What the map of a denomination does to its preaching
Brookhaven Presbyterian Church sits on North Druid Hills Road, in Brookhaven, Georgia. The neighborhood is about as comfortable as American neighborhoods get. Roughly one household in twenty-five lives below the poverty line, and the dominant household type on those streets is the most affluent of the thirty the data recognizes.
Six tenths of a mile away, still inside Brookhaven, still in DeKalb County, there is a neighborhood where nearly a third of households live below the poverty line, where the weight of ordinary hardship ranks among the heaviest in the country, and where the national church data finds no congregation of any kind. It is a working immigrant neighborhood, the kind built by people who arrived recently and took the jobs that were going. Same name on the sign as the church, six tenths of a mile up the road.
None of which is a criticism of Brookhaven Presbyterian, which is doing what a faithful suburban church does. But the whole denomination has that shape, and in Brookhaven you can walk the length of it in ten minutes. This paper is about what the shape does over time, and about the one consequence of it that belongs at a preaching conference more than at a church planting one.
Before you let these numbers carry an argument, know what they are.
The unit is the census tract: roughly four thousand people, about the size of a parish, small enough that the neighborhood inside it is one place. It suits this question and almost no other. I put every congregation in the PCA roster on the tract its building stands in, then read that tract against four things.
Access first. The index counts how much evangelical gospel witness a place has structurally available, working from congregations and their reach, never from anybody's judgment about their faithfulness. High access tells you the ordinary means sit within reach of the people living there. It tells you nothing about whether they have taken them up.
Then pressure, the weight of present hardship, built out of five domains: provision and work, justice and protection, life and health, household and generations, and social fabric. Pressure is a level, a reading of conditions today. A neighborhood carrying heavy pressure is a harder place to live, which is a different claim from calling it a worse one. Four separate instruments measure whether that pressure rises or eases, because one measurement of change is easy to fool and four that agree are not.
Then the fabric of a place, which most people have never seen quantified. Whether there are spots where people end up without planning to. Whether anyone walks between them, whether a civic anchor sits nearby, whether the households around it are mixed or uniform. Beside that sits a classification sorting every American neighborhood into one of thirty types by how its households live rather than by what they earn. That distinction matters more than it sounds. A neighborhood of young renters in garden apartments is a different mission field from an established extended-family neighborhood, and the census can report the same median income for both. Most of the argument in this paper rests on that layer.
Built from congregations and their reach, never from anybody's judgment about their faithfulness.
Provision and work, justice and protection, life and health, household and generations, social fabric.
Somewhere people end up without planning to, paths between them, a civic anchor, a mix of households.
Not by what they earn. Every tract is a blend, so a place that is sixty-forty counts as both.
The data counts none of what you would ask about first: members, attendance, giving, conversions, the quality of anybody's preaching. It counts where the buildings are and what kind of ground surrounds them. A real thing to know, and a small part of what matters.
Three other denominational rosters went through the same process for comparison: the Orthodox Presbyterian Church, the Associate Reformed Presbyterian Church and the Assemblies of God.
Sort every American neighborhood by how its households live and the PCA tracks the country almost the whole way down. Family suburb for family suburb, urban professional for urban professional, it sits about where America sits.
There is one exception and it is large. Immigrant gateway neighborhoods, the places built by and for people who arrived recently, are about a tenth of American neighborhoods. They hold four percent of PCA churches. All four gateway types are short by roughly half, and in the most common of the four there are ten PCA congregations in the entire country.
Three other measurements find the same hole coming at it from different angles: concentrations of limited-English households, concentrations of farmworker households, and the flag the system raises on immigrant communities generally. None of the four shares its inputs with the others. Four instruments built separately, landing in the same place, do not all fail the same way.
The Assemblies of God give the finding its weight. A denomination about six times the size, over-represented in gateway neighborhoods, sitting above the rate at which those neighborhoods occur in the country at all. Whatever one makes of Pentecostal ecclesiology, they have solved a distribution problem the PCA has not.
