ZIP Codes and the Legacy of Redlining: How 1930s Maps Shape Today
Last updated · History & Equity
Between 1935 and 1940, the Home Owners' Loan Corporation (HOLC) created "residential security maps" for 239 US cities. These maps graded neighborhoods A through D — with D-rated areas ("redlined") deemed "hazardous" for mortgage lending. The grades were based heavily on racial composition. Ninety years later, the University of Richmond's Mapping Inequality project digitized every surviving map. When you overlay the HOLC grades onto modern ACS data, the correlations are staggering and statistically significant.
What the HOLC maps actually said
The HOLC appraisers used four grades:
- A (green) — "Best": new construction, homogeneous White populations, professional-class residents, deed restrictions preventing "undesirable" owners.
- B (blue) — "Still Desirable": older but well-maintained, "stable" populations, slight signs of transition.
- C (yellow) — "Definitely Declining": older housing, "infiltration" of foreign-born or lower-income residents, declining property values.
- D (red) — "Hazardous": predominantly Black, immigrant, or low-income populations, older housing, near industrial areas. These neighborhoods were effectively cut off from conventional mortgage lending.
The language in the original appraisal notes explicitly cited race as a negative factor. A neighborhood could have identical housing stock to an A-rated area and receive a D grade based solely on the racial composition of its residents.
Income disparities: then and now
A 2024 study by the National Community Reinvestment Coalition found that formerly redlined areas still have median household incomes 37% lower than formerly A-rated areas in the same cities. The specific data:
- Former A-rated areas: median household income $98,400 (ACS 2019-2023)
- Former B-rated areas: $81,200
- Former C-rated areas: $62,800
- Former D-rated areas: $48,100
This gradient persists even after controlling for city, region, and metro size. The gap has narrowed slightly since 2010 (when it was 42%) but remains one of the most durable geographic inequalities in the country.
In some cities, the correlation is almost mechanical: Detroit, Chicago, Philadelphia, and Baltimore show income gradients that follow HOLC boundaries with block-level precision.
Health outcomes follow the same lines
The CDC Places dataset (Census tract-level health estimates) shows that formerly redlined areas have:
- Life expectancy: 3.6 years shorter than former A-rated areas in the same metro.
- Diabetes prevalence: 14.2% in former D areas vs 8.1% in former A areas.
- Asthma prevalence: 11.8% vs 7.9%.
- Lack of health insurance: 12.4% uninsured vs 5.1%.
Researchers attribute this partly to disinvestment (fewer grocery stores, more food deserts, less green space, more pollution sources like highways and industrial zones) and partly to the cumulative stress of concentrated poverty. The Robert Wood Johnson Foundation estimates that ZIP code is a stronger predictor of health outcomes than genetic code.
Educational attainment and school quality
ACS educational attainment data mapped onto HOLC grades shows:
- Bachelor's degree or higher: 52% of adults in former A areas vs 19% in former D areas.
- School funding: because school funding in most states is tied to local property taxes, lower property values in formerly redlined areas translate directly to lower per-pupil spending. The gap averages $2,300 per student per year nationally.
- School performance: Stanford Education Data Archive shows that test score gaps between former A and D neighborhoods average 1.5 grade levels in both math and reading.
This creates a self-reinforcing cycle: lower property values fund fewer school resources, which produce worse outcomes, which attract fewer families with educational attainment, which depresses property values further.
Digital redlining: the modern version
Research from The Markup and academic institutions has documented that modern infrastructure investment follows patterns eerily similar to HOLC maps:
- Broadband: FCC data shows that formerly redlined neighborhoods are 3x more likely to have only one ISP option, compared to 1.5x for former A areas. Competition drives speed and price; monopoly depresses both.
- Fiber deployment: ISPs like AT&T and Comcast have been documented prioritizing fiber rollout in higher-income neighborhoods. In Cleveland, former A areas had 89% fiber availability vs 34% in former D areas (The Markup, 2022).
- Algorithm-based lending: even without explicitly using race, machine learning models trained on ZIP-level data can replicate redlining patterns because the input variables (income, home value, credit scores) are themselves products of historical discrimination.
What is being done and what the data suggests
Several policy responses target the legacy of redlining:
- Community Reinvestment Act (CRA): requires banks to serve the communities where they operate, including low-income areas. CRA-motivated lending has driven $2.1 trillion into underserved Census tracts since 1996. The OCC proposed updated CRA evaluation criteria in 2023.
- Opportunity Zones: 8,764 Census tracts designated for tax-advantaged investment. Overlap with formerly redlined areas is significant (about 60% of Opportunity Zones contain former D-rated tracts).
- CDBG and HOME programs: HUD block grants that target low-income Census tracts for housing rehabilitation, infrastructure, and economic development.
The data shows that targeted investment does narrow gaps — tracts receiving sustained CRA lending show 12-18% faster income growth than similar tracts without it. But the baseline disparities are so large that convergence at current rates would take 60-80 years.
Frequently Asked Questions
What is redlining?+
Redlining refers to the practice of denying or limiting financial services (particularly mortgage lending) to residents of specific neighborhoods based on racial or ethnic composition. The term comes from the red color used on HOLC maps to mark neighborhoods graded D ("Hazardous"), which were overwhelmingly Black, immigrant, or mixed-race communities.
Is redlining still legal?+
Explicit racial redlining has been illegal since the Fair Housing Act of 1968. However, research shows that algorithmic lending, insurance pricing, and infrastructure investment decisions can produce similar geographic patterns without explicitly using race, because the proxy variables (ZIP code, home value, credit score distributions) carry the legacy of historical discrimination.
How can I check if my neighborhood was redlined?+
The University of Richmond Mapping Inequality project (dsl.richmond.edu/panorama/redlining) has digitized and published every surviving HOLC map. You can search by city and see the original A-D grades overlaid on modern maps. HOLC maps cover 239 cities; smaller cities and rural areas were not mapped.
Do all formerly redlined areas still show disparities?+
Most do, but not all. Gentrification has transformed some formerly redlined neighborhoods — parts of Brooklyn, San Francisco, Washington DC, and Portland now have higher incomes and home values than their former HOLC grade would predict. However, gentrification typically displaces the original residents rather than lifting them economically, so the people affected by the original redlining may not benefit.
How does redlining affect home values today?+
A 2018 Brookings Institution study found that homes in formerly redlined neighborhoods are valued at 52% less than comparable homes in non-redlined neighborhoods of the same city, even after controlling for home size, lot size, and age. This appraisal gap represents billions of dollars in lost household wealth for Black families.
What is the difference between redlining and blockbusting?+
Redlining prevented lending in minority neighborhoods. Blockbusting was the practice of real estate agents buying homes cheaply in White neighborhoods by stoking racial fear, then selling at a premium to Black families desperate for housing. Both were tools of residential segregation, but they operated in opposite directions — redlining kept people out, blockbusting exploited those trying to get in.