Methodology

By Jon Brooks, Co-Founder, Momentum Realty · Last updated Nov 19, 2026 · Method and sources · Corrections

Last updated: Nov 19, 2026 (America/New_York). Prepared for movewithmomentum.com; not yet published.

What this data set covers

  • 67 counties (Duval is consolidated with the City of Jacksonville and appears in state data as a city) plus a State of Florida summary.
  • Files: counties.json (every field with fiscal year and source key), counties.csv (one row per county), state.json, this file.
  • Sources are listed by key in counties.json > sources. Every figure comes from a published source; nothing is estimated by us except the derived ratios described below.

Amendment 3

  • Status: Amendment 3 (CS/HJR 1-F) was adopted by voters on Nov 3, 2026 with 61.98% Yes (4,922,977 Yes vs 3,020,085 No; 60% required; Florida Division of Elections, https://floridaelectionwatch.gov/Amendments) and takes effect Jan 1, 2027 (first applies to the 2027 tax roll; DOR bulletin PTO 26-21).
  • Amendment 3 (CS/HJR 1-F): non-school homestead exemption of $150,000 (2027) and $250,000 (2028), indexed after; non-homestead assessment cap on non-school levies from 10% to 5%; people who were not Florida residents on Dec 31, 2026 get a $50,000 non-school exemption (indexed) for their first four years, then the full exemption; ad valorem uses for counties and cities limited to seven categories. School taxes are not affected.
  • Revenue effect: EDR Revenue Estimating Conference (July 10, 2026), by county and taxing authority, at current millage. Second estimate: Florida Association of Counties (June 2026), different data vintage. Both are shown as a range. County-published estimates are shown where found, with their stated scope.
  • Taxpayer effect: EDR notes the revenue loss equals the reduction in taxes paid at current rates. Median homestead savings come from the StepUp Law database (third-party model on 2026 preliminary rolls and 2025 millage); we checked its arithmetic for Baker County.
  • State replacement money: none enacted as of Nov 19, 2026. Lawmakers have discussed a temporary fund for the 29 fiscally constrained counties.

Key definitions

  • Operating revenue FY2023-24 = total LGFR revenue minus "other sources" (transfers, debt proceeds) minus custodial, internal service, pension/trust, private purpose and component-unit funds. This removes money tax collectors pass through to other governments, which appears in state data from FY2020-21 (GASB 84).
  • Core spending = public safety + transportation + economic environment + human services + culture/recreation (functions not affected by the GASB 84 change), inflation-adjusted to 2024 dollars with CPI-U annual averages, per resident (BEBR April 1 estimates). Growth compares the FY2014-15..FY2016-17 average with the FY2021-22..FY2023-24 average to smooth one-time capital spending.
  • Tax-supported debt = general obligation + special revenue debt reported by the county to EDR for Sept 30 (self-reported).
  • Pension liability = FRS pension + HIS proportionate share (June 30, 2025) for BOCC, clerk, property appraiser, tax collector, sheriff and supervisor of elections. Duval uses the City of Jacksonville ACFR (all plans), which is broader.

Financial health score (same method for every county)

Each measure is converted to a 0-100 percentile among the 67 counties (ties averaged), then weighted: | Measure | Weight | Better is | |---|---|---| | Tax-supported debt per resident | 15 | lower | | Pension liability per resident | 15 | lower | | Core real spending growth per resident | 15 | lower | | Property-tax reliance (2023 county levy / FY2023-24 operating revenue) | 10 | lower | | Taxable value per resident (2025) | 15 | higher | | Room under the 10-mill operating cap (2025) | 10 | higher | | Amendment 3 loss FY2028-29 / FY2023-24 operating revenue | 20 | lower | - Before A3: the A3 measure scores 100 for everyone (no loss). - After A3: A3 measure = 100 x (1 - loss share / 30%), floored at 0. Taxable value per resident is reduced by the share of the levy lost; millage room is reduced by the mills needed to replace the loss. After-values are scored against the before-A3 distribution, so no county's score rises. - Missing measures (Dixie debt; Duval millage room) are dropped and the remaining weights renormalized. - Sensitivity: the after score is also computed with the FAC estimate instead of EDR; both ranks are published. - Not included because no consistent statewide data exist: reserves/fund balance, debt service share, operating surpluses or deficits.

