About this site
data vintage 2026-08-13Florida Housing Data is an independent dashboard of Florida housing-market statistics. It covers all 29 metros (CBSAs, including micropolitan areas), all 67 counties, roughly 1,000 cities and places, and 1,115 ZIP codes — the whole state, at every level the underlying sources publish.
It exists because the housing numbers that circulate publicly are usually a headline without a denominator: a median with no count behind it, a year-over-year figure with no stated base period, or a listing statistic quietly compared against a closed-sale one. This site publishes the number, its source, its period, and how many observations stand behind it — and refuses the comparisons that do not hold.
Florida Housing Data is built and maintained by Kevin Cornwell. The data pipeline, the statistical choices, the site, and every word on it are his work, and mistakes on it are his responsibility.
Questions, corrections and data requests all go to the contact page or straight to hello@floridahousingdata.com.
This is not a brokerage and not a lead-generation site. It carries no advertising, no affiliate links, no sponsored placements, and no paid inclusion — no market can buy its way onto a ranking or off one.
It does not list properties, does not broker transactions, and does not sell or broker the visitor data it collects. Nothing here is an appraisal or investment advice; it is descriptive statistics about markets.
Listing and closed-sale metrics are never blended. A median list price and a median sale price measure different populations at different moments in a transaction. They are charted side by side and labelled, never averaged into one series.
Every figure carries its source and vintage. Each chart names the publisher it came from and the period it covers, and /health/ shows the freshness of every source at all times, including when one is late or has failed to load.
Derived numbers are marked as derived. Where a value is computed rather than published — a metro series assembled from its counties, a list-versus-sale spread — it is labelled as such and never presented as the source’s own figure.
Thin samples are disclosed, not hidden. Statewide rollups record how many counties stand behind them, and rankings apply a minimum-listing floor so a five-sale month cannot top a leaderboard.
Published recaps are frozen. A monthly recap reports the data as it stood when that issue was published and is never rewritten when sources later revise history.
Corrections are made in the open. If a number here is wrong, tell me and it gets fixed. Where a correction changes a published figure, the change is noted rather than quietly swapped.
Data is loaded from Redfin’s Data Center, Zillow’s ZHVI research files, the Realtor.com Residential Listings Data Library, and FRED (Federal Reserve Bank of St. Louis) into a Postgres warehouse, aggregated in SQL, and baked to static JSON that ships with the site. There is no API and no database at runtime — your browser reads exactly the files described in the methodology, and you can download the same ones.
Because everything is static and versioned, any figure on the site can be traced back to the source file and the load that produced it. The full definition of every metric, every known caveat, and the reasons some comparisons are refused are documented on /methodology/.
Source data belongs to its publishers and is used under their terms; see the site footer and methodology for attribution. Figures are provided as-is, without warranty, and may be revised by their publishers at any time.