Introduction
Part 1 of this post explained the central problem in the Ammori litigation. We summarized why the Divisional Court ([2026] EWHC 292 (Admin) thought the Home Secretary’s decision to proscribe Palestine Action was unlawful, and why the Court of Appeal ([2026] EWCA Civ 721)) reversed that decision. We also highlighted how each court approached the balance between freedom of expression and national security under the Convention. This part builds on that explanation and looks forward to the argument in the Supreme Court. It is a more technical discussion, which identifies some deeper problems with the law described in Part 1.
The essence of the Divisional Court’s argument in Ammori was that it is unreasonable to proscribe an organisation to make it easier for the police to disrupt the organisation, and to prevent it from persuading more people to engage in ‘terrorism’ as defined by s. 1 of the Terrorism Act 2000. The essence of the Court of Appeal’s argument is the opposite: that this would be reasonable. We need not feign surprise at this basic disagreement between judges faced with a complex problem. What we do find surprising is the legal edifice built to conceal that disagreement.
Had Ammori been decided a generation ago, we think it would have been framed as a rationality review. Below, we say what that means, and why we think this is a more natural legal expression for the disagreement between the Divisional Court and the Court of Appeal. In brief, our argument is that the approach to fair balance in these cases converts a conventional rights review into a disguised rationality review. To explain why, we need to say more about rationality review. We do this in the next section. We then compare the approach courts take under rationality review with the approach taken under proportionality. In the final section we set out our own argument.
Rationality
When the government gets the law right, follows proper procedure, and infringes no rights, the one form of review left to a court is rationality review. In the end, the court asks a familiar question: is the decision so unreasonable that no reasonable authority would make it? This means the court adopts a degree of critical distance from the decision. It is not deciding the matter anew, on the merits, for itself. It is checking whether the decision falls within the range of decisions that are lawfully available to the decision-maker.
Rationality review has two component parts. Both the outcome and the process of the decision must be reasonable. It’s possible to read the Divisional Court in Ammori as saying that the outcome was rational [29, 149]. The Court of Appeal clearly thought it was. But before a court determines whether the decision produced a reasonable outcome, it should first determine whether that decision was made in a reasonable way. A decision that flunks process rationality is unreasonable, even if the outcome would otherwise be reasonable.
Process rationality requires judges to specify factors that they believe to be essential or unacceptable. This is an original component of the test for reasonableness in public law. Lord Greene’s speech in Wednesbury ([1947] 2 All ER 680) says the decisionmaker ‘must call his own attention to the matters which he is bound to consider. He must exclude from his consideration matters which are irrelevant to what he has to consider.’ With characteristic coyness, judges attribute these ‘matters’ to Parliament: Lord Greene (at 682) says the factors are to be found ‘in the statute conferring the discretion.’ But this is disingenuous: judges claim they are doing Parliament’s bidding, but they largely identify irrelevant factors themselves – as Lord Greene puts it (682) ‘by implication’. When courts identify irrelevant factors this usually involves a value judgment. But this is both commonplace in UK public law and acceptable in a democracy.
As we explained in Part 1, Ammori is an argument about a policy: a body of non-legal rules that regulate the exercise of administrative discretion. The head of review that underpins these conclusions is usually process rationality in part because a policy will always be a relevant factor and, if the authority decides to depart from a policy, that decision must be rational. But the basic disagreement between the two courts in Ammori is not approached as a matter of process rationality and instead as a matter of compliance with s.6 HRA. We believe that the key to making sense of the decisions of both the Divisional Court and the Court of Appeal in Ammori is that a simple argument about relevant factors under the head of process rationality has instead been framed as a complicated argument about policy under s.6 HRA. The problem is not merely that this is abstruse, but that it results in a strange mixture of rationality and proportionality review.
Proportionality
There is a tactical reason why a litigant might want to frame an issue as a rights claim rather than a rationality claim. The duty under s.6 HRA is supposed to be a more searching standard of review than reasonableness. A court engaging in rationality review is supposed to maintain a degree of critical distance from the decision. A court engaged in rights review needs to supply the correct answer to a legal question. Hence, in Keyu [2015] UKSC 69, the Supreme Court decided not to use proportionality beyond the Convention, basically because the higher level of scrutiny raised constitutional questions which the Court was not ready to answer. In particular, Lord Neuberger [133] said this ‘would involve the court considering the merits of the decision at issue.’
