System 11 · Schema designed, unpopulated

Ledger

Developer contributions, made mappable

The Ministry’s own schema holds £1.49bn of developer obligations. Not one record has a location. The property identifier field is populated in 0.0% of planning applications; the address field, in the same records, is populated in 98.1%.

HousingLocal governmentPublic money

The problem

When housebuilders get permission, they pay councils for the schools, roads and affordable housing the development will need. These are Section 106 agreements, and the money is enormous. Roughly £9bn sits unspent across English councils, including about £817m of affordable housing money — some 11,000 homes — with one London borough holding over £250m.

The Ministry already runs a national schema for this. It holds 39,325 contribution records totalling £1,491,818,575, models purposes from affordable housing to education to highways, and even tracks whether money has been received or spent. Councils are being asked publicly why the money is sitting there.

Every one of those 39,325 records has zero geometry. Not sparse — none. And the reason traces to one field: in the linked planning application dataset, the property identifier is populated in 0.0% of 100,627 records, while the free-text address field is populated in 98.1%.

That single contrast is the whole argument. The Ministry has already agreed the identifier belongs there. It cannot fill it. Coverage compounds the problem — 67 of roughly 330 authorities publish agreements, and 4 publish applications.

£1.49bn
Recorded contributions currently carrying no location at all
0.0% / 98.1%
Property identifier versus address field population in the same records
~£9bn
Developer contributions estimated unspent across English councils
67 of ~330
Authorities publishing agreements to the national schema

The system

Ledger resolves the addresses that exist into the identifiers that do not, then attaches contributions to sites. The moment that join is made, £1.49bn of recorded obligations becomes mappable for the first time — by council, by ward, by development, by purpose.

It then answers the question councils are actually being asked: what was promised, what was received, what was spent, and what is sitting. The schema already carries funding status; nothing currently reads it spatially.

It also fixes the identity problem underneath. Agreements link to authority-local application references that are not nationally unique, so the same reference string exists in multiple councils. Resolution must be constrained by authority or the entire dataset silently mislocates.

And it publishes coverage as a headline metric rather than a footnote. A national picture built from a fifth of authorities is useful only if it says so.


Worked examples

Two situations this system answers

EXAMPLE 1
The problem

A community was promised a school alongside a 400-home development. Six years on there is no school, and nobody can show whether the money was collected, spent, or is still sitting.

What Ledger does

Ledger ties the money to the site it was promised for, and shows whether it was ever collected and whether it was ever spent.

EXAMPLE 2
The problem

A council holds around £19m of unspent developer contributions and is asked publicly why. It cannot produce a spatial account of what was promised, for what, and where.

What Ledger does

Ledger puts every promise on a map — by council, by ward, by site, by what it was for. £1.49bn becomes visible for the first time.


Data foundation

Every dataset below is open, or its access constraint is stated

DatasetPublisherWhat it provides
Developer agreement contributionsMHCLG39,325 records, £1.49bn, purpose-coded. Open licence, live API. No geometry on any record.
Developer agreement transactionsMHCLG49,891 records carrying received and spent status — the promised-versus-delivered signal, already modelled.
Planning applicationsMHCLG100,627 records. Address text 98.1% populated, property identifier 0.0%. Four contributing authorities.
Infrastructure funding statementsMHCLG236 records from 172 authorities — but each is a link to a PDF, not data.
Title boundariesHM Land Registry via MHCLG8,220,990 polygons republished under an open licence. Keyed on a spatial identifier, not a title number.
New-build energy certificatesMHCLGAddress-level completion signal, for testing whether contributed-for development actually happened.

Capabilities

01
Address-to-identifier resolution
The 98,713 free-text addresses converted into property identifiers — the specific gap the Ministry cannot close itself.
02
Contribution mapping
Every recorded obligation placed, by council, ward and site.
03
Promised versus delivered
Received and spent status read spatially, answering the question councils are being asked publicly.
04
Purpose analysis
Affordable housing, education, transport, open space and monitoring fees separated and compared across authorities.
05
Authority-constrained matching
Application references resolved within their own authority, because the same reference exists in several.
06
Coverage reporting
Which authorities publish, which do not, and what share of England the picture actually represents.
07
Standards feedback
Resolution failures reported back to the Ministry as data-quality defects, improving the source.

Benefits

For government

  • Answers the unspent-contributions question spatially for the first time, for a Ministry already under pressure on it.
  • Fills a field the Ministry designed and cannot populate, improving its own platform rather than competing with it.
  • Gives councils an audit position on money they hold — an average of roughly £19m each is a real liability.
  • Produces the evidence base for whether developer contributions are delivering the infrastructure they were collected for.

For the public

  • Communities can see what was promised for a development near them, and whether it arrived.
  • Affordable housing money that has sat unspent becomes visible and therefore accountable.
  • Councils under scrutiny can demonstrate where money has gone rather than being unable to answer.

Delivery

Likely sponsor
MHCLG Digital Planning, individual planning authorities, the Home Builders Federation
Procurement route
DOS7 Lot 1, aligned to the planning data standards programme

Phasing

1
Resolve and map
3 months
Address resolution across the published application set, with match confidence reported per record.
2
Contribution layer
3 months
Obligations attached to sites and published by council and purpose.
3
Delivery tracking
4 months
Received and spent status surfaced, with the unspent position by authority.
4
Coverage campaign
ongoing
Work with non-publishing authorities to raise coverage from a fifth toward national.

Risks & mitigations

Coverage, not capability

The limiting factor is that most authorities do not publish. Ledger can resolve everything published and must state plainly how much of England that represents.

Reference collisions

Authority-local references are not nationally unique. Matching must be constrained by authority; failing to do so produces confidently wrong output.

Political sensitivity

Mapping unspent money names councils. Reporting presents context — money is often legitimately committed but not yet drawn — rather than a naked league table.


Sources

All sources checked in August 2026. Figures carry the reference period of their source, which may differ from publication date. Where a figure could not be verified against a primary source it is not used.