Outcome story · smoking

Smoking: deprivation predicts it, culture bends it

No measure tracks neighborhood deprivation more tightly than smoking (ρ = 0.80). Subtract that prediction and what remains is a map of history: the tobacco belt, Utah, and the casino frontier.

Of all 26 measures, none follows neighborhood deprivation as tightly as cigarette smoking: ρ = 0.8with the Area Deprivation Index across ZIP codes. Know a neighborhood's ADI and you can predict its smoking rate remarkably well. Which makes the failuresof that prediction the interesting part — they isolate everything deprivation can't explain.

Smoking against deprivation

Each dot is a ZIP code; the line is the quadratic fit used to compute residuals — hover for the ZIP

Loading…
The relationship steepens at the deprived end: each step deeper into deprivation buys more additional smoking. Residuals below measure each ZIP's distance from this line.

Subtract poverty, and history remains

Map the residual — actual smoking minus deprivation-predicted smoking — and the income map of America disappears, replaced by a cultural one. The contiguous red mass is the tobacco belt: Tennessee (+2.8 points), Kentucky, West Virginia — places that grew tobacco, worked it, and smoke more than their economics alone would predict. Nevada and Alaska join them by a different route (casino and frontier culture). On the blue side, Utah smokes 2.6 points less than its deprivation predicts — the LDS effect — joined by heavily Hispanic border areas and immigrant metros, consistent with the well-documented immigrant smoking advantage.

The residual map: smoking minus what deprivation predicts

Red = smokes more than predicted · blue = less

This is the map of smoking culture — regional norms, religion, tobacco heritage, and policy (tax differentials between neighboring states are visible along several borders).

State residuals, ranked

Population-weighted average residual (percentage points of smoking prevalence)

NV +3.3, AK +3.1, TN +2.8 smoke the most above prediction; UT -2.6, CT -1.8, IL -1.8 the most below. For a measure this strongly determined by deprivation, multi-point state-level departures are large.

The policy reading: deprivation sets the baseline, but the residuals prove smoking is not economically fated. Utah's ZIP codes face the same prices and the same stress gradients as everyone else's — norms moved the number anyway. What culture can do, policy and cessation infrastructure have room to do too, and the red residual states are the map of where that headroom is largest.

Read this carefully. Estimates are CDC PLACES-style model-based small-area estimates, not direct measurements. Every association here is ecological — it describes places, not people, and implies nothing about causation. Cross-measure models are fit on the ~23,800 ZIP/ZCTA areas with complete data on all 26 measures (coverage is limited mainly by the newer social-needs measures); maps and community-type assignments extend to areas observing at least 18 of the 26. Full details on the methods page.