One key difference between the proportionality assessment and rationality review is that the former assesses the decision on the merits. This does not mean making the decision again. It means asking whether the balance struck between the interests protected by the Convention was lawful or unlawful. In Keyu, Lord Neuberger explained the difference between considering ‘the balance which the decision-maker has struck between competing interests … and the weight to be accorded to each such interest’ and the court ‘thereby displac[ing] the relevant member of the executive as the primary decision-maker.’ In an orthodox process of rationality review, the Court does not need to ‘consider’ [CA 160] still less ‘weigh’ [CA 164], these relevant factors. Instead, it determines whether the decision-maker did those things. Weighing these relevant factors is the job of the primary decision-maker.
If this is an orthodox account of proportionality then we think there are two important errors in the approach taken by both the Divisional Court and the Court of Appeal. One error has to do with the things that were weighed under the fair balance part of the test. Another has to do with what it means to ‘weigh’. The court defers to the government’s view of the balance to be struck. In essence it asks whether that balance falls within a lawful range. But there is a reason why we explain rights compliance using a metaphor of balance. If we are weighing the right things, then there is only one balance that can be struck.
What to Weigh?
What does it mean to weigh the right things? In Part 1 we explain the stages of the proportionality test: is there an important objective; is there a rational relation between the interference and that objective; is a less intrusive means available and has a fair balance been struck. In this section we ask what are the interests the court is supposed to hold in balance in the final stage of that inquiry?
In Bank Mellat Lord Sumption [20] says it is ‘the rights of the individual and the interests of the community’ that need to be balanced. Lord Reed [70] describes the whole scheme of the Convention as a search for ‘fair balance between the demands of the general interest of the community and the requirements of the protection of the individual’s fundamental rights.’ This can be misleading. The public interest in speech and association is not merely the right of an individual. It is clearly an interest of the community as well. Later, Lord Reed expresses the same idea more precisely. He says the fourth stage asks whether the ‘impact of the rights infringement is disproportionate to the likely benefit of the impugned measure.’ It balances ‘the severity of the measure’s effects on the rights of the persons to whom it applies against the importance of the objective…’.
Lords Reed and Sumption say the things being balanced are rights. But the word ‘rights’ is used in a special sense here. The right to freedom of expression under article 10 is a qualified right. What it means to say you have a right to speak is that a balance has been struck between your interest in speaking and other interests, including national security and the rights of others. To call this interest a right is to skip balancing altogether – on this account you have a right to shout ‘fire’ in a crowded theatre. The right to free expression exists only to the extent that the interest in speaking is qualified by other relevant interests. The final part of the test checks the balance between these respective interests, which the Convention carefully refers to as ‘freedoms.’ It is constitutive of the right to free expression.
We underline that the court should still give weight to the government’s expert assessment of the threat to national security posed by Palestine Action. The government has special responsibility to ‘determine the danger’ (R v. Hampden 3 How. St. Tr. 826, 1235). It will necessarily do this when forming a view about whether the objective is important at stage one, and whether and to what extent the public interest in national security is engaged at stage four. But the decision of whether the action taken in response to the threat strikes a lawful balance is for the court. Moreover, the thing that receives deference is not the ‘balance struck’ by the government, so much as the assessment of the specific threat posed by this organisation. (Mark Elliot explores this point, in a different frame of reference in ‘From Bifurcation to Calibration’ in The Scope and Intensity of Substantive Review (Hart, 2015).)
Let us turn back to the ‘interests’ weighed by the Court of Appeal in Ammori. This is relatively easy to do, because they are set out in italics as subheadings under the heading ‘factors that weigh in the balance against proscription’ at [152] and later ‘in favour of proscription’ at [159]. They include ‘the rights of law-abiding individuals to freedom of speech and freedom of assembly’ [152] but later ‘The nature of Palestine Action’; ‘The level of the threat from Palestine Action’; ‘The escalating trajectory’; and ‘The key benefit of degrading the organisation and its finances.’ These are not ‘interests’ in the sense of ‘factors listed as qualifying the right to freedom of expression in Art. 10’. What they are, are relevant factors: the things the decisionmaker ought to have in mind when making her decision.
How to Weigh?
