A county that looks fine on paper and isn’t, in fifteen places.

This report is built from public records: church locations, road networks, and census figures about the households in each neighborhood. That means it can see some things well and other things not at all, and the difference matters before you trust a single number in it.

It can see where church buildings are, how far people live from them, how many people live in each part of the county, and what census conditions those households are carrying. It measures structures.

It cannot see faith. It does not know who prays, who has been converted, who is discipled, or whether any particular congregation is healthy or faithful. A neighborhood this data calls thin may be full of Christians who drive fifteen minutes to church. A neighborhood it calls well-served may be spiritually cold. Nothing in this report is a judgment on anyone’s soul or on any church’s ministry.

Hold both of those together and the data becomes useful. It answers one narrow question: where in this county would it be structurally hardest for a person to encounter a gospel-preaching church? That is worth knowing. It is not the same question as where God is working.

Anne Arundel’s reputation is affluent and coastal, and the overall numbers agree with it. Averaged across its 127 scored neighborhoods, the county’s thin-presence score comes in a shade better than Maryland’s and well better than the country’s. Compared against the 338 American counties of similar size, Anne Arundel is better resourced than roughly 85% of them.

mean thin presence · Anne Arundel 45.2 · Maryland 46.2 · national 50.0

Now look at the spread instead of the middle. The distance between Anne Arundel’s best and worst neighborhoods is wider than in two thirds of those same counties. And the severe end runs ahead of the state: a larger share of Anne Arundel neighborhoods sit at 60 or worse than across Maryland as a whole. The county has proportionally more badly underserved neighborhoods than the state it sits in, while posting a better average than that state.

share of neighborhoods at 60+ · Anne Arundel 12.3% · Maryland 9.3%

Both facts are true at once, and only one of them ever gets said out loud.

One thing worth saying plainly, because it cuts against the alarm: Anne Arundel has no Desert-tier neighborhood. The worst score in the county is 69.8. Nothing here reaches 75. A church planter choosing between this county and Baltimore City on severity alone should go to Baltimore City. The case for Anne Arundel is different, and it is about scattering and invisibility rather than depth.

The map is where this becomes a pastoral problem rather than a statistical one. The dark neighborhoods do not sit together in one poor quarter of the county where a benevolence committee would already be looking. They are scattered: on the Broadneck and Pasadena peninsulas, out along the Odenton corridor, on Annapolis Neck, and down in south county. No single church’s natural field of vision covers more than two of them.

GospelGraph sorts American neighborhoods into thirty household types, built from how people in each block group live: what they do for work, who they live with, whether they own or rent, how old they are. Anne Arundel contains 25 of the thirty, which is a wide spread for a single county.

One type dominates. Affluent Owner Family Suburbs holds 25.8% of the county’s households and is the largest single share in 33 of its 130 neighborhoods. Add the next two and you have half the county. The peninsulas and south county are made of it; north county is not.

The parts of the county are made of different things. Nearly half of the peninsulas and of south county is Affluent Owner Family Suburbs. North county’s largest share is Cost-Pressed Family Suburbs, a type whose defining fact is that suburbs built for two adults and two cars are being run by households with one of each. West county leads with Dense Professional and Solo Households, and Annapolis with Professional Multifamily Households.

largest type share by part · peninsulas 42% and south county 46% Affluent Owner Family Suburbs · north county 22% Cost-Pressed · west county 25% Dense Professional · Annapolis 19% Professional Multifamily

That is four different pastoral situations inside one county, and it is the strongest argument in this report against a single county-wide strategy.

Each panel gives the type’s own one-line summary, one thing people in it commonly say that means something other than it appears, one approach the guide warns tends to backfire, and one move it recommends instead. All of it is quoted from the field guide for that type.

The layers are meant to be read in an order, and the order matters more than any single number in them.

Take Ferndale, tract 7508.03, the county’s top-priority neighborhood. The score puts it in the Arid band, the county’s second-worst. The type says Cost-Pressed Family Suburbs. And of the community patterns the fabric layer can read there, most come back broken or missing and only a handful intact.

