Chapter 9.2 · The eight Americas

Eight types of neighborhood

Mean life expectancy for deaths in 2010 to 2015 runs from 74.7 to 82.1 years across eight overlapping types of census tract, built by k-means from 81,343 tracts on 7 measures of income, poverty, education, family structure, tenure, age and density, with no outcome among them.

Author: Vishal Singh
Affiliation: NYU Stern
Published: 9 October 2026
Updated: 28 October 2026

Mean life expectancy for deaths in 2010 to 2015 is 82.1 (95% CI 81.9 to 82.3) years in the census tracts this book calls high-income owners and 74.7 (95% CI 74.2 to 75.1) years in the tracts it calls high-poverty, single-parent. Those are the two ends of eight types of tract, sorted by income, poverty, education, family structure, homeownership, age and density. No measure of health, mobility, race or ethnicity went into the sort, so the gap between the two ends, 7.5 (95% CI 7.1 to 7.9) years, is not built into the types. The names are the book's labels for groups of tracts that look alike on the sorting measures, and they label tracts, not the people who live in them.

The high-poverty, single-parent type is least favorable on all eight outcomes

The types come from 7 inputs measured on 81,343 census tracts. The American Community Survey (ACS) for 2019 to 2023 and tract land area supply the log of median household income, the poverty rate, the college share of adults, the single-parent share of households, the homeownership rate, the share of older residents and population density. Age and density describe how a place is settled, and they set older owners, low-density owners and dense, college-educated renters apart from the other types.

Each input is standardized across tracts, and each tract counts once. K-means then places eight centers, assigns every tract to the nearest and repeats until the assignments settle.

The chapter compares the types on eight outcomes, none of which went into the sort. The first four are life expectancy, the adult income rank of children raised in low-income families, and teen births and incarceration among those children. The other four are diabetes, smoking, frequent mental distress and obesity among adults. The high-poverty, single-parent type has the least favorable mean on all eight, and high-income owners have the most favorable on every outcome except diabetes, where dense, college-educated renters are lower. Among women born in 1978 to 1983 and raised in low-income families, the share who had a teen birth is 13.9% (95% CI 13.1% to 14.7%) in tracts of high-income owners and 38.9% (95% CI 37.8% to 40.0%) in the high-poverty, single-parent type.

Figure 1

The high-poverty, single-parent type has the least favorable mean on all eight outcomes, and high-income owners the most favorable on every outcome except diabetes

Population-weighted mean of tract values by neighborhood type with 95% intervals from a bootstrap over counties, each panel in its own units

Life expectancy and adult income rank are highest and teen births and incarceration lowest among high-income owners, and the reverse holds in the high-poverty, single-parent type

Life expectancy, the adult income rank, teen births and incarceration, population-weighted type means with 95% intervals

Life expectancy at birth (years)

Adult income rank of children from low-income families (percentile)

Teen births among daughters of low-income parents (%)

Incarcerated on Census Day 2010, children of low-income parents (%)

Population-weighted mean of tract values (units in each panel title)

Life expectancy and adult income rank are highest and teen births and incarceration lowest among high-income owners, and the reverse holds in the high-poverty, single-parent type. The order across the other types differs by outcome.

