About this study

Why this exists

This is the second study on this site. The first asked how each Florida county's property tax revenue and spending moved from FY2019 against its own population growth and inflation. This one asks the same question of state government, agency by agency, with the same rules: describe, don't recommend; findings in both directions get the same treatment; every figure comes from a public record; every rule that produces a sentence would fire again on next year's data.

What was pulled

Public sources, downloaded by script from the publishing agency and not edited by hand.

1. Transparency Florida — operating budget and amount spent. The Executive Office of the Governor's budget-and-spending site, fed nightly from the state's accounting system (FLAIR) and appropriations system (LAS/PBS). For each agency and each fiscal year FY2015–FY2027 the General Public view lists every budget entity (the unit the Legislature appropriates to inside an agency) with its operating budget — the year's appropriations as amended during the year by budget amendments, supplementals and reversions — and the amount spent against them. 4,253 budget-entity-years were read. The Budget Analyst detail pages give the operating budget by appropriation category and fund (General Revenue or the named trust fund) for every budget entity, FY2019–FY2027: 3,125 pages, 55,735 category-and-fund cells. The State Positions summary gives authorized, established, filled and vacant positions by agency, each year. Source.
2. Transparency Florida — amount spent by appropriation category. The public view of each budget entity, giving the amount spent by category (salaries and benefits, expenses, special categories, aid to local governments, fixed capital outlay and so on) for each year. This is what lets the study set construction money and fund transfers apart from the rest of spending; every one of these pages was cross-footed against the budget-entity page it belongs to. Source.
3. Transparency Florida — Ten-Year History of Appropriations. The appropriations the Legislature enacted for each year (the General Appropriations Act and other appropriation bills, net of vetoes), by agency and program, split into operations and fixed capital outlay, for FY2016–FY2025 (the edition on the site; the report shows the same ten years whatever year is asked for). Used to compare the operating budget, which is amended during the year and reduced by reversions, with what was enacted. Source.
4. Transparency Florida — ledger pages. One page per budget entity, category and fund, FY2019–FY2026, carrying the original appropriation in the General Appropriations Act, each budget amendment, reversions, releases and disbursements by fund. Used for the split of spending between General Revenue and trust funds: the ledger's shares for each agency and year are applied to the budget-entity figures the pages use (the two downloads were a day apart, and for recent years the nightly postings in between move the totals by a few hundredths of a percent). 49,469 pages, 179,979 dated appropriation lines.
5. Agency Legislative Budget Requests, Exhibit D-3A. From the Florida Fiscal Portal, each agency's budget request for each year FY2018-19 through FY2027-28: 309 request files, 144,904 fund lines, each marked state, state match, federal or recipient money. Requests are what agencies asked for, not what was spent; this study uses them for one thing only, the split of mixed trust funds between state and federal dollars within each budget entity. Source.
6. Fiscal Analysis in Brief. The Legislature's yearly summary of the General Appropriations Act, FY2018-19 through FY2026-27: appropriations by fund source and by program area, and a table of every trust fund with its state and federal dollars (1,538 fund-years). Used to classify trust funds and to cross-foot totals. Source.
7. Population. Florida resident population on April 1 of each year, from the Demographic Estimating Conference of June 22, 2026 — one published series already reconciled to the 2020 Census, so no splice was needed. Each fiscal year is matched to the April 1 estimate inside it (FY2026, the year ending June 30, 2026, to April 1, 2026). The April 1, 2026 figure is the conference's first forecast-year value; the difference between its December 2025 and June 2026 values for that year is under a tenth of a percent. Source.
8. Prices. The Consumer Price Index for All Urban Consumers, All Items, South region (BLS series CUUR0300SA0), monthly, via FRED, averaged over each July-to-June state fiscal year. October 2025 CPI not published by BLS (federal shutdown); average of the 11 published months. Source.
9. The state's Annual Comprehensive Financial Report, FY2025 (Department of Financial Services; opinion by the Auditor General): ten years of governmental expenses by function, used as an independent check on the direction and scale of the Transparency Florida series. Source.

