Department Intelligence — methodology & sources
A public-evidence layer, not a verdict. It records what New Jersey fire departments publicly disclose about money, people, and readiness — additive to, and separate from, the county civilian fire-death score and rankings.
1 · What the Department Intelligence layer is
Department Intelligence aggregates public evidence about New Jersey fire departments — funding, staffing, readiness, ISO/PPC, grants, and transparency. It has two tiers: curated Department Intelligence profiles (individually sourced and verified) and the broader NJ DCA directory records (self-reported survey data). It is additive: it does not change the county civilian fire-death score, the county rankings, any completed-state data, or the methodology behind them.
2 · Curated Department Intelligence profiles
There are 28 curated profiles. Each is built by hand from authoritative public documents — adopted municipal budgets, NJ DCA fire-district budgets, department pages, OpenFEMA grant records, Census population. Every populated value carries a source URL, publisher, document type, year, a confidence label (verified / partial / stale), and a note; anything without a public source is left unavailable (blank). These are the higher-confidence records and the only ones with their own profile page.
3 · NJ DCA directory records
Separately, the NJ DCA Annual Fire Department and Agency Directory (2023) gives 720 department records across 21 counties, parsed with coordinate-based PDF extraction. These are self-reported NJ DCA directory data: FDID, county, municipality, type, roster counts, officers, population, square miles, NFIRS participation, funding category, and an EMS flag, straight from each department’s survey return. You can browse and search all of them.
- Parsed from the 2023 public NJ DCA directory PDF (720 department records, 21 counties).
- Coordinate-based extraction (word x/y positions) recovers the department name reliably — an earlier plain-text parse could not, so it was withheld.
- Per-county roster and department counts mostly reconcile to the directory’s own printed county-summary pages, checked county by county: career firefighter totals match in 19/21 counties, volunteer totals in 14/21, department counts in 12/21, and the career/volunteer/combination type mix in 8/21. 12/21 counties reconcile on all three of department count, career total, and volunteer total (Camden, Cumberland, Essex, Mercer, Middlesex, Morris, Ocean, Passaic, Salem, Somerset, Sussex, Union).
- The counties that do not fully reconcile miss by a small margin — typically one department more or fewer than the printed row, which shifts the volunteer roster total by that department’s membership. Those roster figures stay caveated as self-reported and are never presented as verified. One county — Gloucester — has a printed summary row that is itself inconsistent with its department pages (a defect in the source PDF), so its reconciliation is not meaningful and is flagged separately.
- Each county’s directory panel shows its own reconciliation status, so a reader can see whether that specific county’s totals line up with the printed summary. The curated, budget-verified records above remain the higher-confidence basis regardless.
4 · Why directory records are not full profiles
A directory record is one self-reported survey return; a curated Department Intelligence profile is an independently-sourced, budget-verified record. We do not present directory records as curated profiles, do not give them their own profile pages, and do not treat their self-reported figures as independently verified. The directory adds statewide breadth; the curated set keeps the higher-confidence depth. Where a directory record maps cleanly to a curated one, it links to the fuller profile.
5 · Roster count vs on-duty staffing
Directory staffing figures are a roster count, not on-duty staffing — the number of firefighters a department reported on its membership roster, not how many are staffed on any shift. A career department’s roster and its on-duty complement are very different numbers, and the two must not be conflated. Curated profiles similarly distinguish roster/headcount evidence from on-duty minimum-staffing where a source provides it.
NFPA staffing benchmark
Each department profile carries an NFPA Staffing Benchmark section — a staffing-only comparison against the NFPA deployment standards. It does not assess turnout, travel, or total response-time compliance, and it is not a legal finding.
- Benchmark edition. NFPA 1750:2026 · career-department provisions (Ch. 4–6, §1.3). Career departments are measured against the baseline per-company minimum of 4; the elevated (5 — urban / high-volume / geographically restricted or isolated) and maximal (6 — dense-urban / high-hazard / tactical) company minimums are not assessed without sourced zone/hazard evidence, and are never inferred. Volunteer departments fall under the NFPA 1750:2026 volunteer-department provisions (Ch. 7–8); combination departments (NFPA 1750:2026 combination-department provisions) are assessed on their career companies; fire-district/municipal/unknown records are “NFPA 1750:2026 · applicability unresolved.”
- Baseline only — not full compliance. This assesses baseline company staffing only. The effective response force (low 20/21, medium 28/29, high 42/43) and response-time objectives (turnout 80s fire / 60s EMS; travel 240 / 360 / 480 / 610s, each ≥90%) are not assessed, and department-wide shift staffing is never divided into a per-company figure. A “meets baseline” result is never a claim of full NFPA 1750 compliance.