Show data

Data for The high-poverty, single-parent type has the least favorable mean on all eight outcomes, and high-income owners the most favorable on every outcome except diabetes, Life expectancy, mobility and justice
conf_high	conf_low	estimate	label	n	panel	type
82.30	81.93	82.13	High-income owners	10,184	Life expectancy at birth (years)	High-income owners
80.64	80.10	80.36	Older owners	3,221	Life expectancy at birth (years)	Older owners
79.73	79.47	79.60	Middle-income owners	16,921	Life expectancy at birth (years)	Middle-income owners
77.92	77.71	77.81	Low-density owners	15,201	Life expectancy at birth (years)	Low-density owners
80.62	79.92	80.31	Dense, college-educated renters	6,980	Life expectancy at birth (years)	Dense, college-educated renters
77.58	77.08	77.33	Moderate-income, mixed-tenure	14,393	Life expectancy at birth (years)	Moderate-income, mixed-tenure
75.82	74.88	75.34	Lower-income renters	6,268	Life expectancy at birth (years)	Lower-income renters
75.12	74.22	74.67	High-poverty, single-parent	4,595	Life expectancy at birth (years)	High-poverty, single-parent
50.12	49.23	49.68	High-income owners	10,651	Adult income rank of children from low-income families (percentile)	High-income owners
45.44	44.44	44.94	Older owners	3,543	Adult income rank of children from low-income families (percentile)	Older owners
45.00	44.42	44.71	Middle-income owners	17,392	Adult income rank of children from low-income families (percentile)	Middle-income owners
42.77	42.37	42.57	Low-density owners	15,490	Adult income rank of children from low-income families (percentile)	Low-density owners
45.08	43.73	44.44	Dense, college-educated renters	7,706	Adult income rank of children from low-income families (percentile)	Dense, college-educated renters
40.91	40.34	40.64	Moderate-income, mixed-tenure	14,860	Adult income rank of children from low-income families (percentile)	Moderate-income, mixed-tenure
38.43	37.36	37.89	Lower-income renters	6,757	Adult income rank of children from low-income families (percentile)	Lower-income renters
36.42	35.18	35.81	High-poverty, single-parent	4,884	Adult income rank of children from low-income families (percentile)	High-poverty, single-parent
14.67	13.10	13.86	High-income owners	10,629	Teen births among daughters of low-income parents (%)	High-income owners
19.97	18.31	19.12	Older owners	3,504	Teen births among daughters of low-income parents (%)	Older owners
22.93	21.84	22.38	Middle-income owners	17,382	Teen births among daughters of low-income parents (%)	Middle-income owners
27.88	27.21	27.54	Low-density owners	15,474	Teen births among daughters of low-income parents (%)	Low-density owners
22.10	19.14	20.51	Dense, college-educated renters	7,685	Teen births among daughters of low-income parents (%)	Dense, college-educated renters
30.43	29.15	29.77	Moderate-income, mixed-tenure	14,852	Teen births among daughters of low-income parents (%)	Moderate-income, mixed-tenure
35.04	32.46	33.76	Lower-income renters	6,745	Teen births among daughters of low-income parents (%)	Lower-income renters
39.98	37.77	38.86	High-poverty, single-parent	4,883	Teen births among daughters of low-income parents (%)	High-poverty, single-parent
1.35	1.18	1.26	High-income owners	10,648	Incarcerated on Census Day 2010, children of low-income parents (%)	High-income owners
1.86	1.63	1.75	Older owners	3,533	Incarcerated on Census Day 2010, children of low-income parents (%)	Older owners
1.90	1.76	1.83	Middle-income owners	17,390	Incarcerated on Census Day 2010, children of low-income parents (%)	Middle-income owners
1.98	1.89	1.94	Low-density owners	15,488	Incarcerated on Census Day 2010, children of low-income parents (%)	Low-density owners
2.14	1.69	1.90	Dense, college-educated renters	7,695	Incarcerated on Census Day 2010, children of low-income parents (%)	Dense, college-educated renters
2.59	2.36	2.48	Moderate-income, mixed-tenure	14,858	Incarcerated on Census Day 2010, children of low-income parents (%)	Moderate-income, mixed-tenure
3.25	2.83	3.04	Lower-income renters	6,754	Incarcerated on Census Day 2010, children of low-income parents (%)	Lower-income renters
3.81	3.28	3.54	High-poverty, single-parent	4,884	Incarcerated on Census Day 2010, children of low-income parents (%)	High-poverty, single-parent

The high-poverty, single-parent type has the least favorable mean in every panel of both tabs, and high-income owners have the most favorable in every panel except diabetes. Dots are population-weighted means of tract values with 95% intervals from resampling whole counties, and rows follow the type numbering, not the size of the value. Values in Table 4.

Note: No outcome is an input to the typology. PLACES estimates are modeled partly from ACS covariates such as poverty and education, so the agreement between the types and the adult health measures is partly mechanical. The Opportunity Atlas measures share no file with any input.Note: Opportunity Atlas measures are for children born 1978 to 1983 raised in families at the 25th income percentile, interpolated from 2010 to 2020 tracts by land area, and tract population stands in for the number of children. Life expectancy is for deaths in 2010 to 2015 and covers 77,763 of the typed tracts. PLACES values are modeled crude prevalence for adults, with BRFSS 2022 data. Each panel uses the typed tracts that have the measure.Note: Intervals show how the means would vary with the sample of counties, given the types; they do not include uncertainty about the typology itself (a bootstrap over counties; replicates and seed in the methods note).Source: ACS 2019-2023, USALEEP life expectancy 2010-2015, Opportunity Atlas (children born 1978 to 1983), PLACES 2024 release and ADI 2023; tract values weighted by ACS population; population-weighted, modeled estimates for the PLACES measures

The adult health measures share a source with the inputs. PLACES estimates are modeled partly from ACS measures such as poverty and education, so the agreement between the types and those measures is partly mechanical. The Opportunity Atlas measures share no file with any input.

The Black and Hispanic shares of residents are not inputs either. Both shares are higher in each of the three types lowest on the resource score than in any of the other five. In the high-poverty, single-parent type the Black share averages 35.2% and the Hispanic share 38.3%, against 5.0% and 9.6% among high-income owners.

