Chapin, Connected

We built a complete graph of the Chapin community's street network — 482 miles of streets and 1,510 intersections across the town, the Amicks Ferry peninsula, Timberlake, White Rock, and Ballentine — and measured how it actually connects. (Little Mountain, Peak, and Pomaria are separate municipalities and are excluded.)

Three vital signs, and none of them are close to healthy:

• 38.9% of all street-ends are dead-ends. 963 cul-de-sacs and stubs against 1,510 true intersections. More than one in three streets simply stops.
• Link-node ratio: 1.126. Planners score network connectivity by the ratio of street segments to intersections. A connected grid runs 1.6 or higher. Chapin sits at 1.126 — deeply dendritic. A tree.
• Median detour factor: 1.708. For trips of half a kilometer to 3.5 km, the median Chapin journey drives 1.708× the straight-line distance. Ten percent of trips drive 3.336× — nearly four miles of road for one mile of geography. Traditional towns run 1.2–1.35.

Some of this is geography — Lake Murray severs the map, and no street grid can cross a cove. But the lake explains only the edges. The interior pattern — subdivision after subdivision, each with one entrance, each turned away from its neighbors — is a choice. It was the standard American choice from about 1950 onward. Chapin simply built a lot of it, fast.

Network science has a tool for finding the streets a network forces trips onto: betweenness centrality. Run it on Chapin and the answer is exactly the streets everyone already complains about.

Amicks Ferry Road carries the highest forced load in the study area by a wide margin — the peninsula's single spine, with thousands of homes and no alternative. Then Broad River Road and Dutch Fork Road (the only spines serving White Rock and Ballentine), Chapin Road, Columbia Avenue, Old Lexington Highway. These streets aren't busy because people love them. They're busy because for most trips, the network offers no other legal path.

This is what engineer Walter Kulash called vertical thinking about traffic: when the trunk congests, thicken the trunk. It has a fifty-year track record, and the record is that widened arterials fill again — because the widening does nothing about why every trip needs the arterial. Lateral thinking asks the other question: can trips spread across more paths? That's not a bigger road. It's a better-connected network.

We taught the computer to find every place in greater Chapin where two streets nearly touch but don't connect — dead-ends within a few hundred feet of each other, separated by a strip of woods and a property line.

It found 153 distinct near-misses. After filtering out anything that crosses water, I-26, or the rail line, 42 street-buildable candidates remain. The single best example: two street-ends that sit about 120 feet apart — while the drive between them today is more than two and a half miles, so far around that connecting them would save nearby trips an average of over five miles of driving, each way. A hundred and twenty feet of pavement versus five miles of driving, every trip, every day, forever.

These gaps are not engineering problems. Most are a few hundred feet of flat, buildable ground. They're ownership and intention problems — the connection was never required when the subdivisions were platted, so it never happened. Which means the fix costs almost nothing by road-building standards. It just has to be chosen.

Take just the top ten connectors — 1.35 miles of new neighborhood street, in total. Reconnect the network and re-run every measurement. Here's what changes.

Every arterial in town gets lighter. Chapin Road's forced load falls 46%. Old Lexington Highway falls 23%. St. Peters Church Road falls 43%. Columbia Avenue — the road getting the $80 million — falls 37%. Even Amicks Ferry, pinned by lake geography, sheds 27%.

The town-wide median detour factor drops from 1.708 to 1.678, and the worst-10% trips improve from 3.336× to 2.884×. And 102 more homes come within a one-mile walk of a school, a grocery store, or downtown — a 2.7% gain in walk-access, from pavement measured in feet.

The $80 million project widens the trunk. The 1.35-mile project shrinks the tree. The second protects the first.

Streets aren't the only web. Chapin has a second network waiting in plain sight: its creeks, its rail corridor, and its utility easements — continuous, gently graded corridors that touch nearly every neighborhood.

Overlay them and a town-scale greenway system draws itself: creek arms reaching from Lake Murray up between the subdivisions, the rail line running through the middle of town, existing paths ready to be stitched in. Where a street connector is politically hard, a walking-and-bike connection along a creek is often easy — no cars, no cut-through fear, and it delivers much of the same access. A child who can bike to school on a greenway is a car off Columbia Avenue at 7:40 a.m.

This is the forward-looking layer of the study: a Chapin where the street web and the green web grow together — where growth arrives (it will) onto a network that spreads it instead of stacking it. The town is still small enough to choose this. That window is measured in years, not decades.

A sharp piece of reader feedback sent us back to the data with one more question: when a subdivision gets approved and its traffic lands on the arterials — who actually pays for those roads? In South Carolina, the answer is almost never the government that approved the subdivision.

South Carolina runs the fourth-largest state-maintained road system in America — about 41,000 centerline miles. Roughly 53¢ of every road-mile dollar in this state is a state responsibility, against a national norm around 19%. And it isn't just interstates and US highways: the state system reaches all the way down to two-lane neighborhood collectors. Amicks Ferry, Columbia Avenue, Old Lexington Highway, St. Peters Church Road, Three Dog Road — every one of them is a state secondary route.

Put those three numbers together and the incentive problem comes into focus. The land-use decisions are local: a subdivision plat approved here, a rezoning there. But 90% of the traffic consequence lands on roads the local governments don't maintain. When the entity that approves growth doesn't carry the cost of the congestion it creates, the discipline that once forced careful decisions — can we afford the roads this pattern requires? — simply never activates. The bill is real; it just arrives in Columbia, in the state's pavement budget, instead of at town hall.

It also explains, in one map, why the connectors in Chapter Four never get built. Every candidate connector is a local street — it would be funded by the town, a developer, or a neighborhood — while nearly all of its benefit shows up as relief on state assets: Chapin Road's forced load falls 46%, Columbia Avenue's 37%. The actor who would pay is not the actor who saves. A serious connectivity program here almost certainly means SCDOT participation — the agency with the most lane-miles to protect has the most to gain from a town that stops forcing every trip onto its arterials.

There's a longer story about what happens when a city does pay for its own streets — Charleston studied one annexation for a full year in 1849, when its street bill was its own. We're telling that story in a coming study.