Surprising findings
Short data-journalism pieces built on a stated, checkable comparison — where documented health most departs from what income, geography, or the surrounding data would predict. Every finding names its model and its uncertainty, ships its data as a CSV, and keeps a frozen permalink per data release.
4 findings published · Refreshed with each data release; new findings added as the data earns them
Estimated vs expected
The ZIP codes that outlive their incomes
Income predicts about a quarter of the variation in ZIP-level life expectancy — 2.7 more expected years per income doubling. These are the places that most escape that line, in both directions.
CDC USALEEP 2010–2015 · Census ACS 2020–2024 · 28,882 ZIP codes compared
Neighbor contrast
The ZIP codes that outlive their neighbors
Most ZIP codes sit within ±2.5 years of the ZIP codes around them. This page names the ones that most break from their surroundings — in both directions, tested against the combined uncertainty of both estimates.
CDC USALEEP 2010–2015 · ZipVitals neighbor data · 31,308 ZIP codes compared
Risk-scale contrast
Air toxics and the years between the two ends of the risk scale
The tenth of ZIP codes with the most EPA-modeled air-toxics cancer risk lives 1.36 fewer years than the 7,132 ZIP codes with the least — and the higher-risk tenth is also the higher-income group ($85,553 vs $78,107). Two records about the same places, described together; never a causal claim.
EPA AirToxScreen · CDC USALEEP 2010–2015 · 21,366 ZIP codes compared
Two published standings, side by side
Where the environment and chronic illness point in opposite directions
Some ZIP codes stand in the cleanest tenth of the country on the environmental-hazard domain and the heaviest tenth for chronic disease — and dense city cores mirror it. The same domain standings every ZIP page publishes, read against each other.
ZipVitals domain standings · data release 2026 · 24,186 ZIP codes compared
How a finding earns this page
A stated model, honestly scored
Each finding declares its comparison in one sentence and publishes the fit — including when it is modest. Every table applies a population floor, and where a finding tests a gap against uncertainty — the life-expectancy findings — a ZIP code is named only when its gap is larger than the uncertainty of its own estimate. Each finding’s ledger states the exact gates it applies.
Frozen citations that never move
Cite a finding today and the permalink resolves to exactly these numbers years from now, beside an always-latest version. When a data release moves a finding materially, the finding is re-reviewed before the latest page changes.