Items to review (formerly "red flags"; applied identically to every county; "worth a closer look", never an allegation)

Code Rule Threshold Fiscal year Source
R1 Required annual audit not filed Listed by the Auditor General as not filed by Oct 10, 2025 FY2023-24 Auditor General report 2026-049, Exh. A
R2 Annual audit filed late More than 30 days past the 9-month deadline, or late two years running FY2023-24 AG 2026-049, Exh. C
R3 Audit report items missing Listed in AG 2026-049, Exh. F FY2023-24 AG 2026-049
R4 State finance report not filed No 2025-cycle filing (s.129.03(5), F.S.) 2025 EDR 2025 data set
R5 Spending growth well above population plus inflation Core real spending per resident up more than 25% FY2015-17 to FY2022-24 EDR AFR, BEBR, BLS
R6 High tax-supported debt At least 3x the county median per resident Sept 30, 2025 EDR 2025 data set
R7 High pension liability At least 2x the county median per resident 2025 FRS GASB 68; Jacksonville ACFR
R8 Pension plan funded below 60% Any plan the county sponsors Latest valuation FRS valuation; local plan reports reviewed (Duval only)
R9 Large economic-environment spending At least 3x the median share of operating revenue FY2023-24 EDR AFR, LGFR
R10 Heavy reliance on state and federal money Accounts 331, 333, 334, 335.12, 335.18 at least 40% of operating revenue FY2023-24 LGFR
R11 Large A3 exposure with little rate room Loss at least 15% of operating revenue AND replacement would exceed the 10-mill cap FY2028-29 EDR REC, DOR, LGFR
Checked with no county listed: financial emergency conditions under s.218.503 F.S. and deteriorating financial condition (Auditor General report 2026-056, FY2023-24: none were counties).
Not yet checked county by county (pending): material weaknesses and repeat findings in each county's audit; unrestricted reserves (GFOA-style threshold); repeated operating deficits; incentive agreements.
The State of Florida is tested against the equivalent rules in state.json.

Cross-checks performed

  • EDR authority rows sum exactly to county totals for every county and year; they match an independent public transcription (StepUp) with zero mismatches.
  • EDR vs FAC: statewide county-government FY2028-29 $5.96B vs $5.05B. Counties where the two differ by more than 25% are flagged in their panel (largest: Holmes, EDR 3.2x FAC).
  • EDR AFR FY2023-24 totals equal the LGFR expenditure file for all 66 counties in the county files (Duval is in the city files) (only population definitions differ: with vs without inmates).
  • DOR 2025 county operating levy vs taxable value x millage: within 0.96 to 1.02 for every county.
  • Duval: EDR "Jacksonville" A3 row ($180.3M / $297.9M) vs Council Auditor (~$200M / ~$300M); EDR 2025 debt ($2.35B) vs ACFR bonds and notes ($2.60B, scope differs); FRS share ($5.5M) vs ACFR NPL ($4.04B, local plans).
  • County-reported A3 estimates (Miami-Dade, Orange, Hillsborough, Pinellas, Pasco, Clay, Brevard, Lee, Duval) compared with EDR; scopes differ and are stated.

Corrections made during the build

  1. Duval showed $0 county A3 loss in EDR's county rows; corrected to EDR's "Jacksonville" (consolidated government) row.
  2. Replaced FY2025-26 budget denominators (which vary in what counties include) with FY2023-24 operating revenue from audited AFR data.
  3. Removed custodial/pass-through funds from revenue after finding that state spending-per-resident series jump in FY2020-21 (e.g., Alachua $1,338 to $3,610) because of an accounting change.
  4. Category-level intergovernmental revenue was replaced by named accounts after Citrus County showed a jump from $38M to $290M that reflects pass-through money.
  5. Population years aligned to fiscal years (April 1 within the fiscal year) and spending growth switched to 3-year averages.
  6. Dixie (no 2025-cycle filing) is scored on available measures instead of being penalized for missing data.
  7. The after-A3 score was redesigned so it cannot rise above the before score.

Known limitations

  • Debt and some budget data are self-reported to EDR. Every county audit has been screened for findings (Nov 19, 2026); GF reserves are hand-verified for 20 counties.
  • Adopted-budget deep dives are pending; 14 counties have their own A3 estimates cross-checked (Miami-Dade, Orange, Hillsborough, Pinellas, Pasco, Clay, Brevard, Lee, Duval, Flagler, Sumter, St. Johns, Leon, Volusia).
  • Calendar-year CPI approximates fiscal years.

