Explainer Cities, Urban Development & Infrastructure · Housing and Urban Policy · Issue 2026-W46
Unoccupied homes are not ready supply: what the new cities report changes in a housing decision
The Organisation for Economic Co-operation and Development (OECD) released its cities report on 4 November 2026. Can unoccupied dwellings relieve housing pressure? Answering requires separating existence from suitability, availability and location. We propose a record connecting the statistical count to an investigable unit, before a national stock becomes an unsupported local promise.
The Organisation for Economic Co-operation and Development (OECD) released its cities report on 4 November 2026. Can unoccupied dwellings relieve housing pressure? Answering requires separating existence from suitability, availability and location. We propose a record connecting the statistical count to an investigable unit, before a national stock becomes an unsupported local promise.
The count does not say the key is available
The housing chapter identifies a spatial mismatch between housing pressure and unoccupied dwellings. It defines the latter as homes that are nobody's usual residence at the census, including vacant, seasonal and secondary homes. That category is therefore not equivalent to homes ready to rent in the neighbourhood where they are needed. [1]
The publication year also does not mean every unit was observed in that year. The source annex shows varying data years and sources across countries. Retain the observation date and unit definition beside any number used in a housing memo or city comparison. Otherwise, differences in coverage or definitions can appear to be changes in available supply. [2]
Our editorial proposal is to start the memo with a specific need: housing for which group, near which employment or service destinations, and within which period? Without that boundary, a discussion can jump from distant or seasonal homes to a rapid estimate of a solution in another urban centre. This article provides no estimate of available homes in a Saudi or Gulf city. It offers a way to examine what would be needed before making one.
Record usability instead of adding up stock
We propose four separate states: a statistically observed unit; a unit whose status has been verified; a unit suitable for the intended need; and a unit that can be made available within a known period, cost and authority. Moving between states requires evidence. Missing information should not automatically earn a positive classification. Retain reasons for exclusion so that excluded units do not quietly return to the headline total in the next presentation.
Begin within a limited geographic area by reviewing the data definition and boundaries. Agree a verification method with authorised bodies that respects privacy and property rights. The record does not authorise entering property or requesting personal information without permission. Where a judgment requires structural safety, legal status or suitability for letting, refer it to appropriate specialists and authorities. An aerial image or incomplete database field cannot establish that a dwelling is fit for occupation.
After verification, ask whether the unit meets the intended households' needs in cost, space and access to work and services. Select those criteria for the actual problem rather than assuming a lower rent guarantees a suitable location. Record the availability condition, lead time, cost and decision-maker. Unknown information should be a recognised state in the record because it prevents the project from making a promise it cannot deliver.
A hypothetical example exposes the leap
Suppose 20 unoccupied dwellings are identified in a pilot area. A fictional review places them in non-overlapping groups: eight seasonal homes, five requiring an assessment for repairs, six with unresolved status, and one candidate for further availability checks. The categories sum to 20, but the number of near-term candidates under these assumptions is one, rather than 20. Candidate status still does not establish the owner's consent or an affordable price.
These numbers illustrate the reasoning; they estimate no market's conversion rate. Their value is to expose an assumption hidden inside the stock headline: how many units could actually serve the defined need? If the six unresolved cases cannot be checked, neither add them to available supply nor erase them from the evidence. Retain them as an information gap and decide whether further investigation is worthwhile.
The record may support a decision other than reuse. If potentially available dwellings are distant from work and services, it may be more useful to examine improved access or new supply elsewhere. Those are analytical alternatives, not recommendations for a particular project. Each has costs, timing and effects on existing residents. Comparing them requires their own evidence and constraints, rather than one assumed conversion rate.
The example also clarifies who should receive the results. A planning team may need the distribution of cases and unresolved conditions, whereas an authorised delivery team needs the evidence for a particular unit's next step. Sharing a large list of addresses with every participant is not necessary for every purpose. Agree access and responsibility before expanding the pilot, keeping the investigation proportional to the decision it is meant to inform.
Where the conclusion stops
The report does not establish that one policy for unoccupied housing will work in every city. Nor does this article infer a tax or legal obligation from foreign data. Any proposed instrument needs examination under local rules, implementation capacity, the rights of affected parties and potential effects on residents and prices. Before announcing available-unit totals, ask the accountable body for the source and update date of every transition in the record.
Our suggested pilot success measures are better knowledge of unit status and the time and cost of verification, followed by actual availability if an approved project begins. A count of addresses alone does not establish success. If verification costs exceed the expected value of information, or access authority remains unresolved, pause expansion and reconsider alternatives. That is a stopping point for exploration, not permission to bypass a formal process to complete a target.
At the next housing review, ask for one number describing stock and another describing units that meet the need and availability conditions, with the distance between them explained. The new report offers an opportunity to read the spatial comparison more carefully. Turning a statistical unit into usable housing still requires local evidence and decisions that a headline cannot supply.
Follow peers and make professional contacts through ARPS communities; ordinary membership is free. Which availability condition most often disappears when dwelling counts become delivery plans?
References
References open in a new tab.
Discuss this with practitioners
Create a free account to save articles and join the Specialists community discussion.
Discussion 0 published comments
No published comments yet. Be the first to share your experience.