What happens to heritage case law now the NPPF has changed?

More than fourteen years of heritage case law was built on a set of paragraphs that no longer exists. Between 2012 and this August the courts did the slow work of turning eight paragraphs of national policy into a workable test: what substantial harm means, when the balance is struck, what an officer has to say and in what order. On 17 August 2026 the paragraphs those judgments construe were replaced. The judgments were not.

That leaves an awkward question for anyone with an application in the system. Which authorities still bite, and which have quietly become history? The answer is not uniform, and it turns on a distinction that is easy to state and easy to forget: whether a case was interpreting the statute, or interpreting the policy.

Two piles: the statute and the policy

The statutory duties are untouched. Section 16 of the Planning (Listed Buildings and Conservation Areas) Act 1990 still requires special regard to the desirability of preserving a listed building, its setting and any features of special architectural or historic interest which it possesses. Section 66 applies the same duty where planning permission would affect a setting. Section 72 still requires attention to the character and appearance of conservation areas. The new Framework does not amend the 1990 Act, and could not. It says so itself: footnote 2 records that the Framework “does not displace the statutory obligations placed on decision-makers by virtue of, among other things, section 38(6) of the Planning and Compulsory Purchase Act 2004, sections 66 and 72 of the Planning (Listed Buildings and Conservation Areas) Act 1990”.

National policy is a different animal. It is guidance, and the courts have been careful to say so. Its interpretation is a question of law, but it is not to be read as though it were a statute, and a judgment construing a particular form of words has no life beyond those words. Change the wording and the gloss goes with it.

So the pile a case sits in depends on what it was actually deciding.

What survives: heritage case law built on the statutory duty

East Northamptonshire District Council v Secretary of State for Communities and Local Government [2014] EWCA Civ 137, the Barnwell Manor appeal, is the clearest survivor. Sullivan LJ held that a decision-maker discharging the section 66(1) duty must give considerable importance and weight to the desirability of preserving the setting of a listed building, and that the duty applies with particular force where the asset is of the highest significance. That holding attaches to the statute. Policy HE6(3) now says much the same thing, which reinforces it rather than replacing it.

Catesby Estates Ltd v Steer [2018] EWCA Civ 1697 is in the same position. Its guidance on identifying the setting of a listed building, and on the physical, visual and historical connections that may contribute to it, describes a concept the statute uses and the new Framework retains. Nothing in Chapter 20 disturbs it.

The general reasons authorities also stand. A decision-maker must still give reasons that are intelligible and adequate, and must still grapple with the principal controversial issues. None of that came from the Framework.

What falls away with the paragraph numbers

R (James Hall and Company Ltd) v City of Bradford Metropolitan District Council [2019] EWHC 2899 (Admin) is the starkest example. It held that national policy recognises three categories of harm, substantial, less than substantial and none, and that there is no intermediate bracket between them. One of those three categories has been deleted. A case whose reasoning rests on the existence of a category that no longer exists cannot be applied to the new text.

A larger group goes the same way for a quieter reason. Safe Rottingdean Ltd v Brighton and Hove City Council [2019] EWHC 2632 (Admin), City & Country Bramshill Ltd v Secretary of State for Housing, Communities and Local Government [2021] EWCA Civ 320, R (Wyeth-Price) v Guildford Borough Council [2020] EWHC 3355 (Admin) and R (Save Stonehenge World Heritage Site Ltd) v Secretary of State for Transport [2021] EWHC 2161 (Admin) are all, at bottom, decisions about how paragraphs 193 to 196 of the old Framework fitted together: whether an internal heritage balance must be struck first, what may be weighed, and in what sequence. The reasoning is often transferable. The authority is not, because the architecture it describes has been rebuilt.

Bedford Borough Council v Secretary of State for Communities and Local Government [2013] EWHC 2847 (Admin) is a special case, because it has been displaced rather than merely stranded. It set substantial harm at the point where the significance of an asset would be vitiated altogether or very much reduced. HE5 now defines substantial harm as occurring where a key element of significance would be seriously affected. That is a lower threshold, and it sits in policy, where Bedford’s formulation never did.

The awkward middle

Two authorities are more interesting than either pile allows.

Jones v Mordue [2015] EWCA Civ 1243 held that a decision-maker who works methodically through the NPPF heritage paragraphs will have complied with section 66(1), and that a reference to one of them supports the inference that the rest were considered. The duty survives, because it is statutory. The shortcut does not, because it pointed at a numbered group of paragraphs that has been replaced. Whether working through HE5 and HE6 carries the same inference is untested. Until it is tested, the safer course is to address the duty expressly rather than rely on having cited the policy.