In Public Law after the Human Rights Act (Hart, 2010) Tom Hickman distinguishes two senses in which rights claims involve weighing:
‘On the one hand, ‘weight’ is being used to describe how [the different dimensions of the public interest] are to be reconciled. These competing considerations must be ‘weighed’, which is to say ‘balanced’. On the other hand, the term ‘weight’ is being used to describe the degree to which the court accepts the assessment of the Secretary of State as to the importance of the policy considerations’ (at 133).
Balancing involves weighing in the first sense, but not necessarily in the second sense.
Rationality review is both a variable intensity review and also review at a degree of critical distance. These are related ideas, but they are separate. Rationality review involves the giving of weight to the government in the second sense: how much weight depends on the legal and factual context. In other words, there is a lawful range of decisions available to the decisionmaker and the breadth of that range depends on what is at stake.
Not all examples of variable intensity review involve the giving of weight in this sense. Compare the common law requirement of procedural fairness. This is a question of law: what Canadian public lawyers would call a ‘correctness standard’ (Vavilov [2019] SCC 6). The question before the court is ‘what procedure does the law require?’ not ‘was this procedure within a lawful range of procedures?’ Nonetheless, what the law requires varies according to what is at stake. A decision to take something away demands more than a decision not to give something in the first place. In turn, it depends on how important that thing is.
As Lord Reed says in Bank Mellat at [70], ‘the intensity of review under EU law and the Convention varies according to the nature of the right at stake and the context in which the interference occurs.’ It is possible for the state to interfere with a penumbral expression of a freedom and a core expression of that freedom. Article 10 protects core political speech in hustings. It also protects (some) explicit sexual imagery, but with a lower intensity. At least in theory, proportionality does not involve a degree of critical distance from the government’s decision, as the reasonableness standard does. If it did, then there would be no reason for the Court to stay its hand in Keyu.
In Part 1 we underlined that the critical part of the proportionality test in Ammori is the final ‘fair balance’ part of the assessment. The Court of Appeal held that the government is ‘better placed than the court to adjudicate the balance of the various rights and interests engaged in the context of national security’ [116]. We think the court is giving weight in Hickman’s second sense, not the first. It is not balancing so much as placing a thumb on the scale.
As we note in Part 1, the Court of Appeal in Ammori refers to this weighing as the ‘margin of appreciation.’ Strictly, the margin of appreciation is the deference that the Strasbourg Court does to national authorities in the translation of the Convention into a domestic legal system. It is unhelpful and confusing to use this language in this context. Both court and the government are national authorities. Both are responsible for the translation of the Convention into UK law.
In Ammori the Court of Appeal first found that the decision ‘may have been borderline’, then concluded that this does not ‘reduce the latitude to which the decision-maker is entitled’, given the constitutional and democratic legitimacy of the executive government in matters of national security [180]. This is a strange sort of deference in which the court works out the correct answer, legally speaking, then rubs it out and writes in a new, more deferential answer. Trevor Allan critiqued this approach to deference ‘as either empty or pernicious’ – empty where it doesn’t change the law, pernicious when it does (‘Judicial Deference and Judicial Review’ [2011] 127 LQR 96). Having conceded, in terms, that a borderline case has been pushed over that line, the Court of Appeal claims, hollowly, that the ‘court remains the ultimate arbiter’ [116].
We think the Court of Appeal took this approach because it thought the decision of the Supreme Court in Shvidler requires a court to give weight to the ‘balance struck’ by the primary decision-maker. Shvidler was not a unanimous decision. In his dissent, Lord Leggatt warned that by incorporating a wide margin of appreciation in the ‘fair balance’ stage of the test, the majority risked abstaining from the critical decision: whether the interference is lawful. In a recent case note on Shvidler, Tom Hickman describes this as ‘the disappearance of proportionality’ [2026] 85(2) CLJ 314.
The approach of the Court of Appeal to the fair balance part of the test in Ammori amplifies Lord Leggatt’s warning. Either the Court determines the fair balance for itself to a correctness standard, or it defers to what the government decided. It cannot do both without collapsing the distinction between rationality and proportionality policed by Keyu. This is all the more important given that the Home Secretary argued before both courts that articles 10 and 11 were not engaged at all. No human rights impact assessment was conducted by the government or provided when the Order proscribing Palestine Action was laid before Parliament. No balance was struck. It is not clear what exactly there is to pay deference to. As Greene, Lock, and Murray put it, the judiciary simply ‘exits the scene’.