Ferndale 7508.03 · thin presence 65.9, Arid · 22 of 41 patterns read: 4 present, 4 thin, 10 broken, 4 absent

Two of the broken ones sit at the top of its priority order. Neighborhood Tissue is marked absent: “Families here are civil, scheduled, and unknown to each other.” Activity Nodes is marked broken: “Parents in the pickup line know each other’s cars and not each other’s names.”

Both come with first moves that are ordinary church work. Open the building one weekday for something with no religious ask attached, and hold it for a year before evaluating it. Map who in your congregation lives in the neighborhood and ask them to do one weekly errand inside it. Pick one day and one hour a week and keep it without exception for twelve months, because a node that is sometimes open is not a node.

And both come with checks you can run this week, which is the point of the whole sequence:

This data cannot see whether a church is healthy, so nothing here is a report on any congregation. What it can do is name what a county made of these household types tends to be good at, and what the field guides warn happens when a church is good at exactly that.

Anne Arundel churches are, on the evidence of the county’s composition, likely to be strong at the things capable and well-resourced congregations are strong at: welcoming visitors properly, running programs that work, recruiting volunteers, and funding what needs funding. Half the county is made of households that are good at those things and expect them.

Every one of those strengths appears in the field guides as a failure mode.

Each of these is either something the field guides name directly or something this study’s own numbers contradict. None of them is an accusation. They are the assumptions most likely to be quietly true in an Anne Arundel church, and each has a check attached.

Fifteen of the county’s 130 neighborhoods contain no church at all in the records this study uses. Three of those are the blank spots discussed later, mostly water and industrial land. The other twelve are ordinary places where people live, sleep, raise children, and have no congregation inside the boundary.

The distances are the part worth sitting with. In Annapolis Neck, tract 7026.01, six thousand people live 1.2 miles from the nearest church. In Arnold, tract 7311.07, out on the Broadneck peninsula, nearly four thousand people are 1.3 miles from one. Nobody would call these remote. They are places where the subdivisions went up and nothing was planted alongside them.

Ten of the twelve populated ones are led by Affluent Owner Family Suburbs or High-Achievement Homeowners. The guide for the first warns that these suburbs are “friendly, organized and easy to join” and that newcomers are “greeted, invited and slotted into things fast,” which looks like belonging and is not. A neighborhood with no congregation in it and households who join things easily can look perfectly healthy from a distance.

Churches usually find need by looking for poverty. In Anne Arundel that method walks straight past the worst neighborhoods in the county.

Eleven neighborhoods score 55 or worse on thin presence while fewer than one household in twenty is below the poverty line. Four are under one in twenty-five. The worst-scoring neighborhood in the whole county, Lake Shore, tract 7313.06, out on the Pasadena peninsula, pairs one of the county’s lowest poverty rates with a wellbeing reading that is comfortable by any standard.

Lake Shore 7313.06 · poverty 3.2% · wellbeing 63.9

Two are stranger still. Shady Side, tract 7070.01, in south county, reports almost no poverty at all and has eleven churches, and still lands in the worst band, because the nearest of those eleven is three quarters of a mile from where its people live and the neighborhood is large. And Odenton, tract 7407.03, reports the same near-zero poverty and one church.

Shady Side 7070.01 · poverty 0.37%, thin presence 62.8 · Odenton 7407.03 · poverty 0.28%

Nine of the eleven are led by one of two types, Affluent Owner Family Suburbs or High-Achievement Homeowners, and the guides for both say the same thing about outreach here: there is no material need to find, so a ministry that screens for need will screen these neighborhoods out. That is this finding arriving from the pastoral side rather than the statistical one.

The wellbeing score is built from five separate areas, weighted and combined into one number. Anne Arundel’s combined score is healthy, which is exactly why a county-level reading of this county tells you almost nothing. Split it into its five parts and something specific appears.

Against Maryland, Anne Arundel is ahead on work and provision, on safety and justice, on family and household, and on community life, with the widest lead on safety. On health it is behind by ten points.

vs Maryland · work +6.4 · safety +12.4 · family +6.8 · community +9.0 · health −10.0

Health carries the second-largest weight of the five, so this is not a rounding detail. It is the largest single structural weakness this data attributes to Anne Arundel, and it is invisible in the combined score because four other areas are carrying it.