- Roster is not on-duty staffing. A verdict is computed only from a cited public source stating actual minimum on-duty staffing, staffing per company, or per apparatus. Career/volunteer roster headcounts, station counts, apparatus counts, ISO ratings, SAFER awards, budget, and population are never used to reach a verdict.
- Verdict logic. Meets / Below appear only when a qualifying on-duty figure exists to compare. Where the standard is known but on-duty staffing is not published, the result is “Not publicly disclosed.” Where the governing standard itself can’t be established, it is “Insufficient public evidence.”
- Benchmark, not mandate. NFPA 1750:2026 is a consensus standard. It binds a department legally only where its Authority Having Jurisdiction has adopted it, so no record is called “legally noncompliant.”
- Source quality is ranked. For a verdict we accept only, in priority order: official department/municipal/fire-district pages; adopted budgets, ordinances, contracts, and annual reports; CPSE/CFAI Standards of Cover or accreditation documents; and NJ government fire records. News, social media, job postings, third-party directories, and vendor pages are treated as contextual only and never used to reach a verdict.
In this New Jersey pilot, only departments that publish an actual per-company on-duty minimum (e.g. Jersey City’s “4 per company”) receive a meets/below baseline company staffing result; every other record honestly reports the governing provisions with staffing “not publicly disclosed,” and higher-risk, ERF, and response-time dimensions stay “not assessed.” See the NFPA staffing benchmark dashboard for the statewide summary and transparency gap.
6 · Budget basis definitions
Budget figures are public evidence, not fully normalized spending comparisons.
New Jersey publishes fire money on incompatible bases, so every curated budget carries a basis stating what the dollars cover. Two figures are only comparable when their basis matches. The key bases:
- Fire S&W + OE salary_wage_and_operating
- Adopted-budget Fire line: Salaries & Wages + Other Expenses. Usually excludes pension/health, which sit in citywide lines.
- Fire district adopted budget fire_district_adopted_budget
- An NJ fire district's total adopted appropriations (published under NJSA 40A:14-70.2).
- Member contributions (reconstructed) member_contributions_reconstructed
- Not a published figure. The regional department publishes no budget of its own, so this is the sum of what each member municipality appropriates toward it, added up across several separate budget documents. Treat it as a LOWER BOUND — the department may also hold grant, fee or surplus revenue that no member appropriates.
- Full fire appropriation full_fire_appropriation
- The whole fire agency appropriation (personnel + other expenses).
- Grant award only grant_award_only
- Grant dollars only — not an operating budget.
- Stale/secondary budget evidence stale_secondary
- A real but secondary or older figure (e.g. a news-reported budget); a primary source would supersede it.
- Basis not established unknown
- No basis established — the figure's coverage is not known.
Two further bases exist in the model for completeness — personnel_cost_only (Personnel cost only) capital_only (Capital only) — no current record uses them.
7 · Fire district & regional department caveats
Many New Jersey townships are served by multiple independent fire districts, and some departments are regional (e.g. one department serving several municipalities). We never blend them: a fire-district record represents that specific district (never a blended township total), and a regional department is one record for the whole system. A regional system that publishes no budget of its own is shown on a distinct member_contributions_reconstructed basis and flagged as a reconstructed lower bound, never as an appropriation the body itself published.
8 · FDID matching rules
The FDID is the canonical NJ DCA/NJFIRS 5-digit fire-department identifier. A curated profile is tagged with an FDID only when the directory maps to it unambiguously — a single-municipality match whose county is verified. Multi-district townships (where several districts share a municipality), regional systems, and cross-county name collisions are left unmatched rather than assigned a guessed FDID. So a curated record with no FDID means “no clean 1:1 directory match,” not missing identity.
9 · County directory summaries
Each county page shows a NJ DCA directory intelligence panel: directory department count, type mix, career and volunteer roster totals (with how many departments reported a number), NFIRS participation, EMS first-responder count, how many map to a curated profile, and how many are directory-only. These are self-reported roster figures for statewide context, kept visually and semantically separate from the curated county panel above them. They describe disclosure and coverage, never fire-safety performance.
10 · What missing data means
Missing does not mean zero. A blank field — in a curated profile or a directory record — means no public source reported it, not that the value is zero, not that a department lacks funding or staff, and not that a service is failing. An explicit reported 0 (for example a volunteer department that reported 0 career firefighters) is a known zero and is shown as “0”; an unreported field is shown as “—”. The two are kept distinct end to end.