Types built without outcomes spread life expectancy less widely

The chapter's earlier typology put life expectancy, mobility, diabetes, smoking and the Black and Hispanic shares among its inputs. Refit on the same tracts, its types span 10.0 (95% CI 9.8 to 10.2) years of life expectancy from the highest to the lowest type mean, and the types here span 7.5 (95% CI 7.1 to 7.9) years. Part of the earlier spread came from sorting tracts on life expectancy itself and on measures that move with it. The types here still sort life expectancy, from measures of resources and settlement alone, and the two sets of types agree only in part. The report gives every estimate with its interval and sample size.

How the types were built and checked

The inputs are ACS values for 2019 to 2023, while the outcomes come from earlier years. Life expectancy is the USALEEP estimate for deaths in 2010 to 2015, and the income rank is the Opportunity Atlas rank for children born in 1978 to 1983. An outcome can therefore predate the conditions it is compared with. Tracts that lack an input, institutional tracts, tracts of 200 or fewer residents and the tracts of Puerto Rico are left out, and a typed tract need not have an outcome. Each outcome is compared over the typed tracts that have it.

Eight types is the book's choice, kept from the first edition, and the data do not select it. The silhouette described below is highest for a fit with 2 types. Refits on resampled tracts and with population weights show that the partition is one of several that the data support. The intervals from resampling whole counties describe how the means would vary with the sample of counties given the types, not with the choice of types.

The types are not sharply separated

The silhouette scores how well tracts fit their own type against the nearest other type. For the typology it is 0.19. K-means run on simulated data that have the correlations of the inputs and no clusters still finds eight groups, and their silhouette is 0.14. The tract data sit a step above what a cloud without clusters gives, and the margin is small at every number of types the chapter tries. The benchmark is a guide and not a proof, because the simulated margins are symmetric and tract incomes, densities and rates are not. In a random sample of tracts, 6.6% have a negative silhouette, which means that on average they sit closer to the tracts of another type than to those of their own.

Figure 2 places sampled tracts on the first two principal components of the inputs. The first is a combined resource score, and the types are numbered along it. The ranges from the fifth to the ninety-fifth percentile of tracts overlap for every pair of neighboring types along that score. The overlap suggests that the types are bands of a continuum, and a tract near a border could have been assigned to the neighbor by a small change in the inputs.

Figure 2

Neighboring types overlap along the first component, and 6.6% of 20,000 sampled tracts sit nearer another type than their own

A random sample of tracts on the first two principal components of the 7 standardized inputs, colored by type, with each type's centroid labeled

Loading data

Neighboring colors shade into one another along the horizontal axis, the resource score, which suggests a continuum. Each dot is a tract in a random sample of typed tracts, colored by its type, and each labeled dot is a type's population-weighted centroid. The vertical axis separates denser, more renting tracts from older, low-density, owner-occupied ones. Values in Table 5.

Note: The dots are a random sample of 6,000 of the typed tracts drawn with a fixed seed (replicates and seed in the methods note); the components, the types and the centroids use every typed tract. Components are in standardized units, and their signs are set so that log income and log density rise with the first and second component.Note: The first component weights income, college share, homeownership and the share of older adults positively and poverty and the single-parent share negatively. The second component weights density and college share positively and homeownership and the share of older adults negatively.Source: k-means types of tracts on seven ACS 2019-2023 measures and 2020 land area; principal components of the standardized inputs; unweighted sample of tracts, centroids population-weighted

Between the two ends the types do not line up in one order. Dense, college-educated renters, numbered fifth, have a life expectancy 2.5 (95% CI 2.1 to 2.8) years above that of low-density owners, numbered fourth, although the two types sit at about the same place on the resource score. They also live longer than middle-income owners, numbered third.

What the types describe

The types describe bundles of resources and settlement, and the outcomes are compared with them afterward. They show how closely conditions and outcomes move together across tracts and do not show that a tract's conditions produced any outcome. A second clustering, built from deprivation measures, is in Chapter 10.2, Families of places. Chetty and Hendren used federal tax records for children born in 1980 to 1988 whose families moved once between commuting zones, and compared children who moved at different ages.[1] Children's adult income ranks move toward those of children of the destination's permanent residents with parents at the same income. The move is 0.04 percentile for each year of childhood spent there and each percentile by which the destination's ranks are higher. It does not test the eight types, and it does not say that moving a child from one type to another changes an outcome by the gap between the types' means.

On the county map, low-density owners are the most common type in most counties of the South, the Midwest, the Plains and the Mountain West. Middle-income owners are the most common type in most counties of Connecticut, Delaware, Hawaii and Rhode Island, and the other types lead mainly in counties with large populations. Across all counties with a type, 42.9% of typed residents live in a tract of their own county's modal type, which leaves more residents in a tract of another type.

