State of the Field: Assemblies of God

The roster, and what joined

The 2024 roster carries 12,700 rows. 11,008 of them, 86.7 percent, landed on a 2020 census tract that the index scores. The 1,692 that did not fall into three groups, and none of the three is random.

The 11,008 occupy 9,650 distinct tracts. Most hold one Assemblies of God congregation; 1,132 hold two or more, and those tracts account for 2,490 congregations between them.

The roster also carries two structural columns that the rest of this report cuts by. Sixty-six district councils and networks, thirteen of them organised around a language or a culture rather than a territory, holding 1,515 of the placed congregations. And three affiliation classes: 6,428 General Council-affiliated and self-governing, 3,107 District-affiliated and under district oversight, 1,473 Parent-affiliated daughter works.

Two layers cover roughly half the roster, and every figure from them below is a share of the tracts they score, not of all 9,650. A blank in either is an absence from that layer, not a zero.

Where the churches are

All fifty states and the District of Columbia hold at least one placed congregation. Two states hold more than a thousand.

Seven of the fifteen are in California. The language and ethnic districts run the other way: Texas holds 344 of their 1,515 congregations, Florida 212, New York 202, New Jersey 98, and California only 57, because the two large California geographic districts absorb that work directly.

Access

The Gospel Desert Index scores how much evangelical gospel witness a tract has structurally within reach, from 0 to 100, higher meaning less access. It is built from congregations and their reach, never from a judgment about faithfulness. The roster sits almost exactly on the national mean.

The tier distribution is the national one to within three points in every band. Where the roster departs from the country is density: the average AG tract holds 9.7 other churches against a national 6.8, and only 2.6 percent of AG tracts hold no church at all against 11.6 percent nationally.

Settlement type, built form and viability

The archetype classifies the physical form of a tract. The Ministry Viability Index reads whether that form could hold a gathering: how connected the streets are, how many third places and public rooms there are. Posture combines viability with need into four postures the platform names Plant, Partner, Strengthen and Sustain.

Eleven points more suburban than the country, and under-represented in every other form. The rural figure is worth stating in its own terms: 2.8 percent of AG tracts against 4.6 percent nationally, and the 1,563 unplaced rows lean rural, so this is a ceiling on the shortfall rather than a floor.

Viability sits exactly on the national mean while intersection density sits five points under it, which is the suburban skew showing through: wider blocks, fewer corners, with third places at the national rate.

Need pressure and the eight domains

Need pressure is the weight households are carrying now, built from five Shalom domains. The eight ministry-alignment domains read the same conditions as a call: a higher score means the tract's conditions ask for more of that kind of work. All are scaled so the nation sits at 50.

Six of the eight domains sit above the nation, family stability and economic opportunity furthest. The two that sit below, housing stability and reconciliation, are the two that separate the language districts from the geographic ones most sharply, which section 14 returns to.

Shalom, level and change

The Shalom index is a composite wellbeing score by era, 0 to 100, higher meaning better, with three pillars: living, health and hope. It is scored on 95,837 tracts, so its national column is a different base from the church-fit index. Both 2020 and 2024 are read here.

The roster stands on ground scoring two and a half points below the country on wellbeing, and the shortfall is carried by health, four points down, and living, three points down. Hope sits half a point above. The direction of change is the country's: both fell by about seven tenths of a point across the era and just over half of tracts in each fell at all.

The three clocks

GospelGraph measures change in access three ways, because one measurement of change is easy to fool. Gospel Shift compares two eras. GIM v3 fits a slope across twelve overlapping ACS windows. The GDI trajectory reads four vintages of the access score itself. A fourth reading, the field instruments' direction call, resolves on about half the roster.

All three clocks agree, and they agree that the roster's ground is moving at the country's pace. The one departure is the GDI trajectory slope, 0.027 against 0.005 a year: the access score under AG tracts has been drifting up, meaning slightly less access over the four vintages, by about a sixth of a point in total. It is small, it is consistent with the denomination sitting in suburbs where church counts are stable while population grows, and it is not a trend anyone should plan around yet.

