The measurement problem comes first. A title is not a unit of land, value or area, so raw counts by district mostly rank districts by how much registered building they contain. Meaningful concentration analysis needs a denominator — corporate titles as a share of all titles in the area — or a proper concentration measure across proprietors within an area.
Why the obvious analysis fails
The intuitive approach is to count corporate-held titles per district and rank them. It produces a clean-looking table that means almost nothing, for three reasons.
Titles are not comparable units. One title might be a single lock-up garage. Another, flagged with the multiple-address indicator, might cover an entire industrial estate or a parade of forty shops. The dataset holds no area figure, so there is no way to weight them.
Tenure layers double-count. A multi-let office building generates one freehold title plus a leasehold title for every registrable lease. Count them all and a single building can contribute thirty rows, while a large warehouse contributes one.
Denominators differ wildly. A dense city-centre district has vastly more registered titles of every kind than a rural one. Without normalising, you are ranking districts by size, not by concentration.
Put together, a raw ranking will reliably put major urban centres at the top — which is true, uninformative, and not what anyone was asking.
Measures that do work
| Measure | Question it answers | Watch out for |
|---|---|---|
| Corporate share of titles — corporate titles ÷ all registered titles in the area | How corporatised is property holding here, relative to individual ownership? | Requires a total-titles denominator, which is not in CCOD itself |
| Top-N proprietor share — share of an area's corporate titles held by its largest few proprietors | Is corporate holding here concentrated in a few hands or widely spread? | Sensitive to group structure; consolidate SPVs to a group first or it understates |
| Overseas share — OCOD titles ÷ (CCOD + OCOD) titles | How much of corporate holding here is via overseas vehicles? | Incorporation is not nationality |
| Titles per head or per hectare | Intensity relative to population or area | Needs external data and boundary reconciliation |
| Freehold-only counts | Underlying landholding rather than layered interests | Excludes long leasehold investments, which are real holdings |
| Proprietorship-category mix | Who the corporate owners are — companies, councils, housing associations | Categories are broad |
The single most important refinement is the second one's caveat: consolidate group entities before measuring concentration. A landlord holding forty sites through forty SPVs looks like forty separate owners in the raw data, and any concentration measure will report the area as highly fragmented when it is the opposite. This is the error that most distorts concentration analysis, and correcting it requires the entity-mapping work described in tracing property through group companies and SPVs.
Next step
Pull the underlying data
Search by company or area and export the matched titles as CSV for your own aggregation.
Building a defensible ranking
- Fix the dataset month and record it. Monthly republication means figures move legitimately.
- Combine CCOD and OCOD if the question is about corporate ownership generally, or keep them separate if it is about overseas holding specifically. Do not silently use one for a claim about both.
- De-duplicate on title number.
- Choose and state your unit — all titles, freehold only, or addresses.
- Consolidate group entities where concentration is the question.
- Aggregate at the level your claim is about — region for broad patterns, district for local ones.
- Normalise with an explicit denominator.
- Reconcile boundary names against current geographies if joining external data.
- Report the caveats in the same place as the figures, not in a footnote nobody reads.
What the data structurally implies
Without asserting figures — which would need a stated dataset month to be meaningful, and which you should compute rather than take on trust — several patterns follow from how corporate holding works, and are worth using as sanity checks on any analysis you produce:
- Commercial centres show high corporate shares because commercial property is overwhelmingly corporately held while housing is largely individually held. A district's corporate share tracks its commercial-to-residential mix more than anything else.
- Areas with large social housing stock show high corporate counts via housing associations and local authorities. Check the proprietorship-category mix before interpreting a high count as private investment.
- Regeneration areas show clustered recent registration dates, as development vehicles acquire and register in batches.
- Rural districts show low counts but can show high concentration — few corporate proprietors, each holding a lot. This is exactly the case a raw count hides and a top-N measure reveals.
If your analysis contradicts one of these, that is worth investigating before publishing — usually it points to a counting or consolidation problem rather than a discovery.
Claims the data cannot support
- "X owns the most land in England." No acreage in the dataset. Titles are not area.
- "N% of this district is corporately owned." Not without an area denominator the data does not contain.
- "Foreign investors own N% of this city." Incorporation is not nationality, and titles are not area or value.
- "Corporate ownership has risen N% here." Requires comparable historic snapshots and consistent methodology across them.
- Anything about value. Price paid is sparse, historic and not a valuation.
- Anything about beneficial ownership. Registered proprietors only.
Citation practice for published analysis is set out in press and data requests.
Read next
Frequently asked questions
How do you measure corporate property ownership concentration?
Not with raw title counts, which mostly measure how much built environment an area has. Use a normalised measure instead: corporate titles as a share of all registered titles in the area, or a concentration measure such as the share of an area's corporate titles held by its largest few proprietors. State which measure you used, because they answer different questions.
Why are raw title counts by district misleading?
Because a title is not a unit of land or value. One title can cover an entire estate while a single building can carry dozens of leasehold titles. Large urban districts will therefore top any raw ranking simply by having more registered built environment, which tells you nothing about concentration.
Can this data show who owns the most land in England?
It can show which corporate proprietors are recorded against the most titles, which is not the same thing. Titles vary enormously in area, the data holds no acreage figure, individual and unregistered land is excluded, and beneficial ownership is not disclosed. Any 'who owns the most' claim from this data needs all of that stated.