Budget pressure signals

A statewide, mechanism-specific read of where New Jersey school districts show budget pressure — typed signals with stated evidence, computed the same way for every district. It is deliberately not a health score, a ranking, or a red/yellow/green diagnosis.

Signals sit in four panels that are never blended into one number. Financial (F1–F4): budgeted-data trends that fire only when a district is both beyond the statewide 90th-percentile adverse threshold and moving that way consistently. Structural (S1–S4): exact facts — state-aid collar status, aid dependence, enrollment, levy behavior. Reliability (R1–R2): how far a district’s actual results track its budget, which governs how much to trust the financial panel. (Observed events — the fourth panel — are a separate manually-audited registry, not computed here.)

The statewide picture

Each strip is the distribution of one financial signal across all evaluated districts. The dashed line is the frozen 90th-percentile threshold; a district fires only if it clears the line and its year-over-year direction is consistent, so the navy dots (fired) are a subset of the shaded side. The 14 labelled ticks are the representative districts the pilot tracks by hand.

Frozen thresholds this run: F1 ≤ -1.7pp · F2 ≥ +5.92pp · F3 ≥ +2.31/yr · F4 ≥ +11.6%/yr · R1 variance ≥ 40% of budget. Per signal: F1 fired 11, F2 23, F3 24, F4 22 districts. F1’s low count is expected — pandemic-era reserves broadly grew and are only now normalizing, so few districts show a sustained multi-year decline.

Every district

All evaluated districts, plus the below-scale representative districts (shown for their exact collar facts). Click a column heading to sort. A dot (●) marks a representative district.

Sorted by district name. “Signals” counts fired financial signals (F1–F4) only; structural and reliability panels are shown but not summed into the count — by design, they are context, not a score.

The 14 representative districts, in detail

The pilot tracks these by hand — chosen to cover every crisis subtype plus stable controls. Each carries its full panel read. Note how districts under real pressure can show a quiet financial panel: cap-squeeze and governance failures leave little mark on budgeted balance sheets, and surface instead in the structural and reliability panels.

What this will not claim

• No 0–100 health score.
• No best-to-worst ranking of districts.
• No red/yellow/green or “healthy/at-risk” label.
• No crisis probability modeled from a handful of cases.
• No implied all-clear: “no signal” means none was detected in the available data, not that a district is fine.
• No governance accusation inferred from ratios.

Method and frozen definitions: the edie education pilot (plans/budget-entrypoints/health-signals/definitions-v1.md). Sources: NJDOE User-Friendly Budgets, GBM state-aid district details, and TGES Indicators 20/21, via the njschooldata warehouse. A district shown with missing years never renders “no signal.” Reliability note: a persistent same-sign gap between planned and actual reserve draws is near-universal — conservative fund-balance budgeting is standard practice — so only unusually large variance (top decile) is highlighted. State aid is measured net of Department-of-Education loans against state aid — a loan is debt, not aid, so counting it would inflate a district’s state share and distort the aid-share signals (this materially affects Lakewood, which finances much of its budget through such loans). Three district-years NJDOE omitted from the statewide files (Lakewood 2023-24 & 2024-25, Jackson 2023-24 — all fiscally-stressed districts) were backfilled from primary sources (per-district UFB PDFs; Lakewood 2024-25 from the audited ACFR, an audited-actual basis, flagged per district). This is a draft; figures and thresholds may change.