Health is also the weakest of the five in 50 of the county’s 130 neighborhoods, more than any other area. The floor is in north county, which reads well below the peninsulas and further still below Annapolis, and which carries nearly ninety thousand people at an overall wellbeing score roughly twenty points below the peninsulas’.

health by part · north county 36.9 · peninsulas 51.7 · Annapolis 64.1 · north county wellbeing 46.5 across 87,835 people

The area covers access to health care and health outcomes together. This data does not publish which specific indicator moves for a given neighborhood, so the honest statement stops at the name of the area. What a pastor can do with it is still concrete, because health access has an address.

North county’s leading household type sharpens this. Cost-Pressed Family Suburbs is defined by families running a two-car suburb on one car and one adult, and its field guide names weekend hospital and care-home shifts as a fixed feature of the week. A part of the county whose largest type works weekend care shifts, and which also scores worst in the county on health, is not a coincidence worth ignoring.

The published surface gives one health score per neighborhood, so the obvious question, whether this is mental health or physical illness, looks unanswerable. It is not. The model underneath splits Life & Health into four constructs, and the build artifacts carry a score for each. Reading them changes what this finding means.

Three of the four are Anne Arundel strengths. The county beats the national mean by double digits on longevity and child survival and on disease burden, and sits at exactly the national average on environmental health. The entire deficit is in one construct: sobriety and freedom from dissipation, where Anne Arundel scores eleven points below the national mark.

vs national · longevity +12.7 · disease burden +14.0 · land 0.0 · sobriety 40.0 vs 51.0

And the disease-burden construct can be opened further, because it is built from two weighted halves, a physical burden index carrying the larger share and a mental distress index the rest. Reading the underlying CDC PLACES prevalences for Anne Arundel’s 340 block groups against the national set answers the mental-versus-physical question directly.

disease burden weights · physical 60% · mental distress 40%

Anne Arundel is better than the national average on both mental-health measures, on depression by a little and on self-reported poor mental health by enough to land it better than roughly seven American block groups in ten. On the eight physical measures it beats the national figure on seven. The single exception is obesity, where it runs a point and a half above.

depression 20.6% vs 21.4% national · poor mental health 15.0% vs 16.9% (28th percentile) · obesity 34.6% vs 33.0%

So the answer to both halves of the question is no. Anne Arundel’s health weakness is not chronic disease and it is not mental distress. It is alcohol and overdose.

Averaging the drinking score across a county this varied hides the thing a pastor needs. Split it by part of the county and the health finding comes apart into two problems that sit in different places and want different responses.

North county has the illness problem. Its health score is the county floor, its longevity reading sits more than twenty points below the peninsulas’, and it is the only part of Anne Arundel where chronic illness runs above the national rate. Its drinking score, though, is the best in the county outside Annapolis, whose higher figure is an average of two halves that disagree and is taken apart below.

north county · health 37.4 · longevity 41.9 · sobriety 44.9

The peninsulas have the drinking problem. Broadneck and Severna Park sit on the county’s sobriety floor while running some of its longest lives, with chronic illness a full point below the national rate. People there live longer and get sick less than the country does, and drink more.

peninsulas · sobriety 33.5 · longevity 65.1

The two move against each other across the whole county, not in those two corners alone. Sobriety and disease burden correlate at r = −0.34, which in plain words means the neighborhoods that drink are not the neighborhoods that are sick. A single county-wide health ministry aimed at the average would miss both.

The same split shows up in the personality types, and it points where nobody looks first.

Four types in this county are affluent owner-occupier suburbs: Affluent Owner Family Suburbs, High-Achievement Homeowners, Affluent Mature Owner Couples, and Affluent New-Build Families. Together they cover 64 of the county’s residential neighborhoods, they score seven points worse on sobriety than every other classified neighborhood, and their poverty rate runs at less than half of everyone else’s.

affluent owner types · sobriety 38.2 vs others 45.2 · poverty 3.5% vs 8.4%

That is not an artifact of age or of income measured badly. Holding age composition fixed, the poverty rate and the sobriety score still move together, and both run so that higher is better. In plain words: across Anne Arundel, the poorer a neighborhood is, the better it scores on drinking and overdose. Holding poverty fixed, older neighborhoods score better than younger ones.

partial correlations · poverty and sobriety r = +0.52 with age held · age and sobriety r = +0.59 with poverty held

This lands on the peninsulas hardest because that is what the peninsulas are made of. Thirty-one of their 39 neighborhoods are led by one of the three affluent owner types. The part of the county with the best houses, the longest lives and the least illness is the part carrying the county’s worst drinking score.