Sixty-nine PCA congregations stand in gateway neighborhoods. Read that list and a pattern comes off the page.
Sa-Rang. Nachimban. Il Shin. East Sarang. Hana. Areumdown. Mision Vida Nueva. Dios Con Nosotros. Iglesia La Vid. Comunidade Crista Presbiteriana. The Korean Presbyterian Church of Southern New York, the Philadelphia Korean Presbyterian Church, the Korean Faith Presbyterian Church of Houston.
One in five of the congregations in gateway neighborhoods carries a Korean, Spanish, or Portuguese name outright. Across the denomination the figure is one in twenty-five, so the concentration runs more than fivefold, and the English-named churches on that list include several more that anyone who knows those presbyteries would recognize as immigrant congregations.
Names are not ethnicity. Check the list against presbytery records before you repeat any of this from a platform, because I am reading congregation names off a roster and nothing more. I say it anyway, with that caveat attached, because it changes what the earlier finding means. Appendix page 03 lists the fifteen, where all sixty-nine sit, and every PCA congregation in the largest gateway type.
The PCA's presence in immigrant neighborhoods did not come mostly from the denomination's church planting machinery. Immigrant Presbyterians built it, planting among their own, and the denomination received those congregations more than it sent them.
That makes the finding harder to answer, not easier. Sixty-nine congregations settle the question of whether the PCA can reach gateway neighborhoods. They leave open why the part of the denomination that plans and funds church plants has produced so few of them, and whether it has learned anything from the churches that did.
The other finding is the one people half expect.
The average PCA congregation stands in a neighborhood already holding about eleven churches, where the average American neighborhood holds under seven. The denomination is thinly present where there are no churches and heavily present where there are many. Across all of South Carolina, exactly one PCA church is the only congregation in its tract.
This does not mean a church in a crowded neighborhood is failing. A congregation drawing from ten miles out is not serving the four thousand people it happens to sit among, and a tract full of churches can be as unreached as an empty one.
The denomination tells itself a story about expanding access. The map describes something closer to expanding options.
Cut it the other way and it stings more. A few hundred neighborhoods in America carry heavy pressure with no church in them at all, and four out of five already have a PCA congregation within fifteen miles. Distance is not what keeps the denomination out of them, any more than distance keeps Brookhaven Presbyterian from the neighborhood up the road.
GospelGraph published a separate finding last year, in The Desert Moved, that changes what the gap in the previous section means. For twenty years the picture of a gospel desert in America has been urban: the sold church building, the neighborhood a denomination left in 1974. That picture is no longer where the numbers point hardest.
Hold every tract that is measurably improving to the worst two access tiers and to a built form that could hold a gathering, and 2,168 remain. Almost none of them are inner cities. They are new subdivisions, exurban build-out and university towns whose edges filled in. The steepest riser in the country is a Nashville commuter suburb that holds no congregation at all. These places are not unreached because a church left. Ten thousand households arrived over eight years into a place with a road, a school and a grocery store, and the ordinary mechanism by which a neighborhood acquires a church, which is somebody already being there, never ran.
Running the PCA roster against that set is the shortest way to state the problem.
Twenty-seven congregations. The denomination is on one percent of the ground its own data calls the fastest-opening unreached field in the country.
The Orthodox Presbyterian Church makes the number easier to read rather than easier to bear. A smaller Presbyterian body, 338 congregations against the PCA's 1,913, it is in two of the 2,186. One tenth of one percent, and three and a half of its churches sit in the hardening mirror for every one on the frontier, against the PCA's two.
Run the crossing the other way and the PCA has twice as many churches in the mirror set, the 3,185 tracts that are hardening with a congregation already inside them, as it has on the frontier. Fifty-four against twenty-seven. That is not a criticism of the fifty-four, who are mostly doing the harder job. It is a description of where the denomination's attention already sits.
The household layer makes it sharper, and this is the finding I would put last because it is the one that will not leave.
