Tract-level ministry data: reading a city one neighborhood at a time
Small enough to show the street-by-street difference. Stable enough to compare. Public enough to check.
Ministry decisions go wrong when a whole city is read as one place. Census tracts let you see the block that is carrying more than the one beside it.
What is tract-level ministry data?
Tract-level ministry data is neighborhood information, measured for each U.S. census tract, that a church can use to decide where and how to serve. A census tract is a small statistical area drawn by the Census Bureau to capture a socially coherent neighborhood of roughly 1,200 to 8,000 residents, identified by an 11-digit code: 2 digits for the state, 3 for the county and 6 for the tract. The United States has 84,414 of them, and GospelGraph scores every one it can on the same scales.
A tract is not a neighborhood in the full human sense. Its edges follow roads, rivers and rail lines rather than the places people say they live. But it is a strong unit for comparison: smaller than a city, ZIP code or county, stable across public datasets, and large enough that published survey estimates protect individual privacy.
Why use census tracts instead of ZIP codes or cities?
ZIP codes are built for mail delivery, not for describing neighborhoods, and they can span very different places. City and county averages are worse: a county can be "good on average" while particular neighborhoods inside it carry heavy weight. The Anne Arundel County field study makes exactly that point, reading all 130 of the county’s neighborhoods for the churches that would act on it.
For a church, the practical gain is comparison. Two areas five minutes apart can carry very different need, very different access to an evangelical church, and very different kinds of households. At tract level you can see that before you send a team.
What does GospelGraph measure at tract level?
Each tract carries several readings that are kept separate and never blended into one score.
- Gospel Desert Index: the structural Gospel-access gap, 0–100. 83,344 tracts are scored; 1,070 are marked Unscored rather than given a zero.
- Need pressure: 100 minus the latest Shalom composite. 83,381 tracts have a 2024 composite.
- Gospel Impact Metric: the change in Shalom from 2020 to 2024, available where both eras are scored.
- Shalom domains: Provision & Work, Justice & Protection, Life & Health, Household & Generations, and Social Fabric, each 0–100.
- Church access: distance to the nearest church, churches per square mile, and the count of churches inside the tract, from four deduplicated sources.
- Neighborhood personalities: measured one level down, by census block group, and composed up to the tract by household.
Where does the tract data come from?
Neighborhood conditions come from the Census Bureau’s American Community Survey 5-year estimates. The five-year window substantially reduces sampling error at tract level, which is why the platform scores tracts rather than smaller units. Among the ACS tables read are total population, poverty, housing vacancy, labor force and unemployment, and race and ethnicity. Daytime life comes from the Census Bureau’s LODES employment data, which links payroll records to blocks. Church locations come from OpenStreetMap, the IRS 990 Business Master File, the ChurchesUSA directory and Overture Places. The data sources page explains each one and its limits.
Everything is aggregate. GospelGraph does not build profiles of individual people, and no layer reads a household’s beliefs.
Is tract-level data precise enough for churches?
Precise enough to compare places and to find patterns worth investigating. Not precise enough to pronounce on a single street. Survey estimates carry margins of error, and domain scores are ranks against the national distribution, so a 70 means higher than 70% of the country. A rank near the middle can move several points from one year to the next without anything changing on the ground. Treat small gaps between two tracts as noise.
Some tracts rest on thin ground: fewer than 200 residents, fewer than 50 housing units, a large workplace with almost no residents, or mostly water. Those tracts carry a flag wherever their score appears, and the pipeline declines to score many of them at all. A missing value always comes with its reason; it is never filled in with a guess or a zero.
Household mix is the one place a tract is too coarse. Close to a third of tracts have no built-form state that covers more than 60% of their households, so GospelGraph reads the Field Guide by block group and hands a reader at an address the guide composed for their own block group.
How should a church use tract-level data?
Start from an address you care about: your building, a member’s street, a candidate planting site. A free field report resolves it to its tract and reads the ground. Then open the map explorer and look at the tracts around it, because the neighborhood your people call home may span three tracts that carry very different weight.
Bring the numbers to the people who know the place. Tract-level data can reveal patterns, but it cannot tell you a neighbor’s name, history, wounds or hopes. It exists to send a church into a neighborhood more humbly, not to excuse it from the slow work of presence. For the national picture, read State of the Field, which covers 83,344 tracts and more than 125 million households.
Common questions
What is a census tract?
Why not use ZIP codes?
Is tract-level data precise enough for churches?
Does GospelGraph use personal data?
Find your tract
Type any U.S. address into a free field report and see the tract it sits in, read for a church.