Urban population characteristics and their correlation with historic discriminatory housing practices
Introduction
This project seeks to understand the long-term effects of discriminatory housing policy. Housing discrimination, on the basis of race, religion, ethnicity, and other characteristics, exists worldwide and the negative consequences may impact the residents well into the future. This research began about 50-years after racially based housing discrimination was prohibited in the USA, and we investigate whether and how those practices have marked cities in the USA. We took an exploratory, data-driven approach to analyze the spatial distribution of population characteristics like health, education, and income along with the historical housing policy of those areas. We used publicly available data provided by the Center for Disease Control (CDC) and the American Census Bureau to study health outcomes like chronic heart disease and diabetes prevalence, along with other key indicators including poverty status, unemployment rate, and health insurance rates, and we used historic neighborhood evaluation maps created by the Home Owners Loan Corporation (HOLC) to capture discriminatory policy and practices. Understanding and remembering the deep roots of residential inequality and discrimination is important as we progress further away from the days of legalized segregation1 and into less evident forms of discrimination. This progression comes with a need to understand the systemic foundation that has been laid to keep people of color, particularly Black people, out of white neighborhoods and the negative and persistent consequences those systems carry. Critical thinking about the deliberate choices made in the past and the relationships those have to current outcomes is crucial as we take steps forward and develop policy for the future. Our objective in this work is to explore and compare modern data across HOLC neighborhood boundaries, presenting a broad view of the state of intra-city inequality in the US today by capturing differences across 12 metrics and 14 cities.
This paper is structured as follows. Section 1.1 provides a brief overview of housing discrimination in the United States and the historical context for the HOLC appraisal maps. Section 1.2 connects the history of discriminatory practices health and population level outcomes, including how neighborhood development and segregation may impact populations. Section 1.3 reviews existing literature linking outcomes to housing discrimination and segregation, with specific emphasis on those linked to HOLC districts and maps. Section 2 details our methods and data sources, followed by results in Section 3. Section 4 discusses our results in context, along with potential for expansion of this work, and Section 5 concludes. While the HOLC maps used in this study are unique to the United States, the presence and lasting impacts of discriminatory housing practices on the basis of race, religion, and ethnicity are not. In our discussion, we include several examples of housing discrimination in non-US contexts and the potential implications this work has outside of the US.
The term “redlining” has been defined as “discriminatory practice in real estate, typically involving lenders that refuse to lend money or extend credit to borrowers in certain areas of town,” (Weintraub, 2018) and the origin of the term is commonly attributed to a series of color-coded neighborhood appraisal maps created in the 1930s by the Home Owners Loan Corporation (HOLC). Neighborhoods with “undesirable” populations were given a rating of D, colored red on the map, even if the financial state of the neighborhood or its objective real estate value would suggest otherwise. These maps give us a clear indication of where many discriminatory policies and practices were being implemented at many levels of the system because of the residents who lived in them. In this paper, we use HOLC neighborhood boundaries to study areas that most likely faced racial discrimination in the 1930s and in varying forms since; these neighborhoods are sometimes referred to as “redlining districts,” and consist of the neighborhoods given the lowest appraisal rating by the HOLC. We provide more context on the history of the housing system and some examples of the racial discrimination within it in the following paragraphs.
The idea that racial minorities and immigrants moving into a neighborhood decreased property value reached mainstream prominence in the early 1900s among those heavily involved in the housing process, including mortgage bankers, real estate agents, appraisers, and brokers. This idea led to the establishment of deed restrictions based on race, racially restrictive covenants, and “neighborhood improvement organizations,” in an attempt to keep neighborhoods from being integrated (Gotham, 2000). These ideas were amplified so far as to be written into organization documents: Article 34 of the 1924 National Association of Real Estate Boards’ official code stated that, “a Realtor should never be instrumental in introducing into a neighborhood … members of any race or nationality … whose presence will clearly be detrimental to property values in that neighborhood” (National Association of Real Estate Boards, 1924). Additionally, the National Association of Home Builders pushed for land use and zoning regulations to keep neighborhoods segregated. This effort to keep neighborhoods from integrating continued through the market crash in 1929, and the New Deal legislation passed in response to the Great Depression had clear influence from this effort (Gotham, 2000).