The Desert Moved sorts all thirty household types by how over-represented they are among the unreached, and finds that exactly three of the thirty are scarce among both kinds of unreached tract. Those three are the whole of what the data is prepared to call reached.
The PCA is over-represented in all three.
Now the growth edge, the types over-represented among the unreached on both counts:
Read the two lists together. The PCA is above its share in every household type the data calls reached, and below its share in every growth-edge type except one. The exception is Affluent New-Build Families, at 1.43x.
Set that against Working New-Subdivision Families at 0.82x. These are the same kind of place. Both are houses that went up after 2000 on ground that was a field, both have no institutions and no history, both sit at the top of the unreached list. The difference between them is the household income of the people who moved in, and the PCA is present in one at nearly double its share and absent from the other.
The word that predicts whether the PCA is there is not urban or rural, and it is not new or old. It is Working.
Which is the legibility argument again, arriving with numbers attached. The affluent new-build subdivision is a room the denomination knows how to read. The working one, forty minutes further out, is the same age and the same emptiness and a different set of questions, and the PCA has not learned them.
The article usually written off this data says the PCA has become a wealthy denomination planting among the wealthy. On the money that claim is thinner than it sounds. The gap between the average PCA neighborhood and the average American one in household poverty is about a percentage point.
The neighborhood types make the stronger version. The PCA is over-represented in the affluent suburban types, though only modestly, and short by half in every gateway type. Its neighborhoods have measurably better civic fabric than the country, with more of the informal architecture that produces repeated contact between strangers, and thinner cultural mixing.
Poverty sits near the edge of this story. The PCA plants among people it can already read, in neighborhoods already equipped for the kind of ministry it knows how to do.
Go looking for the decision that produced this map and you will not find one. No committee voted to skip gateway neighborhoods. A thousand smaller decisions made the map instead, and each one made sense to the people making it.
A core group forms where people already know each other. A plant goes where a mother church has members willing to drive. A presbytery approves the work with the clearest path to self-support. A candidate takes the call he can picture himself doing. Every one of those is a defensible judgment, and together, across the denomination's fifty-odd years, they produce a map.
Legibility is the thread through all of them. Ministry costs less where you can already read the room, where you know the questions before anyone asks them and the neighborhood's own fabric does part of your gathering for you. The PCA has optimized for that without ever deciding to, and has gotten good at it.
The trouble is that legibility compounds. You can read the neighborhoods you have already learned to read, and you learn by being there. Which is also how the gateway churches came to be planted by immigrant Presbyterians: they could read those rooms on arrival.
The Field Guide carries a section for every household type called The Gospel on Their Terms. It states the doctrine, then states it again in whatever vocabulary that household already uses for debt, standing and worth. Reading the four types that define the frontier next to each other turns up something a Presbyterian conference ought to hear.
They land on the same doctrine through four different doors.
Exurban Working Homeowners hold the largest share of the frontier, eighteen percent of it. Their guide opens with the record. The county keeps one, everybody knows their own entry, and the town remembers the man at nineteen for the rest of his life. God keeps a fuller file than the county does, and what Jesus did with it was not to argue that it was wrong. He paid it and closed it.
Car-Oriented Working Gateway Families run at twice their national share out there. Their guide opens with work, and with the observation that God is not keeping the same ledger their employer keeps. Labor with the hands is honorable and settles nothing. Then a line worth reading twice: it asks nobody to leave behind the language or the parents they came with.
Working New-Subdivision Families are also at twice their share, and they get the deed. Buying out here was an achievement and Scripture is not sniffy about shelter. The arithmetic that bought the house does not reach as far as God. What Jesus obtained was paid once and in full, never has to be serviced monthly, and cannot be lost in a bad quarter.
Cost-Pressed Family Suburbs get the household, and the guide is strict about the order. Start with God calling himself father to the fatherless, not with what a family is meant to look like. Sin is not a household arrangement. He does not ask anyone to complete a family. He asks them to let go of a life.
The county keeps one. God keeps a fuller file, and Jesus paid it and closed it.
18.2% of the frontierFather to the fatherless. He does not ask anyone to complete a family.
