The framework published on 17 August 2026 has not made speculative housing easier or harder in general. It has changed what the argument is about.
Under the old framework the argument was about weight. A council that could not demonstrate five years of housing land had its protective policies downgraded, and the decision-maker then weighed harm against benefit knowing the scales were tilted. Under the new one the argument is about category. Land is either inside a settlement or outside it, and if it is outside, a housing scheme has to fit one of ten permitted kinds of development before anything else is considered.
For anyone objecting to a field on the edge of a Cotswold town, that is a different fight with a different set of holds.
What a speculative promoter now argues.
A promoter with an option on a greenfield site outside a settlement has one realistic route, and it is policy S5(1)(j).
That policy says development which would address an evidenced unmet need should be approved outside a settlement. The framework then states in brackets what an evidenced unmet need includes for housing: the local planning authority cannot demonstrate a five year supply of deliverable sites, or it scored below 75% in the most recent Housing Delivery Test.
Cotswold District cannot currently demonstrate five years. So the gate opens, in the same way and for the same reason as before.
What happens next is not the same. There is no downgrading of the council's policies. Instead two conditions attach to qualifying at all. The development must be physically well related to an existing settlement. And it must be of a scale that can be accommodated taking into account the existing or proposed availability of infrastructure.
Only once both are satisfied does the balance run, and the balance is the ordinary one: approve unless the benefits would be substantially outweighed by any adverse effects.
What an objector gains, and what an objector loses.
Both are worth being honest about.
The gain is that two arguments which used to be weighed have become conditions. Whether a site is physically well related to the town, and whether the roads, schools, sewers and water supply can take it, are no longer factors thrown into a balance where the thumb is already on the scale. They decide whether the scheme is in the category at all. A site that sits beyond a gap, or a mile from the last house, is now arguing about its own admissibility.
Infrastructure capacity is the sharper of the two. An objection built on a highway authority's figures, a water company's consultation response or a school place forecast is now attacking the qualifying condition rather than adding weight to one side of a balance.
The loss is the weight argument itself. Under paragraph 11(d) a great deal of energy went into whether a council's policies were out of date and how much weight they retained. That question has gone. In its place sits Annex A, paragraph 2, which says that development plan policies materially inconsistent with the new national policies get very limited weight, and which will be used by promoters against old local plans across the country.
The framework does guard against the widest version of that. It says that other development plan policies should not be given reduced weight simply because they were adopted before publication. So the argument is about inconsistency, policy by policy, not about age.
Allocation and permission move the settlement boundary.
This is the provision that will surprise people, and it sits in the glossary rather than in a policy.
A settlement includes cities, towns, villages and other predominantly built-up areas. It also includes land which is allocated or has permission for development and which will form part of the built-up area once that development is complete.
A field with planning permission on it is part of the settlement. Not when the houses are finished. When the consent is granted.
There is a companion provision that stops this being read too widely. Policy S5(1)(i) treats the development of land allocated for that purpose in the development plan, where the allocation lies outside settlements, as its own separate category. So an allocation that will not form part of the built-up area, an employment site in open countryside for instance, stays outside. The glossary catches allocations that will become part of the town. S5(1)(i) catches the rest.
For a town like Lechlade, the practical effect is straightforward and uncomfortable.
The Wern has already enlarged Lechlade.
In June 2026 an inspector allowed up to 54 homes at the Wern, on the edge of Lechlade, over her own findings that the scheme would harm the rural character of the site and the setting of a listed farmhouse, and that it conflicted with both the local plan and the neighbourhood plan. She allowed it because the council could demonstrate only 1.8 years of supply. That decision was correct under the framework in force, and it stands.
Under the framework now in force, that consent does something further. The Wern is land with permission which will form part of the built-up area once it is built. On the plain reading of the glossary, it is now inside the settlement of Lechlade.
Which moves the edge of the town outward by 54 homes, on paper, today.
