Eighteen NIBRS offense classes. Every reporting agency in the United States. Population-adjusted so small jurisdictions are not over-rewarded by one-off rates. Public methodology summary included.
Used by relocation desks, P&C underwriters, newsroom data desks, multifamily investors, and policy researchers. One dataset, five framings. The free sample is below.
Composite safety index, 51 states + DC, 2020-2025 source coverage. Top five (safest): Vermont (90), New Hampshire (88), Maine (85), Idaho (82), Rhode Island (80). Bottom five: New Mexico (39), Alaska (38), Louisiana (42), Arkansas (45), Mississippi (50). Population-adjusted on a log-10 curve.
Change in composite score, 2024 → 2025. The seven biggest movers in each direction.
The Vermont statehouse reporter has thirty minutes before the 6 p.m. budget meeting. The story is whether last session's bill moved the needle on assault rates. She needs a one-page state brief her editor will let her cite by URL.
Stacked share of the total crime penalty. Pick a state to repaint. Six rolling categories shown here; the full 18 ship in every Excel edition.
Cosine similarity on the full 18-offense vector. Relocation desks use this to find lookalike markets when the original one prices out.
Janet runs the relocation desk at a brokerage in Austin. For ten years, every buyer asked about crime and she said the same thing: "I can't legally answer that." In April HUD said she could. She needs an answer she can read in two minutes and a methodology page she can link to without losing the next listing.
Hatched fill = states with agencies reporting below the 80% threshold. We do not impute the difference. We disclose it on the front page.
NIBRS adoption is uneven. Some states report at 99%+, others below 60%. Every visualization on this page is gated to states meeting our 80% threshold; the rest are flagged, not estimated.
This is the disclosure most aggregators skip. We publish it under the hero.
No black-box claims. Pick a state and see how the inputs move. The public method summary lives at /methodology.
A P&C underwriter at a mid-size carrier is preparing a rate filing for the state DOI. The crime-risk vendor she has been using will not explain its sources or limitations. The DOI reviewer will not approve a filing whose inputs cannot be defended in writing. She needs a crime score with source notes, coverage notes, and a documented audit path.
Density = how thoroughly a state's data is covered. The denser the column, the deeper county- and city-level breakdowns can go without imputation.
Choropleths reward large states. The hex cartogram treats Rhode Island and Wyoming as equals. The eye finds different stories.
Before you buy, you should see the format. Every paid suite ships as a PDF for reading and an XLSX or CSV for working. Programmatic access is custom-scoped only.
SAFETY ATLAS · VOL. I · 2026 Vermont · Composite 63.49 · Rank 1 ───────────────────────────── Personal 85.4 (national 50.0) Property 63.8 Community 56.9 ───────────────────────────── TOP 5 SAFEST SCORE Rutland Town 90.9 Stowe 88.2 Manchester 86.4 Hartford 83.7 Williston 81.6 ───────────────────────────── Cite as: Safety Atlas (2026). "Vermont 2026." Vol. I, p. 51.
CITY_ID POP ADJ. HOM/100K ASS/100K BURLINGTON_VT 44,649 47.09 2.24 389.71 S_BURLINGTON_VT 20,488 57.77 0.00 170.83 RUTLAND_VT 15,747 57.17 12.70 393.73 RUTLAND_TOWN_VT 15,747 90.94 0.00 38.10 … 12 sheets per state · 63 columns per city Cover · Cities · Top 50 · Bottom 50 Watchlist · Leaderboards · Neighborhoods NIBRS Victims · Offenders · Weapons NIBRS Hotspots · Methodology
city/VT/burlington.json
{
"city_id": "BURLINGTON_VT",
"population": 44649,
"adj_score": 47.09,
"raw_score": 43.11,
"categories": {
"personal": 55.32,
"property": 27.70,
"community": 35.71
},
"coverage_note": "reported agency coverage disclosed",
"offenses": { "HOM": {...}, ... },
"as_of": "2025-12-31",
"methodology": "v1.0"
}
PDF and XLSX ship with every suite · CSV/JSON exports are custom-scoped.
Each suite is a bundle of views, packaged for one buyer. The data is the same. The framing — and the depth — is what differs.
Single state, national bundle, or the all-five Atlas. Single payment, no subscription, 7-day refund by reply.
For realtors, brokerages, and corporate mobility teams.
For brokers, underwriters, and rate-filing committees.
For data desks, state papers, and syndication services.
For PropTech, REITs, multifamily owners, and CRE analysts.
For NACo, NGA, academics, and expert witnesses.
All five suites, one license. Custom scope and pricing.
An analyst at a multifamily REIT is screening secondary markets in the Southeast. The pitch deck needs an Econ × Crime quadrant — high median income, top-quartile safety, growth permits up. No vendor sells that combination openly. She has a Tuesday meeting.
Provenance entries appended on every ingest. Methodology changes flagged separately. Errata trigger automatic refunds for the prior 30 days.