12.4% of the frontierGod is not keeping the same ledger their employer keeps.
5.5% of the frontierNever serviced monthly. Cannot be lost in a bad quarter.
3.3% of the frontierA record paid and closed. A ledger the employer does not keep. A deed transferred instead of a payment made. A place secured and handed over rather than assembled. That is justification, four times, in the vocabulary of a mortgage and a county file.
This ground is built to hear the doctrine the PCA takes for its own inheritance. The denomination is preaching it forty minutes back up the road.
The guides also name what will fail out there, and the first of the two is a habit of Reformed preaching.
A sermon that makes the Christian life sound like effort lands badly on these streets. These are households who already work hard and are still one bad month from trouble. Tell them the answer is to try harder and you have either insulted them or crushed them, and either way you have preached the opposite of what all four guides say they need.
The second is a matter of emphasis. Racial reconciliation is the theme the PCA has talked about most in the last decade, and on this ground it is the least called for. Of the eight ministry-need domains it scores lowest on the frontier, at 49.5, and it is the only one of the eight where the neighborhoods the PCA already occupies call louder than the frontier does. That is no argument against the work. It says the frontier will not be entered through it.
Preaching is not a general activity. Every sermon goes into a particular place, to people whose questions were formed by the street outside, and a preacher gets better at answering the questions he actually hears. Over a career that shapes him. Over a generation it shapes a denomination, and over two it shapes what the denomination's preaching is even capable of.
So a tradition that plants in neighborhoods legible to it hears the questions of the legible. Its preachers sharpen on those and stay dull on the others, and the commentaries and the illustrations and the conference sessions all follow the congregations that exist. Nothing corrupt happens anywhere in that chain.
At the end of it you have a pulpit tuned to a narrow slice of the country, held by people with no particular reason to notice.
The sixty-nine immigrant congregations sharpen the point. Somewhere in the PCA are pastors who have spent whole careers preaching in gateway neighborhoods and know things about it you cannot get out of a commentary. The denomination counts them, but it has not sat at their feet, and doing so would cost less than a church planting initiative.
Where the PCA plants decides what its preachers get practice at, and almost nobody inside the denomination is watching that happen.
Start with the bind. A quarter of these neighborhoods hold no church of any kind, so there is no session to particularize under and nobody to elect, because there is nobody yet. The sacraments wait on a minister, and the road to becoming one runs through three years in a seminary town. The ground that most needs a visible church sits furthest from the thing that constitutes one. A mission church under a commission is the existing answer and a good one. On ground this thin it can also mean being governed from a long way off for a long time.
Then the money. Poverty runs at almost seventeen percent out there and the households are working ones. A launch team carrying a full-time call to self-support inside five years assumes a giving base these tracts do not have. Three things would change that, and the denomination can pick any of them.
A presbytery can carry a frontier call twice as long as it plans to now, and budget for that at the start rather than discovering it in year four.
It can stop being the fallback for men who could not raise support and become how a church normally begins on this ground.
Supplied and supervised for as long as it takes, with nobody treating that as a failure to launch.
Then the week. Forty-four percent of frontier neighborhoods stand empty from eight to six, against eleven percent of the ground the PCA already holds. The data measures the commute and not the shift pattern, so what follows is a question and not a finding. A congregation that gathers on Sunday morning and nowhere else has already selected for the households whose employers give them Sunday morning off. Reformed conviction about the Lord's Day is one thing and a single service at eleven is another, and only the second is up for discussion. Somebody should work out which of the two is doing the selecting before they plant out here.
Then distance. Forty-two percent of the frontier is exurban or rural. The courts still meet. The connection they exist to provide gets thin when the nearest sister church is an hour off and presbytery costs a Saturday.
Then the instinct, which the Field Guide names outright for the largest household type on the frontier. Run the church's life through the families already inside it, the guide says, and things get done faster than any other way. The cost is that everyone you reach arrives through somebody who was already in, and the households connected to nobody stay connected to nobody. Work against your own efficiency here.