The Home Owners' Loan Act of 1933 created the Home Owners' Loan Corporation (HOLC) as an emergency relief system targeted at mortgages that were in default or at risk of foreclosure after the 1929 market crash (Home Owners' Loan Act of 1933 and H.R. 5240, 1933). The HOLC used government-funded bonds to buy back mortgages from banks. By 1934, roughly one year after its creation, 20 percent of all mortgages in the United States were owned by the HOLC, and in their three-year lifespan, they lent around $350 billion, or about $750 billion in today's dollars (Roosevelt Institute, 2012). There are those who maintain a positive view of the HOLC for its contribution to reinvigorating the housing market from collapse, and some called for an updated version to be created in 2008 after the market crash (Pollock, 2008). However, the 1980s discovery by historian Kenneth Jackson of the blatantly racist neighborhood evaluation system created by the agency reminds us of the fact that improving the housing market for white neighborhoods did not necessitate the same improvement levels for other neighborhoods (Kazmi, 2017). Black, immigrant, and Jewish populations were not given the same access to government aid when attempting to enter the housing market. In fact, neighborhood descriptions from HOLC maps indicate that Black, Jewish, Asian, Italian, Polish, Hungarian, Czech, Greek, Mexican, Russian, Slavic, and Syrian populations were consistently used to lower the grade of neighborhoods across the country (Nelson et al., 2021).
The Federal Housing Act was passed in 1934 as a complementary New Deal attempt to rebuild the home building and finance industries and bring the country out of the Depression (Gotham, 2000). Although many conversations about New Deal programs center around social welfare, the impacts to housing were significant. The Federal Housing Administration (FHA), created by the Federal Housing Act, played an integral role in America's suburbanization by reducing home down payments, providing regulations for homebuilding, and subsidizing home financing in a stated attempt to make home ownership accessible to the middle- and working-class. However, the FHA also enforced racially restrictive housing covenants on housing insured by the government and often refused to insure homes in areas comprised primarily of racial-minorities. The FHA guidelines were created in an effort to impose standardization across the country on lending practice as well as to create a network of connected mortgage lenders under the federal government (Gotham, 2000).
The HOLC provided the standardized approach to mortgage restructuring that FHA desired, an approach based on individual applicant characteristics and predicted future property value (Crossney and Bartelt, 2005). The neighborhood appraisal maps provided these predicted property values at a neighborhood level. Some have doubted the impact and widespread use of these maps, stating that the HOLC was not “the first organization to include racial criteria as part of an appraisal or survey,” and was not the primary lending institution of the US at the time (Crossney and Bartelt, 2005). However, several studies indicate that their use was widespread, particularly by the FHA, and indicate strong communication between government lending agencies and private institution practice (Aaronson, Hartley, & Mazumder, 2018; Greer, 2012; Woods, 2012). Others have documented that while Black Americans were not completely excluded from HOLC lending, the loans were used to reinforce existing residential racial segregation and to further the wealth holdings of predominantly white building owners and mortgage lenders of Black-owned homes (Michney and Winling, 2020). Neighborhoods were appraised in more than 200 cities between 1935 and 1940, with factors like population dynamics, economic class, percent of foreign families, percent of Black families, building characteristics, and area descriptions used for evaluation. Fig. 1 shows a digitized HOLC map from Mobile, Alabama, and Fig. 2 shows an excerpt from a neighborhood description sheet in Detroit, Michigan, made available by the University of Richmond's Digital Scholarship Lab in a collaboration with three other universities (Nelson et al., 2021). Neighborhoods were given one of four grades: 1st – “Best,” 2nd – “Still Desirable,” 3rd – “Definitely Declining,” or 4th – “Hazardous.” The maps displayed this information by color-coding neighborhoods according to their grades. In this paper, as is common practice with these maps, we refer to the grades using a letter or color as follows: 1st: A, green; 2nd: B, blue; 3rd: C, yellow; 4th: D, red. In this paper, we investigate differences in long term outcomes by neighborhood grade, building on existing work to understand connections between discriminatory housing practices and outcomes on individuals, neighborhoods, and racial groups.