Figure 3

Low-density owners are the most common type in 2,207 of the 3,110 counties with a type, and middle-income owners in 438

Neighborhood type that holds the most residents in each county, tracts typed by k-means on 7 measures, ACS 2019-2023 population

Loading county values

Most common neighborhood type

No data (3,143 counties)

Select a class to show only those places.

Find a county

Hover or tap a county for its value, or search by name. Zoom with the buttons on the map, then drag to move it.

Each county takes the color of the type that holds the most of its typed residents, so the color hides the mix inside the county. Gray counties have no type, and the tooltip says why. Values in Table 3.

Note: A county gets a type when its typed tracts hold at least 50% of its residents. 13 counties have no typed tract (a tract that is not institutional, has every input and has more than the floor of residents) and 20 have typed tracts that hold a smaller share, so they are gray with the reason in the tooltip. A county with a tie in residents takes the lower-numbered type, and 1 county has a tie. 485 counties with a type rest on fewer than 3 typed tracts.Note: Connecticut is drawn as its eight legacy counties, with tracts assigned through the exact tract crosswalk. Puerto Rico is outside the typology and is not drawn.Note: Colors mark the eight types and have no order.Source: k-means types of tracts on seven ACS 2019-2023 measures and 2020 land area, summed to counties by ACS population; population-weighted

What this cannot tell you

Every comparison here is between tracts. The mean life expectancy of the high-poverty, single-parent type is an average over tracts that says little about any one resident of one (ecological inference). A name such as high-poverty, single-parent summarizes two inputs of a profile, and it does not describe the people who live there. The life expectancy and Opportunity Atlas values are interpolated from 2010 tracts by land area, not by population. Tracts that lack an input are not typed, and they are poorer on average than the typed tracts, so the residents of typed tracts are not a random sample of the country.

Go deeper

The report: every estimate with its interval and sample size.
Tables: every figure of the report as a table.

Data sources

American Community SurveyACS 5-year estimates 2019-2023 for 2020 tracts: income, poverty, education, households, tenure, age and population (the inputs), and race and ethnicity (compared by type, not an input)Years: 2019-2023 (5-year) · Weights: ACS 2019-2023 tract population · License: Public domain (U.S. Census Bureau)
Census TIGER/Line tracts2020 tract land area for population densityYears: 2020 · Weights: none (land area) · License: Public domain (U.S. Census Bureau)
CDC/ATSDR Social Vulnerability IndexCDC/ATSDR Social Vulnerability Index 2022: share of residents in group quarters, for the institutional-tract flagYears: 2022 release (ACS 2018-2022) · Weights: none (flag) · License: Public domain (CDC/ATSDR)
USALEEP life expectancyNCHS small-area life expectancy at birth, deaths in 2010 to 2015, 2020 tracts (area-weighted); an outcome compared by type, not an inputYears: 2010-2015 · Weights: ACS 2019-2023 tract population · License: Public domain (NCHS, CDC)
Opportunity AtlasAdult income rank, teen births and incarceration of children born 1978 to 1983 raised in families at the 25th income percentile, 2020 tracts (interpolated by land area); outcomes compared by typeYears: 1978-1983 birth cohorts (2018 release) · Weights: ACS 2019-2023 tract population · License: Public use (Opportunity Insights and the U.S. Census Bureau)
CDC PLACESCDC PLACES 2024 release (BRFSS 2022): diabetes, smoking, frequent mental distress and obesity among adults, modeled crude prevalence; outcomes compared by typeYears: 2024 release (tracts); 2025 release (ZCTAs, 2023 data year) · Weights: ACS 2019-2023 tract population · License: Public domain (CDC, U.S. Department of Health and Human Services)
Area Deprivation IndexArea Deprivation Index 2023 national rank, a profile column and not an inputYears: 2023 release (version 4.0.1; built on ACS 2019-2023) · Weights: ACS 2019-2023 tract population · License: Free for non-commercial use; cite the Neighborhood Atlas and Kind and Buckingham
Published estimates from cited studiesThe exposure effect quoted in the causal box (Chetty and Hendren 2018)Years: as reported in each study · Weights: as reported in the study · License: Quoted from published research for commentary

All data sources

References

Chetty, Raj; Hendren, Nathaniel (2018). The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects. The Quarterly Journal of Economics 133(3): 1107-1162. doi:10.1093/qje/qjy007

How to cite

@misc{singh2026eighttypesofneighborhood,
  author = {Singh, Vishal},
  title = {Eight types of neighborhood},
  year = {2026},
  url = {https://vishalsingh.org/zipcode/eight-types-of-neighborhood},
  note = {In ZIP Code Destiny. Published 9 October 2026; updated 28 October 2026.}
}