The lowest-scoring type with enough neighborhoods to table is a different story: Multifamily First-Timers, the new-build townhouse stock in west county and Pasadena where people buy early and often alone, scores in the twenties. Below the tabling threshold, three renter types score lower still and are named here rather than hidden: Garden Apartment Young Adults, Car-Oriented Creative Renters, and Young Urban Working Renters, the first two in single digits. At the other end, Young Multifamily Workers and Established Blue-Collar Homeowners score far above everything else in the county, both across two neighborhoods.

sobriety by type · Multifamily First-Timers 28.4 in 6 tracts · Garden Apartment Young Adults 5.6 in 1 · Car-Oriented Creative Renters 9.8 in 3 · Young Urban Working Renters 16.6 in 1 · Young Multifamily Workers 79.8 in 2 · Established Blue-Collar Homeowners 73.5 in 2

The neighborhood personality types are not decoration on this finding. The two parts of the county with the two problems are led by different types, and the guides for those types describe exactly the conversational move each problem requires.

On the peninsulas, the leading type is Affluent Owner Family Suburbs, and its first protective mask is “We are on top of it, thanks,” which the guide reads as “Usually true, and always the end of the talk.” Another is “This season is just insane for us,” which it calls “the only trouble it is acceptable to admit here. Nobody in this neighborhood says anything is going badly.” A household managing somebody’s drinking in that neighborhood has a ready-made and socially perfect way to never say so, and it is the same sentence everyone uses about work.

The guide’s own first action is ask twice: “Everybody here answers the first question with a version of fine. That is manners, not evasion. The second question, asked slowly, is where this neighborhood opens.” Its fourth is make the struggling child sayable: “Talk openly about your own children’s hard patches. Where progress is the public account, somebody has to make the other kind of report survivable.” Both were written about a neighborhood type, before anybody looked at its drinking score, and both are the correct move for it.

In north county, where the problem is illness rather than drinking, the two leading types are Cost-Pressed Family Suburbs and Middle-Income Established Homeowners. The guide for the second names “We are very fortunate, really” as a mask that “heads off any suggestion that comfort is worth a look,” and “It is just the three of us now” as the divorce nobody mentions, in a place where “the street has no words for it.” Its practical action is bring skill to old houses: “Four in five of these homes predate 1980 and their systems are failing together. Honest help with a furnace or a roof is welcome and is a real way in.” A furnace and a chronic condition are managed by the same household on the same budget.

Multifamily First-Timers, the lowest-scoring type with enough neighborhoods to name, carries the mask the guide calls busy with no witnesses: “Real work, done alone, in a room. Nobody saw it happen.” Its action list includes “Somebody here can disappear entirely with nothing appearing to be wrong. Keeping track of who you have not seen is a real pastoral act.” Written about isolation, and it is the single most useful sentence in this report for a drinking problem, because drinking alone in a townhouse looks like nothing from the street.

The wellbeing model carries a sixth domain, Peace, which measures reduction of violence. It is excluded from the composite everyone is scored on, and the published glossary describes it as an inactive pilot covering a small fraction of the country.

That pilot is Anne Arundel. Every tract in the United States carrying a Peace score is in this county, and no other county has one. Its scores here run nearly the full range with a median just above the middle, so it is discriminating between neighborhoods rather than returning a flat value.

Peace pilot · 129 tracts, all in Anne Arundel · range 1.6 to 98.8 · median 60.3

Nothing in this report uses it, because it is excluded from the composite by design and because one county is no basis for calibration. It is recorded because Anne Arundel churches should know that their county is the place where a sixth measure of neighborhood peace is being tested, and that they are the only people in the country who could tell anyone whether it reads true.

