Methodology & Sources

Data

• Census Bureau, American Community Survey (ACS) 5-year estimates, 2009–2013 through 2020–2024, from the Bureau's bulk summary files: median real estate taxes paid (B25103), median home value (B25077), median household income (B19013) and owner-occupied units (B25003). These are U.S. Government works and in the public domain. Values are what owner-occupants report in the survey. Dollars are nominal (not inflation-adjusted). Each period covers five years of survey responses, labeled by its last year.
• Florida Department of Revenue, Florida Ad Valorem Valuation and Tax Data Book, "Millage and Taxes Levied" report for 2025 (Data Extract: November 2025). Average total millage = total taxes levied ÷ county taxable value. School millage is computed on school taxable value. This is a Florida public record, attributed to the Department.
• Census Bureau county adjacency file (2024) for neighboring counties.

Definitions

• Effective rate = median real estate taxes paid ÷ median home value. It summarizes what owners actually pay. It is not a rate levied on any property.
• Margin of error (MOE): Census publishes 90% MOEs for each estimate. MOE of the rate uses the Census ratio approximation: MOE(X/Y) = √(MOE_X² + (X/Y)²·MOE_Y²) ÷ Y.
• Ranks: 1 = highest, among 67 counties.
• Tax as % of income = median tax ÷ median household income (all households, not only owners). This is an indicator, not a household-level burden.
• New-buyer estimate: median home value taxed at the county's average 2025 millage with the 2026 homestead exemption.

Indexing and quality rule

A county page is submitted to search engines only if the median-tax MOE is at most 15% of the estimate, the effective-rate MOE is at most 20% of the rate, the county has at least 2,000 owner-occupied homes, and no value is top-coded. Pages that fail stay available to readers with a data-caution note, carry a noindex tag, and are left out of the sitemap. Currently 9 of 67 counties fail: Dixie (effective-rate margin of error is 21% of the estimate (limit 20%)), Franklin (median tax margin of error is 16% of the estimate (limit 15%); effective-rate margin of error is 24% of the estimate (limit 20%)), Gulf (median tax margin of error is 20% of the estimate (limit 15%); effective-rate margin of error is 23% of the estimate (limit 20%)), Hamilton (median tax margin of error is 18% of the estimate (limit 15%); effective-rate margin of error is 21% of the estimate (limit 20%)), Hardee (median tax margin of error is 16% of the estimate (limit 15%); effective-rate margin of error is 23% of the estimate (limit 20%)), Hendry (median tax margin of error is 19% of the estimate (limit 15%)), Liberty (effective-rate margin of error is 26% of the estimate (limit 20%)), Madison (effective-rate margin of error is 21% of the estimate (limit 20%)), Taylor (median tax margin of error is 15% of the estimate (limit 15%); effective-rate margin of error is 24% of the estimate (limit 20%)).

Limits

• Survey medians mix homestead and non-homestead owners and long-time and recent buyers.
• Millage is a county average; cities and special districts differ.
• Non-ad valorem assessments are excluded.
• Figures are updated when the Census Bureau releases new ACS data (each December) and the DOR publishes its Data Book.

Build date: 2026-10-24. Data reproducible with the open build scripts (fetch_acs.py, fetch_dor_fl.py, gen.js).