I. Introduction
Readers of this article no doubt already know that certain UK courts have the power to make declarations of incompatibility under section 4 of the Human Rights Act 1998 (“HRA”). They will likely also be aware that this means that any of the courts listed in section 4(5) HRA may make a declaration if a provision is incompatible with a right contained in the European Convention on Human Rights (“ECHR” or “Convention rights”).Footnote 1 This article considers a logically prior question that, it is argued, has no clear answer as yet: what does it mean for a provision to be “incompatible” with a Convention right? The answer to this question is not only vital for the proper operation of section 4 HRA, but in at least three other circumstances too. First, legislation from the devolved legislatures that is incompatible with Convention rights is “not law”.Footnote 2 As devolved legislation depends on compatibility for its validity, the test used to determine incompatibility is vitally important to ascertain. Second, delegated legislation can be struck down, or declared incompatible, by the courts if it is incompatible with Convention rights.Footnote 3 And third, under section 19 HRA, the minister in charge of a bill must either make a statement that, in their view, the bill is compatible with the Convention rights or (more unusually) make a statement to the effect that although they are unable to make such a statement, the Government nevertheless wishes to proceed with the bill. Following that statement, compatibility will also be considered during parliamentary scrutiny of the bill (including by the Joint Committee on Human Rights (“JCHR”)). Once again, it seems crucial for there to be a clear understanding of what an incompatibility is if ministerial statements and parliamentary scrutiny are to be meaningful and if a healthy dialogue between the executive, Parliament and the judiciary – the ultimate arbiter of whether a provision is in fact compatible or not – is to be maintained.
The question of what it means for a provision to be incompatible with a Convention right has received relatively little scholarly analysis, although there is a growing body of scholarship critiquing the courts’ review of legislation under the HRA and querying whether section 4 assessments should be made on the facts of the instant case or whether the courts should be performing a more “expository” role.Footnote 4 This article adds to the existing scholarship by providing a thorough treatment of the most relevant judicial decisions, not just in relation to section 4 HRA, but also in the context of compatibility (and therefore validity) of devolved legislation, as well as linking the discussion to section 19 HRA. In particular, the article draws attention to the contradictions in the judgments, as well as lamenting the fact that cases have tended to fixate unduly on how many people the legislation might affect (and have set that threshold too high), whilst neglecting explicit consideration of the equally important question of how likely (or foreseeable) the breach might be. The article will first establish the importance of clarifying the test used to establish incompatibility. It will then critically examine the relevant case law on the matter, revealing a lack of clarity in those cases. The article will then go on to provide much-needed clarification by developing a two-part test, namely considering first, how many people must likely be affected before a provision may be deemed “incompatible” and, second, the foreseeability, or likelihood, of them being affected. It will conclude by highlighting the additional clarification that can be brought to the notion of who is a “victim”, or potential victim, of a breach of Convention rights. The test proposed in this article would simultaneously better protect rights whilst honouring the intended purpose of the HRA, by better facilitating dialogue between the branches and layers of the state.
II. The Importance of a Clear and Consistent Test for “Incompatibility”
Before trying to solve the difficulties inherent in assessing compatibility, it is important to first explain why such an exercise is essential. It will be argued in this article that, despite the different contexts, the test used to establish incompatibility should be consistent across each of the four areas mentioned – section 4 HRA, section 19 HRA, the competence of devolved legislators and the legality of delegated legislation – in order to achieve maximum clarity. It is important to note that all four circumstances concern the compatibility of legislative provisions, in contrast to the compatibility of decisions or actions of public authorities towards one specific person or group of people (as would be challenged under section 6 HRA). The courts have generally not adapted with ease to their role under the HRA of reviewing legislation, rather than specific actions. Courts tend to be reluctant to engage in abstract reasoning, because of concerns about the constitutional propriety of “expound[ing] a principle [without] a particular factual context”.Footnote 5 In addition, Janet Hiebert has pointed out that, before the HRA, “rights were thought of as being protected by Parliament, and not from it”.Footnote 6 Concerns about reviewing legislation as a whole do not apply when the court is looking at a specific application of the facts under section 6 HRA, nor when using the power of interpretation in relation to the instant case under section 3 HRA. Where the provision itself is deemed to be compatible, the alternative for the challenger is to argue that the compatible legislation has nevertheless been operated in an incompatible way by a public authority. In other words, it is crucial to determine whether the fault lies with the legislator or the decision maker. The focus of this article is on the approach to cases of the former type.
There are dangers and difficulties in the potentially more abstract reasoning involved in reviewing legislation as opposed to specific administrative decisions and the courts’ current sparing use of section 4 HRAFootnote 7 may be the reason for section 4 being relatively uncontroversial in attempts by the previous Conservative Governments to repeal or reform the HRA. Alternatively, section 4 HRA might just be “politically less contentious” than section 3 HRA, section 4 having fewer “unexpected side-effects” than the immediate remedy of section 3.Footnote 8 Whilst successive Conservative Governments – including in the abortive Bill of Rights Bill of 2022–23 – planned to repeal (or at least reduce) the interpretation power under section 3 HRA, section 4 HRA generally escaped scrutiny and controversy.Footnote 9 Under the previous Conservative Government, the HRA latterly became at risk of more insidious erosion by provisions such as Illegal Migration Act 2023 (“IMA”), section 1(5), which was designed to remove the application of section 3 HRA from provisions “made by or by virtue of” the IMA. The IMA would not have removed the operation of section 4 HRA. That section 4 HRA was destined to survive unscathed from attempts by the previous Conservative Governments to reform or repeal HRA is to be celebrated. However, it is possible that the planned reforms would have had an incidental impact on section 4 HRA. Without the section 3 HRA power of interpretation doing the majority of the work as it currently does,Footnote 10 it would have been possible that more section 4 declarations were made, with the potential that section 4 itself attracted controversy. Under the current Labour Government, the HRA may be safe for now, but, should either of them form the next Government, both the Conservatives and Reform UK propose not only to repeal the HRA, but also to withdraw from the ECHR.Footnote 11 In any event, it should not be assumed that section 4 HRA is working perfectly. Some clarity is needed on how “incompatibility” is assessed and it is argued here that the courts can afford to take an approach that is more in line with the drafters’ intention without potentially jeopardising the future of section 4 HRA.