As we underline above, under ‘what to weigh’, some deference to the executive’s decision may be required. It is both appropriate and legally correct for the court to give weight to the government’s assessment of the threat to national security that the interference seeks to mitigate. That is no less than establishing what is at stake. But the court has a special responsibility to determine the law. There is no need for the court to adopt a degree of critical distance in this context as it does when conducting a rationality review.
Is this what fair balance looks like?
Proscription is a serious interference with articles 10 and 11 of the Convention. Any statement supportive of Palestine Action is now a criminal offence. The gravity of the interference is reflected in the very large numbers of people who have deliberately engaged in civil disobedience by inviting their arrest under the Terrorism Act for holding signs or articles expressing support for the group. Those offences may not affect the legality of proscription. But they are concrete and precise examples of the political freedoms protected by the Convention, juxtaposed against the broad and sweeping prohibitions that flow from proscription.
Palestine Action is not ISIS or the Provisional IRA, or even the suffragettes. Only three actions crossed the s.1 threshold, as the Divisional Court emphasised. But one of the actions carried out in their name resulted in serious injury and several resulted in serious damage to property, not least the Brize Norton action which precipitated the proscription. Others involved vandalism of buildings and organisations owned by Jewish people, at a time when the threat of violence against the Jewish community is very high. No-one representing Palestine Action appears to have much control over what is done in its name, and both lower courts were critical of the lack of transparency provided about the group in evidence.
With these factors in mind, even on a proper approach to proportionality, we suspect the ultimate outcome is unlikely to change. The Supreme Court can find more than enough material in the government’s case to rule that this was not a disproportionate use of the power. The argument presented here does not seek to pre-empt that. Rather, we think it is a mistake to maintain a critical distance from the government’s view when striking the lawful balance under the proportionality test. We stress that proportionality still leaves room for deference. However, the weight is given to the government’s assessment of the threat to national security, not to the government’s view of the balance struck between the public interest in security and the public interest in expression.
We reject the idea that the court should defer to the government’s view of the balance, and we expect many people will find this indefensible and undemocratic. The government frequently asks citizens to make substantial sacrifices for the common good. What possible mandate could a court have to gainsay that? Surely these constitutional concerns animate the decisions in Shvidler and Ammori?
We concede that there is a good, democratic argument that questions of policy are questions for the government, not the courts. But this is an elementary argument in UK public law: it is the backbone of common judicial review. Rationality review grants precisely this sort of deference to the government, for precisely this reason. But we think, until recently, everyone thought rights review was different. We think Parliament demanded that different approach when it placed judges under a statutory duty to decide whether a balance is incompatible with the Convention. A generation of judges recognised the democratic challenge this presents, not least the Supreme Court in Keyu.
We think the decision in Shvidler wishes this problem away. It assuages political discomfort about the legitimacy of judicial review by turning rights review into a disguised form of rationality review. It restores the basic principle that questions of policy are questions for the government, so judges need to keep their distance. This is a major change to the scheme Parliament contemplated when it enacted s.6 and it has been achieved without amending the HRA. This is exactly the sort of judicial intervention the Supreme Court would not contemplate in Keyu. It is ironic that it has been achieved by a court that is notionally committed to the democratic centrality of Parliament.
This is why we hope the Supreme Court will take a more orthodox approach to rights review. The way forward may involve a revision of the approach of the majority in Shvidler, which is apt to distort the balancing test, especially as it cascades through inferior courts. The court should be clear that the balancing of different factors is not the same as giving weight to the government’s view. These are different metaphors reflecting different processes. It must also identify what exactly is weighed. To count off a shopping list of relevant factors is to make the decision anew. The things that need to be held in balance are, on the one hand, the public interest in the freedom to speak and associate, and on the other hand, the public interest in national security and public safety. The weight given to these dimensions of the public interest is constitutive of the right to speak freely.
The authors wish to thank Paul Scott, Ben Yong, Colm O’Cinneide, Rick Rawlings, Jonathan Hall and an anonymous reviewer for their helpful comments. Errors are, of course, our own.
(Suggested citation: B. Keenan and E. Smith, ‘Proscription, Proportion, and Palestine Action – Part 2’, U.K. Const. L. Blog (3rd November 2026) (available at https://ukconstitutionallaw.org/))
Leave a Reply