"Community-Driven Exclusion Mapping: Examining the Discriminatory Impacts of Housing Segregation Across Urban, Rural, and Suburban Geographies,"by Peter Gilbert May/June 2014 issue of Poverty & Race
Community-Driven Exclusion Mapping: Examining the Discriminatory Impacts of Housing Segregation Across Urban, Rural, and Suburban Geographies
Mapping with Geographic Information Systems (GIS) is now firmly established as a necessary tool in understanding, documenting, and combating housing segregation. Unfortunately, most of the tools and information are designed with only sophisticated users in mind and may not provide data most relevant for individual communities, especially rural communities. The UNC Center for Civil Rights Inclusion Project provides one model for interactive mapping of housing segregation that is both large in scale and designed to be used directly by impacted communities.
HUD’s Proposed RuleRegarding the requirements of Affirmatively Furthering Fair Housing (AFFH), HUD’s proposed rule includes a prototype web-based map that combines basic demographic data (age, race, disability status, English proficiency, and poverty) with the location of existing subsidized housing and voucher users, and with generalized “community assets and stressors.”1 The map is designed specifically to be used by local governments, states and public housing agencies in order to conduct a required Analysis of Fair Housing (AFH), which is necessary to receive various HUD grants.2 One of the goals of the new rule is to provide data directly so that governments and agencies can focus more on community engagement and less on data-gathering, to allow users to “spend less time gathering information and more time engaged in conversation with the community.”3 The tool is not primarily designed for communities to use directly.
Most of the remedies for Fair Housing issues addressed through HUD grants impacted by the new rules concern the location of new affordable housing, such as the HOME and HOPWA programs. Both HOME and HOPWA provide some funding for the rehabilitation of existing communities, as does the CDBG program. However, much of the emphasis is on new construction. HUD also requires an Analysis of Impediments to (AI) as part of AFFH to identify barriers to fair housing, which does require looking at historic patterns of segregation, but redressing them is primarily through providing access to affordable housing in more affluent or white neighborhoods.
HUD’s prototype mapping tool reflects this underlying purpose of providing affordable housing in areas with greater opportunity. The map displays generalized data about “community stressors,” like failing schools and health hazards, but does not display information on individual schools or polluters that affect a particular neighborhood.4 This type of data will be useful to governments and policy makers when planning for siting affordable housing in appropriate areas, but does little to assist existing communities in identifying or combating already existing impacts of segregation and exclusion.
Other civil rights advocates have pioneered the use of GIS technology in challenging entrenched segregation patterns. The Opportunity Mapping initiative of Ohio State University’s Kirwan Institute for the Study of Race and Ethnicity pursues the dual goals of understanding “where opportunity-rich communities exist (and assess who has access to these communities) and to understand what needs to be remedied in opportunity-poor communities.”5 Half of the focus, like that of the HUD tools, is therefore on access to “opportunity-rich communities,” while simultaneously attempting to understand needs of “opportunity-poor communities.” This approach has been focused in primarily urban areas such as Detroit, Jacksonville, Baltimore, Chicago and San Francisco, and in studies of more densely populated states like Connecticut and Massachusetts. The studies have been used by non-profit advocates to document housing segregation and direct resources to low-opportunity communities.
The UNC Inclusion ProjectRather than looking for broad patterns in urban areas, the Inclusion Project at the UNC Center for Civil Rights grew out of direct community representation with the goal of creating an interactive map and easily accessible data for communities to use in direct advocacy, as well as being available for local governments and advocates. The Center’s direct representation of excluded communities revealed recurring patterns of communities underbounded from municipal limits, kept out of the best schools, and burdened with landfills and other environmental hazards. The Project attempted to document these patterns empirically, identify new communities with similar impacts, and, most importantly, provide a resource for these communities to study and communicate their own situation.
Starting with the hypothesis that concentrated communities of color face disproportionate impacts of housing segregation, the study examined data in five key areas: environmental justice, housing, political exclusion, education, and access to infrastructure. Many data points were the same as those used in the more urban-focused opportunity mapping; others, such as access to infrastructure and exclusion from municipal boundaries, are more particular to rural and suburban forms of exclusion.
