Redlined then, shorter lives now
Do neighbourhoods the 1930s appraisers graded “hazardous” still live shorter lives?
Federal appraisers graded city neighbourhoods A to D in the late 1930s. Beside each 1930s map, the same extent today: census tracts coloured by life expectancy at birth. Point at a place to outline it on both maps.
In Nashville, people in the 7 tracts graded A now live 82.6 yrs on average; in the 23 tracts graded D, 70.4 yrs — a gap of 12.2 years (population-weighted).
1930s HOLC grades
Neighbourhoods as the appraisers graded them, 1935–40.
- A “Best”
- B “Still desirable”
- C “Definitely declining”
- D “Hazardous”
Life expectancy today
2020 tracts that are mostly inside graded land, 2010–15 estimates.
Point at a 1930s polygon or a 2020 tract: its outline appears on both maps. Tab into a map to step through places with the arrow keys.
The grade is a tendency, not a sentence: every graded tract in the country
Boxes span the middle half of tracts in each grade, whiskers the 5th to 95th percentile, the thick tick the median; dots are the payload's fixed sample of 220 tracts per grade. A tract you point at on the map lands here on its grade's row.
Where the shadow is longest: 29 cities
How to read the maps. Both panels are fitted to the same extent through one projection, so a place sits at the same spot in each; pointing at either map outlines the same place on both. The left map is the Home Owners' Loan Corporation's security grades as digitized by Mapping Inequality. The right map shows 2020 census tracts that the build assigned a grade because at least half their land lies inside graded polygons; other tracts are not drawn. Each tract takes the grade that covers most of its land. Graded tracts with no life-expectancy estimate were left out of the book's city maps, so they show as gaps rather than as no-data fills.
Measures. Life expectancy at birth is the NCHS small-area estimate (USALEEP) for 2010–15, carried onto 2020 tracts. The income rank is the Opportunity Atlas's mean adult household-income percentile for children raised at the 25th percentile, birth cohorts 1978–83. City gaps compare population-weighted life expectancy in A- and D-assigned tracts, for cities with at least five A and ten D tracts. Gaps here are the difference of the rounded endpoints; the book's published gaps, taken before rounding, differ by 0.1 in five cities (Nashville: 12.3) and are listed in the table. In the dumbbell's tooltip, “Earlier” is the D value and “Later” the A value.
What this cannot show. The appraisers drew red lines around neighbourhoods that were already poorer, more crowded and more often Black, so the gap mixes whatever the maps did with the disadvantage they recorded. Boundary studies find real but smaller causal effects. This is the descriptive ledger, not a causal estimate.
HOLC polygons: Mapping Inequality, University of Richmond Digital Scholarship Lab, CC BY-NC-SA 4.0. Life expectancy: NCHS USALEEP. Mobility: Opportunity Atlas (Chetty, Friedman, Hendren, Jones & Porter). Grades and colours are the appraisers' ranking, reproduced to show what they did.
Nashville tracts · 41 rows
| Tract (GEOID) | 1930s grade | Life expectancy (years) | Adult income rank of children raised poor (percentile) |
|---|---|---|---|
| 47037017701 | A | 79.6 | 47 |
| 47037017800 | A | 80.1 | 45 |
| 47037017901 | A | 81.9 | 43 |
| 47037017902 | A | 82.3 | 58 |
| 47037018500 | A | 83.7 | 47 |
| 47037018000 | A | 84.6 | 47 |
| 47037017702 | A | 84.9 | 50 |
| 47037011400 | B | 71.6 | 30 |
| 47037011700 | B | 73.6 | 31 |
| 47037011200 | B | 74.4 | 33 |
| 47037016800 | B | 79.8 | 46 |
| 47037016700 | B | 83.3 | 43 |
| 47037016900 | B | 84 | 38 |
| 47037011300 | C | 70.8 | 29 |
| 47037016600 | C | 72.9 | 41 |
| 47037017100 | C | 77.8 | 31 |
| 47037012100 | C | 79 | 35 |
| 47037017000 | C | 79 | 35 |