Presbyterians are built for the thing the guide warns about. Membership, office and hospitality all run through established households, and out here that is the strength and the trap at once.
Out here there is no existing network to build through. These are houses that went up after 2000 on ground that was a field, and everybody on the street arrived at about the same time, so nobody has the standing to introduce anybody. Social belonging reads broken in eighty-seven percent of these neighborhoods and neighborhood tissue in ninety-five. You gather a launch team from people who already know each other, and on this ground those people do not exist yet. Building the network is the planter's first job here. Everywhere else it comes with the ground.
A quarter of the frontier holds no church at all. No session to particularize under, and the sacraments wait on a minister.
Poverty at almost seventeen percent. A full-time call to self-support in five years assumes money these tracts do not have.
Forty-four percent of it stands empty from eight to six, against eleven percent of PCA ground.
Forty-two percent exurban or rural. The courts still meet; the connection thins.
Elections, a written constitution, a court outside itself. It cannot be owned by the man who started it.
Thirty households answering to nobody are alone. Thirty attached to a presbytery are not.
Civic anchor broken in 74%, social belonging in 87%. Nothing public exists to join.
The denomination does not have to invent any of it. It has to pay to get a minister there.
The polity is not the liability out there. It is the asset.
Civic anchor is broken in seventy-four percent of these neighborhoods. Social belonging in eighty-seven. Neighborhood tissue in ninety-five. Nobody has built an institution yet and there is nothing public to join. Into that a Presbyterian church brings a body that keeps minutes, holds elections, writes its constitution down, answers to a court outside itself, and cannot be owned by the man who started it. In a subdivision with no public life none of that is overhead. It is the church providing something vital the neighborhood does not otherwise have.
Thirty households in Rutherford County answering to nobody are alone. Thirty households attached to a presbytery are not. That is the right instrument for a place with no institutions and the denomination already owns it. What has to give is what it costs to get a minister onto the ground, and what it takes to constitute a church once he is there.
Denominational findings make for good conference talks and bad Monday mornings, so the underlying data carries something at the scale of one church.
Every congregation in it has an address attached now: the nearest neighborhood holding no church at all and carrying heavy pressure, with how far away it is and what kind of place it is. For hundreds of PCA churches that address is a few miles off. For some, as in Brookhaven, it is a walk. A session can go and look at it this month, which is a smaller and more useful act than adopting a strategy about it.
The same rows carry the fabric of your own neighborhood, worth knowing anyway. Does anyone walk past your building? Is there a civic anchor inside the radius your people actually move in? If your outreach assumes repeated informal contact with neighbors, you can now check that assumption, and in a good many places it comes back false. Plenty of what looks like a strategy that failed was a strategy nobody ever had.
Two things, and the first belongs at the front of the room rather than in the footnotes.
This is one frame of film. The roster carries no planting dates, so nothing here can say whether the denomination has been moving toward harder ground or away from it. If the churches planted in the last decade sit on emptier and less legible ground than those planted two generations ago, the correction is already underway and this paper describes history. One column added to the denominational roster would settle it, and the same column would answer the sharper question, which is when the gateway congregations arrived and whether anyone has planted alongside them since.
The second reads more as a warning. A neighborhood with no church in the data is not thereby unreached, and a neighborhood with three churches is not thereby reached. Buildings are the easy thing to locate, and this dataset locates buildings.
Every claim above has a figure behind it. The eleven pages below carry them, at the precision the data was computed to rather than the precision the argument needed, so a few numbers here sit a decimal point away from their rounded form in the text. Underneath them sits one row per congregation across fourteen data layers, so you can check any figure and reproduce any claim. Where the numbers disagree with the argument, take the numbers.
Two findings survive checking, and neither is the one usually written.
| Finding | PCA | National | Assemblies of God |
|---|---|---|---|
| Congregations in an immigrant gateway neighborhood | 4.0% | 9.4% | 11.6% |
| Churches standing in the average tract | 10.9 | 6.8 | 10.0 |
Seventy-six PCA churches are the only congregation in their tract. Two hundred thirty-six sit in tracts holding twenty or more. The affluence reading usually written instead rests on a one-point poverty gap, which will not carry it.