Section 19 HRA is a provision that has received relatively little academic scrutiny.Footnote 12 As Hiebert has pointed out, section 19 HRA was envisaged as a mechanism to “improve rights protection by Parliament” and, along with other HRA provisions, has often been described as creating a culture of rights, or culture of justification.Footnote 13 But the operation of section 19 HRA is limited by its failure to require the minister to give reasons for their belief in compatibility,Footnote 14 which results in “formulaic” statements.Footnote 15 These deficits are partially alleviated by the role of the JCHR in probing further into reasons for apparent compatibility and fostering more debate on potential incompatibilities.Footnote 16 As such, a clear and consistent test for “incompatibility” would be beneficial in the context of both section 19 HRA and subsequent parliamentary scrutiny. In her qualitative research, Hiebert uncovered some worrying examples, such as positive statements of compatibility being made even where government legal advisers had identified a “high level of risk” of incompatibility,Footnote 17 and the unlikelihood of a negative JCHR report resulting in changes to a bill, unless the Government thought the bill would be defeated.Footnote 18
Most of the time section 19 HRA operates fairly uncontroversially, with statements being made as a matter of course. Until the latter years of the last Conservative Government there had only been five instances of a statement of compatibility not being made.Footnote 19 There were then two examples in quick succession of bills being passed where no statement of compatibility could be made, namely the aforementioned IMA and the Safety of Rwanda (Asylum and Immigration) Act 2024. The aforementioned Bill of Rights Bill also proposed to remove the section 19 requirement altogether. It is to be hoped that we are now back in a phase of “business as usual” with section 19 HRA, given that a negative statement was intended to be “a rare and exceptional event”.Footnote 20 However, it is important to consider what test a minister applies (or should apply) when determining whether or not they think a provision is compatible. Without considering what “compatibility” means, this exercise is largely performative, and puts a healthy dialogue between the political branches and the judiciary in jeopardy if they do not share the same definition of “incompatibility” across both sections 4 and 19 HRA. A shared understanding of compatibility does not mean that no declarations would be made, not only because of the existence of pre-HRA legislation, but also because a minister and a court may apply the same test and reach a different conclusion, just as two different courts might apply the same test and reach a different conclusion.Footnote 21 But the current lack of shared understanding as to the test to be applied compromises dialogue between the courts and political actors when a provision that has been subject to a positive section 19 statement is later ruled by the court to be incompatible. Governmental/parliamentary relations are also currently jeopardised, as it is unclear whether ministers and the JCHR, or parliamentarians generally, share the same understanding of what “incompatibility” means.
It is true that the political branches and the courts will reason about human rights in different ways.Footnote 22 Ministers will consider whether the provision is likely to survive a legal challenge in terms of “risk”,Footnote 23 their legal advisers having assessed whether, in their view, the provision will survive a legal challenge,Footnote 24 whereas the court will be considering whether the legislation will “always, or almost always” give rise to a breach.Footnote 25 Despite the different viewpoints, it would be useful to share the same test. After all, the goal should be the shared one of not allowing legislation incompatible with Convention rights to pass or remain on the statute book, unknowingly, and of course the advice given to a minister will be “heavily” based on the existing case law.Footnote 26 In any event, without prior consideration of what “compatibility” means,Footnote 27 what precisely are political actors looking for? Adopting the test set out in this article would make the task more straightforward and transparent as the questions asked can be clear: might the legislation breach anyone’s rights and how likely might that breach be/how foreseeable is it? Of course, reasonable minds may differ as to the answers to those questions,Footnote 28 but agreeing on the questions would be an improvement on present practice. Transparency is vital for facilitating “democratic dialogue” and “constitutional collaboration”.Footnote 29
In relation to the compatibility of devolution legislation too, it is axiomatic that the meaning of compatibility is vitally important. After all, it will determine the validity or otherwise of a piece of devolved legislation. Again, it seems useful to consider how incompatibility should be ascertained, to foster a harmonious relationship between the UK and devolved layers of the constitution and to allocate power at both levels appropriately. The same applies to delegated legislation and relations between the executive and the courts. In either case an overly broad judicial conception of “incompatibility” risks inappropriate erosion of power delegated to the relevant body (whether devolved legislature or executive) whereas an overly narrow conception risks allowing those same decision makers to act beyond their powers. A clear understanding of how incompatibility is assessed allows all branches and layers of the state to know where they stand.
It is therefore clear that the question of what “incompatibility” means must be answered in relation to section 4 HRA, section 19 HRA, the competence of devolved legislators and the legality of delegated legislation. But like many fundamental questions, that answer is both very important and very difficult to answer. The difficulty is particularly evident in the many instances where the right engaged is qualified rather than absolute and thus requires to be balanced against other considerations and/or other rights, meaning that the provision’s compatibility is not clear-cut. Some options for assessing compatibility were discussed in relation to devolved competence in Re Abortion Service (Safe Access Zones) (Northern Ireland) Bill.Footnote 30 Lord Reed, giving the judgment of the court, explicitly rejected Lady Hale’s dictum in Re McLaughlin’s Application for Judicial Review – a section 4 HRA case – that incompatibility “in a legally significant number of cases” would suffice for the finding of incompatibility.Footnote 31 Instead, Lord Reed preferred a higher threshold – that incompatibility would only be established if a provision would give rise to an unjustified interference with a Convention right “in all or almost all cases”.Footnote 32 To be fair, neither definition is without problem. In either case, the court must grapple with probabilities and scenarios beyond the instant case, as well as considering what threshold is required for a number to be “legally significant” or “almost all”.Footnote 33 But, it is argued here, that something closer to Lady Hale’s definition can be preferred – in relation to devolved and delegated competence, section 4 declarations and section 19 statements – without leading to inappropriately abstract reasoning (particularly important in a judicial context). Indeed, this article will go further in suggesting that such a test in fact allows for less abstraction because it does not involve having to think in hypotheticals, provided the likelihood, or foreseeability, of the breach is given proper consideration in addition to the number of people potentially affected. In Re Abortion Services, the court fixated on the number of people potentially affected, without explicitly considering the equally important question of the likelihood or foreseeability of a potential breach (is it, for example, likely, highly likely or inevitable?). The correct answer, it is submitted, is that if anyone’s rights are inevitably violated, the legislation is incompatible. But furthermore, if a more widespread breach can be ascertained, that the threshold of likelihood may be lowered to “likely” rather than “inevitable”. In other words, there is no hard and fast rule as to the precise number of people affected or the likelihood of them being affected. Both considerations must be taken into account in conjunction with one other and the overall context of the case (e.g. the right engaged) – a higher “likelihood” might require a smaller number of affected persons and vice versa.
It is important to clarify, however, that the argument of this article is not that section 4 HRA should be deployed indiscriminately. The purpose of this article is to attempt to facilitate a long-overdue discussion as to how incompatibility should be assessed. It is to encourage us to think more carefully about how to fulfil simultaneously the roles of highlighting human rights violations, allowing the political actors sufficient discretion as to whether and how to remedy those violations, and providing devolved and executive lawmakers (and those scrutinising their decisions) with a clear framework for how to test for compatibility. The same logic of fostering dialogue can be transferred to the other contexts mentioned above, namely section 19 HRA and the compatibility of devolved and delegated legislation.Footnote 34 Although the stakes may be higher for the devolved nations, in the sense that incompatible legislation is void ab initio, that in itself does not justify a different test being used. Legislation should still be considered in terms of both the number of potential people affected and the likelihood/foreseeability of a breach, alongside the context. The test for incompatibility preferred in this paper could allow policy-makers freedom to refine policies to render them Convention compliant, without always having to overhaul them radically.