| 47037019300 | D | 66.3 | 27 |
| 47037017200 | D | 67.6 | 37 |
A-vs-D gap by city · 29 rows
| City | A tracts (years) | D tracts (years) | Gap shown (years) | Published gap (years) | A tracts | D tracts |
|---|---|---|---|---|---|---|
| Nashville | 82.6 | 70.4 | 12.2 | 12.3 | 7 | 23 |
| Louisville | 80.5 | 70.5 | 10 | 10 | 6 | 15 |
| Milwaukee Co. | 84.7 | 74.8 | 9.9 | 9.9 | 5 | 37 |
| Indianapolis | 81.1 | 71.3 | 9.8 | 9.8 | 7 | 38 |
| Pittsburgh | 82 | 73.1 | 8.9 | 8.9 | 7 | 33 |
| Cleveland | 80.3 | 72 | 8.3 | 8.3 | 24 | 44 |
| Columbus | 78.9 | 71.5 | 7.4 | 7.4 | 10 | 13 |
| Oakland | 85.2 | 77.8 | 7.4 | 7.4 | 11 | 39 |
| Chicago | 83.3 | 76.5 | 6.8 | 6.8 | 13 | 213 |
| Oklahoma City | 78.1 | 71.2 | 6.9 | 6.8 | 9 | 11 |
| St. Paul | 82.5 | 75.7 | 6.8 | 6.8 | 6 | 12 |
| Des Moines | 79.7 | 73 | 6.7 | 6.7 | 7 | 12 |
| St. Louis | 80.4 | 73.8 | 6.6 | 6.6 | 17 | 15 |
| Miami | 82.4 | 75.9 | 6.5 | 6.5 | 8 | 31 |
| Baltimore | 78.3 | 72.4 | 5.9 | 5.9 | 11 | 37 |
| Essex Co. | 82.3 | 76.6 | 5.7 | 5.7 | 22 | 37 |
| Philadelphia | 79.6 | 73.9 | 5.7 | 5.7 | 27 | 129 |
| San Antonio | 79.9 | 74.2 | 5.7 | 5.7 | 6 | 23 |
| Minneapolis | 81.7 | 76.2 | 5.5 | 5.5 | 8 | 17 |
| Fort Worth | 80.3 | 75 | 5.3 | 5.4 | 5 | 16 |
National tract distribution by grade · 4 rows
| Grade | 5th pct | 25th pct | Median | 75th pct | 95th pct | Tracts | Population-weighted mean |
|---|---|---|---|---|---|---|---|
| A “Best” | 73.5 | 78.3 | 81.1 | 83.4 | 86.5 | 673 | 80.91 |
| B “Still desirable” | 71 | 75.8 | 79 | 81.7 | 85.7 | 2,247 | 79.11 |
| C “Definitely declining” | 69.5 | 74.4 | 77.8 | 80.8 | 84.3 | 5,061 | 77.95 |
| D “Hazardous” | 67.8 | 72.4 | 76 | 79.5 | 83.6 | 2,986 | 76.78 |
Mapping Inequality (University of Richmond DSL); NCHS USALEEP 2010–15; Opportunity Atlas · ZIP Code Destiny, “The Ninety-Year-Old Map”
Reading notes
Descriptive, not causal: appraisers drew red lines around neighbourhoods that were already poorer and more often Black. Each 2020 tract takes the grade covering most of its land; tracts without a life-expectancy estimate are left out and show as gaps, not no-data fills.
Source & method
Kept the national tract life-expectancy distribution by HOLC grade (5th/25th/50th/75th/95th percentiles and the payload's fixed-seed sample of 220 tract estimates per grade, both over the 10,303 of 10,967 graded tracts that have an estimate), the 29 cities' population-weighted life expectancy in A- and D-assigned tracts with tract counts and the published gap, and each grade's national tract count, population, life expectancy and kfr. Dropped income, poverty, homeownership, ADI and CDC PLACES outcomes, which this figure does not draw. The city list (with each city's gap looked up by name) indexes the three per-city geometry fixtures. Values are as published, rounded to 1–3 decimals; the published gap is computed from unrounded estimates, so it can differ by 0.1 from the difference of the rounded endpoints.
Capture provenance
- box:birth_death / articles/_build/data/s16_redline.json
- Modified 2026-06-12 · SHA-256 39eb131afbcbb71671a6dc70759873d1f7bffd286ebd1092409a971dd3a490b0
- Captured
- 2026-09-12
- Fixture SHA-256
- c426e3be757f11a50abe345d20f0812ce9c27d191f347999709edd4988a9e6a7
- Reuse note
- HOLC polygons: Mapping Inequality: Redlining in New Deal America, University of Richmond Digital Scholarship Lab (Nelson, Winling et al.), distributed under CC BY-NC-SA 4.0 — attribute on every use, non-commercial only, share derivatives alike (licence to be re-verified against the project's current terms). Tract life expectancy: NCHS U.S. Small-Area Life Expectancy Estimates Project (USALEEP), 2010–2015, public domain. Upward mobility (kfr): Opportunity Atlas, Chetty, Friedman, Hendren, Jones & Porter (2018), public release. Tract assignment and aggregates: the book's build (ZIP Code Destiny, story 16).
