Association for Public Policy Analysis & Management

Eighty years ago, the Home Owners’ Loan Corporation (HOLC) created “Residential Security” maps of major American cities. The maps have been interpreted by some as an indication of federal complicity in “redlining” practices (Jackson 1987; Massey and Denton 1993). Following the work of Hillier (2003) and Crossney and Bartelt (2005) we interpret the HOLC maps as documents of how loan officers, appraisers, and real estate professionals evaluated mortgage lending risk during the era immediately before the surge of suburbanization in the 1950’s. Neighborhoods considered high risk, or “hazardous” were often “redlined” by lending institutions, denying them access to capital investment which could improve the housing and economic opportunity of residents. This study examines how neighborhoods were evaluated for lending risk in the HOLC maps, then compares their recent social and economic conditions with city-level measures of segregation and economic inequality.

Spatial analysis of the HOLC residential security maps and recent census data was performed using GIS raster calculations. Digitized images of the original HOLC maps were used to compute the percentage of area in the original maps of each city graded A, B, C, or D (corresponding to green =“Best”; blue = “Still Desirable”; yellow = “Definitely Declining”; red = “Hazardous” designations respectively). The digitized HOLC classification areas were then overlaid and compared with map layers reflecting the current economic status and racial/ethnic composition of census tracts in those areas using raster analysis. For economic status, the median family income (low-to-moderate and middle-to-upper) based on Federal Financial Institutions Examinations Council (FFIEC) 2016 criteria was used, adjusted by the median family income of the MSA for each city in the analysis. The racial/ethnic composition was quantified by taking the Non-Hispanic white population of each census tract into consideration and classifying it as either majority white, or majority-minority. Statistical analysis using ordinary least squares regression was then used to compare MSA level measures of segregation (Massey and Denton 1988, 1989), and economic inequality (Gini index) with the level of economic and demographic change in the HOLC classified areas.

Results of the analysis indicate that nationwide, 74% of the areas graded “hazardous” by the HOLC are low-to-moderate income, and 64% are majority-minority now. In the HOLC “best” graded areas, 91% are middle-to-upper income and 86% majority Non-Hispanic white. Regional factors were evident in the distribution of cities with less change in HOLC graded areas. Cities in the South (80.41%) and Midwest (81.22%) have the highest percentage of “hazardous” areas that are low-to-moderate income. In terms of demographics, the South had the most HOLC “hazardous” graded areas that are now majority-minority (72.07%). Suggesting that the most economically disadvantaged and segregated Southern cities have undergone less change than other regions of the country. Bivariate correlations suggest significant relationships between less change in the HOLC D graded neighborhoods and indicators of Black and Hispanic segregation, and greater economic inequality.