III. The Existing Case Law
A. A Lack of Clarity in the Cases
In relation to declarations of incompatibility, I have argued elsewhere that a requirement that rights need to be breached in “all or almost all cases” sets too high a bar that unduly risks rights protection by leaving legislation that is known or highly suspected to be incompatible in some cases on the statute book.Footnote 35 This article seeks to apply those arguments beyond section 4 HRA to devolved and delegated competence and section 19 HRA. It also highlights that the current test used by the courts concentrates on the number of people potentially implicated whilst neglecting explicit consideration of the equally important issue of the likelihood or foreseeability of a breach. Re Abortion Services is one of the most recent cases to consider the question of what test should be used to determine whether a measure should be considered incompatible with Convention rights. It is the culmination of a line of cases discussed in the judgment and here, namely R. (on the Application of Bibi) v Secretary of State for the Home Department,Footnote 36 Christian Institute v Lord Advocate Footnote 37 and Re McLaughlin.Footnote 38
The question of how many people need to be potentially affected cannot be answered in isolation from the question of the likelihood or foreseeability of the potential breach, although the courts have tended to neglect explicit consideration of the latter question. The relationship between the two considerations is well illustrated by the case of R. (Bibi) v Secretary of State for the Home Department.Footnote 39 It concerned a challenge to the validity of an Immigration Rule (“the Rule”) whereby non-European spouses/partners of British persons or non-European settled persons had to pass a test of competence in the English language before emigrating to the UK. Like Re Abortion Services, the case started as an ab ante challenge to the Rule, although, unlike Re Abortion Services, the Rule was in force by the time of the judgment.Footnote 40 The Rule was challenged on the basis, inter alia, that it was contrary to Article 8 ECHR, or Article 14 ECHR in conjunction with Article 8 ECHR. The Guidance given to caseworkers on how to apply the Rule provided for very limited exceptions that could be made to the Rule, including the fact that, for example, difficulties in travelling to a test centre were not typically relevant reasons for waiving the fluency requirement.
Lady Hale (with whom Lord Wilson agreed) described the challenge to the Rule as “a difficult task” because, whilst it might be easy to show that a rule was incompatible with an individual’s Convention rights, it would be “much harder to show that the Rule itself is inevitably unlawful”.Footnote 41 Lady Hale observed that the Rule, as applied under the Guidance, set out pre-entry language requirements that posed “very substantial practical problems” for some spouses (e.g. in terms of affordability and/or availability of tuition and test centres in their country of origin) and that therefore “there are likely to be a significant number of cases in which the present practice does not strike a fair balance as required by article 8”.Footnote 42 She continued, however, “[t]his does not mean that the Rule itself has to be struck down” because “[t]here will be some cases in which the interference is not too great”.Footnote 43 But Lady Hale goes on to say that there “may well be some benefit […] in declaring that [the Rule’s] application [i.e. under the Guidance] will be incompatible with the Convention rights of a UK citizen or person settled here, in cases where it is impracticable without incurring unreasonable expense for his or her partner to gain access to the necessary tuition or to take the test”.Footnote 44 She makes no final decision on this point since finding an incompatibility in the Guidance as opposed to the Rule was not the remedy sought. But crucially, she was open to a finding of incompatibility (of the operation of the Rule under the present Guidance) on the basis that it “is likely to be incompatible with the convention rights of a significant number of sponsors”.Footnote 45
By contrast, Lord Hodge (with whom Lord Hughes agreed) said:
I think that there may be a number of cases in which the operation of the Rule in terms of the current Guidance will not strike a fair balance. But there may also be many cases in which it will. The court would not be entitled to strike down the Rule unless satisfied that it was incapable of being operated in a proportionate way and so was inherently unjustified in all or nearly all cases.Footnote 46
Lord Neuberger, however, went so far as to say that the Rule “seems to be bound to result in [A]rticle 8 rights being infringed on a number of occasions”.Footnote 47 Like the other Justices, he was concerned about the Guidance that was used to interpret the Rule, going on to say that it appeared to him “virtually certain that there will be a significant number of cases where application of the Guidance will lead to infringement of [A]rticle 8 rights”.Footnote 48 He went the furthest in being tempted to make a declaration in respect of the Guidance, making the important point it would then give a “formal record” of the court’s concerns, as well as lamenting the needless cost and delay of issuing further proceedings.Footnote 49 In the event, the fact that that was not the remedy sought weighed against him making the declaration.
R. (Bibi) v Secretary of State for the Home Department clearly demonstrates a lack of consensus on the number of people affected (a significant number, or even just “a number” vs. all or almost all), but it also shows disagreement about the likelihood of that group facing an incompatibility (likelihood vs. virtual certainty vs. inevitability). One occasionally wonders what import may be placed on a particular choice of word in a judgment (“may” or “likely” or “bound”, for example). Without explicit explanation, or even emphasis, one can only hope that it has been chosen with precision in mind. However, examining legislation is not an exact science. Some provisions might plainly be inevitably incompatible with Convention rights, but only in very few cases. Some other provisions might appear to affect a much greater number of persons – but the potential for a breach in all of those cases might be more “likely” than inevitable. The messy truth is that there is no hard and fast number that can be set. The courts should consider the two questions in conjunction with one another, alongside the context (e.g. the right impinged): how many people are likely to be affected and what is the likelihood/foreseeability that they will be affected? A provision may therefore be incompatible because it inevitably breaches a small number of persons’ rights or because it is likely to breach everyone’s rights – or something in between. Cases where there is a breach of one solitary person’s rights are usually more properly dealt with by a challenge to the application of the legislation in question. It may, of course, be seen without too much of a stretch that the breach is not as isolated as it first seems, but then the case would not be a true example of a solitary breach. The test advocated here would, however, allow for an incompatibility to be found even if there were a true case of a solitary breach of one person’s rights, if that breach was inevitable, in the same way as it would allow incompatibility to be found in the case of a more remote likelihood of a breach to a larger group of people. If there is an inevitable, unavoidable breach of even one person’s rights then the legislator is at fault, rather than the body applying the law and such blame should be allocated appropriately.
To take another example, Christian Institute v Lord Advocate Footnote 50 concerned a challenge to the Children and Young People (Scotland) Act 2014. The relevant provisions were designed to foster collaboration and the sharing of information between relevant persons and bodies (e.g. local authority, health authority and schools) in the interests of promoting the well-being of children. In particular, every child in Scotland was to be designated a “named person” with responsibility for their well-being, which would include sharing confidential information about the child and their parents, as well as providing advice to the parents about the child’s well-being. The data-sharing and data-retaining aspect of the legislation was challenged on the basis, inter alia, that it breached the Article 8 ECHR rights of parents and/or children and young people. The court acknowledged that the information sharing required by the legislation would clearly interfere with Article 8 ECHR. The court therefore moved on to consider whether that interference was justified. Immediately, the court held that the provisions were not “in accordance with the law” because they failed to comply with rule-of-law standards of accessibility and foreseeability. For example, the court was concerned about the “logical puzzle” of how the Children and Young People (Scotland) Act 2014 and the Data Protection Act 1998 interacted.Footnote 51 In terms of proportionality, the court found it necessary to “distinguish between the Act itself and its operation in individual cases”.Footnote 52 The relevant passage is worth quoting in full:
This court has explained that an ab ante challenge to the validity of legislation on the basis of a lack of proportionality faces a high hurdle: if a legislative provision is capable of being operated in a manner which is compatible with Convention rights in that it will not give rise to an unjustified interference with [A]rticle 8 rights in all or almost all cases, the legislation itself will not be incompatible with Convention rights: R. (Bibi) v Secretary of State for the Home Department [2015] UKSC 68; [2015] 1 W.L.R. 5055, paras. 2 and 60 per Lady Hale, para. 69 per Lord Hodge. The proportionality challenge in this case does not surmount that hurdle. Nonetheless, it can readily be foreseen that in practice the sharing and exchange of information between public authorities are likely to give rise to disproportionate interferences with [A]rticle 8 rights, unless the information holder carries out a scrupulous and informed assessment of proportionality.Footnote 53
One problem with this paragraph – taken from the judgment of the court (Lady Hale and Lords Reed and Hodge, with whom Lords Wilson and Hughes agreed) – is that it in fact misstates what Lady Hale actually said in R. (Bibi) v Secretary of State for the Home Department. It is true that a piece of legislation ought not to be deemed incompatible with Convention rights simply because it can be operated in a way that is incompatible – the quoted text makes it clear that in such cases the onus is on the public authority to be “scrupulous” in ensuring that the provision does not breach rights. If every piece of legislation were rendered incompatible just because it could be operated in a way that breached Convention rights, we would be left with no compatible legislation at all. Where, however, there are cases in which no matter how scrupulously the public authority applies the legislation, rights are likely to be, or will inevitably be, breached, the onus is not in fact on the public authority but on the legislator. The court then needs to consider how widespread the potential issue is and what the precise likelihood of breach is, in order to decide whether the legislation is incompatible. The “high hurdle” the court speaks of is in fact too high. If a public authority cannot operate legislation compatibly with Convention rights for a certain group of people (however small), then the legislation is incompatible and should be declared as such. That may eventually mean the legislation is recast more narrowly, or indeed that it is not amended at all – but it is the proper role of the court to suggest that reconsideration takes place.