The widest gap in this dataset is not income. It is immigrant gateway neighborhoods, and four independent measurements point at it.
| Core segment | PCA | National | PCA churches |
|---|---|---|---|
| Car-Oriented Working Gateway Families | 0.6% | 2.5% | 10 |
| Dense Gateway Families | 0.7% | 1.8% | 12 |
| Transit-Dependent Gateway Households | 0.8% | 1.6% | 14 |
| Logistics-Corridor Homeowners | 0.6% | 1.4% | 11 |
| Measure | PCA | National |
|---|---|---|
| Limited-English concentration | 3.0% | 5.4% |
| Farmworker concentration | 0.4% | 1.9% |
| Immigrant-community situation tag | 1.1% | 2.1% |
Four independent measurements pointing the same way is not a data artifact. The over-representation is where you would guess: Affluent Owner Family Suburbs 15.5% against 12.9%, High-Achievement Homeowners 5.9% against 3.7%, and the university-town modifier on 4.3% of PCA tracts against 2.3% nationally.
Sixty-nine PCA congregations stand in a gateway neighborhood. A fifth carry a Korean, Spanish or Portuguese name outright.
| State | Churches | State | Churches | State | Churches |
|---|---|---|---|---|---|
| CA | 24 | NJ | 6 | PA | 2 |
| FL | 11 | TX | 6 | WA | 2 |
| NY | 9 | NC | 2 | Nine more | 1 each |
| Congregation | City | Need | Churches in tract |
|---|---|---|---|
| Il Shin Presbyterian | Buena Park CA | 53.4 | 24 |
| El Centro Korean Church | El Centro CA | 55.7 | 11 |
| Victory Presbyterian | Los Angeles CA | 62.4 | 6 |
| Servant Church of San Diego | San Diego CA | 64.7 | 9 |
| Areumdown Presbyterian | Cutler Bay FL | 57.8 | 2 |
| Hope Chapel | Miami FL | 57.1 | 7 |
| Ouachita Presbyterian | West Monroe LA | 79.1 | 17 |
| Dios Con Nosotros | Edinburg TX | 62.0 | 2 |
| Iglesia La Vid | Laredo TX | 56.4 | 1 |
| Grace Community Presbyterian | Bremerton WA | 49.4 | 9 |
This is a heuristic on congregation names, not a measurement of anyone's ethnicity. The English-named congregations among the sixty-nine include more immigrant churches than these fifteen. Check it against presbytery records before repeating it.
Access runs 0 to 100 and sorts every tract into five tiers. Higher means less access.
| State | PCA churches | State | PCA churches | State | PCA churches |
|---|---|---|---|---|---|
| DC | 1 | RI | 2 | ID | 3 |
| ND | 1 | WY | 2 | ME | 3 |
| VT | 1 | AK | 3 | NV | 3 |
Arid is where the shortfall lives: nearly a fifth of American tracts are Arid and a ninth of PCA congregations sit in one. A denomination shaped like the American suburb has 26 churches in the entire rural United States. Nevada is the one to look at twice. It carries 30 of the country's 547 churchless high-need tracts and the PCA answers with three congregations.
Half of PCA congregations sit in a tract holding eight or more churches. A quarter sit in tracts with fourteen or more. The densest holds 94.
| State | PCA churches | In tracts with 20+ | Sole church in tract |
|---|---|---|---|
| South Carolina | 107 | 50 | 1 |
| Mississippi | 89 | 40 | 0 |
| Georgia | 113 | 23 | 2 |
| North Carolina | 105 | 17 | 3 |
| Tennessee | 77 | 16 | 2 |
| Florida | 155 | 13 | 13 |
| Alabama | 90 | 13 | 2 |
| Virginia | 99 | 12 | 3 |
| Texas | 96 | 4 | 11 |
| Pennsylvania | 108 | 2 | 3 |
In South Carolina 47% of PCA churches stand in a tract with twenty or more other churches, and exactly one PCA church in the state is the only church in its tract. Mississippi has none. Florida and Texas run the other way, holding 24 of the denomination's 76 sole congregations between them.