Phase 2 additions (Nov 19, 2026)

Audit screen and new red-flag rules (R12-R16)

All 66 county audit filings with the Auditor General were read (FY2025; FY2024 for Bay, Dixie, Taylor). Duval's city ACFR is not yet screened. - R12 Material weakness in internal control over financial reporting, county or any constitutional officer, latest filing. Federal-award-only weaknesses and prior-year findings are not counted. - R13 Finding also made in the two preceding audits (management letter). - R14 Unrestricted General Fund balance below 16.7% (GFOA two-month minimum) of GF expenditures plus transfers out, or a deficit. Hand-verified counties only (20). - R15 GF total fund balance fell 20% or more in the latest audited year. Hand-verified counties only. - R16 Unfavorable financial condition assessment or deteriorating condition reported in the management letter. Automated extraction of GF figures was unreliable (wrong fund blocks, budget columns), so only hand-checked values are used; other counties are "not yet evaluated", not "passing".

Strongest/weakest on management measures

Separate from the health score. Weights: audit on time 15; no material weakness 20; no three-year repeat 10; GF reserves 15 (<0 → 0, <16.7% → 50, 16.7-60% → 100, >60% → 85); GF result 10 (≥−5% → 100, ≥−10% → 60, ≥−20% → 30, else 0); real core spending growth per resident 15, tax-supported debt per resident 7.5, pension liability per resident 7.5 (percentiles, lower is better). Missing measures are dropped and weights rescaled; coverage is published. GFOA award and millage stability are not scored (award page often missing from filings; millage history not compiled).

Where each $1 goes

FY2023-24 county AFR account-level expenditures (EDR "Unadjusted-Total Exp by Cnty"). Slices: public safety 521-529; debt service 517; pensions paid directly 518; transportation 541/544/549; parks, libraries, culture 571-579; health and human services 561-569; general government and admin 511/512/514-516; economic development and housing 551-559; courts 600-769; environment 534/537-539. Excluded: 513 and 519 (pass-through money coded there, e.g., Sarasota 519 $1.9B, Alachua 513 $758M), 58x-59x transfers/non-operating, fee-funded utilities, airports, ports, parking. FRS contributions stay inside each function's payroll. Duval has no county account detail in the file. State donut: FY2026-27 appropriations by area (Fiscal Analysis in Brief 2026).

What could be cut, and why (options the data supports reviewing)

Population tiers: under 50K, 50K-200K, 200K-500K, 500K-1M, 1M+. Listed when spending per resident is ≥1.25x the tier median (≥1.5x for public safety, with tradeoff). Amount = (county per resident − tier median) × population: an upper bound, not a savings estimate. Ratios >5x are "check coding" and not totaled. Hospitals (561) and solid waste (534) excluded from peer comparison. GASB 77 abatements listed as incentive items. Vacancies, refinancing and incentive performance: not reported in state data.

County responses

Commission agendas/minutes, county websites, local news, FAC and League of Cities documents, filed bills, checked through Nov 19, 2026. Secondary-tracker items are labeled "primary source to confirm". Where nothing was found: "none found as of 2026-11-19".

Incentive ROI benchmark

EDR, Economic Evaluation for Select State Economic Development Incentive Programs (2026): CITC −0.75 on $306.4M; Job Growth Grant Fund 0.04 on $152.6M; High-Impact Sector Performance Grant −0.90 on $3.3M; QTI 1.29 on $33.2M (QTI repealed July 1, 2023, ch. 2023-173).

New export files

management_ranking.csv, a3_hit_rankings.csv, county_responses.csv, statewide_responses.csv, review_options.csv, dollar_pie.csv, audit_screen.csv, reserves.csv, phase2.json; counties.json gains per-county management, audit_screen, general_fund_verified, dollar_pie_fy2324, review_options, a3_hit, incentives_audit, county_responses and a top-level phase2 block.