In Christian Institute v Lord Advocate on the narrow proportionality question as to whether the rights infringement was disproportionate to any benefit to be gained, the court accepted that it would not give rise to disproportionate interference in all cases.Footnote 54 However, taking into account “individual case[s]” the court was concerned about the “very wide scope of the concept of ‘wellbeing’” and the fact that it was not clear that advice and services might be voluntary rather than compulsory.Footnote 55 In summary, the task for authorities was “daunting” in particular because of the overly “relaxed approach to disclosure” and the use of “very broad criteria”.Footnote 56 The court therefore ultimately concluded that the provisions “may in practice result in a disproportionate interference with the [A]rticle 8 rights of many children, young persons and their parents, through the sharing of private information”.Footnote 57 No firm conclusion needed to be reached because the provisions were beyond the competence of the Scottish Parliament on the basis of the earlier conclusion that they were not in accordance with law. But the case is a useful example of a situation where both the likelihood and extent of the potential incompatibility may be described as “medium” (i.e. something less than an inevitability and something less than “all or almost all”). It is submitted that this is enough for an incompatibility. As well as providing enough protection for citizens, legislation of the sort impugned in this case would be very difficult for administrators to navigate to ensure that they do not cause a rights breach.Footnote 58
Another good example can be found in Re McLaughlin, which concerned a challenge to the compatibility of the Social Security Contributions and Benefits (Northern Ireland) Act 1992 with Article 8 ECHR. Ms McLaughlin had cohabited with her partner for 23 years before his death but, because they were not married, she was not entitled to the bereavement payment and widowed parent’s allowance under the scheme that she would have been entitled to had they been married. The court held that the discrimination in the enjoyment of the right conferred by Article 8 ECHR justified a declaration of incompatibility. It was, as the court noted, “not a difficult conclusion to reach on the facts of this case”, given, for example, the length of the cohabitation and the fact that they shared four children together.Footnote 59 However, the court went on to note that “[i]t does not follow that the operation of the exclusion of all unmarried couples will always be incompatible”.Footnote 60 The example is given of a child for whom the deceased was not responsible, having only been recently introduced into the household, for example, a hypothetical scenario in which Ms McLaughlin recently obtained custody of a hypothetical child who had, until recently, resided with their father (who was not the deceased). Lady Hale (with whom Lord Mance, Lord Kerr and Lady Black agree) goes on to say that “the test is not that the legislation must operate incompatibly in all or even nearly all cases. It is enough that it will inevitably operate incompatibly in a legally significant number of cases”.Footnote 61 Curiously, paragraph 88 of Christian Institute v Lord Advocate is cited as authority for this proposition when the test set out in that paragraph was the “high hurdle” of “an unjustified interference […] in all or almost all cases”. However, as mentioned above, that test cites paragraphs of R. (Bibi) v Secretary of State for the Home Department that say completely different things: Lord Hodge invokes the “all or nearly all” test,Footnote 62 whereas Lady Hale suggests a “significant number of sponsors” test.Footnote 63 In any case, Re McLaughlin is a good case for demonstrating the fact that pondering “all or almost all” cases is neither helpful nor appropriate – it may always be possible to imagine a person whose rights would not be infringed by the legislation, but that does not mean that a declaration ought not to be made. Quite the contrary – it shows that the legislation might need to be tweaked, even if it is the case that it does not need to be completely rewritten.
B. The Abortion Services Case
Re Abortion Services arose in the devolution context, although as is argued in this article, the test used to assess incompatibility should be common to sections 4 and 19 HRA too, as well as determining the validity of other legislation lacking sovereign status. The case concerned the Abortion Services (Safe Access Zones) (Northern Ireland) Bill (now Abortion Services (Safe Access Zones) Act (Northern Ireland) 2023 (“Abortion Services Act”)), which was passed by the Northern Ireland Assembly. The Abortion Services Act prohibits anti-abortion protest in “safe access zones”Footnote 64 around places that provide abortion services (“protected premises”)Footnote 65 with a view to protecting patients attending such premises, their companions and the staff who work there (“protected persons”).Footnote 66 The crime created by section 5(2) of the Act is broad.Footnote 67 D (defined by s. 5(1) as “a person who is not a protected person”) may exhibit intention or may be reckless as to whether their behaviour has one of the following effects: influencing a protected person directly or indirectly, preventing or impeding their access, or causing them harassment, alarm or distress. The provision therefore covers a wide range of behaviour, from that which would be not be protected by the relevant Convention RightsFootnote 68 in any event (e.g. spitting at people or chasing them)Footnote 69 to more gentle protest activity (e.g. which neither actually causes, nor intends to cause, distress, but which nevertheless does exert influence on a protected person). Of particular note was the part of clause (now section) 5(2)(a) of the Bill (now Act) that makes it an offence “to do an act in a safe access zone with the intent of, or reckless as to whether it has the effect of – (a) influencing a protected person, whether directly or indirectly”. The question for the court was whether this provision disproportionately interfered with the Article 9, 10 and 11 Convention rights of the anti-abortion protestors, in particular because there is no defence or excuse to the offence it creates. Thus, it was argued, there was no possibility of conducting a proportionality assessment of the interference with any one protestor’s rights in any one individual case. The court ultimately rejected this line of argument on the basis that “questions of proportionality, particularly when they concern the compatibility of a rule or policy with Convention rights, are often decided as a matter of general principle, rather than on an evaluation of the circumstances of each individual case” and furthermore “whether an interference with a Convention right is proportionate is not an exercise in fact-finding”.Footnote 70 It was ultimately held that the Bill was not incompatible with the Convention rights and was thus within the Assembly’s competence. Incompatibility must of course also be assessed in other contexts, namely in interpreting legislation under section 3 HRA and assessing the actions of a public authority under section 6 HRA.Footnote 71 In both of these instances, the court must proceed on the individual facts before it, therefore not encountering the same problem of how systematic the breach of rights must be in order for incompatibility to be established. Say, hypothetically, a different challenge is later made to section 5(2)(a) Abortion Services Act. Let us imagine a case is brought by P, a protestor charged and convicted with the section 5(2)(a) offence for protesting in a safe zone in the prohibited manner. P may challenge the compatibility of the legislation itself but, assuming the hypothetical court followed the view of the court in Re Abortion Services, that challenge would be unsuccessful. In the alternative, there is still the possibility that P’s specific treatment was incompatible with their Convention rights. This is because the actions of public authorities may be incompatible in an assessment under section 6 HRA. In such cases, the court does not need to trouble itself with how systematic the breach needs to be – they are purely looking at the case of this particular individual. For example, perhaps P was mistakenly thought to be an anti-abortion protestor when in fact they were taking part in an industrial action picket line outside their place of work. Or perhaps they were protesting against abortion, but 300 metres from the protected premises and thus outside the safe zone. Or perhaps P is an anti-abortion protestor within the safe zone, but argues that in their particular case they believe their individual Convention rights have been breached by the application of this legislation, notwithstanding the fact that the legislation is not incompatible according to the two-part test advocated here. To be clear, it is not that the meaning of “incompatibility” differs; it is just that it is only when reviewing Acts (as in legislation) rather than acts (as in actions) that the court might need to think beyond the instant case.
