London planning data can sharpen housing debates, but it will not settle them

By Eleanor Marsh

London planning data is now good enough to challenge lazy claims about housing, transport and development — but not good enough to replace judgement. That distinction matters because London’s planning arguments often turn on a single number: homes approved, homes completed, schemes delayed, affordable units promised, or land identified for growth.

The strongest use of the data is not to declare a simple winner between City Hall, boroughs, developers and residents. It is to ask better questions: where permissions are being granted, whether homes are being built, how affordable housing is counted, and whether policy targets are being matched by infrastructure and delivery.

Why London planning data matters

London’s planning system is layered. The Mayor sets strategic policy through the London Plan, boroughs decide most applications, developers control many delivery timetables, and national planning rules shape the tests applied to local plans and housing need. A single development can therefore appear in several places: a borough planning portal, the Greater London Authority’s referral records, the Planning London Datahub, the London Development Database, and sometimes national planning datasets.

That creates an opportunity for public scrutiny. Residents can move beyond anecdote and check whether a borough is resisting new homes, approving them but not seeing delivery, or relying on a few large sites that may take years to complete. Councillors can test whether a claim about “overdevelopment” is matched by completions. Journalists can separate planning permission from construction. Campaigners can track whether affordable housing promises survive from committee report to built scheme.

It also creates a trap. Planning data is not neutral in the way a thermometer is neutral. It is structured around legal categories, reporting systems and administrative habits. If the question is badly framed, the spreadsheet will still produce a confident-looking answer.

What official sources show

Several official sources are now central to reading London’s development pipeline.

The Greater London Authority describes the Planning London Datahub as part of its digital planning work, intended to improve the collection and sharing of planning information across the capital. The older London Development Database has long been used to monitor planning permissions and development progress. The London Datastore provides a wider public catalogue for city datasets, while the UK Government’s Planning Data platform aims to standardise planning data nationally.

The London Plan remains the policy anchor. It sets strategic objectives for housing, transport, design, climate resilience and town centres. When data is used to assess whether London is delivering enough homes, or whether growth is being directed to suitable places, the Plan is usually the reference point.

The public benefit is clear: the official data landscape is richer than it was a decade ago. The risk is equally clear: richer data can make weak analysis look more authoritative.

Competing readings of the same evidence

Take a borough with a large number of approved homes but fewer completions. One reading is that the planning authority is doing its job and the delivery problem lies with developers, finance, construction capacity or infrastructure dependencies. Another is that the permissions are too concentrated on complex sites, making the local plan look stronger on paper than it is on the ground. A third is that viability negotiations have produced schemes that technically count towards supply but do not answer local affordability pressures.

All three readings can be plausible. The data alone will not tell readers which one is right.

The same problem appears with affordable housing. A committee report may describe a percentage of affordable provision. A later permission may include conditions or legal obligations. A revised scheme may change tenure mix. A completed development may be delivered in phases. If a reader compares only the first public claim with the final building, they may miss the steps in between.

There is also a political split over what should count as success. City-wide policy tends to reward aggregate delivery: more homes, more density near transport, more efficient use of land. Local objections often focus on immediate impacts: daylight, school places, GP access, construction disruption, loss of industrial space or pressure on buses. Both sets of concerns can be real. Data can show the scale of growth, but it cannot by itself decide how much change a neighbourhood should absorb.

What remains unclear

The biggest uncertainty is not whether London needs better planning data. It does. The uncertainty is whether the data will become consistent enough to support fast public comparisons between boroughs, schemes and policy outcomes.

Several caveats should stay visible.

First, application data is not delivery data. A permission may be implemented slowly, redesigned, sold, stalled or never built. Secondly, net additional homes are not the same as gross new homes, because demolitions and conversions matter. Thirdly, affordable housing categories are policy definitions, not a direct measure of whether a household on an ordinary local income can rent or buy nearby. Fourthly, planning records do not always explain why a project moved slowly. A stalled site may reflect interest rates, land assembly, remediation, utilities, legal challenge or a developer’s commercial timing.

There is also a geography problem. London does not function only by borough boundary. Housing markets, commuting patterns, town centres and infrastructure corridors cross administrative lines. A borough-by-borough league table can be useful, but it can also conceal the role of strategic sites, Opportunity Areas, transport capacity and land constraints.

Practical checks before using the numbers

Readers should treat planning data as a starting point for a chain of evidence, not the end of the argument.

Start with the London Plan to understand the policy test being applied. Check whether the question is about targets, permissions, starts, completions or affordability. Use the Planning London Datahub or London Development Database for the development record, then compare it with the relevant borough planning portal and committee papers. For larger or strategic schemes, check whether the Mayor had a referral role. If the issue is wider than one site, look for supporting datasets on the London Datastore or national Planning Data platform.

The most useful questions are practical:

• Is the figure counting applications, approvals, homes started or homes completed?
• Is the number gross or net of demolitions and losses?
• Does the affordable housing figure describe tenure, discount, eligibility or only a planning category?
• Has the scheme changed since the first committee report?
• Are infrastructure constraints recorded in the planning documents, or merely asserted in debate?

London’s open planning records are improving the quality of urban argument, but they do not remove the need to read carefully. The best public scrutiny will combine official datasets with committee reports, legal agreements, site visits and local context. When those sources point in different directions, the honest answer is not to force certainty. It is to say exactly what the data shows, what it does not show, and what must be checked next.