What was calculated

Every figure on every page is produced by the same script from the sources above. The definitions:

1. Spending. The amount spent as recorded on Transparency Florida for each budget entity and fiscal year, summed to the agency. It is money disbursed against that year's operating budget, including amounts certified forward and paid in the first months of the next year. The main series on every page — “spending excluding construction” — leaves out two kinds of recorded spending, both shown separately: fixed capital outlay (appropriation categories 08 and 14: construction and land, appropriated in one year, spent over several and recorded against the year appropriated, so that recent years would otherwise be understated for construction-heavy agencies), and transfers to a state trust fund or the General Revenue Fund (the Transfers category, and any category whose title begins “Transfer” and names a trust fund or the General Revenue Fund), because the same dollars appear again as spending when they are drawn from the receiving fund. Payments to another agency for services (human-resources services, Capitol Police) are not transfers in this sense and stay in. Budget entities, flags and the fund estimates are all on the same basis.
2. Operating budget. The year's appropriations as amended: the General Appropriations Act plus supplementals and budget amendments (including federal money received during the year), less vetoes and, for a closed year, the amounts that reverted at year-end. Because reversions are already out, spending equals the operating budget to the dollar for many budget entities; "budget not drawn in year" is the operating budget minus spending — authority still being spent in later years or never drawn, and each page names the categories it sits in from the category pages (fixed capital outlay, continuing appropriations, lump sums, an undrawn transfer line). An agency whose recorded spending is under half its budget in every year with little construction (the Legislature) gets a data note saying its series is compared only with itself.
3. The baseline. For any figure and any later year: the FY2019 figure × (population in that year ÷ population in FY2019) × (CPI in that year ÷ CPI in FY2019). FY2019, the year ended June 30, 2019, is the last full fiscal year before the pandemic. "Growth above baseline" is actual growth minus that, in percentage points. "At the baseline" for a year is the FY2019 figure grown to that year; "cumulative above the baseline path" is the sum, over FY2020–FY2026, of each year's actual minus its at-baseline figure. A series missing any year in that window shows no cumulative figure.
4. Fund cuts. General Revenue is the state's general tax money and is always state money. Trust funds are split into state and federal by rule: where the Legislature's own classification of a fund (Fiscal Analysis in Brief, state and federal dollars by fund) is at least 90% one or the other, the fund is treated as wholly that; for a mixed fund, the agency's own budget request gives the split for that budget entity and year (that year or a neighbouring one), failing that the statewide split, and failing that the entity's requests in any year or the fund across all agencies. For the main series the request lines outside fixed capital outlay are used, because a fund's federal share can differ sharply between construction and the rest (the State Transportation Trust Fund carries federal highway money in the Work Program and almost none in the department's operating lines); the operating-budget bars, which cover all categories, use every line. Spending by fund comes from the ledger pages, which record each year's disbursements by budget entity, category and fund: the General Revenue and trust-fund dollars are the ledger's own, and only the federal share of each trust fund is estimated by the rule above (for the main series, from the ledger lines outside fixed capital outlay and fund transfers). The ledger's shares are applied to the budget-entity figures the pages use, so the parts always sum to the whole. Every page labels the federal and state-trust-fund figures as estimates.
5. Per resident. Each year's figure divided by that year's population. "In FY2019 dollars" divides by the CPI ratio as well.
6. Positions. Established positions from the State Positions summary (positions set up in the personnel system on the day the summary was taken), compared with population growth and shown per 10,000 residents. Filled and vacant positions are shown alongside.
7. Budget entities and reorganizations. Entities are tracked by their appropriation code, which outlasts renamings. An entity that first appears or last appears inside the period, and is material (the larger of $10 million and 2% of the agency's spending), gets a data note. Within an agency, when the entities that ended or fell sharply in a year and the entities that started or rose sharply the same year move by matching amounts (within 20%), each is marked ↔ and the observation says the change coincided with an opposite move. Across agencies, an entity of $5 million or more that ends in one agency the year another starts elsewhere, by a matching amount (within 20% with a word in common in the names, or within 10%), is noted on both pages as a possible move between agencies.
8. Flags and data notes. Rules, not judgment, decide what is marked. Flags: a budget entity up 25%+ or down 30%+ in a year, by at least $10 million and at least 2% of the agency's spending (both, so that large agencies are not flooded with small changes); an entity 25+ points above baseline over FY2019–FY2026; a jump in an entity whose budget was at least half federal money. The three largest single-year changes by dollars are printed as observations; every flag is in the review workbook. Data notes: an agency's total doubling or halving in a year; a material entity first or last appearing inside the period (a year with no budget and no spending counts as absent); entities that ended or fell and entities that started or rose by matching amounts in the same year (within 20%); trust-fund dollars that could not be classified; a difference between the site's statewide table and the sum of an agency's entities, or between an agency's budget-entity pages and its detail pages; every complete year in which 25% or more of the operating budget was not drawn; recorded spending under half the budget in every year; and the year in progress. 83 data notes appear across the 32 agency pages.