The Northern Ireland Act 1998 provides that an Act of the Assembly is “not law” if it is outside the competence of the AssemblyFootnote 72 and, further, that one reason for a provision being outside competence will be if it is incompatible with any of the Convention rights.Footnote 73 The court accepted the Lord Advocate’s argument (as intervenor, as similar legislation was being considered for Scotland)Footnote 74 that to be incompatible and thus beyond competence legislation must “always, or almost always” interfere with a Convention right.Footnote 75 Similarly, the Northern Ireland Human Rights Commission (“NIHRC”) argued that, as Convention rights were not breached by the bill in “all or nearly all cases”, the bill was valid.Footnote 76 With validity (and therefore devolved autonomy) at stake, it is vitally important to have a clear definition of “incompatibility”.
Re Abortion Services does not provide the necessary clarity on the meaning of “incompatibility” because it demonstrates a difference of opinion as to the prevailing authority on the relevant test. It was argued by the Attorney General for Northern Ireland that the “always, or almost always” test in Christian Institute v Lord Advocate had been superseded by the “less demanding”Footnote 77 “legally significant number of cases” test in Re McLaughlin.Footnote 78 In other words, that incompatibility need only be found in a legally significant number of cases, rather than all or almost all cases. Tellingly, the court in Re Abortion Services noted that “[n]either party presented any analysis of the authorities, or any arguments as to why one approach or the other ought in principle to be preferred”.Footnote 79 The analysis is particularly defective because of its focus on the number of potential affected parties, without properly considering the likelihood of incompatibility resulting. This much is evident when Lord Reed states that “[t]he difference” between the two tests “is between Lady Hale’s words, [in Re McLaughlin] ‘in a legally significant number of cases’ and the earlier words, ‘in all or almost all cases’” without any commentary on the propriety or import of the “inevitability” part of the test found in Re McLaughlin or any potential alternatives.Footnote 80
Working through the authorities and citing his own judgment together with Lady Hale and Lord Hodge (with whom Lord Wilson and Lord Hughes agreed) in Christian Institute v Lord Advocate, Lord Reed in Re Abortion Services notes that a “high hurdle” is required for a challenge to the validity of devolved legislation,Footnote 81 thus the correct test is the “all or almost all” test.Footnote 82 Later in the judgment, Lord Reed accuses Lady Hale (“[w]ith respect”, of course)Footnote 83 of “not stat[ing] the test accurately”Footnote 84 in Re McLaughlin when she said “a legally significant number of cases” would suffice for incompatibility to be established.Footnote 85 It is argued here, however, that (with respect) the law is not as settled as Lord Reed suggests and Lady Hale’s view in McLaughlin was not a case of her misspeaking. This revelation does not establish that something akin to Lady Hale’s test is preferable – that will follow. But what it does prove is that the idea that Lady Hale “depart[ed] from” any settled test in Re McLaughlin is erroneous.Footnote 86 If a misstatement of the law happened, it happened in paragraph 88 of Christian Institute v Lord Advocate, which gave a false impression of a unified judicial mindset that, in fact, did not exist.
It is submitted that, although the outcome of the case was correct, the better argument was an alternative submission by the NIHRC that the legislation’s interference with the anti-abortion protestors’ rights was “inherently proportionate”.Footnote 87 First of all, it is relevant to consider the limited geographical area of the safe access zones (100 metres around the entrances and exits to protected premises or up to a further 150 metres upon application to the Department of Health).Footnote 88 The evidence heard in the case also supports the idea that it was not possible to ban certain types of anti-abortion protest while allowing some others. Although certain anti-abortion protestors were guilty of especially egregious conduct (e.g. spreading misinformation that abortion increases the risk of breast cancer or showing extremely graphic and upsetting pictures), even protests that took the form of silent prayer vigils encouraged various behaviours potentially detrimental to protected persons’ health (such as seeking abortions at later gestational stages, procuring medication online, taking two doses of medicine simultaneously so as to avoid a repeat visit to the clinic or even attempting suicide).Footnote 89 Furthermore, it was clear that the protests were occurring “on a daily basis” such that it would not be possible for a protected person to seek to avoid a particular day or time.Footnote 90 It was therefore fair to say that there was no less intrusive way of dealing with the protestors and that only their “complete removal” would suffice.Footnote 91
In Re Abortion Services, the court cites Strasbourg authority on the wide (but not unlimited)Footnote 92 margin of appreciation afforded to location, time and manner of protests.Footnote 93 There are obvious counterarguments to that line of jurisprudence.Footnote 94 It seems fair to say that restrictions as to the place of protest can significantly impede a protestor’s clout. It is true that the protestors may still protest their message, but the curtailment of where they can do so is a significant impediment to their effect. Protestors may choose a protest location because it is meaningful to the cause or because it will generate a lot of attention to their message, or both. However, the particular context of anti-abortion protests and particularly the countervailing rights of protected people should be noted. In this case, those countervailing rights include confidentialityFootnote 95 and the very grave potential effect of the protests, including, at their extreme, meaning that protected persons were not able to avail themselves of vital health services.Footnote 96 In addition, the protests were severely affecting those employed by service providers too, meaning that the Article 8 rights of both patients and staff were engaged. Essentially the inherent controversy about the subject matter, the strength of feeling of the protestors and the vulnerabilities of those seeking abortion services meant that a restriction on protestors’ rights was not disproportionate. For protected persons seeking medical treatment, confidentiality and privacy are always relevant. Furthermore, some protected persons will be especially vulnerable, for example because of being sexual abuse survivors or minors.Footnote 97 The rights of all protected persons make it impossible to allow anti-abortion protests, of any type of conduct, within the vicinity of abortion service providers. There is simply no more proportionate way around it.