Batch update (Nov 19, 2026, evening)

  • County responses: the 9 tracker-only rows (Hillsborough, Osceola, Palm Beach, Martin, Lake, Monroe, Brevard, Pinellas, Seminole) now cite meeting records, county websites/presentations, city agenda documents or news quoting officials. Volusia's tracker row was removed (no primary source; county records show impact modeling and FY2026-27 rollback rates only). Unconfirmed tracker items removed: Osceola "higher fire assessment", Safety Harbor "about $3M". Monroe Lower Keys Fire & Ambulance figures corrected to the county's Oct 20, 2026 presentation ($423,271 FY2028; $596,682 FY2029). New rows: Martin hiring pause, Osceola FY2027 budget cuts, Seminole nonprofit grants held back, Melbourne fire assessment hearing (Dec 8), Governor's Nov 4 special-session statement (statewide).
  • County-source figures (a3.county_published) added for Lake, Osceola, Martin, Nassau, Alachua, St. Lucie, Polk, Seminole and Monroe (23 counties total). Broward's reported $326M was not found in the cited Axios article; Broward stays pending.
  • Dollar pie and peer comparison: Sumter account 512 ($513.8M) and Columbia account 511 ($87.8M) excluded after audit check (audited general-government spending $30.0M and $15.0M). The admin peer median for the 50K-200K tier fell from $99 to $35 per resident; Walton admin is newly over 5x (check coding).
  • Check-coding: Bay and Monroe economic development matched to tourist development tax spending in their audits (real, restricted; still excluded from totals). Remaining unverified: Collier, DeSoto, Dixie, Taylor environment; Hamilton, Hardee, Walton economic development; Walton and Liberty admin; Monroe health and environment; Wakulla parks.
  • Debt: Pinellas $0 confirmed (FY2025 audit: no bonded governmental debt; $38.5M leases/SBITAs). St. Johns: EDR 2025 file shows $0 tax-supported debt ($70.0M in 2024) while the FY2025 audit shows about $240.5M governmental revenue bonds and notes plus $55.2M state loans. Score still uses EDR for consistency; method decision pending (field debt_audit_check).
  • Reserves hand-checked (GF unrestricted = committed + assigned + unassigned, over expenditures + transfers out): Palm Beach 25.8%, Polk 35.4%, Lee 15.8% (R14), Brevard 33.6%, Marion 28.9%, Hillsborough 43.4%, Osceola 44.5%; Columbia ratio 25.2%. 27 counties now hand-checked.

Batch 2 (Nov 19, 2026)

  • New rule R17 "State filing differs materially from the audit": applied where a hand check found a >25% gap between an EDR-reported figure and the audit. St. Johns flagged (EDR 2025 $0 tax-supported debt vs ~$240.5M governmental revenue bonds and notes plus $55.2M state loans in the FY2025 audit). Scores keep the EDR figure for every county; the panel shows both.
  • Reserves (R14/R15) hand-checked for 26 more counties plus Duval (City of Jacksonville GF/GSD, FY2025 ACFR): 54 total. Method: GF unrestricted (committed + assigned + unassigned) over GF expenditures + transfers out; the operating-statement column is chosen by matching its ending fund balance to the balance-sheet total. Pending: Dixie, Franklin, Holmes, Jackson, Jefferson, Lafayette, Miami-Dade, Sarasota, St. Lucie, Taylor, Union, Walton, Washington.
  • Duval audit screen: material weakness 2025-001 (financial statement presentation) -> R12. No 3-year repeat.
  • Homestead saving spread: when p25 = p75 (Monroe) or p75 = median (Collier, Manatee, Martin, Nassau, Palm Beach, Sarasota, St. Johns, Sumter, Walton) the saving is capped by the full added exemption; labelled on the panel (not a data error).
  • Homestead before/after bill: AV = StepUp median; 2025 DOR millage by authority (school; county operating+debt+dependent+LTCW; MSTU avg; ISD avg; municipal avg). Before: school taxable AV-25K; non-school AV-25K-min(max(AV-50K,0),25K). After: non-school max(0, AV-250K). Illustrative only.
  • County responses: new kind "plan" (planning step), excluded from the "who picks up the bill" list. Added Miami-Dade (FY2026-27 budget reductions; contingency report motion), Orange (mayor-elect options; fire fee / stormwater study), Palm Beach (parks/library 50% planning instruction), Hillsborough (Jan 2027 workshop; staff options), Broward (workshop recommended). County figures added: Broward (BCPA $333.8M yr 2 via Brizaga), Palm Beach (PA $221M/$324M; Fire Rescue $147M yr 2; administrator >= $110M GF; Stet News).
  • Debt by purpose shown for Monroe, Manatee, Osceola (FY2025 audits).