A town that loses an appeal now loses something extra. The consent enlarges the settlement, and the field beyond it becomes easier to describe as well related to the town.— The Editor · Ground Level, August 2026
The next field out has not moved. It is still outside the settlement, and it still needs a category under S5. But the condition it has to satisfy, that it be physically well related to an existing settlement, is measured against a settlement that is now larger and closer to it than it was in July.
This is not a loophole and it is not an accident. A framework that decides development by where the built-up area is has to say what happens to land that is about to become built-up area. But town and parish councils should understand the consequence. An appeal loss is no longer just the homes. It is the line.
The 150 homes at Lechlade.
Application 25/04047/OUT, for up to 150 homes on land east of Cotswold Bungalow off Burford Road, was submitted with a determination target date of 17 April 2026. It has not been determined. No officer report has been published and no committee date has been set.
It now falls to be decided under a framework that did not exist when it was submitted. Three things follow.
First, on the evidence available it is a site outside the settlement, so policy S5 applies and the promoter needs S5(1)(j). The council's supply position opens that gate. The scheme then has to be physically well related to Lechlade and of a scale the infrastructure can take. Fairford Town Council's objection of 18 March 2026, that the A417 through Fairford would be overwhelmed by these 150 homes together with the 54 at the Wern, is an objection to a qualifying condition rather than a makeweight. So is Thames Water's response, and so is the groundwater evidence residents assembled.
Second, it is not a strategic site. The framework defines one as typically having capacity for at least 1,500 dwellings, implemented in phases with significant infrastructure requirements and a masterplan. So policy HO13(4), which says development proposals should not be inconsistent with the scale, location and phasing of strategic sites in an emerging plan however advanced, does not reach this application. That route is closed, and it is better to know that than to argue it and lose.
Third, prematurity. Policy DM4(2) allows refusal on prematurity grounds only where the proposal is so substantial that granting it would undermine plan-making by predetermining decisions about scale, location or phasing that are central to an emerging plan, and where that emerging plan is at an advanced stage. The framework then says a plan is unlikely to be at an advanced stage if it has not yet been submitted for examination. Cotswold's Regulation 19 plan was approved by Full Council on 12 August 2026 for consultation. It has not been submitted. So the prematurity argument is weak today and becomes materially stronger on submission, which is a matter of timing that both sides can read as easily as we can.
The protection Cotswold may not reach.
Annex A contains a provision that every council in England will now be working towards.
Where a local planning authority can demonstrate a five year supply of deliverable housing sites with the appropriate buffer, and its Housing Delivery Test result is above 75%, a local housing need figure higher than the housing requirement in its adopted plan does not count as evidence of unmet need for the purposes of S5(1)(j). The protection runs for five years from the date the plan is adopted (Annex A, paragraph 3).
In plain terms: adopt a plan, hold your supply, and the gap between your plan's requirement and the standard method stops being a weapon for five years.
Cotswold District's own evidence says it will not reach that on adoption. Its Housing Land Supply and Stepped Requirement Topic Paper models 4.7 years of deliverable supply at 1 April 2028, the intended adoption date, and paragraph 9.11 states that a five year supply cannot be demonstrated at that date. The trajectory recovers to 5.2 years by 2029/30.
The protection is written for a council that can demonstrate five years. On the council's own projection, adoption alone will not deliver it, and the gate at S5(1)(j) stays open into the first years of the new plan.
That is the position as the evidence stands today, and there is a consultation and an examination between here and adoption. It is the number to watch.
What to do with an objection this week.
Stop citing paragraph 11(d). It does not exist.
Argue the category before the balance. On a site outside a settlement, the first question is whether the scheme qualifies under one of the ten kinds of development in policy S5(1) at all, and the second is whether it is physically well related to the town and of a scale the infrastructure can carry. Those are now conditions, and a well-evidenced objection to either is worth more than it was in July.
Ask your district where the settlement boundary runs, and on what evidence. Policy S2 requires plans to identify settlements and their boundaries. Where the adopted plan uses a development boundary drawn for a different purpose in a different decade, that question does not have a clean answer yet, and the answer will decide applications.