NPPF Hotspots in England Where House Prices Could Rise
- A new National Planning Policy Framework was published on 17 August 2026 — the first full rewrite since 2012 — and its policies apply to decisions from the day of publication, with no transition.
- The hotspot is defined in the document, not by commentators. Land within around 800 metres of a station, where that station sits in a top-80 travel-to-work area by GVA and runs at least four trains an hour, or two an hour in one direction.
- Inside that radius the effects are concrete: outside settlements such schemes should be approved unless the harm substantially outweighs the benefit, in the Green Belt they are no longer inappropriate development, and a minimum of 35 dwellings per hectare applies — 45 where frequency is double.
- The frequency test has two limbs and they are not equivalent. Lichfields’ modelling of the two thresholds differs by roughly 770,000 homes, and most coverage quotes only the stricter one.
- Housing Delivery Test results were published the same day: 93 of 301 English authorities delivered under 75% of their requirement, which triggers the presumption in favour of development in those areas too.
A new National Planning Policy Framework was published on 17 August 2026 — the first full rewrite since 2012 — and it does something unusual. It defines its own hotspots.
Not in commentary, not in a press release: in the glossary. If you want to know whether your area is one, there are three tests and you can apply them yourself.
What counts as an NPPF hotspot?
Three conditions, all of which must hold. The framework calls the result a “well-connected station”, and land near one is treated differently from land anywhere else in England.
| Test | Threshold |
|---|---|
| Where | A top-80 travel-to-work area by GVA |
| Frequency | 4+ trains an hour, or 2+ one way |
| Distance | Around 800m, or a 10-minute walk |
The document’s own words for the second and third of those:
“served… throughout the daytime by at least four trains or trams per hour overall, or at least two trains or trams per hour in any one direction”
“this should be considered to be around 800 metres, or around 10 minutes’ walk time if topography, route availability and quality or physical barriers would prevent or discourage walking from up to 800 metres away”
Two details that matter and are easy to miss. The GVA ranking is frozen — the framework specifies 2023 ONS data, held fixed until the day after the 2028 figures appear, and fixed in five-year blocks after that. And if only part of a site falls inside the walking distance, only that part gets the benefit, which the glossary states explicitly.
What actually changes inside the radius?
Three separate policies switch on, and together they are a large change.
| Policy | Effect near a qualifying station |
|---|---|
| S5(1)(h) | Outside settlements, should be approved |
| GB7(1)(h) | In the Green Belt, not inappropriate |
| L3(2)(c) | Minimum 35 dwellings per hectare |
The Green Belt one is the sharpest. Development in that category “should not be regarded as harmful to the Green Belt or be required to demonstrate very special circumstances” — the test that has defeated Green Belt housing schemes for decades simply does not apply.
On density, the framework sets a floor rather than a ceiling: at least 35 dwellings per hectare, rising to at least 45 where service frequency is at least double the qualifying minimum. Exceptions are allowed only for sites below the major-development threshold where the standard is inappropriate or impossible, and for traveller sites.
The phrase to be careful with
Coverage of this framework describes a “default yes”. That phrase does not appear in the document. What S5 actually says is that listed development “should be approved, unless the benefits of doing so would be substantially outweighed by any adverse effects”.
That is a strong presumption. It is not an automatic grant, and the difference will be argued at appeal for years.
How much land does this cover?
Nobody knows precisely, because the qualifying test has two limbs that produce very different answers.
Lichfields modelled the scenarios in December 2025, before the final framework. Their figures for land outside settlement boundaries, using an 800-metre radius:
| Scenario | Homes potential |
|---|---|
| Top 80 areas, ~4 trains/hour | 719,200 |
| Top 80 areas, ~2 trains/hour | 1,487,000 |
| Top 60 areas, ~4 trains/hour | 632,600 |
| Top 60 areas, ~2 trains/hour | 1,314,000 |
Read the top two rows together. The final policy qualifies a station on either limb — four an hour overall or two an hour in one direction — so the real figure sits between them, and the gap is about 770,000 homes.
Most coverage has quoted the stricter reading. The looser limb is in the same sentence of the same definition.
Note also that the consultation proposed the top 60 travel-to-work areas and the final framework went to 80. That expansion happened between draft and publication, and on Lichfields’ own numbers it is worth roughly 87,000 to 173,000 homes depending on which frequency limb applies.
The second kind of hotspot
Housing Delivery Test results were published the same day, and they identify a different set of places. An authority scoring below 75% faces the presumption in favour of sustainable development — the same practical effect, arrived at by failing rather than by geography.
| Score | Consequence | Authorities |
|---|---|---|
| Below 75% | Presumption | 93 |
| 75–85% | 20% buffer | 19 |
| 85–95% | Action plan | 35 |
| 95%+ | None | 154 |
Just under a third of English authorities are in the first row. Their combined three-year shortfall is 158,693 homes.
The national picture looks far healthier than that distribution suggests: across all 301 authorities, 707,734 homes were delivered against 772,163 required — 91.7%. The shortfall is concentrated, not general.
Where delivery fell furthest short
Two different lists, because “lowest percentage” and “biggest shortfall” are not the same question.
Lowest scores:
| Authority | Score |
|---|---|
| City of London | 4% |
| Castle Point | 11% |
| Portsmouth | 11% |
| Southend-on-Sea | 20% |
| Basildon | 23% |
| Eastbourne | 23% |
| Gosport | 25% |
| Sevenoaks | 25% |
| Hertsmere | 26% |
Largest shortfalls, in homes:
| Authority | Score | Short by |
|---|---|---|
| Birmingham | 57% | 9,299 |
| Tower Hamlets | 47% | 6,197 |
| Bristol | 41% | 5,942 |
| Bournemouth, Christchurch and Poole | 35% | 5,211 |
| Greenwich | 42% | 4,928 |
| Leicester | 39% | 4,459 |
| Coventry | 47% | 4,199 |
| Newham | 61% | 3,829 |
| Sheffield | 58% | 3,779 |
The two lists barely overlap. The lowest percentages sit in smaller southern districts — Thames estuary, south coast, London fringe — while the largest absolute gaps are in major cities. A reader checking their own area needs to know which question they are asking.