Batch 3-5 and final check (Nov 19, 2026)

  • Reserves (R14/R15) now hand-verified for all 67 counties. Added Dixie (FY2024 B-to-A), Franklin, Holmes, Jefferson (restated beginning), Lafayette (net transfers), Taylor (FY2024 MD&A, unassigned only), St. Lucie and Sarasota (shift-encoded text decoded: chr(ord+29)), Jackson, Union, Walton, Washington (read from rendered page images), Miami-Dade (Clerk FY2025 ACFR, $000s). New flags: Holmes R14, Jackson R14+R15, Dixie R15, Washington R15.
  • Rank effects: Holmes out of best-run top 10 (#15), Walton in (#9). Miami-Dade out of bottom 10 (#53), Hardee in (#58).
  • Check-coding: all 12 compared with audits (CODEV on the Page). Function totals are not like-for-like (EDR includes capital and enterprise funds), so R17 is applied only to like-for-like items (debt totals).
  • Debt by type/purpose (DEBTP): Monroe, Manatee, Osceola, Miami-Dade, Hillsborough, Orange, Broward.
  • Rounding fix: review_options.csv now replicates the Page's reviewOptions() (rounded EFFO values, 1.5x/1.25x thresholds, coding >5x, abatement rows) with unrounded values in extra columns; growth items moved to review_growth_items.csv. a3_hit_rankings.csv now carries the Page's ranks (unrounded sort, stable ties).
  • Top-10 re-verification: App re-rendered; management top/bottom 10, hit hardest/least (by % of revenue) and all 66 parsed health-rank rows match the data.

Batch: R17 marker and incentive deals (Nov 19, 2026)

  • R17 marker: St. Johns shown with ⚑ in top/bottom-10 tables and the full ranking table; debt cell shows "$0 (state) ⚑". Alternative rank computed by rerunning phase2 scoring with tax-supported debt = $295.7M / 348,336 residents (~$849 pc), all other inputs unchanged: score 81.4, rank #27. Scores used on the Page are unchanged (same state-reported input for every county).
  • GASB 77 abatement amounts (FY2025 audits) added for Broward ~$169,000; Charlotte $382,330; Gadsden $0; Hamilton $0 (begins 2026); Manatee $330; Orange $1,391,763; Polk $10,509,533; St. Johns $539,465; St. Lucie $1,646,298; Martin $3,994,259; Hillsborough $768,822; Palm Beach $1,218,115; Duval $16,616,418 (City of Jacksonville audit Note H). Review-option label now "Property tax abated or rebated for businesses (GASB 77 note)"; legal text lists the statute by program.
  • Incentive deals (INC): 10 counties. Sources: FY2025 audits; s.125.045(4) EDO reports (OEP, GFLA, PCED, Beacon Council); Miami-Dade R-1009-25; Palm Beach agenda 3I-1 (Apr 12, 2022); Osceola news Dec 2, 2024; EDR LFY2020-21 incentive compilation (econincentives21.pdf). Cost per job: Hillsborough $433,555/1,367 verified QTI jobs = ~$317 (one year's county share); Miami-Dade Santander max $5M/250 = $20,000 per promised job. FloridaCommerce Annual Incentives Report/portal blocked automated access (Azure WAF) and was not used.

Batch: Why the budget grew + spending to review (Nov 19, 2026)

  • Base: EDR county government expenditures by account code (annual financial reports), LFY2014-15 (cntyfiscal/2015expendituretotalsbycounty.xlsx) and LFY2023-24 (2024expendituretotalsbycounty.xlsx, unadjusted sheet). Duval = City of Jacksonville from the municipal files (includes JEA). FY2024-25 county files not published yet (404), so LFY2023-24 is the latest year.
  • Comparable spending = all functional accounts except general administration (511, 512, 513, 519), court-related (6xx/7xx) and transfers/non-operating (58x/59x): from FY2020-21 the state files include custodial pass-through money in admin and court accounts (statewide 513: $38.1B unadjusted vs $4.75B adjusted).
  • Bridge: population effect = T0·(P1/P0−1); inflation effect = T0·(P1/P0)·(CPI24/CPI15−1), CPI-U 237.017 → 313.689; real change = T1 − T0·k, k=(P1/P0)·(CPI ratio). Per-line "above population and inflation" = v1 − v0·k. Components add to the total exactly for every county.
  • 23 lines grouped into 9 areas; top 3 drivers = largest positive per-line excess. Coding check when the FY2014-15 base is ≤0 or the line grew more than 4×, and the excess exceeds 1% of the base total.
  • Context: federal grants (331*, 333, 332.00), ARPA (331.51) FY2023-24; salaries and benefits share from EDR 2025 county metrics (Dixie not reported).
  • Data checks (GBQC): Hillsborough (sheriff/jail $380.0M → $20.8M, utilities $308.9M → $112.7M; audit FY2025 public safety $1.23B vs $368M state FY2023-24), Flagler (FY2014-15 debt service 37% of base; totals below population growth), Glades (totals below population growth), Monroe (utilities fell by more than half), St. Johns (FY2014-15 debt service 28% of base). Their drivers are kept out of the spending-to-review ranking.
  • spending to review list per county: non-coding peer outliers (reviewOptions), business tax abatements labelled with incentive reporting status, growth drivers (excluding coding-check, storm and fee-funded utility lines, and data-check counties), sorted by $; then items to review and audit findings without $. Items overlap and are not summed.
  • Exports: budget_growth.csv (67), budget_growth_by_area.csv (603), potential_waste.csv (373 items).
  • Pending: budget-book reasons for each driver (pay contracts, headcount, capital projects), starting with the largest counties; vacancies and admin overhead not measurable from state data.