Mean PCA need pressure is 48.4 against a national 50.2. The mean tells you almost nothing here.
Those 114 congregations are already doing the hardest work the denomination does, and the denominational average erases them. Any state of the field that leads with 48.4 has told the PCA it is a comfortable church, and told those 114 sessions their experience is anecdotal.
The pattern layer scores twenty-four features of neighborhood fabric per tract. Ten are risk-named, so the polarity is swapped before any of them are counted together.
The mean PCA tract carries 13.5 unhealthy patterns of 24 against a national 14.2. Four run the other way, and the largest is a mosaic of subcultures: cultural mixing is the one piece of fabric thinner where the PCA is, which is the gateway finding arriving through a different door. Partner is the hardest ground GospelGraph names, thin common life and heavy need at once. It is the posture the PCA is least present in and the Assemblies of God most, at 26.0%.
Four separate instruments measure movement, and they agree that nothing denominationally distinctive is happening.
| Instrument | What it fits | PCA | National |
|---|---|---|---|
| GIM | Shalom, 2020 against 2024 | 19.0% hardening | 17.5% |
| GIM v3 | Slope over twelve ACS windows | 19.0% hardening | 17.6% |
| GDI trajectory | Four access vintages | 66.3% stable | 64.4% |
| Field instruments | Resolved direction, 919 tracts | 35.9% hardening | 34.4% |
Four instruments, no denominational drift, and that is worth saying plainly rather than hunting a trend that is not there. One hundred and seven PCA congregations sit on ground that is hardening and already at need pressure 60 or above. Those are getting harder from a start that was hard, and a church there will feel its ministry getting more expensive over five years without anything changing in the church. Note the pillar gaps: three points on living conditions, just under three on health, one and a half on hope. The denomination has not insulated itself from the thing hardest to fix.
Five hundred forty-seven tracts hold no church in the merged data and sit above need pressure 70. They cluster in Texas, California, New York, Florida, Arizona and Nevada.
| Denomination | Within 5 mi | Within 15 mi | Share within 15 |
|---|---|---|---|
| PCA | 264 | 439 | 80% |
| OPC | 33 | 109 | 20% |
| ARP | 35 | 131 | 24% |
| Assemblies of God | 496 | 520 | 95% |
| Congregation | Miles | Nearest churchless high-need tract | Need | Poverty |
|---|---|---|---|---|
| Tenth Presbyterian, Philadelphia | 0.4 | Philadelphia County, PA | 70.2 | 24.9% |
| Christ the King, Dorchester MA | 0.4 | Suffolk County, MA | 72.0 | 24.9% |
| Redeemer West Side, New York | 0.4 | New York County, NY | 78.4 | n/a |
| Korean Open Church of Tampa Bay | 0.5 | Hillsborough County, FL | 76.5 | 36.5% |
| Hope Church, Scranton PA | 0.5 | Lackawanna County, PA | 72.4 | 20.4% |
| Brookhaven Presbyterian, GA | 0.6 | DeKalb County, GA | 71.8 | 29.9% |
Four out of five of the hardest empty tracts in the country already have a PCA church close enough to drive to on a Tuesday. Whatever keeps the denomination out of them, it is not the distance. These are straight lines between tract centroids, so they give direction and rough scale, never drive time: a gap eight miles away across a river is further than one twelve miles up a road.