Unfortunately, discrimination by those in positions of power on the basis of race, ethnicity, and religion are not exclusive to the United States. Case studies of Johannesburg, South Africa, after the end of apartheid, or the systematic government separation of racial groups from 1948 to 1964 (Seekings, 2008), found limited success had been achieved in racial equality and integration, even under a Constitution advocating for inclusiveness and equality (Fincher et al., 2014). Like in the United States, deeply entrenched systems of racial hierarchy and exclusion are difficult and slow to extinguish, and represent a fight against the status quo. In housing and planning, changes to status quo can be particularly difficult to implement (Carey & Harvey, 1975). Additional examples of present day racial inequality and discrimination in housing outside the United States are explored further in the discussion.
There are many influences on how neighborhoods may develop and change over time. Some of the key influences relevant to this study include changing lending practices and policy, neighborhood segregation, gentrification, and environmental hazards.
A case-study of Sacramento, California detailed three primary epochs of redlining practices by using HOLC maps and spatial data by census tract, extending beyond the introduction of the 1968 Fair Housing Act, which made housing discrimination on the basis of race illegal (Civil Rights Act of 1968, 1968; Hernandez, 2009). The three epochs described indicate that the first redlining period of the 1930s–1950s held the most brazen discriminatory practices, including from the HOLC, while the 1950s–1980s were a time of urban renewal programs, highway construction projects, and continued redlining of integrated neighborhoods, practices which were also damaging to many communities. The final period explored by the study, the 1980s–2000s, included subprime lending with higher interest rates concentrated in areas with more racial minorities (Hernandez, 2009). Other efforts to understand the history of residential segregation by race in the United States described three periods from a different perspective (Kramer and Hogue, 2009). The first, from 1890 to 1940, initiated segregated areas for Black residents in urban environments. In 1890, the average Black urban resident lived in a neighborhood with 27 percent Black residents, and in 1940 that number had increased to 43 percent (Kramer and Hogue, 2009). The second period, from 1940 to 1970, saw the increase of segregation due to policy and practices detailed in the previous section, along with increasing Black urban populations and racial tensions (Kramer and Hogue, 2009). In 1970, what some consider the peak of neighborhood racial segregation, the average urban Black person lived in a neighborhood that was 68 percent Black (Kramer and Hogue, 2009). In this most recent era, the introduction of the Fair Housing Act was intended to stop racial discrimination in housing and encourage integration, though its impacts have been limited (Larkin, 2007; Massey, 2015; Robinson, 1995). In some areas, most predominantly in the South and West US, integration has been primarily of middle-class Black families into historically white neighborhoods, while lower income Black families have become more isolated and faced infrastructure disinvestment (Kramer and Hogue, 2009). Theories about why integration has been slow to occur despite legislation encouraging it include preference by racial groups to live in neighborhoods that already have some threshold of residents of their racial groups, along with de facto racial discrimination still in place (Schelling, 1969).
Gentrification is one mechanism by which neighborhood racial composition and economic status change over time, though explicitly defining gentrification can be complicated (Easton et al., 2020). Some have distinguished gentrification as conflict and displacement, separate from positive terms like redevelopment, renewal, revitalization, regeneration, and reinvestment, but this distinction is debated by others (Newman and Wyly, 2006). Classical definitions referred to gentrification as the invasion of the middle classes into the working classes, renovating residences and taking over the area until the social character is changed, while more modern definitions have expanded to include new-build development alongside renovations as a part of this process (Davidson and Lees, 2005). In this work, we reference a measure of gentrification based on: increase in residents holding bachelor's degrees, increased median home value (inflation adjusted), and home value percentage increase relative to the metro area (Freeman, 2005; Gentrification in America, 2020). According to Newman & Wyly (Newman and Wyly, 2006), one of the most common outcomes of displacement due to gentrifying is for residents who are pushed out of their homes to double or triple up with friends or family, which may explain why some gentrified area demographics have yet to change as much as may be expected.