The case also alerts us to the fact that the proportionality of the restriction was given full consideration in the debates on the bill. In terms of the specific issue about the lack of excuse/defence to the crime introduced by the Abortion Services Act, this was considered fully during the debates on the bill but rejected on the basis that it would hollow out the protections offered.Footnote 98 In terms of the specific extent of the safe zones (in other words, the precise number of metres from the entrances and exits to abortion service providers that protest would be banned) this was also debated “extensively”.Footnote 99 Finally, the maximum penalty is non-custodial, currently a maximum £500 fine.Footnote 100 Perhaps most convincing is Lord Reed’s rebuttal of the argument that the protests contribute to public debate. As he so rightly says, protests in the vicinity of protected premises serve only to attempt to influence “the personal decisions of individual women” as opposed to aiming to influence law reform.Footnote 101 The latter would be far more effectively carried out by the other means Lord Reed lists such as protesting in other locations (e.g. in a city centre), writing books or articles, appearing on TV or posting on social media.Footnote 102 None of these methods are restricted or forbidden by the legislation. It is further true that protest rights do not give a right to a “captive audience” and indeed that, in this case, protected persons had no ability to avoid the message being propounded.Footnote 103
IV. Developing a Clear and Consistent Two-Part Test
Unlike R. (Bibi) v Secretary of State for Home Department, Christian Institute v Lord Advocate and Re Abortion Services, Re McLaughlin concerned primary legislation passed by Westminster.Footnote 104 What is not clear from the line of judgments, especially given the misunderstandings already noted, is whether the court intends for different tests to apply depending on whether they are determining validity or declaring incompatibility. Lord Reed’s mention of the need for a “high hurdle” in relation to the former might suggest that there could be such an intention,Footnote 105 but it is not crystal clear that that is what he intended. It is argued here that the test applied when considering “incompatibility” ought not to differ depending on whether one is dealing with determining the validity of devolved legislation or deciding whether to make a declaration under the HRA (or indeed making a statement of compatibility under section 19 HRA or assessing the compatibility of delegated legislation). To have different tests for incompatibility jeopardises a healthy dialogue between the various branches and layers of the state (political/legal as well as UK/devolved). There is no distinction made in the various pieces of legislationFootnote 106 and evidently the courts have struggled to bring clarity to the area. Given that the political actors assess the risk of the legislation being found incompatible by the courts, it makes sense for there to be a shared understanding. It also makes sense for the minister declaring their belief that legislation is compatible with Convention rights to share the understanding of “compatibility” with the parliamentarians who go on to scrutinise it. Without a shared understanding, it is hard to see how there can be meaningful assessments of the compatibility of any provision.
On the one hand the stakes appear much higher for devolved legislation (or delegated legislation), as it may be struck down as opposed to merely declared incompatible. A declaration of incompatibility under section 4 HRA, of course, “does not affect the validity, continuing operation or enforcement” of primary legislation passed by Westminster.Footnote 107 The consequences, therefore, appear very different for devolved provisions held to be “not law” for want of compatibility and for legislation merely declared incompatible. But, on the other hand, Lord Reed himself in Re United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill described section 4 HRA as a qualification of Westminster sovereignty.Footnote 108 Space does not permit a full exploration of this curious statement, but it is assumed that he was referring to the 100 per cent “success rate” of declarations of incompatibilityFootnote 109 and the evident political limitation that declarations effectively place on Parliament. It is not accepted here that section 4 HRA constitutes a qualification of sovereignty per se. But section 4’s practical impact does demonstrate that the stakes are ultimately not as drastically different as it first seems between primary legislation declared to be incompatible and legislation lacking sovereign status deemed to be “not law”.
In relation to section 19 HRA, although the vetting of the legislation undertaken by a minister is very different in nature to a judicial pronouncement, a shared understanding of compatibility would be very useful. When a minister pronounces that, in their view, the legislation is compatible with Convention rights (or indeed that such compatibility cannot be pronounced) that view is inextricably linked to what they think the judiciary might ultimately determine. A shared test is in line with the initial plans for how sections 4 and 9 HRA would operate. Initial hopes were that the HRA “would ensure that the Convention rights are taken more fully into account in the development of new policies and of legislation”.Footnote 110 It has been pointed out that those hopes were “ambitious” and arguably at odds with the HRA being pitched as more “modest” than a conventional Bill of Rights.Footnote 111 But it should follow from the HRA’s focus on “regular parliamentary engagement” with compatibility that it affords an easier opportunity for dialogue than most Bills of Rights.Footnote 112 As Murray Hunt describes it, the HRA gives the potential for a “more coherent” system of rights protection than the binary alternatives of legislative or judicial supremacy.Footnote 113 Alison Young has described section 19 HRA as being “designed to facilitate democratic discussion”Footnote 114 and she has also observed that much of the HRA allows for “inter-institutional interactions”.Footnote 115 In short, if the proposals here are thought to be overly “ambitious and optimistic”Footnote 116 then they are in line with the aims of the HRA, not contrary to them.
But, as yet, none of this militates in favour of any particular threshold, as opposed to favouring clarity and consistency in whichever threshold is chosen. It is argued here that the “always, or almost always” threshold preferred in Re Abortion Services should be rejected. This is for three main reasons.
First, to set the threshold lower is not to advocate for judicial supremacy – quite the opposite. It is the “always, or almost always” test that requires the court to try to think of all the different ways in which the legislation may be compatible or not, rather than just focussing on the people it most obviously affects. “Legally significant number of cases” might mean that the legislation breaches (or is likely to breach) a certain group’s – or subset of that group’s – rights, for example, certain cohabitees as compared with spouses, as in Re McLaughlin.Footnote 117 Crucially, the test should rely equally on the likelihood or foreseeability of that breach taking place. A provision that will inevitably breach a certain group (or person)’s rights is as much candidate for a declaration of incompatibility as a provision that will likely (but not inevitably) breach the rights of a larger group of people. The particular right at stake ought also to be taken into account in this assessment. For example, a provision that breaches an absolute right, such as Article 3 ECHR, the prohibition of torture, might be incompatible even if it affects a smaller number of people and/or has a more remote risk of harm. The judicial role envisaged here is a constitutionally appropriate one and accords with the intention behind section 4 HRA to review legislation rather than specific actions.Footnote 118 It is true that to adopt the approach in this article would likely lead to more declarations of incompatibility being made (and more legislation being found incompatible in the other contexts discussed in this article), although it is impossible to give a precise number. It is also true that an increase in the use of section 4 HRA could see it attracting more controversy in the way that the courts’ more robust use of section 3 HRA has. That said, there is no guarantee that such controversy would definitely result, given that section 4 HRA may simply be inherently “politically less contentious”.Footnote 119 Although an average of two declarations are made per year, this obviously varies from year to year; for example, 11 declarations of incompatibility were made in the year 2023–24.Footnote 120 Although these 11 declarations were made in only five separate cases, years with a higher-than-normal number of declarations do not seem to attract particular controversy. Defining “incompatibility” in the way advocated here could take some of the heat out of declarations of incompatibility. Instead of being seen as a condemnation of legislation that “always, or almost always” breaches rights, it can be seen as a signal that the legislation merely needs adjusting, to remove a more limited incompatibility. Of course, all of this may be “wishful thinking”Footnote 121 and would require a huge change of mindset. But it should logically follow that a declaration of incompatibility should not be viewed as dramatically if the legislation is not being condemned as “always or almost always” breaching rights.
