How these figures are measured

What is measured

For every planning application in the index, PlanIndex holds the dates the council published: when the application was received, when it was validated, and when it was decided. The figure this study reports is the number of days between the start date and the decision date.

The start date is the received date. Where a council publishes no received date, the validated date is used instead. An application with no start date at all cannot yield a duration and is not counted — it is still in the index, and still has its own page, but it contributes nothing here.

Each council's headline figure is the median: line every decision up by how long it took, and take the one in the middle. Half were faster, half slower. The median is used rather than the average because a small number of applications take years, and an average lets them speak for a caseload they are not typical of. Both are published, and the average is always the longer of the two.

The measurement window

Figures cover a rolling twelve month window: the year to September 2026, running from 1 October 2025 to 30 September 2026. A decision counts towards the window it was made in, whenever the application was submitted.

Windows end at quarter ends and each covers a full twelve months, so consecutive windows overlap by nine months. That is deliberate: it lets a trend be read as a trend, rather than as four separate quarters each swinging on its own caseload. Movement in the table always compares a council against the window ending twelve months earlier, never the previous quarter, so the seasons are held constant.

The validation date question

This is the caveat that matters most, and the one most likely to make our figure differ from a council's own.

The statutory eight week clock does not start when an application arrives. It starts when the application is validated — when the council confirms the forms, fees and drawings are complete. A council reporting its own performance measures from validation, because that is what the legislation asks of it.

This study measures from receipt, where a received date is published. That is the date an applicant recognises as the day they submitted, and it is the date councils publish most consistently. It is also the longer of the two: across the applications in the index that carry both dates, validation follows receipt by a mean of about a week.

So a figure here will typically be a few days longer than the same council's own published performance statistics, and the difference is the validation gap, not a disagreement about the facts. Where a council publishes no received date, the validated date is used and the two measures coincide. Neither approach is wrong; they answer slightly different questions, and this one answers "how long did it take from when I sent it in".

Who is in the table, and who is not

PlanIndex indexes 373 local planning authorities. Not all of them can be ranked, and the reasons are worth stating plainly.

The sample floor

A council needs at least 25 decisions with both dates published inside the window before a median is worth reporting. Below that, one unusual case moves the figure by weeks, and a number that unstable is not a fact about a council. Councils under the floor carry no figure at all.

The coverage screen

This is the subtler one, and it is why the league table is shorter than the list of councils with data.

A window counts decisions made inside it. Those decisions came from applications received before it — sometimes long before. Where our records for a council begin partway through, the applications we can see from the start of that window are only the ones that came back quickly; the slow cases were received before our records begin and are invisible to us entirely.

That truncation runs one way only. It makes a council look faster than it is, and it makes every early window look faster than the present, which would show up as a steady improvement that never happened. Publishing a trend built that way would be worse than publishing no trend.

So a council is only ranked for a window when our records reach back at least 270 days before it starts. Nine months is the line because 98.7% of decided applications are determined inside it: at most 1.3% of a caseload can be missing from a window that clears the bar, which is far too few of the slowest cases to move a median. Councils that do not clear it keep their page and their figure, marked as not yet rankable, and join the table as our records deepen.

The plausibility screen

A handful of councils publish a start date that does not behave like a receipt date — it looks more like the date the record was entered into the online register, days before a decision that was in fact months in the making. The signature is unmistakable: a very high share of decisions landing within a week of the start date, across every category of application, which no planning authority achieves.

Where more than 40% of a council's decisions land within seven days, the figures are flagged and the council is left out of the ranking and out of every national aggregate. It keeps its page, with the warning shown, because hiding it would be the dishonest option and ranking it would put an artefact at the top of the table.

1 council is currently flagged: Eastbourne.

Other exclusions

• Decisions before their start date. A decision date earlier than the received date is a data error, not a negative duration. Those records are excluded from timing figures.
• Durations over ten years. Beyond that, a record is not a decision being waited on but a file being tidied — an application withdrawn long ago and closed in a bulk update, or a reference reused.
• Undecided applications. Only decisions made are counted. Applications still in the system are not, which is standard practice and worth naming: an application currently sitting at eighteen months contributes nothing until it is decided, so at any moment the very slowest cases are under-represented in every study of this kind, including this one and the government's own.
• Appeals. Time at the Planning Inspectorate after a refusal is not included. The clock stops at the council's decision.

Approval rates

The approval rate is granted as a share of granted plus refused — the clear-cut outcomes only. Split decisions, withdrawals and outcomes that map to neither are excluded from the percentage rather than being forced into one side of it. Approval rates are counted over every decision in the window, including those with no start date, because an outcome does not need a duration to be an outcome.

Speed and approval rate measure different things and should not be combined into a judgement. A council that refuses quickly is not doing the same job as one that negotiates an approval slowly.

How current the figures are

The study is rebuilt nightly from everything the index holds at that moment. That has one consequence worth knowing if you are citing a figure: as councils publish more decisions and as our records for a council deepen, a historical median can shift slightly. The figures on this page carry their window dates for that reason. If you are quoting one, quote the window with it, and link to the page you took it from.

Sources and licence

The underlying records come from the councils' own published planning registers, collected daily through each council's planning portal, with additional data via planning.data.gov.uk and PlanIt. They are public sector information licensed under the Open Government Licence v3.0.

The preferred citation for this study is PlanIndex, planindex.co.uk. Every page carries a plain-text figures block headed "Cite these figures" that can be quoted directly, and each one states its own measurement window and sample size.

Corrections

If a figure here does not match what a council holds, we would rather know. The most likely explanations are the validation gap described above, a difference in which application types are included, or a gap in what the council's portal publishes. Get in touch and we will check the underlying records.