation, Baltimore County Public Schools, October 2, 2018). The hostile resistance came in coded racist language about families "whose lifestyle may not have the same values" (Personal conversation, Baltimore County parent, September 18, 2018). In the end, the constrained structures of the public participation processes in both counties fostered inequities in engagement that reflect (rather than disrupt) the segregated status quo.

Discussion
This study reveals that school district rezoning policies provide a weak regulatory framework for desegregating schools. In Maryland, county-wide school districts provide a better opportunity to use school zoning tools to achieve greater integration than in smaller districts. Despite this context, our analysis of the structural and institutional constraints operating on districts exposed the limitations of these policies-on their own-to address the composition of schools. It showed that while school rezoning policies ostensibly provided a framework for rezoning that could be used to encourage greater integration, structural and institutional constraints-regulatory processes and normative mechanisms (Scott, 2008)-pushed districts to focus on school capacity needs rather than school composition.
How does this happen? First, school zoning decisions are not made in a vacuum but rather are shaped by policies and actions taken by other actors in a multi-level and multi-sectoral governance structure. While we often hear about advocacy for school desegregation through better housing policy (Ayscue et al., 2013), we identified complex layers of policy that included land use controls, growth caps, and zoning that come from both the state and local level action. In both districts we studied, districts responded to population trends and development patterns that conflated school capacity with school boundary adjustment proposals, rather than using school rezoning as a proactive approach to managing school composition. This put boundary changes at the tail end of policy decisions emanating from higher levels of government across education and non-education arenas. Maryland growth management polices prioritized protecting the environment and preserving rural areas of the state, which directed growth to older, more densely populated and diverse areas of the counties. Without incentives or policy coordination across policy sectors to encourage desegregation, counties had little inducement to link development policy to its impact on school composition. Tools such as APFO that imposed capacity tests that could limit new development were easily manipulated so that development could go ahead or co-opted by residents to avoid school rezoning.
Second, the rules governing school zoning policies themselves prioritized capacity over desegregation. The school zoning policies included mechanisms that confined possible rezoning alternatives to those that favored the segregated status quo. When mechanisms to foster desegregation were available to school boards and staff, such as establishing desegregation goals, they did not use them. As a result, boundary changes deviated very little from the existing demographic composition of schools, often at the expense of addressing capacity imbalances between schools. By using enrollment projections and prioritizing capacity issues to initiate a boundary process, these policies legitimized the process by demonstrating that the district was responsive to the community by addressing overcrowding while, at the same time, ensuring the compositional status quo. The application of these policies contributed to consolidating the underutilization of certain schools, which was conducive to reproducing school segregation (Bonal, 2012, p. 413). In other words, to gain legitimacy, policymakers do what is expected of them and reified particular modes of participation that decoupled (Ray, 2019) the attendance boundary rezoning process from racial and socioeconomic segregation.
Third, our observations affirm the ways that regulatory processes and normative mechanisms structuring the public engagement process influenced which groups mobilized and how conflict was managed. Ultimately, these institutional arrangements privileged opposition to rezoning and provided political leaders cover from difficult decisions that would have favored desegregation; districts could "maintain legitimacy and appear neutral or even progressive while doing little to intervene in pervasive patterns of racial inequality" (Ray, 2019, p. 42). Our findings build on other research that examines how families take advantage of their cultural, political, and social capital to oppose redistricting (see e.g., Lareau et al., 2018). These accounts focus on how the geographic concentration of families though the housing market led to a pooling of resources to oppose rezoning. We open up another avenue and demonstrate how people activate their political, cultural, and social capital through institutional and structural mechanisms governing the rezoning and other land use processes.
Our findings illustrate the ways in which school districts and other public agencies enable segregation and thereby legitimize the unequal distribution of social and material resources in racially disparate ways, even "in the absence of conscious discriminatory intent" (Ray, 2019, p. 34). The distance between the public commitments to equity and the implementation that is required to realize this goal is vast, and is a hallmark of organizations that have institutionalized race into organizational policies and procedures in ways that maintain the status quo (Bonilla-Silva, 2015; Ray, 2019).

Conclusion
What alternatives may be available to these processes or close the gap between rhetoric and practice? First, as Bonal (2012) argues, political and legal tools are available within the educational policy arena. That is, school rezoning processes are not entirely constrained by external factors and could be used address school composition. For example, school level segregation indices could be a trigger for rezoning along with school capacity concerns. Following the failure of HCPSS's rezoning process (the focus of our study), the HCPSS superintendent launched his own plan for boundary changes that articulated an equity (economic) desegregation goal-reduce the number of schools with high poverty rates-in addition to a utilization goal-reduce the number of schools that are over or under capacity. This plan prioritized balancing capacity utilization across the system and addressed economic segregation by taking into account the distribution of students by socioeconomic status (Martirano, 2019). Likewise, elsewhere in Maryland, the Montgomery County Public Schools (MCPS) articulated an equity goal that would affect the distribution of students and completed a study of boundary adjustments (although as of this writing, the board of education had taken no formal action on implementation). Students in particular mobilized opposition to the district's current levels of segregation and called for a county-wide, comprehensive boundary study that considers composition of students as a central axis of analysis (St. George, 2018).
Notwithstanding these efforts, our findings suggest that using school rezoning to intentionally desegregate schools is more complex than simply redrawing school boundary lines based on student demographic information. School rezoning requires school leaders to develop a deeper understanding of how non-school policy arenas such as growth management and zoning interact with school boundaries and the ability to work across policy sectors to craft policies that address segregation. Since growth management and school attendance policies co-exist and interact across state, county, and district levels, jurisdictions need to pursue multi-faceted policies that work across governmental levels and provide incentives for interagency cooperation.
Second, the state could go further and institute closer oversight of school construction and associated boundary changes for impacts on segregation, creating a tighter accountability and regulatory framework for reducing segregation. For example, under its current program, Maryland's Interagency Commission on School Construction uses four criteria to prioritize state school construction funds to local districts: building age, concentration of low-income students, volume of portable classrooms, and building utilization rates. The state could tie building construction money to school desegregation metrics as well (The National Coalition on School Diversity and PRRAC [Poverty & Race Research Action Council], 2020). In addition, state and county planning documents could articulate social equity goals that consider the composition of neighborhoods. At the federal level, enforcement of fair housing laws could encourage analysis of school boundaries at the local level. In particular, the U.S. Department of Housing and Urban Development's Affirmatively Furthering Fair Housing process offers an avenue for local jurisdictions to coordinate housing, land use and growth management policies with school policies and data to understand and mitigate segregation.
Desegregation policy is part of a political and relational process between advocates, elected leaders, families, and youth. Advocates thus have a role in articulating how cross-sector policies contribute to disparate outcomes. They can also provide support for political leaders willing to take on controversial positions in support of desegregation. Public officials could be more cognizant of the ways that school rezoning policies and their implementation processes structurally privilege some residents over others. For example, public engagement processes could be redesigned to address the asymmetry in participation. At the state and local level, linking education policy to land use policy, an often-overlooked driver perpetuating segregated schools, is necessary to undo generations of harm from segregation. Finally, further research inquiry should examine the interaction of school and land use policies to identify how seemingly race-neutral institutional mechanisms, norms, and regulatory processes contribute to and maintain segregated schools.