Households: thirty segments

The v4 segmentation sorts every American neighborhood into one of thirty core household types by how people live rather than by what they earn, grouped into ten LifeModes. Every tract is a blend, so shares below are weighted by full composition and by household count: a sixty-forty tract counts as both, and a tract of four thousand households counts for four of one thousand. Lift is the AG share divided by the national share; 1.00x is the national rate.

The shape is a working-household denomination. Every segment above 1.20x is a working or gateway type: Low-Income Solo Holdouts 1.46x, Highway Corridor Commuters and Car-Oriented Working Gateway Families 1.45x, Established Blue-Collar 1.42x, Logistics-Corridor 1.41x, Industrial-Fringe 1.34x, Exurban Working 1.25x. Every segment below 0.60x is affluent or professional: High-Achievement Homeowners 0.41x, Affluent Transit-City Creatives 0.46x, Professional Multifamily 0.49x, Garden Apartment Young Adults 0.52x, Affluent Mature Owner Couples 0.57x.

Two gateway types run under the national rate and are worth naming so the gateway story is not overstated: Transit-Dependent Gateway Households at 0.68x and Dense Solo Gateway Renters at 0.78x. The denomination reaches gateway families who drive. It is thinner among gateway households who do not.

Farmworker concentration is the modifier the roster over-indexes on most, at more than twice the national rate, followed by public housing and chronic-health burden. It under-indexes on university towns and HBCU-adjacent tracts by about half.

Field signals

Field signals are a composite watch score with a tier and a set of flags, scored on 57,412 tracts nationally and 5,852 of the roster's, which is 60.6 percent. Every share in this section is of the tracts the layer scores.

The roster runs cooler than the country on this layer: fewer hot tracts, more positive ones, and ten points less daytime isolation, which is what a suburban roster with good church density looks like from here. New activity runs two points above the nation.

Twenty-four neighborhood patterns

Each tract is read for twenty-four neighborhood patterns, each in one of four states. Ten of the twenty-four are risk-named, where present means the hazard is realised rather than that the fabric is intact; those ten are swapped before counting, so every row below is the share of tracts where the pattern is unhealthy. Risk-named patterns are marked.

The fabric under the roster is the country's fabric, to within a few points on every row. The one pattern that departs by more than five is network of paths: 64.6 percent of AG tracts lack a walkable network of routes against 57.7 percent nationally, seven points worse. That is the suburban skew from section 4 read from the ground. Caregiver pressure and family pressure sit about four points worse; under-resourced sits three and a half points better.

Situation tags

Situation tags are threshold flags raised when several layers agree on a condition. They resolve on 95,837 tracts nationally and 9,435 of the roster's, 97.8 percent.

This is the layer where the roster looks least like the country. Gospel absence is raised on AG tracts at 1.74 times the national rate, health crisis at 1.52, mercy entry at 1.50, rising need at 1.46, chronic disinvestment at 1.38. All five are flags for hardship. The tags that run under the national rate, thin church ecology, daytime isolation and accessible desert, are flags for emptiness. The roster stands where things are hard and where churches already are, and this layer says both at once.

The nearest churchless high-need tract

547 tracts in the country hold no church in the merged church data and carry need pressure above 70. For every congregation, the pipeline finds the nearest one, centroid to centroid. Distance is straight-line, so it gives direction and rough scale rather than a drive time.

A single tract in Pierce County, Washington is the nearest churchless high-need tract for 624 Assemblies of God congregations, because the Northwest holds few of the 547 and many AG churches. Read this table as a map of where the targets are sparse rather than as a queue.

The opening frontier

The Desert Moved defined the fastest-opening ground in the country: tracts that GIM v3 confirms are opening, sit at GDI 60 or above, and score 50 or above on viability. There are 2,186. Its mirror is the ground hardening with a church already inside: 3,185 tracts.

The household lift below is the one figure in this report weighted by congregation rather than by tract, because it is quoted from The frontier the PCA is not on on that paper's basis so the two can be read together. The differences from the tract-weighted table in section 8 are small and in the second decimal.

Two ways to cut the roster

Every table above is the roster as one body. The roster carries two columns that divide it, and both divisions are larger than most of the differences between the roster and the country. The argument built on them is in Field Study No. 3; the figures are here.