What was produced

1. 32 agency pages, one per agency in the state budget, each built to the same plan: snapshot and baseline, spending against the baseline by year, where the money comes from (General Revenue, state trust funds, federal), where it goes (budget entities, appropriation categories, named special categories), positions, the agency's place among the others, observations, questions, data notes and sources. Administered Funds — the statewide holding line for pay adjustments and reserves, moved to agencies by budget amendment — has a page but is left out of the rankings.
2. One comparison page with every agency on the same years: the state figure, the ranking against the baseline, a table with the state and federal estimates and the budget-to-spending gap, positions, and the structural facts that shape the figures.
3. This page.

All text on the agency and comparison pages is generated from a fixed set of sentence templates and a word list that bars characterizations ("waste", "excessive", "surge", "should" and the like; the one allowed use of "should" is the statement that the baseline describes the arithmetic, "not what the agency could or should have done"). Observations state figures; questions are phrased so an agency or legislative staff member could answer them. The pages name no individual official.

How it was checked

1. Every Transparency Florida detail page was cross-footed against its own printed totals (3,125 pages, zero mismatches), and the sum of every agency's budget entities was matched to the site's statewide table for every year; the agency-years that differ (two, by at most $2.0 million) and the one agency-year whose detail pages differ from the budget-entity page (by at most $20.0 million) carry a data note.
2. Every amount-spent-by-category page was checked against its own printed total (4,185 pages; 0 off by more than $10) and its total matched to the budget-entity page it belongs to, so the construction and transfer amounts set aside are exactly what the site records. The Ten-Year History was checked program by program against each agency total, operations plus fixed capital outlay against the same total (33 agency pages — the 32 agencies and the report's own statewide line — over 10 years), and its statewide page against the sum of the agency pages and the site's own comparison page (0 mismatches).
3. Every budget-request file was cross-footed against its own printed category and issue totals (42,718 totals, 0 mismatches), and the Fiscal Analysis in Brief fund tables against the Legislature's own totals in every year.
4. Every figure on four agency pages and the comparison page, and the first observation on every agency page, was recomputed from the source pages by an independent, automated review that had not seen how the pages were built; its findings were fixed and the recomputation repeated.
5. Every page passes the banned-word check and a structural check (no missing sections, no broken links, no placeholder text), and was viewed at phone width and in print.

What it cannot show

1. Why a figure changed. The records show the change; they do not show the reason. Caseloads, hurricanes, federal rules, new facilities and new laws are all legitimate answers to the questions the pages ask, and the pages leave those answers to the agencies and the Legislature.
2. What an agency could or should have spent. The baseline is a yardstick, not a budget.
3. Exactly how much of a trust fund's spending was federal. The state/federal split is an estimate by the rule above, and every page says so; the General Revenue figures are exact.
4. Spending outside the appropriated operating budget: the Lottery's prize payments, the state universities' own spending, local school districts' spending of the state money they receive, and pension benefits paid from the retirement trust fund.
5. Where pandemic relief money the state deposited in the General Revenue Fund in FY2021 ends up in the split: it is counted as state money there because that is how the state recorded it; the pandemic years are marked on every page.
6. FY2027, which is in progress, and anything after the state's accounting system changes to Florida PALM (planned for January 2027 in the project's July 2026 steering-committee presentation) until the first rebuild confirms the source pages are unchanged.

Who made it

Beachside Automation, with software-assisted analysis and human review. The pipeline — fetch the public pages, parse them, compute the measures, write the pages, check the pages — runs end to end by script, so the study can be rebuilt each year when the fiscal year closes. Feedback is welcome by email at [email protected]. If an agency believes a figure misstates its record, the correction will be made and the change noted on the page.