R (Kinsey) v London Borough of Lewisham [2021] EWHC 1286 (Admin) has gone the other way. It held that an officer who finds less than substantial harm without saying where within that range it falls has understated the harm. The category has gone, but the requirement has been promoted: HE5 now says in terms that the degree of harm should be identified. Kinsey’s principle is stronger this week than it was last, even though the words it was construing have vanished.

Appeal decisions were never binding, and still are not

It is worth being precise here, because the point is often put loosely. There is no doctrine of binding precedent in the English planning system. Each application for planning permission or listed building consent falls to be determined on its own merits, in accordance with the development plan unless material considerations indicate otherwise. Earlier decisions of a local planning authority or an inspector bind nobody and confer no entitlement to a like outcome. They are, however, capable of being material considerations, and consistency in decision-making is desirable, so a decision-maker departing from an earlier decision on a materially similar proposal should be able to explain why.

What has changed is how much help they give. An inspector’s reasoning from 2019 on whether a scheme crossed from less than substantial into substantial harm was applying a test that has since been redrawn. That decision remains evidence of the level and character of change found acceptable at a particular building, which is often the reason it is worth citing at all. It is no longer a reliable guide to how the question will be framed.

Where Local Plans now stand

This is the part that has had the least attention and may matter the most.

Section 38(6) of the Planning and Compulsory Purchase Act 2004 is unchanged. Applications are determined in accordance with the development plan unless material considerations indicate otherwise. The Framework is not the development plan. It is a material consideration, and the new one describes itself as one of critical importance.

Annex A deals with this directly, and it repays close reading, because it cuts both ways. Paragraph 1 provides that the policies in the Framework “are material considerations which must be taken into account in decision-making from the day of its publication”. There is no saving for applications already in the system.

Paragraph 2 then provides that development plan policies, “or parts of those policies”, which are materially inconsistent with the national decision-making policies “should be given very limited weight”, the only exception being where they have been examined and adopted or made against this Framework. So far, so severe. But the same paragraph goes on: “Other development plan policies should not be given reduced weight simply because they were adopted prior to the publication of this Framework.” Footnote 67 confirms that made neighbourhood plan policies sit inside that protection.

That second sentence is the one to have to hand, and it has been rather lost in the commentary. The test is material inconsistency, not age. An adopted heritage policy is not devalued because it predates 17 August 2026, and an argument that does no more than point at the adoption date should not succeed. Note too that the paragraph bites on parts of policies. A heritage policy with four limbs may lose one and keep three.

So where does that leave the average adopted heritage policy? It was drafted to mirror the Framework as it then stood. It very probably uses the phrase less than substantial harm, sets the public benefits balance up in the old shape, and may reproduce the great weight formulation that HE6(1) has replaced with substantial weight. Whether reproducing a superseded label makes a policy materially inconsistent with HE5 and HE6, or merely leaves it differently worded while pulling in the same direction, is precisely the argument that will be run at inquiries this autumn.

Our own view is that a policy applying the same protective approach in older vocabulary is not materially inconsistent with anything, and that the burden should sit with the party asserting inconsistency to identify a conflict in outcome rather than in wording. Annex A asks whether the policy is inconsistent with national policy, not whether it is written in the same words.

Two practical consequences follow. If you are relying on a local heritage policy, expect to be asked whether it still carries weight, and be ready to answer limb by limb rather than policy by policy. And if the policy was examined and adopted against this Framework, say so at the outset, because Annex A paragraph 2 exempts it outright.

What to do with a live application

1. Re-read the heritage evidence against HE5 and HE6 rather than the paragraphs it was written for. A Heritage Impact Assessment concluding less than substantial harm is using a category national policy no longer recognises.
2. Address the section 16 or section 66 duty expressly, rather than relying on having cited the policy and leaving the inference to be drawn.
3. Check every authority the case rests on and ask what it was construing. If it turns on the statute it stands. If it turns on the old paragraph structure, argue the reasoning rather than cite the case.
4. Read the development plan policies you are relying on for wording lifted from the old Framework, and be ready on the question of weight. Annex A paragraph 2 turns on material inconsistency, not on the age of the policy.

None of this is settled. The first challenges to decisions taken under Chapter 20 will take months to reach a hearing, and until they do a good deal of this is argument rather than answer. What is already clear is that a heritage case built on the old architecture, and cited in the old order, is going to look dated in front of an inspector this autumn.

This is the second of two pieces on the new Framework. The first, on how Chapter 20 redefines harm and what that does to the assessment itself, is here.

Heritage Unlimited prepares Heritage Impact Assessments, setting assessments and listed building consent advice for owners, architects and planning consultants across England. If you have an application in the system that was prepared under the old Framework, we can review the heritage evidence against Chapter 20 and tell you what needs to change. If the weight to be given to a local heritage policy is going to be contested, we would be glad to talk it through.