The first published study documenting the link between health and racial residential segregation appeared in 1950 (Yankauer, 1950), and connections between race, housing, and health have been studied by many fields since (Mohai and Bryant, 1991; Institute of Medicine (US) Committee on Environmental Justice, 1999; Northridge et al., 2003). One of the major links comes from exposure to environmental toxins, including industrial pollution, toxic materials, landfills, and incinerators, and exposures to these hazards are not equally distributed by racial group in the US (Kramer and Hogue, 2009). In a 1988 study of race and ecological disparities, Black children were 2–3 times more likely than white children to suffer from lead poisoning, after controlling for income (Bullard, 1993). A 1987 study found race to be among the most significant factors associated with the location of commercial hazardous waste facilities, with 60 percent of Black and Hispanic residents living in communities with one or more uncontrolled toxic site. While there is plenty of evidence demonstrating racial disparities in exposure to environmental hazards, the direction of causality is less clear (Cutter, 1995; Fincher and Iveson, 2012). However, whether the siting decisions happened before the residents moved in or after, our understanding of the de jure and de facto segregation demonstrates the challenges of Black and immigrant populations seeking to move away from these environmental hazards.
Even with insight on the lending and appraisal practices of the twentieth century and neighborhood dynamics, we still have more to learn about how discriminatory practices over time connect to present day outcomes. We know the importance of financial institutions on the wellbeing of a community, in particular for their ability to provide access to equity and generational wealth (Chetty and Hendren, 2016; Coates, 2014; Squires and Velez, 1987). The president of the National Community Reinvestment Coalition called homeownership the “number-one method of accumulating wealth,” (Jan, 2018) yet the homeownership gap between white and Black Americans is now larger than during the Jim Crow era (Glantz & Martinez, 2018). Several studies have explored modern mortgage and housing outcomes and their relation to both HOLC redlining and general discriminatory practices (Aaronson, Hartley, & Mazumder, 2018; Brownlow, 2006; Pulido, 2000). For example, a 1987 study of Milwaukee, Wisconsin found a “strong bias in favor of suburban and white neighborhoods against inner-city and minority communities” when evaluating homeowner's insurance policies from 1985 and suggested that the interactions of redlining and race were prominent even when controlling for socioeconomic indicators (Squires and Velez, 1987). This work compared policy reports, major insurance company earnings, and demographic data over zip codes and used a correlation matrix to understand relationships between their datasets. A study comparing outcomes including income rank, family structure, incarceration rates, and geographic mobility found large and significant causal effects of HOLC boundaries for cohorts born in the 1970s and 1980s (Aaronson et al., 2021). Significant work has also been contributed in this field by Amy Hillier who has explored spatial effects of redlining, particularly on mortgages and interest rates (Hillier, 2003a, 2003b, 2005a, 2005b).