Both are reasonable guesses in a county with a large retired population and three major installations. Both are wrong, and both are wrong in the same direction.

Across the 127 neighborhoods where both can be measured, military adjacency and the health score move together, not apart. The quarter of the county scoring worst on health is mostly non-military; the quarter scoring best is overwhelmingly military-adjacent. Elderly share behaves the same way: more residents over 65 goes with a better health score, not a worse one. Neither association is explained by income; controlling for poverty moves both by less than one hundredth.

military and health r = +0.51 · worst-health quartile 12% adjacent, best 80% · elderly and health r = +0.41

What does track the health floor is the working-age share (r = −0.41). The neighborhoods scoring worst on health are the ones with proportionally the most people between 18 and 64.

That fits the construct finding above rather than contradicting it. Binge drinking and overdose mortality are concentrated in working-age adults, not in the retired and not in children. A county whose health deficit is entirely a sobriety deficit should show exactly this pattern: worse where working-age adults are concentrated, better where the population skews older.

There are two ways to ask whether a neighborhood’s conditions are improving. The short measure compares 2024 against 2020. The long measure fits a line across twelve overlapping windows running from 2009 to 2024. Same question, different amounts of history.

In Anne Arundel they disagree about 33 of the 128 neighborhoods where both can be calculated. And the disagreement leans hard one way: 25 neighborhoods look like they are getting worse on the short measure and better on the long one. Only 8 run the other direction.

That lopsidedness is the useful part, and it is a fact about the instrument rather than about Anne Arundel. The short measure subtracts a 2016-2020 census reading that sits about two points high for every neighborhood in the country, because of collection problems during the pandemic. Twelve windows average that out. Two points cannot.

Anne Arundel shows the predicted pattern exactly. Its average short-measure change is −0.44 and its average long-measure change is +1.09 per four years. Sixty-one neighborhoods look like they are hardening on the short measure; on the long measure only 26 do.

Four neighborhoods combine heavy through-traffic with a poor thin-presence score. Thousands of people pass through them daily. Three of the four are up along the Baltimore line.

The through-road figure is what makes this pastoral rather than statistical. Brooklyn Park, tract 7501.01, has a church 600 feet away, which sounds like a solved problem, and more than 3 miles of major road running through it. The Brooklyn Park tract beside it, 7502.04, carries 12 miles of major road, and its nearest church is two thirds of a mile off.

A church 600 feet away on the far side of Ritchie Highway is not 600 feet away to a mother with a stroller. The road figure is the closest thing this data has to a measure of that.

This study set out to find a pattern that would matter a great deal if it were real: neighborhoods with plenty of congregations but few that have lasted. Churn rather than roots.

On the surface Anne Arundel appears to have it badly. Sixty-eight neighborhoods hold four or more churches while showing one or none that the records place there twenty years ago. County-wide, 864 church records yield only 60 that clear twenty years and 5 that clear fifty.

Then you look at how many of those records carry a founding date at all. 215 of 864, which is 24.9%.

Out of 130, the priority scoring marks 13 as something other than ordinary. Most of them are flagged because they combine real need with enough existing structure that a congregation could work there. One is flagged as a crossroads, where movement through the neighborhood is heavy enough that presence there reaches well beyond the residents. Two are flagged as isolated: badly served and far from help.

Ferndale, tract 7508.03, up on the Baltimore line, leads the list. Nationally it ranks 205th out of every scored neighborhood in America. It is the only Anne Arundel neighborhood in the top thousand of anything in this dataset, and it holds nearly seven thousand people at the highest poverty rate in this group.

Ferndale 7508.03 · national rank 205 of 84,410 · 6,782 people · poverty 14.2%

Four of the thirteen have no church inside them at all. Three are in Annapolis and the Neck, four in west county, five on the peninsulas, and one in south county. Once again they refuse to cluster.

Each of the thirteen carries a household type and a fabric reading of its own, which is where the packets stop being a ranking and start being usable. Ferndale 7508.03, worked through earlier, is the template: a score, a type, a set of broken patterns, four first moves and four checks you run on foot.

