Housing policy rests on assumptions. The record lets us test them.

A policy is introduced because somebody expects it to change what people do. Applicants respond, officers decide, schemes proceed or fail.

Those responses leave a record. Perfect Scale uses it to test what happened.

London does not operate one planning system for small sites.

Once differences in the applications each borough receives are accounted for, a wide gap in outcomes remains. Between the most restrictive authority and the most permissive it is 43.6 percentage points.

City of London: outside the 32 authorities with complete covariates.Barking & Dagenham: −16.4 points. 34.7% approved against 51.1% expected on its own caseload, n=144 decided.Barnet: −8.7 points. 43.0% approved against 51.6% expected on its own caseload, n=284 decided.Bexley: +18.9 points. 71.1% approved against 52.1% expected on its own caseload, n=363 decided.Brent: −7.6 points. 46.8% approved against 54.4% expected on its own caseload, n=622 decided.Bromley: −6.4 points. 47.6% approved against 53.9% expected on its own caseload, n=368 decided.Camden: +12.2 points. 75.8% approved against 63.7% expected on its own caseload, n=236 decided.Croydon: −21.8 points. 30.5% approved against 52.3% expected on its own caseload, n=824 decided.Ealing: +15.5 points. 69.0% approved against 53.5% expected on its own caseload, n=629 decided.Enfield: +7.3 points. 57.3% approved against 50.0% expected on its own caseload, n=844 decided.Hackney: +7.1 points. 63.2% approved against 56.1% expected on its own caseload, n=204 decided.Hammersmith & Fulham: +16.3 points. 75.5% approved against 59.2% expected on its own caseload, n=318 decided.Haringey: +13.8 points. 67.4% approved against 53.7% expected on its own caseload, n=390 decided.Harrow: −3.2 points. 47.9% approved against 51.1% expected on its own caseload, n=361 decided.Havering: −21.4 points. 29.5% approved against 50.9% expected on its own caseload, n=383 decided.Hillingdon: −1.7 points. 50.4% approved against 52.1% expected on its own caseload, n=345 decided.Hounslow: −13.1 points. 41.9% approved against 55.0% expected on its own caseload, n=296 decided.Islington: +11.3 points. 71.5% approved against 60.2% expected on its own caseload, n=214 decided.Kensington & Chelsea: +15.1 points. 80.8% approved against 65.7% expected on its own caseload, n=182 decided.Kingston: −10.5 points. 43.0% approved against 53.6% expected on its own caseload, n=244 decided.Merton: +8.7 points. 62.2% approved against 53.5% expected on its own caseload, n=331 decided.Newham: −14.6 points. 38.2% approved against 52.7% expected on its own caseload, n=317 decided.Redbridge: −0.5 points. 51.0% approved against 51.5% expected on its own caseload, n=798 decided.Richmond: +21.8 points. 79.5% approved against 57.7% expected on its own caseload, n=244 decided.Sutton: +15.0 points. 66.8% approved against 51.8% expected on its own caseload, n=226 decided.Waltham Forest: −15.3 points. 35.6% approved against 50.9% expected on its own caseload, n=382 decided.Westminster: +13.9 points. 78.8% approved against 64.8% expected on its own caseload, n=311 decided.Greenwich: −0.4 points. 55.2% approved against 55.6% expected on its own caseload, n=362 decided.Tower Hamlets: −8.8 points. 51.8% approved against 60.7% expected on its own caseload, n=164 decided.Lambeth: −16.8 points. 41.6% approved against 58.4% expected on its own caseload, n=286 decided.Lewisham: −1.9 points. 53.8% approved against 55.7% expected on its own caseload, n=292 decided.Southwark: +15.5 points. 71.7% approved against 56.2% expected on its own caseload, n=205 decided.Wandsworth: +4.0 points. 68.3% approved against 64.2% expected on its own caseload, n=520 decided.CroydonRichmondWestminsterEnfieldBexleyHavering
−220+22 pp
Fewer approvals than its caseload predictsMore
Decided applications
11,689
Authorities
32
Adjusted for
4 factors

Outside the 32 with complete covariates.

Descriptive, not causal. How this is calculated

Figure 1Mix-adjusted outcome differential by borough: each authority’s approval rate against what London-average decision behaviour would have produced on its own caseload, holding scheme type, unit count, conservation status and transport accessibility constant. Hover a borough for its figures.
Croydon
−21.8
Havering
−21.4
Lambeth
−16.8
Barking & Dagenham
−16.4
Waltham Forest
−15.3
Newham
−14.6
Hounslow
−13.1
Kingston
−10.5
Tower Hamlets
−8.8
Barnet
−8.7
Brent
−7.6
Bromley
−6.4
Harrow
−3.2
Lewisham
−1.9
Hillingdon
−1.7
Redbridge
−0.5
−24−120+12+24
Greenwich
−0.4
Wandsworth
+4.0
Hackney
+7.1
Enfield
+7.3
Merton
+8.7
Islington
+11.3
Camden
+12.2
Haringey
+13.8
Westminster
+13.9
Sutton
+15.0
Kensington & Chelsea
+15.1
Ealing
+15.5
Southwark
+15.5
Hammersmith & Fulham
+16.3
Bexley
+18.9
Richmond
+21.8
−24−120+12+24
Figure 2The same differential, ranked. Percentage points, from the most restrictive authority to the most permissive. 11,689 decided applications across the 32 authorities with complete covariates.
Read the research

The London Small Sites Dataset

Every residential planning application for one to nine homes across London’s 33 planning authorities, recorded to a common method. Decisions from January 2023 onwards, refreshed quarterly from the Planning London Datahub.

Applications
14,600
Decisions
11,694
Planning authorities
33
Decisions from
2023

Explore the dataset·Download the data

Latest research

AnalysisCroydonAugust 2026

What Croydon actually refuses on

Transport is Croydon’s most-cited refusal category, and the first-listed reason in 31 of 483 refusals. Design is the leading primary ground.

All research

Research programme

  • Do policy thresholds change what applicants propose?
  • What causes small housing schemes to be refused?
  • Do design codes change outcomes?
The full programme

Questions are welcome.

If you are working on housing or planning and there is a factual question the record might help answer, get in touch.

Work with us