Second, to set the threshold at “always, or almost always” reduces human rights protections. As Christopher Rowe has pointed out, the “always, or almost always” test “could immunise almost any measure from Convention challenge by the domestic courts, so long as it was passed in a legislative form, rendering Convention rights nugatory”.Footnote 122 This led Rowe to go so far as to describe the current test as “absurd”Footnote 123 – a statement that is endorsed here. When faced with legislation that breaches, or is likely to breach, the rights of any person or people, the courts should aim to protect them. The result of the “always, or almost always” test is that the court may condone legislation that it knows or suspects affects someone’s rights.Footnote 124 Even if such groups or persons might ultimately find their rights protected by challenges to the individual application of the law to them, the fault and responsibility should be identified as lying with the legislator, rather than those tasked with applying the law and further litigation should not be necessitated.
Third and finally, and segueing on from the previous point, to fail to make an authoritative ruling on provisions that breach the rights of a legally significant group of people could potentially put decision makers in a difficult position, particularly where the affected group is less readily defined than, for example, the cohabitees in Re McLaughlin. This problem is demonstrated well by the case of Beghal v D.P.P. where the court declined to make a declaration of incompatibility in relation to Schedule 7 of the Terrorism Act 2000 because, inter alia, Mrs. Beghal’s detention under the legislation had only been for 30 minutes and her information was not retained.Footnote 125 The court, however, expressed some concern about detentions of the maximum duration allowed under the legislation (six hours) or where information was retained (also allowed for under the legislation).Footnote 126 Without a declaration and a probable subsequent legislative amendment, essentially decision makers are left to rewrite the legislation themselves to decide what the maximum detention period “really” is – despite what the statute says. This is clearly undesirable for a Holy Trinity of parliamentary-sovereignty, rule-of-law and separation-of-powers reasons. It could also cause a plethora of subsequent individual challenges to legislation, with consequent continued human rights breaches in the interim as well as resource implications in bringing unnecessary cases.Footnote 127 Even worse, not all cases concern public authorities at all, potentially leaving those claimants without an alternative remedy.Footnote 128
It is true that declarations of incompatibility can equally be said to put the political branches in a difficult position, as can arguably be evidenced by the majority’s refusal to make a declaration in the notorious assisted suicide case of R. (on the Application of Nicklinson) v Ministry of Justice.Footnote 129 The case concerned the compatibility of section 2 of the Suicide Act 1961 (which makes it a crime to encourage or assist suicide) with the right to respect for a private life under Article 8 of the ECHR in the context of those who seek to end their lives due to terminal or chronic illness, but are unable to do so without assistance. This complicated and sensitive case resulted in nine separate judgments, with seven Justices ultimately refusing to make a declaration of incompatibility. Four Justices (Lords Clarke, Sumption, Reed and Hughes) found that the subject matter was better left to Parliament and not suited to a declaration of incompatibility. Of the remaining five Justices who were more willing to entertain a compatibility assessment, only a minority (Lady Hale and Lord Kerr) would have made a declaration. The remainder (Lords Neuberger, Mance and Wilson) felt it would be “institutionally inappropriate at this juncture” to make a declaration, including for the reason that Parliament was already considering changes to the law (which, in the event, were not made at that time).Footnote 130 Interestingly, two Justices noted that a declaration would be “legitimate” even though the legislation would only “sometimes” breach Convention rights (i.e. the potential breach concerned only those who had a firm desire to end their lives, but were unable to do so without help, despite the potential merits of a crime of assisting suicide more generally).Footnote 131
It could be suggested that a shift of mindset is required in relation to declarations of incompatibility and that they are, in a way, a victim of their own success. Declarations need not be viewed as an “unexploded bomb” that must be urgently defused.Footnote 132 For example, a (hypothetical) declaration of incompatibility in R. (Nicklinson) v Ministry of Justice that was met with a reasoned response that the Government had decided to let the incompatibility continue would be, it is argued, a satisfactory outcome. As Lady Hale pointed out in her dissenting judgment in that case, Parliament is free to “do nothing” in response to a declaration of incompatibility.Footnote 133 At the time of writing the Terminally Ill Adults (End of Life) Bill is currently going through Parliament, over a decade on from R. (Nicklinson) v Ministry of Justice. Had a declaration been made in R. (Nicklinson) v Ministry of Justice and had nothing happened until now – and even if the current bill does not pass – the argument here is that that is acceptable. Some declarations might take a long time to be remedied and some might not be remedied at all. They should never be ignored but, it is argued, that in some contexts a “thanks but no thanks” response to a declaration of incompatibility would be better than the judiciary arguably working backwards from the assumption that the law would have to be changed.Footnote 134 This would be of a piece with a culture of justification, which does not necessarily require every incompatibility to be remedied, but rather “rational explanations for why a particular action or decision has been taken, or why there has been an omission to act”.Footnote 135 It is true that Strasbourg might eventually force the Government into action from an international law perspective, but it should not be assumed that such a ruling would follow as a matter of course and, even if it did, the Council of Europe has accepted wide margins of appreciation and very minimal – and very slow – legislative change.Footnote 136 Even countries with “strong form” judicial review might fail to honour a judicial ruling on legislative competence, for example, due to disagreement or misinterpretation.Footnote 137 In sum, the test for “incompatibility” advanced here could encourage the proper use of declarations of incompatibility as a discretionary remedy designed to draw attention to potential (and potentially modest) incompatibilities, not as something to be feared by either the judges making them or the political actors receiving them. Even the judges who were least willing to entertain making the declaration in R. (Nicklinson) v Ministry of Justice might have reached a different conclusion had they taken the approach advocated here. The benefit would then be the court exercising its proper role in “formally record[ing]” the incompatibility,Footnote 138 whatever the ultimate outcome.
It seems, then, that the correct approach is still very much lacking in the case law. It is suggested here that the court needs to think more carefully about the two parts of the test and how they interact. First, how extensive is the potential breach? The all or almost/nearly all test should be rejected in favour of a lower threshold. In practice this is likely to be either: (1) a readily identifiable group, or part thereof, such as, in Re McLaughlin, cohabitees or (2) as in Beghal v D.P.P., a group but in a less readily identifiable sense, such as those treated according to the upper limits of the legislation. Separately but relatedly, the court needs to consider the likelihood or foreseeability of the rights violation. This is the crux of the issue that can allow the court to defend human rights violations without fearing being overly-interventionist – the incompatibility is not hypothetical where it is inevitable and the legislation need not be radically altered if the incompatibility is not widespread. In cases such as Beghal v D.P.P. where the legislation can be operated in both compatible and incompatible ways, the assessment is less clear-cut, but the argument here is that the court should make the declaration in the interests of properly allocating responsibility for the incompatibility with the legislator where the legislation can plainly be seen on its face to have foreseeably incompatible applications. The alternative is to allow potentially over-inclusive legislation to remain knowingly on the statute book, with consequent difficult decisions for public bodies to make and potential individual applications to the court. By contrast, in cases such as Re Abortion Services, application of the test indicates a finding that the legislation itself is compatible, notwithstanding the fact that then there may be incompatible applications of the law in particular, unforeseeable, individual cases.
