The Impact of Taxing Vacancy on Housing Markets: Evidence from France

Summary

Published in the Journal of Public Economics 185 (2020), article 104079, by Mariona Segú (Université Paris-Sud / Paris-Saclay), this is the first rigorous causal evaluation of a tax on vacant housing. (Quotes on this page are verified against the 2018 working-paper version, "Taxing Vacant Dwellings: Can fiscal policy reduce vacancy?", MPRA Paper 89686.) Framing, from the abstract: "Vacancy is a common phenomenon across developed countries. For policymakers, vacancy is undesirable as it challenges housing affordability, especially in large cities."

Design and Findings

France introduced the Taxe sur les Logements Vacants (TLV) in 1999 in urban municipalities over 200,000 inhabitants with tight housing markets. Using exhaustive fiscal microdata on every dwelling in France, 1995–2013, Segú compares taxed and untaxed municipalities with difference-in-differences plus propensity-score matching:

• "Results suggest that the tax was responsible of a 13% decrease in vacancy rates between 1997 and 2001" — 0.8 percentage points off an average 6.32% vacancy rate in taxed municipalities.
• "In absolute terms, there were 40,000 less vacant dwellings in treated municipalities than there would have been without the implementation of the tax."
• "The impact is specially concentrated in long-term vacancy. Results also suggest that most of the vacant dwellings moved to primary residences" — the freed homes were occupied, not shuffled into other idle uses.
• The effect was about 50% larger in municipalities with initially high vacancy.

Why It Matters

The finding does double duty. First, it establishes that a meaningful slice of urban vacancy is not frictional: dwellings idle for years returned to use when holding them empty became more expensive, which is exactly the behaviour the speculative-vacancy and option-value account (Cunningham 2006) predicts. Second, it is direct policy evidence that a holding cost on idle property changes owner behaviour — the mechanism underlying LVT's dampening effect on land speculation, tested here on buildings rather than land.

Limits

• The TLV is a tax on vacant dwellings, not a land value tax; extrapolation to LVT is by mechanism, not identity.
• The headline effect, while robust, is modest in absolute terms (0.8pp); a vacancy tax is not presented as a housing-supply policy on its own.
• Selection into treatment (the taxed cities were chosen for tight markets) is handled by matching and robustness tests, but the design is quasi-experimental, not randomized.

See Also

• Land underuse and speculative vacancy persist in high-demand cities — the outcome page this supports
• Speculative Vacancy · Prosper Australia, Speculative Vacancies 11
• Cunningham (2006), Seattle real options
• LVT dampens land speculation

Sources

1. Mariona Segú (2020), "The impact of taxing vacancy on housing markets: Evidence from France," Journal of Public Economics 185, 104079. DOI · open working-paper version (MPRA 89686, 2018) — used for all quotes and findings on this page (B-claims, quasi-experimental; quotes verified against the working-paper text, accessed 2026-07-10; the published version's wording may differ slightly).