11 · What Fireground Index does not claim
- It does not claim more money or more staff causes safer outcomes, or that less public data means worse performance.
- It does not imply misuse of funds anywhere.
- It does not claim roster counts equal on-duty staffing.
- It does not present directory records as independently verified curated profiles.
- It does not change the fire-death score, the rankings, completed-state data, or methodology.
12 · Source list
- Adopted municipal budgets — state-format Current Fund Appropriations (FCOA 25-265 “Fire” Salaries & Wages + Other Expenses), from each city’s official budget PDF.
- NJ DCA / DLGS fire-district adopted budgets — total appropriations and tax levy for independent fire districts (published under NJSA 40A:14-70.2).
- NJ DCA Annual Fire Department and Agency Directory (2023) — the self-reported directory (FDID, roster, population, square miles, NFIRS, funding, EMS), coordinate-parsed.
- OpenFEMA — Non-Disaster Firefighter Grants (AFG / SAFER / FP&S) award records.
- USAspending.gov — federal award records, used to disambiguate grant recipients.
- U.S. Census Bureau — 2020 Decennial and ACS / Population Estimates figures (via data.census.gov and Census Reporter) for population served.
- Official department & city pages — station lists, staffing, ISO/PPC statements, and annual reports where published.
The county civilian fire-death score and rankings use a separate, independently-documented dataset — see the New Jersey data sources page. Department Intelligence never feeds that score.
Illustrative example (fictional)
A fictional department shown only to demonstrate the curated-profile card framework. It is not one of the 28 real records.
Example Fire District (illustrative sample)
17/22 evidence fields have a public source · latest source 2025. Blank fields mean no public source was found — never that a value is zero.
NFPA Staffing Benchmark
Staffing-only · not a response-time or legal assessmentIllustrative sample value (fictional department).
Disclosed on-duty staffing (4 on duty) meets or exceeds the NFPA 1750:2026 baseline of 4 per company (applied to the career companies), per Example Township (fictional). Higher zone/risk company minimums (elevated 5, maximal 6), the effective response force, and response-time objectives were not assessed. Benchmark comparison only — not a legal compliance finding.
- Higher-risk (zone/hazard) staffing requirement not assessed
- Effective response force not assessed
- Response-time performance not assessed
Baseline company-staffing benchmark only. Higher zone/risk company minimums (elevated 5, maximal 6), the effective response force, and response-time objectives were not assessed. NFPA 1750 is a consensus standard unless adopted by the local Authority Having Jurisdiction.
NFPA 1750:2026 is a consensus deployment standard, not automatically a legal mandate. A department is legally bound only where its Authority Having Jurisdiction has adopted it. This assesses baseline company staffing only — not higher zone/risk staffing, effective response force, or response-time performance — and is not a legal finding.
Operational facts found (7 facts)
- Active roster — careerroster, not on-duty staffing
- 18 personnelVerified↗
- Active roster — volunteer / union membershiproster, not on-duty staffing
- 42 membersPartial↗
- Stations
- 3 stationsVerified↗
- Annual incidents/calls
- 3100 calls/yrVerified↗
- EMS share of calls
- 64% %Partial↗
- Adopted fire-department budget
- $4,200,000 USDVerified↗
- Population / jurisdiction served
- 24,000 residentsVerified↗
Funding
Public funding identified — blank means no public source found, not $0.
- Operating budget
- $4.2MVerified↗
- Department revenue
- No public source found
- Donations / fundraising
- $85,000Partial↗
- IRS Form 990 revenue
- $310,000Verified↗
- Grants received
- $240,000Verified↗
- Capital / apparatus
- No public source found
- Funding per resident
- $175Partial↗
Illustrative sample value (fictional department).
Budget figures are public evidence, not fully normalized spending comparisons. Each figure is labelled with the basis it was published on — budgets are only comparable where the basis matches.
Staffing
Public staffing evidence — blank means no public source found.
Readiness
Public readiness evidence.
ISO / PPC
ISO Public Protection Classification is not nationally complete; blank means no public rating found.
Grants
Grant awards identified in public records — an empty list means none were found, not that none exist.
Transparency
What this department publicly publishes — a data-availability measure, separate from any fire-outcome score.
- Budget published
- Staffing published
- Call volume published
- ISO/PPC published
- Annual report published (not published)
- Grant data found
- Form 990 found