Three Anne Arundel neighborhoods have no thin-presence score at all. Dropping them from the tables would make the county look better covered than it is, so they are listed here instead.

Two are unremarkable. One is open water with nobody living in it, and one holds seven people on industrial land.

The third is a genuine gap and it matters. Jessup, tract 7404, has 5,535 residents, no thin-presence score, and a reported poverty rate of exactly zero. It also has no household type: the personality layer marks it insufficient household base. Three separate layers lose the same 5,535 people.

What it does have is the lowest wellbeing score in the entire county and one of the steepest long-measure declines anywhere in Anne Arundel. The one neighborhood with no access score is also the neighborhood with the worst conditions in the county. It drops off every ranked list precisely because it needs to be at the top of one.

Jessup 7404 · wellbeing 38.0 · long measure −7.3 per four years

One cause, and the data already found it

Jessup holds a state correctional complex. That is local knowledge rather than something this report derived, but the data agrees, and it agrees in a specific place. At block-group level, 240037404001 carries the modifier Institutional facility, whose own description reads: “Most of the people counted here live inside an institution, not in houses along a street.” It is the only block group in Anne Arundel that carries it.

That single fact explains all three anomalies at once. A poverty rate is calculated over the household population, so a tract whose residents are counted in institutional group quarters returns zero rather than a real figure. A household-weighted personality read has almost no households to weight, so it returns insufficient base. And the access score, which measures churches against the people who live on the streets around them, has no such population to measure.

None of that is wrong. What is wrong is that the explanation is invisible at the level everything else is reported. The tract roll-up returns a bare insufficient household base and drops the modifier entirely, so the one tag that makes sense of Jessup never reaches the tract layer, the ranked lists, or the map.

One person, this week

Walk the distance the data calls short. In Brooklyn Park, start at the church 600 feet from tract 7501.01 and walk into the neighborhood. If it takes more than twelve minutes, or means crossing four lanes without a signal, treat that neighborhood as underserved no matter what the map says (section 09).

Ask the health question that fits your part of the county. In north county, ask how many households are managing an ongoing condition without a regular doctor. On the peninsulas, ask how many are managing someone’s drinking, and ask your leadership before you ask anybody else. The two questions are not interchangeable, and both are falsifiable inside a month against your own rolls (section 07).

Run the one assumption check that stung. Six assumptions about your own strengths sit above, written to be tested. Five cost nothing but attention, and one person can settle any of them in a week (section 04).

One congregation’s leadership, this quarter

Choose a parish. Twelve neighborhoods with residents have no congregation in them, and every one is within 1.4 miles of a church. The move is not a plant. It is a decision by the nearest congregation about which single neighborhood next door it will treat as its parish rather than its catchment, and the nearest congregations are now named (section 05).

Put a weekday evening inside the comfortable eleven. The neighborhoods where need hides will not respond to a food pantry. They are short on proximity and unhurried time, so the move is a presence inside the neighborhood on a weekday evening rather than an invitation out of it on Sunday (section 06).

A network or association, because nothing smaller can

Put a name beside all thirteen priority neighborhoods this month. Thirteen is small enough for a county network to cover and large enough that no single congregation should hold more than two. Begin with Ferndale (section 11).

Send somebody to Jessup without waiting for its score. The worst wellbeing reading in the county belongs to the one neighborhood that appears on no ranked list, because the people there live inside a correctional complex. If anyone in the county runs prison or reentry ministry, this is their neighborhood (section 12).

Rebuild any worsening list with both change measures. A ranking built from the recent-change map alone leans pessimistic. Set the two measures side by side and treat the 33 neighborhoods where they disagree as unranked until somebody visits (section 08).

Date the 649 undated church records. This is a paperwork job that a denominational office or a historical society could largely finish in a few weeks from yearbooks and incorporation records, and it would repair the most broken measure in the county’s data (section 10).

None of this asks you to believe a spreadsheet over your own eyes. Every move ends in a walk, a visit, or a question asked out loud, and each of those can prove this report wrong about your neighborhood, which would be a result worth more than agreement. And none of it waits on anyone new arriving. Every move above already belongs to a congregation, a diaconate, or a network that exists in Anne Arundel today.