There have also been attempts to understand links between health and employment outcomes to redlining practices and housing segregation. In particular, links between housing discrimination through segregation and health disparities between Black and white Americans have been demonstrated over a number of health outcomes (Williams and Collins, 2001). This disparity increased between the 1950s and the 1990s for mortality due to coronary heart disease, cancer, diabetes, and infant mortality (Williams, 1999; National Center for Health Statistics US, 2000). An additional study of the Milwaukee, Wisconsin metropolitan area explored the relationship between redlining and breast cancer survival rates in Black women relative to white women and found mixed results indicating a possible but uncertain connection between the two (Beyer et al., 2016). Numerous studies of pregnancy outcomes have found racial disparities related to residential segregation (Bell et al., 2006; Collins et al., 1998, 2004; Yankauer, 1950). Employment rates have also been linked to both racial discrimination and spatial discrimination, with more and better opportunities often located outside of the city center and away from concentrations of Black populations (Zenou and Boccard, 2000). When controlling for education and skills, Black persons in less segregated cities are less likely to be unemployed than those living in more segregated cities (Dickerson, 2007). Increased residential segregation has been linked to consuming less fruit (Dubowitz et al., 2008) and being less physically active (Lopez, 2006). Interested readers can find additional research into the spatial effects of discrimination for particular region-outcome case studies in Appel and Nickerson (2016); Farley et al. (1997); Jacoby et al. (2018); Zhang and Ghosh (2016). One of the most comprehensive studies we have found on the outcomes related to HOLC redlining was undertaken by the National Community Reinvestment Coalition. They studied 115 cities and compared economic indicators and racial composition and found clear evidence of persistent inequality over both measures in areas once drawn red by the HOLC (Mitchell & Franco, 2018). Similarly, a nationwide comparison of HOLC boundaries and land surface temperatures revealed that 94 percent of the 108 urban areas studied have higher average land surface temperatures in their D rated areas than A rated areas (Hoffman et al., 2020). This work extends current research by providing a broader view centered on HOLC redlining districts; we seek spatial trends over wide-ranging cities and outcome variables rather than focusing on one specific location or outcome case-study.
Section snippets
Data
This study incorporates data from four sources to achieve a broad view of the connections between the redlining system and current state of 14 cities where HOLC maps were available across the United States. We selected three of the largest US cities, three of the most gentrified large cities, three of the least gentrified large cities, and two of the most redlined cities. We supplemented this list with four additional cities to provide wider geographic coverage and varied political climates. A
Results
Because of the complex history that has influenced all of our outcome variables, we do not presume that the HOLC was the sole cause of any of the outcomes presented (i.e. this is not a pure causal analysis). However, the data shows us clear evidence that the pockets of inequality observed do closely follow historic HOLC lines, which indicates that there is likely a relationship.
Discussion
Our study represents just one piece of a larger understanding of how inequality is maintained and distributed in the US by investigating its patterns through a static spatial lens. We recognize that the conversation is incomplete without also acknowledging the dynamic nature of this topic with factors including gentrification, new policy, and increased neighborhood mobility creating change, both positive and negative, along the historic neighborhood lines we analyzed (Glantz & Martinez, 2018;
Conclusion
Although our results are in line with the expectations of most scholars within the social justice community, the severity and consistency of inequity found is striking. Black populations were historically prevented from living in A, B, and most C graded neighborhoods, while neighborhoods with Jewish and some immigrant populations were typically appraised as no better than B or C neighborhoods. Today, we still generally see these same A, B, and C graded neighborhoods populated with more white
Author contribution
AG White: Conceptualization, data curation, formal analysis, investigation, methodology, software, validation, visualization, writing – original draft preparation, writing – review and editing.
SD Guikema: Conceptualization, funding acquisition, investigation, methodology, resources, supervision, validation, writing – review & editing.
TM Logan: Conceptualization, data curation, writing – review & editing.
Acknowledgement
This work was partially funded by University of Michigan through Rackham Merit Fellows program and the NSF through grant #1638197. This support is gratefully acknowledged. The findings and opinions are those of the authors and do not necessarily represent those of the funding agencies.
References (100)
- et al.
The long-run effects of the 1930s HOLC “redlining” maps on place-based measures of economic opportunity and socioeconomic success
Regional Science and Urban Economics
(2021 Jan 1) - et al.
Structural racism and health inequities in the USA: Evidence and interventions
The Lancet
(2017 Apr 8) - et al.
Birth outcomes among urban African-American women: A multilevel analysis of the role of racial residential segregation
Social Science & Medicine
(2006) - et al.
New spatially continuous indices of redlining and racial bias in mortgage lending: Links to survival after breast cancer diagnosis and implications for health disparities research
Health & Place
(2016 Jul 1) An archaeology of fear and environmental change in Philadelphia
Geoforum
(2006 Mar 1)- et al.
Individual and neighborhood differences in diet among low income foreign and US-born women
Women's Health Issues
(2008) - et al.
Planning in the multicultural city: Celebrating diversity or reinforcing difference?