When does any of this apply?
Immediately, for decisions. The framework states that its policies “are material considerations which must be taken into account in decision-making from the day of its publication”. There is no transition period on that.
More consequential for anyone with a live interest: local plan policies that are materially inconsistent with the new national decision-making policies “should be given very limited weight” from the same date. The only exception is policies examined and adopted against this framework — which, on day one, means almost none.
For neighbourhood plans the line is drawn at submission. Those submitted on or before 17 August are prepared under the December 2024 framework; anything not yet submitted must comply with the new one.
What about the Green Belt specifically?
Beyond the station route, major housing on land released from the Green Belt or on Green Belt sites must meet the Golden Rules. Until local policies catch up, the affordable housing contribution is 15 percentage points above the highest existing requirement that would otherwise apply.
There is a carve-out worth knowing: the Golden Rules do not bite on land released from the Green Belt through plans adopted before 12 December 2024, or on Green Belt permissions granted before that date.
So where could prices actually move?
This is the part where honesty is worth more than confidence.
What is established, and checkable today: where the new rules apply, how strong the presumption is, what densities are required, and which 93 authorities already fail the delivery test. That is a map of where development pressure is about to concentrate, and it is a real map — you can apply the three tests to any postcode in England.
What is not established: what any of it does to prices, anywhere. The framework was published on the day this was written. No study has measured its effect, because there is nothing yet to measure. Permission is not completion, and the gap between the two is exactly what the Housing Delivery Test exists to expose — those 93 authorities are, after all, places where housing was already required and did not get built.
Anyone offering you a list of towns whose prices will rise because of this document is working from the same information as everyone else, which as of today is a 130-page policy and no outcomes.
How to check your own area
- Find your travel-to-work area and its GVA rank. The framework points at one specific ONS dataset — UK GVA and productivity estimates for other geographies — and freezes it at 2023.
- Count the trains, not the platforms. The test is a normal weekday daytime timetable: four an hour overall, or two an hour in one direction. Planned upgrades count if there is a reasonable prospect.
- Measure 800 metres, and remember it can be partial. If half your site is inside the radius, half your site gets the policy.
- Check your authority’s Housing Delivery Test score. Below 75% and the presumption applies regardless of where the nearest station is.
- Use the government’s Connectivity Tool, which the density policy explicitly directs decision-makers to.
- Watch what gets built, not what gets permitted. That distinction is the whole reason the delivery test exists.
Sources
| Source | What it supports here |
|---|---|
| National Planning Policy Framework (17 August 2026) | Every policy reference, definition, distance and density quoted, and the implementation provisions in Annex A |
| Creating a clear, rules-based planning system | The publication date, the expansion from top 60 to top 80 travel-to-work areas, and what else was published alongside |
| Housing Delivery Test: 2025 measurement | Every delivery figure, score and band, computed from the published spreadsheet |
| Lichfields: how the new NPPF might unlock growth around rail stations | The homes-potential scenarios, published 16 December 2025 by Edward Clarke, Dominic Bowers and Matthew Spry |
Checked 17 August 2026. No affiliate links, and no payment was received for any link on this page.
How we verified this
The policy wording here was read from the 130-page framework itself, not from coverage of it. The document was published on 17 August 2026 and every quoted definition, distance, density and policy reference was taken from its own text and glossary. Where this page quotes, it quotes verbatim.
One phrase in wide circulation does not appear in the document. Reporting describes a “default yes” for development near stations. That phrase appears nowhere in the framework. What the text actually says is that such development “should be approved, unless the benefits of doing so would be substantially outweighed by any adverse effects” — a presumption with a stated exception, not an automatic grant. This page uses the document’s wording.
The Housing Delivery Test figures were computed from the government’s own spreadsheet and checked two ways. For all 301 authorities with a computable score, the published three-year subtotals equal the sum of the three annual figures, and the published measurement equals homes delivered divided by homes required. The banding computed here also reproduces the government’s own consequence column exactly — 93 Presumption, 19 Buffer, 35 Action plan, 154 None — which is a second independent check on the parsing.
Three rows were excluded and the reason is stated rather than hidden. The Isles of Scilly had a requirement of zero, so no percentage exists. Somerset and Cumberland appear as headings reading “consists of the authorities below” with no figures of their own. Excluding them is why the totals here refer to 301 authorities rather than 304.
The highest scores in that dataset are arithmetic artefacts and are not presented as achievement. Copeland was required to deliver four homes over three years and delivered 455, which prints as 11,375%. Any ratio with a denominator that small says more about the requirement than the delivery, so this page reports the lowest scores and the largest shortfalls instead, which are the two cuts that mean something.
The homes-potential figures are Lichfields’ modelling, not government estimates, and they predate the final framework. They were published on 16 December 2025 by named authors and modelled the consultation version. They are quoted as a range because the final policy’s two-limb frequency test straddles two of their scenarios.
On prices, this page reports what is known and says plainly what is not. No published analysis yet measures what this framework does to house prices anywhere in England, because it was published on the day this article was written. What can be established — where the rules bite, how hard, and which places already fail the delivery test — is set out above. Where prices go from here depends on how much of this permission converts into completions, and nobody has that evidence yet.
A source we could not open is not cited. One widely referenced analysis of the 2012 framework’s effect on land values returned HTTP 403, so it does not appear here even as background.