Unlike prior studies, the maps were generated with individual neighborhoods as the basic unit. The study identified all census blocks in North Carolina that are 75% or more non-white or Latino, and then clustered those census blocks that were contiguous. Removing any clusters that had 25 people or fewer left about 3,200 clusters across N.C. These populations ranged in size from a few dozen to many thousands of people, but the average size was about 400 people and the vast majority of clusters were between 100 and 1,000 people—a good approximation for a neighborhood. The clusters are defined empirically and therefore often are both over- and under-inclusive of actual neighborhoods defined through historic and cultural ties.
The Impacts of ExclusionIdentifying individual communities and examining the impacts of exclusion at the neighborhood level reflects how the impacts are experienced. School assignment, exposure to environmental hazards, political boundaries, and access to infrastructure all depend on where you live more than who you are. When a neighborhood is overwhelmingly one particular race, all of the residents face any impacts of that segregation, regardless of their own circumstances or identity.
Using individual communities as the basic structure for the map also makes the mapping tool more useful to individual communities. All community members must do is find their community on the map, and all the basic information for that community—the demographics, exposure rates to solid waste facilities or EPA-monitored polluters, homeownership and vacancy rates—is one click away. Similarly, school data on racial composition or free and reduced lunch data are immediately accessible.
Clusters are then compared to state and county averages; every resident is either in a supermajority non-white cluster or is not. While this binary helps identify issues facing discrete neighborhoods, it simultaneously oversimplifies patterns of segregation. Most maps display racial demographic data as a spectrum, as does Kirwan’s opportunity mapping and HUD’s prototype tool. The spectrum allows more specific data and perhaps a clearer statistical analysis of whether purported impacts are directly tied to racial exclusion. Because the 75% non-white cutoff for the clusters is inherently somewhat arbitrary, many census blocks or communities that face impacts of housing segregation are left out of this analysis.
The Community-Centered ApproachDespite the binary limitation, the community-centered approach revealed startling results for North Carolina. Dramatic disparate impacts were found in the areas of environmental justice, education, and housing. Residents of clusters were exposed to solid waste facilities at almost twice the state average. Residents of majority African-American clusters were exposed to solid waste facilities at a rate of 10.4%, as opposed to 5.3% for the state average. Similarly, for EPA-registered polluters, the odds that cluster residents were within a mile of a facility were 41%, but only 24% for the state average. Cluster residents were also twice as likely to have their closest school to be failing, twice as likely to be high-poverty, and significantly more likely to be racially identifiable. Homeownership rates were also dramatically lower: rental rates for cluster residents were 54% as opposed to 32% across the state.
This community-centered approach balances these broad results across a large area, with the ability to easily reveal specific information about a community, or a particular impact. Presenting the data in an easily useable format is crucial for a community’s ability to use the data. Users can access the data either through the map, easily identifying communities, schools and polluting facilities, and clicking on them for the necessary data, or through customizable charts. The charts display aggregated data in each of the metrics of exclusion, such as rental rates or exposure to solid waste facilities. The data can be aggregated to compare cluster totals to county and state averages, or across regions of the state.
Community-centered mapping like the Inclusion Project is not a substitute for other mapping techniques. Other techniques better present data necessary for identifying where new affordable housing projects should or should not be located, or for broad statements about disparate impact across larger areas. This model instead presents an approach to mapping that is designed with the community as the end-user.
1 http://www.huduserorg/portal/affht_ pt.html#summary-tab
2 Affirmatively Furthering Fair Housing 79 Fed. Reg. 43710 (July 19, 2013) — to be codified at 24 C.F.R. pts. 5, 91, 92, 570, 574, 576, 903.
4 User Instructions for the Prototype Geospatial Tool for the Proposed Affirmatively Furthering Fair Housing Rule 5 http://kirwaninstitute.osu.edu/opportunity-communities/mapping/
State of Exclusion Report
HUD’s prototype geospatial tool
Kirwan’s Opportunity Mapping Project
EPA’s EJView Environmental Justice Mapping Tool
Peter Gilbert is an Equal Justice Works Fellow sponsored by the Norflet Progress Fund working at the UNC Center for Civil Rights. He directs the research for the Inclusion Project and authored its State of Exclusion report. firstname.lastname@example.org
|Poverty & Race Research Action Council | 740 15th St. NW, Suite 300, Washington, DC 20005|
©Copyright 1992-2018 Poverty & Race Research Action Council