Section summary
Per resident is the wrong yardstick for infrastructure money. Per resident, the smallest fifth of towns received 4.9 times their population share of competitive awards. Measured against what each program pays for, the figure is 1.47. In core infrastructure the smallest fifth as a group received 1.29 times its share and the largest fifth 0.99; the money is concentrated in a few winners, and the median small town received 0.73.
Mount Washington keeps up 17.4 miles of town road for 193 residents. Boston keeps up 792 miles for 666,442. That is about 90 miles of road per 1,000 residents in Mount Washington and 1.2 in Boston.
Divide competitive grant money by residents and Mount Washington looks like a winner: $4,402 per resident over ten years, seventh in the state. Boston received $220. That comparison measures population. It does not measure what the grants pay for.
A mile of road, a bridge or a culvert costs about the same to rebuild whether 20 people use it or 2,000. The smallest fifth of towns, 71 towns with 90,673 residents, maintain 2,942 miles of local road: 32 miles per 1,000 residents, against 2.6 in the largest fifth. They own $7,455 of town bridge replacement cost per resident; the largest fifth own $583. Bridges here are scored, not counted: each of 3,771 town-owned structures is valued at its deck area times the federal replacement cost per square foot, for $5.42 billion in all, $603 million of it in poor condition. State and federal bridges are excluded.
Counting the right thing
Each program in the award catalog pays for something specific. We matched each to its own measure: road miles for Complete Streets, road-stream crossings for culvert grants, town-owned high- and significant-hazard dams for dam repair, shoreline for seawall grants, acres of floodplain for resilience grants, jobs for MassWorks, residents for libraries. Then we asked what each town would have received if every program had divided its money in proportion to its own measure. A town that received exactly that scores 1.0.
On that basis the smallest fifth of towns scores 1.47 and the largest 0.92. Small towns still come out ahead, but the gap shrinks from about six to one per resident to about 1.6 to 1. The correlation between award dollars and population falls from −0.54 per resident to −0.14 on the burden index. Most of the size effect is a denominator effect.
In the core infrastructure programs alone, roads and streets, culverts, dams, seawalls, flood resilience and energy, the typical small town receives less than its share. The median small town scores 0.73; the median large city, 0.92. As groups, the smallest fifth received 1.29 times its core-infrastructure share and the largest fifth 0.99: a few small towns that won large grants pull the small-town total above 1.
The link to need gets stronger, not weaker. Per resident, award dollars track income need weakly (rank correlation 0.25) and tax-base need not at all (0.04). On the burden index both are positive: 0.29 and 0.18.
The rankings move. Dense cities and inner suburbs rise because they carry jobs, pupils and cost-burdened households: Revere from 133rd to 22nd, Everett from 148th to 34th. Hill towns fall: Middlefield from 115th to 334th, Tyringham from 139th to 328th. Mount Washington falls from 7th to 155th.
The per-resident figure is not wrong. It answers a different question, and a Boston taxpayer is entitled to ask it. Some programs are capped per community by design, which helps small towns on purpose. A town that keeps a road few people drive makes a choice the state need not underwrite at the same rate as an arterial. Those points argue for publishing both figures.