Progress in Planning
(2014 Aug 1) - et al.
The enduring impact of historical and structural racism on urban violence in Philadelphia
Social Science & Medicine
(2018 Feb 1) - et al.
The dynamics and implications of residential segregation in Nairobi
Habitat International
(2007 Mar 1) - et al.
Racial discrimination and redlining in cities
Journal of Urban Economics
(2000 Sep 1)
Validation of multilevel regression and poststratification methodology for small area estimation of health indicators from the behavioral risk factor surveillance system
American Journal of Epidemiology
The effects of the 1930s HOLC “redlining” maps
The Federal Reserve Bank of Chicago
Pockets of poverty: The long-term effects of redlining
Race and environmental justice in the United States
Yale J. Int'l L.
Definitions: Labor force
Geographical Review
Racial discrimination widespread in the private rental market, advocates warn
ABC News AU. [Internet]
Overweight and obesity: Causes and consequences
Underlying cause of death 1999-2013 on CDC WONDER online database, released 2015. Data are from the multiple cause of death files, 1999-2013, as compiled from data provided by the 57 vital statistics jurisdictions through the vital Statistics cooperative program
National center for chronic disease prevention and health promotion, division of population health. [cited 2018May 10]
Database: 500 Cities Project Data
Cigarette smoking among adults--United States, 2000
MMWR. Morbidity and Mortality Weekly Report
The impacts of neighborhoods on intergenerational mobility: Childhood exposure effects
Quarterly Journal of Economics
Urban segregation in post-apartheid South Africa
Urban Studies
Pub.L 90-284, 90th Congress
The case for reparations
The Atlantic
Very low birthweight in African American infants: The role of maternal exposure to interpersonal racial discrimination
American Journal of Public Health
African-American mothers' perception of their residential environment, stressful life events, and very low birthweight
Epidemiology
The legacy of the home owners' loan corporation
Housing Policy Debate
Race, class and environmental justice
Progress in Human Geography
New-build ‘gentrification’ and London's riverside renaissance
Environment & Planning A
Black employment, segregation, and the social organization of metropolitan labor markets
Economic Geography
Land and housing as crucibles of racist nationalism: Asian Australians' experiences
Geographical Research
Measuring and mapping displacement: The problem of quantification in the battle against gentrification
Urban Studies
Light Gray Canvas map” [basemap]. Scale not given. ”World Light Gray Canvas Base”
The residential preferences of blacks and whites: A four[U+2010]metropolis analysis
Housing Policy Debate
Justice and injustice in the city
Geographical Research
Displacement or succession? Residential mobility in gentrifying neighborhoods
Urban Affairs Review
Urban gentrification: Evaluating alternative indicators
Social Indicators Research
Type 2 diabetes and persons at high risk of diabetes
Gentrification in America report methodology
Modern-day redlining: How banks block people of color from homeownership
Chicago Tribune
Racialization and the state: The housing act of 1934 and the creation of the federal housing Administration
Sociological Perspectives
Race and mortgage redlining in the United States
Redlining revisited: Mortgage lending patterns in Sacramento1930–2004
International Journal of Urban and Regional Research
raster: Geographic data analysis and modeling. R package version2.6-7
Spatial analysis of historical redlining: A methodological explanation
Departmental Papers (City and Regional Planning)
Redlining and the home owners' loan corporation
Journal of Urban History
Searching for red lines: Spatial analysis of lending patterns in Philadelphia, 1940–1960
Pennsylvania History: A Journal of Mid-Atlantic Studies
Residential security maps and neighborhood appraisals: The home owners' loan corporation and the case of Philadelphia
Social Science History
The effects of historical housing policies on resident exposure to intra-urban heat: A study of 108 US urban areas
Climate
Cited by (9)
Party affiliation and the legacy of HOLC redlining maps
2024, Political GeographyHistoric redlining and health outcomes: A systematic review
2024, Public Health NursingWhy are You Here? Modeling Illicit Massage Business Location Characteristics with Machine Learning
2024, Journal of Human Trafficking