Conflict fixes from the independent audit (Nov 19, 2026)

  • Wakulla FRS/HIS: added WAKULLA COUNTY BOARD OF COUNTY COMM (employer 75003; FRS $9,094,586, HIS $2,766,405; FRS GASB 68 schedules, raw/frs2025a.txt). Total $28,468,558. raw/frs_county.json and stage1.json patched; final.py/flags.py/export.py/phase2.py rerun. 58 counties' health score fields shifted (relative percentiles); management list: Calhoun #1, Lake #2; Wakulla #59 (weakest 10), Hardee #57 (out).
  • St. Johns R17: governmental debt $231.2M bonds/notes (+$33.6M loans, leases, subscriptions); $295.7M utility debt excluded. Alternative score 82.0, rank #27.
  • Collier MW attributed to the Property Appraiser. DAR pension medians $1,684 per capita, 36.7% of own-source revenue. Duval R8 47.21% (FY2025 ACFR). Reserves text: all 67.
  • County A3 estimates: Hillsborough $367M (Axios) / $343M (Florida Politics), Lee $109M/$144M (Gulfshore Business Nov 18) with WGCU $121M/$159M, Duval about $300M a year (Action News Jax). Orange fire $46M (2027)/$72M (2028).
  • New responses: Duval, Clay. Removed unsupported claims (Sanford details, Safety Harbor costs, Seminole 24 nonprofits, Volusia $129M, Hillsborough 'half of calls', Citizens peak).
  • Labels: Spending to review; Items to review; Savings options to review. Statewide pie from raw dollars. Revenue measure includes enterprise funds (disclosed).
  • export2.py now reads the live page source (p2/page_new.tsx) instead of p2/data.tsx.

History, FY2014-15 to FY2024-25 (added Nov 19, 2026)

Sources (the same for every county and year) - Revenue and spending: EDR county revenue and expenditure files by account code and fund type (LFY2005-06 to LFY2024-25; Duval = City of Jacksonville consolidated file). Yearly totals equal EDR's category file exactly, and all-funds spending equals the EDR metrics file. - Operating revenue = accounts 311-379 in governmental and enterprise funds (no transfers or debt proceeds). Property tax = 311. Sales taxes = 312.6x plus 335.18 (half-cent). State shared = 335 except 335.18. Grants = 331-334 and 337. Charges for services = 341-349; permits, fees and special assessments = 32x; fines and other = 35x-36x. - Spending by function = all funds, by 51x-57x function, plus 58x other uses. Core functions = public safety, physical and economic environment, transportation, human services, culture and recreation, court-related and general government (custodial accounts 513 and 519 noted from FY2020-21). - A county-year with zero revenue and zero spending in the files is a gap, not zero (Jefferson FY2017-18). FY2024-25 is preliminary: 42 counties have revenue and 56 have spending in the state files so far. - Debt: EDR debt outstanding at Sept 30 (total, and tax-supported = general obligation + special revenue). Dixie is excluded (inconsistent filings). Tax-supported debt reported as $0 while total debt was not, after a prior year above $1M, is left blank (DeSoto FY2022-23 and FY2024-25, St. Johns FY2024-25). - Taxable value: DOR county taxable value history; roll year Y-1 funds fiscal year Y-1 to Y. - Millage and county levy: DOR Data Book millage files for the 2018-2025 rolls (FY2018-19 onward). Earlier files were not available, so earlier years are gaps. Duval's county millage is not comparable (consolidated) and is left blank. - Population: EDR/BEBR April 1 estimates. CPI-U annual averages for inflation adjustment. - Events: OpenFEMA Disaster Declarations Summaries and Public Assistance Funded Projects Details (federal share obligated), plus documented events from audits, budget books, ordinances and named news reports. - FEMA amounts count local applicants only. Excluded: State of Florida agencies (applicant IDs 000-; OpenFEMA files their "Statewide" projects, about $7.0B, under Leon's code) and AdventHealth's system-wide COVID-19 applicant filed under Seminole (117-U8MYV-00, about $810M). Declarations under $1M are not shown on the page but are in projects_timeline.csv. - Each event goes to the fiscal year (Oct-Sep) that contains its date.