V. Who Is a “Victim”?
Considering more carefully what it means for a provision to be incompatible should also provide some assistance in establishing who has standing to challenge such provisions. With regard to challenging the validity of Acts of the devolved legislatures, the devolution statutes provide that a person will not have the ability to raise an action unless they would be a victim for the purposes of Article 34 of the ECHR, except for actions brought by the law officers.Footnote 139 Whether the action is brought by a victim or a law officer (and regardless of the timing of the challenge) a lower threshold for incompatibility than the courts have tended to prefer is warranted. A higher threshold means that devolved legislation that breaches, or is likely to breach, the rights of someone – just not “all or almost all” people – would knowingly be allowed onto, or be left on, the statute book. Although the stakes are high for the devolved legislatures, they are even higher for individual citizens.
The courts have historically shown confusion as to the standing requirements of section 4 HRA, and in particular whether a case may only be brought by a “victim” of a rights breach.Footnote 140 But reading the test for victimhood under section 7(1) HRA makes it patently clear that it only applies to cases brought under section 6 HRA, to challenge the actions of a public authority as opposed to challenging the legislation.Footnote 141 As the regime under sections 3 and 4 HRA is entirely separate from the regime under sections 6 through to 9 HRA, there is no prerequisite to find victim status before a declaration of incompatibility can be made. Indeed, as I have argued elsewhere, it is not in fact possible to have a “victim” under section 4 HRA, strictly speaking, as there is no “unlawful act” under legislation that is incompatible but, by virtue of the terms of the HRA, still valid.Footnote 142
Nevertheless, section 4 HRA is not an actio popularis.Footnote 143 A person cannot raise an action under section 4 HRA merely as a concerned citizen, but can only do so as someone who believes that their rights are, or would be, breached by the legislation. Although it may not happen frequently, the courts should feel at liberty to make the declaration even if they do not think that the person is a victim or are unsure whether they would be.Footnote 144 This is for at least two reasons. First of all, it is how the Act was envisaged to work by its drafters. In previous work, I have argued that during the debates on the bill, it was made clear (inter alia, by way of rejected amendments) that a declaration of incompatibility should be able to be made even where it was not necessary to determine the instant case.Footnote 145 Second, the courts should not be pondering the state of victimhood too far as it involves too much conjecture over how the breach may be remedied, if at all. Given that declarations of incompatibility have no legal effect (other than triggering the possibility of using section 10 HRA to remedy the incompatibility) and that provisions remain in force unless or until they are remedied, as noted above, technically speaking there is no unlawful behaviour on the part of the public authority and therefore strictly speaking no victim (because the legislation is perfectly lawful). Someone who the court thinks may be a “victim” (in inverted commas) may end up not being a victim in the end because the replacement legislation is framed in a way that does not alter their position.
Guidance can be found from Strasbourg’s interpretation of victimhood. For example, in the secret surveillance case of Klass v Germany,Footnote 146 the German Government considered the claimants to be “seeking a general and abstract review of the contested legislation”.Footnote 147 The claimants, by contrast, argued that victimhood was established because every person “who could be presumed to have contact with people involved in subversive activity” risked being subject to secret surveillance and, even if such surveillance did not actually take place, the risk alone restricted their free communication.Footnote 148 Furthermore, because the surveillance was by its very nature secret, it would be difficult for a concrete instance to be proven.Footnote 149 The court agreed with the claimants’ reasoning, albeit it ultimately held that there was no breach because the legislation was “necessary in a democratic society in the interests of national security” and “for the prevention of disorder or crime” in accordance with Article 8(2) ECHR. The case is a useful one for illustrating that the notion of victimhood is not as binary as it might first seem. It is not that every single German person was at risk of surveillance, but a significant portion of them were unavoidably impacted. And even if the risk did not eventuate, they were modifying their behaviour just in case they were being surveilled. In other words, victimhood may be established where a provision does not affect “all, or almost all” people and/or where the impact is not inevitable.Footnote 150
Another good example is the prisoner voting saga. Under the new laws changed as a result of their cases, neither William SmithFootnote 151 nor John HirstFootnote 152 would have been allowed to vote (had those new laws been in place at the time when they were prisoners who wished to register to vote). In Smith’s case, section 4 HRA did not require him to be a victim anyway. In Hirst’s case, Article 34 ECHR did require him to be a victim, but the case elucidates what “victimhood” means. The majority of the Grand Chamber in Hirst v United Kingdom (No. 2) did not accept the Government’s argument that Mr. Hirst’s status as a “lifer” meant that the court was effectively reviewing the legislation in abstracto, because he was extremely likely to remain excluded from voting even if the law was changed for certain prisoners.Footnote 153 Instead the majority was satisfied that the question for the court was whether the legislation as it stood “directly and immediately” affected him and not how potential future changes might affect him.Footnote 154 A case involving a blanket ban is perhaps a particularly “easy” example, but in any event it is worth remembering that the courts are not the ones to decide whether and how to remedy the incompatibility. A similar argument could be made with regard to incompatible and therefore invalid devolved legislation. As the Chamber in Hirst v United Kingdom noted, it was not for them to “speculate” as to whether Mr. Hirst might still have been deprived of his right to vote had the ban not been universal.Footnote 155 Similarly, it is argued that the court in Smith v Scott were quite right to make the declaration notwithstanding Hirst v United Kingdom and not to speculate about whether a declaration was needed to “goad the government into action”.Footnote 156
VI. Conclusion
Since the HRA came into force in October 2000, much academic commentary has been written about provisions that have been (or should have been) deemed incompatible with Convention rights. In the case of the devolved legislatures, such incompatibility means the provisions are “not law”. In the case of delegated legislators, their provisions may be struck down by a court, or declared incompatible. In the case of a declaration of incompatibility, it will seemingly inevitably (although sometimes slowly and modestly)Footnote 157 result in a provision’s reform. In recent years human rights scholars have increasingly paid attention to ministerial statements of compatibility, particularly because of the rare occasions when they have not been given. Yet, perhaps strangely, the same body of scholars, the courts and the political actors involved have neglected the prior question of how to test for incompatibility. This neglect threatens relations with the devolved nations given the extreme consequences of incompatibility for their legislation. Likewise, assessing the compatibility of delegated legislation involves a delicate calibration of judicial and executive power, including with regard to the choice of remedy (quashing or declaration).Footnote 158 In the context of section 4 HRA, there are good reasons why judges have used declarations of incompatibility sparingly and there are equally good reasons why politicians have not ignored them. It may be queried, however, whether declarations of incompatibility have, in a sense, become a victim of their own success. In an ideal world, a declaration would not be seen as a “last resort”Footnote 159 or as a judicial attack on the political branches. Instead, it would be seen as a healthy dialogue between the courts, the Government and the legislatures. That is not to say that exponentially more declarations ought to be made, nor that they ought to be ignored. But it ought to be possible to view a declaration as a signal to rethink a provision, rather than a complete condemnation.
In sum, in terms of sections 4 and 19 HRA, as well as in relation to the competence of devolved legislators and the legality of delegated legislation, reasonable minds may differ as to both the substantive policy to be pursued, as well as to the proper role of the various branches, and layers, of the state. But the dialogue would surely be healthier if we all at least agree on what “incompatibility” means.
Open access