Three other denominational rosters, run through the same pipeline against the same tract index.
| Measure | PCA | OPC | ARP | AG | National |
|---|---|---|---|---|---|
| Congregations joined | 1,734 | 334 | 256 | 11,008 | — |
| Mean access gap | 46.0 | 46.2 | 48.2 | 50.3 | 50.0 |
| Mean need pressure | 48.4 | 45.2 | 51.0 | 52.9 | 50.2 |
| Mean Shalom, latest era | 51.6 | 54.8 | 49.0 | 47.1 | 49.0 |
| Mean churches in tract | 10.9 | 8.5 | 17.6 | 10.0 | 6.8 |
| Mean tract poverty | 12.2% | 10.0% | 12.7% | 14.8% | 13.2% |
| Mean ministry alignment | 48.2 | 45.1 | 49.8 | 53.3 | 50.4 |
| Gateway dominant | 4.0% | 4.8% | 4.7% | 11.6% | 9.4% |
| Posture: Partner | 16.6% | 11.1% | 25.4% | 26.0% | 20.9% |
| Field signal hot | 13.8% | 10.0% | 27.9% | 21.1% | 22.7% |
| Limited-English modifier | 3.0% | 2.7% | 3.5% | 8.0% | 5.4% |
| Unhealthy patterns of 24 | 13.5 | 14.2 | 14.2 | 14.5 | 14.2 |
The ARP is the most crowded roster of the four at 17.6 churches per tract, which is what a denomination concentrated in upstate South Carolina and western North Carolina looks like. The OPC runs the other way on nearly every measure. It has the lowest need pressure of the four, the highest Shalom, the lowest poverty, the fewest hot signals and the smallest share of the hardest Partner ground, and at 8.5 churches a tract it is the least crowded of the three Presbyterian bodies. It is the calmest map of the four, and it is also the one least present on the frontier: two tracts of the 2,186. Where it beats the PCA is the gateway share, 4.8% against 4.0%, which is still about half of the 9.4% those neighborhoods occur at. The uncomfortable comparison is still the AG at 11.6%. Across 11,008 congregations it also sits above the national mean on need pressure, poverty and ministry alignment, and inside fifteen miles of 95% of the hardest empty tracts in the country.
Ministry alignment here is recomputed as the mean of the eight domains so the column is derived one way for all four; that lands within a few tenths of the figure the print canvas carried. Every other number in this table reproduces it exactly.
Five limits, and the first two belong at the front of the room rather than in the footnotes.
One hundred sixty-four carried no coordinates in the roster; fifteen are Connecticut churches on a tract vintage the index does not hold. They are not spread evenly. Mississippi loses 23, Alabama 20, Georgia 20, South Carolina 13, with names like Boligee, Faunsdale, Lowndesboro and Little Sandy Ridge. That is the rural Deep South, so the 1.5% rural figure is a floor and the suburban share is overstated by the same amount. Getting coordinates for those 164 rows is the highest-value fix available to this dataset.
The layers cover different tract counts, and every share here is taken against its own layer's scored base rather than against 1,734. Field instruments resolve for 919 PCA tracts, field signals for 1,068, the segmentation for 1,733. Anywhere a share looks surprising, check the base.
The Shalom v2 five-domain profile is served one request per tract rather than from a national cache, so 1,734 congregations would mean 1,734 requests. The dump carries the three Shalom pillars and the steepest falling domain per tract instead.
Thirty-seven PCA congregations sit in tracts the church-point data says contain zero churches. Immanuel Church in Philadelphia is one, and its nearest churchless high-need tract is its own, at zero miles. That is a gap in the church data, not a finding.
Nothing here counts members, attendance, giving, conversions or health. It is where the buildings are and what kind of ground surrounds them. A church on hard ground is not thereby faithful and a church on comfortable ground is not thereby compromised. Access is a fact about a place, not a verdict on a session.
What would turn this into a trend: the roster carries no planting dates. If the Administrative Committee holds an organization date per congregation, the same pipeline answers the question that matters. For churches planted since 2015, what is the mean need pressure, the mean church count and the gateway share of the tract, against churches planted before 1990? If the newer cohort sits on emptier ground, the denomination has already corrected and this report describes its history. That is one column, and it is worth asking for.
Published alongside this at gospelgraph.com
Every GospelGraph resource · the glossary, the methodology and the rest of the field reports.