How this tool works

A plain-language look at the data behind the numbers: what Amendment 3 changes, how we estimate its impact on a tax bill and a government budget, how we check our work, and what the model does not capture.

What Amendment 3 changes

Amendment 3 raises the homestead exemption for non-school property taxes only. It needs 60% voter approval on the November 2026 statewide ballot to take effect.

The new exemption replaces the existing homestead exemptions rather than stacking on top of them. Non-ad valorem assessmentsFixed charges on a tax bill for services like fire, garbage, or stormwater. They are not based on a property’s value, and Amendment 3 does not change them. are unaffected. See About Amendment 3 for the full timeline and details.

How we estimate the impact

For every homesteaded property, we start with its assessed value from the county property tax roll. We calculate what its taxable value would be under today’s exemption and under Amendment 3’s larger exemption, and the difference between the two is how much taxable value comes off the books. We multiply that reduction by the non-school tax rate that applies to the property to get the estimated dollar impact, then add every property in a city or county together to get the citywide and countywide totals.

How we check our work

We check our modeled totals against each government’s adopted millage rates and the county’s certified tax roll, and we cross-check city and county General Fund and expenditure totals against independently audited financial statements.

Our model is calibrated against certified 2025 tax rolls. Run against those rolls, it reproduces countywide taxable value to within 0.4%, and every individual city to within 2%. Those differences are routine timing effects, such as assessment appeals still in process, and they fall slightly below the certified totals rather than above them.

The figures published here are built on the preliminary 2026 tax rolls rather than the certified 2025 ones. A preliminary roll has not yet been certified, but it is the more accurate representation of what property is worth and what it raises today, which is the question this tool exists to answer.

What the verified mark means

Every figure on this site comes from what each government reported to the state, worked through the method set out on this page. That is true of every page here, with or without a mark.

What the mark does not cover. The revenue loss is ours. We work it out parcel by parcel from the state tax roll, and we do not ask any government to endorse it — the sheet we send says so in those words. A government carrying the mark has confirmed the budget its loss is measured against, not the projection itself. Some have run the projection independently and told us how close we landed; that is a check worth having, and it is still not the same as an endorsement.

There is only the one mark. Most pages do not carry it, and that is the ordinary state rather than a shortcoming. We write to every government in this analysis and give it the chance to check its own figures, but no government is obliged to reply and many will not. A page without the mark is our work, sourced and shown. A page with it has additionally been through that government’s own finance office.

Where a government reviewed its page and sent us corrections to the figures it reported, we make them, and the page then carries the mark with the date. A government that told us where we were wrong checked our figures more closely than one that had nothing to add, not less, and the mark should not say otherwise. Where what it questions is a figure of ours rather than one of its own — a count we took from the tax roll, say — we record what it told us and say so, and we do not quietly change a number the rest of the page is calculated from.

What the cut reaches

The property-tax cut only reaches a government’s General Fund, the part of the budget property taxes actually support. Restricted money, like grants, gas tax revenue, or dedicated fees, cannot backfill that loss. So the General Fund is the one figure we measure every loss against, and the only percentage any page reports.

The General Fund figure we use is the whole appropriation, and each jurisdiction page shows what it is made of: money spent directly on services, and money the government moves into its own other funds. Those transfers are ordinary budgeting, typically funding capital projects, debt payments, or a fund that does not cover its own costs. We count them because the same commission votes them in the same annual budget, so they are part of what that board decides each year. Leaving them out would make the General Fund look smaller than the amount actually under its control.

Each jurisdiction page lists every General Fund line that government reported to the state, under the state’s own account names and with its own figures, largest first. We do not group them, rank them by what is likeliest to be cut, or mark any of them as off limits.

There is one exception, and it belongs to the government rather than to us. Where a government tells us in writing that two of the accounts it filed describe one thing, we report them as one and say so in plain words on that page, naming the government as the party that asked. We do not make that change on our own judgment, and it can never move money between what a page calls obligated and what it calls available: both accounts have to sit in the same category before we will combine them, so no figure, share or projection changes. Only the label does.

That is a limit on what we claim, not an omission. Predicting which services a government would reduce forecasts how a commission will vote. Declaring which it must fund is a legal opinion, and the facts do not support one: Florida law lets a government reduce even a required debt levy from money on hand, and a charter duty to provide police is written in qualities, never in dollars. We publish the arithmetic and the source. Which lines move is decided by elected officials in public, and you can model it yourself for your City or County.

What this model does not capture

• This is a static snapshot of the current tax roll. It does not model future growth, new construction, or how a government might change its tax rate in response.
• Amendment 3 also lowers the annual cap on assessment increases for non-homesteaded property from 10% to 5%, beginning in 2027. That changes how fast the tax base grows in later years rather than what the current roll raises, so it is outside a static model of the current roll. The loss shown here is the effect of the exemption change alone.
• Some smaller special taxing districts are not yet included, which slightly understates the total impact.
• Municipal senior and other local-option exemptions are not fully netted out for every jurisdiction, which can slightly overstate a city’s estimated loss.
• Budget data and the tax roll come from different fiscal years, so ratios describe a government’s overall structure rather than an exact same-year comparison.

Sources: Florida Department of Revenue property tax roll and non-ad valorem assessment roll; each county property appraiser’s tax roll and millage tables; adopted millage rates; state Economic and Demographic Research municipal fiscal reports; audited city and county financial statements; the enrolled amendment text.