Yearly health score - The same percentile method and weights as the published before-A3 score, recomputed each year against all counties with data that year. The Amendment 3 component is 100 (pre-amendment). - Inputs: tax-supported debt per resident, core real spending growth (3-year average for Y-2..Y vs Y-9..Y-7, per resident, CPI-adjusted), property tax reliance (account 311 / operating revenue), taxable value per resident, and millage headroom (10 minus county operating millage; FY2018-19 onward). - Pension per resident is left out every year (no consistent yearly series yet). Missing inputs are listed in each row's note and weights are rescaled. - Check: correlation with the published FY2023-24 before-A3 score is 0.82.

Not yet compiled - yearly pension and reserves; - millage before the 2018 roll; - ARPA allocations by county; - bond issues, stadium, arena and convention deals and large incentives for most counties; - state-level series beyond the sum of reporting counties ("All counties").

Debt history and debt by purpose (added Nov 19, 2026)

  • Coverage: all 67 counties, Sept 30, 2015 to Sept 30, 2025 (FY2014-15 to FY2024-25). 737 county-years: 326 from ACFR ten-year debt schedules (33 counties), 74 from audit long-term debt notes (34 counties), 329 self-reported to EDR, 8 gaps.
  • Why audited first: EDR's self-reported debt often leaves out leases, subscriptions and premiums and sometimes whole categories (Bay FY2018-19: $29.3M self-reported vs $394.6M audited; Lee FY2024-25: $469.3M vs $1,685.1M). Where a county's chart switches from self-reported to audited, part of the jump is the source change; the panel says so.
  • Image-only PDFs (Broward, Flagler, St. Lucie, Walton, Jackson, Washington, Union) were read from rendered page images and checked against row and column totals.
  • Broward's printed FY2022-23 and FY2023-24 totals equal the business-type subtotal only; we show governmental + business-type.
  • A $0 self-reported year between two non-zero years is treated as a gap (Columbia FY2021-22, Taylor FY2015-16).
  • Debt by purpose is shown only where the schedule names the purpose or pledged revenue (24 counties, FY2024-25). No pension obligation bonds appear in any schedule read. "Other obligation" items (Sumter's Villages agreement, Miami-Dade's AA Arena liability) are listed but not added to totals.
  • Debt service (governmental funds, principal + interest) is from audited statements for 15 counties; the EDR account 517 series is incomplete and is not used for this.
  • The health score is unchanged; it uses state-reported tax-supported debt for every county.

ARPA, incentives and timeline events (added Nov 19, 2026)

  • ARPA: Treasury's county allocation table (May 10, 2021) gives the amount for all 67 counties ($4.17B in all). Obligated and spent amounts and the largest category come from each county's latest Project and Expenditure report in Treasury's April 2026 public data release; 14 counties have completed closeout and 4 are in closeout. Duval's report covers the consolidated City of Jacksonville's county and city allocations ($343.7M; Treasury's 2025 Recovery Plan for the City), so it is not comparable to the county-only allocation.
  • Incentives: EDR's LFY2020-21 compilation is the only statewide, same-method source; it relies on county survey answers, 10 counties did not respond, and EDR flags that it could not verify Orange's $238.3M. FY2024-25 audit abatement amounts (GASB 77) and the 10-program promised-vs-delivered review are shown separately and never added to the EDR figures.
  • Bond and stadium events: taken from the FY2024-25 debt schedules, so amounts are balances on Sept 30, 2025. Original issue amounts and dates are not in those schedules and are not estimated.