What share of vacant homes are uninhabitable or condemned versus simply unoccupied?
Executive summary
Official U.S. vacancy counts show millions of homes empty, but they largely exclude dwellings that are literally uninhabitable or condemned, meaning the vast majority of “vacant” units reported are simply unoccupied for reasons such as seasonal use, being on the market, or short-term vacancy rather than being structurally unlivable [1] [2] [3].
1. How “vacant” is defined — and why that matters
The Census Bureau’s operational definition treats a housing unit as vacant if no one is living there at the time of the survey, explicitly excluding units exposed to the elements or with positive evidence of demolition or condemnation; in short, the official vacancy pool is not supposed to include truly uninhabitable structures [2] [1].
2. The headline numbers: millions empty, but not necessarily ruined
Across U.S. datasets cited by media and analysts, roughly 14–16 million homes have been labeled “vacant,” which translates to roughly 9–10% of housing stock in different reporting periods — a large headline figure that feeds narratives about idle housing even as it mixes many different categories of emptiness [4] [1] [3].
3. One‑third seasonal or recreational — a sizable slice of “vacant”
Analysts who break down vacancy categories find that a significant portion of vacant units are seasonal, recreational, or occasional-use properties; one report highlighted that nearly one‑third of vacant homes can be seasonal/recreational, especially concentrated in vacation markets where a home can be “vacant” by definition even when used periodically [5] [4] [3].
4. Short‑term absences and market friction inflate vacancy counts
Census‑based vacancy statistics make no distinction between short‑term gaps (a house empty for a week while between occupants) and long‑term, perpetually idle units; that data quirk means many counted vacancies are transitory and not evidence of dereliction [2] homelessness.ucsf.edu/blog/vacancies-are-red-herring" target="blank" rel="noopener noreferrer">[6].
**5. Abandoned, foreclosed and condemned units — included sometimes, excluded often**
While abandoned or foreclosed homes can appear in vacancy tallies, units that are explicitly condemned or physically exposed to the elements are excluded from the primary Census vacancy series; therefore, the share of vacancies that are literally uninhabitable is intentionally minimized in that official count [2] [1].
6. Investors and off‑market ownership muddy the picture
Data and reporting suggest investors own a large portion of unoccupied properties in some datasets — for example, investor ownership accounted for major shares of vacant houses in ATTOM reporting — highlighting vacancies held for investment, renovation, or other non‑occupancy reasons rather than because homes are condemned [7].
7. What the numbers do and do not tell policymakers and the public
Because the principal national vacancy measures exclude blatantly uninhabitable units and do not distinguish short‑ from long‑term vacancies, policy arguments that equate the raw vacant‑home totals with a stock of move‑in‑ready housing for the unhoused are misleading; experts caution that vacancy rates are a blunt instrument and that homelessness and housing supply problems require nuance beyond headline vacancy counts [2] [6] [3].
8. Bottom line and data limits
The bottom line from available reporting: most of the 14–16 million homes called “vacant” are simply unoccupied for reasons like seasonal use, being for sale, or short gaps between residents, not condemned or uninhabitable — and the Census framework intentionally excludes clear cases of structural exposure or official condemnation from vacancy counts [1] [2] [3]. Precise nationwide shares of units that are condemned or structurally uninhabitable versus merely unoccupied are not provided in the cited sources, so an exact percentage split cannot be derived from the reporting assembled here [2] [1] [6].