THE Supreme Court was confronted in In re United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill [2021] UKSC 42, [2021] 1 W.L.R. 5106 with some fundamental constitutional questions. It answered them in sometimes surprising ways, casting unexpected light on key constitutional legislation and on the doctrine of parliamentary sovereignty. It was argued before the court that the Scottish Parliament had exceeded its legislative authority when it sought to give effect in Scottish law to two treaties that had not been incorporated in UK law. This note focuses on the Court's treatment of legislation – the United Nations Convention on the Rights of the Child (Incorporation) (Scotland) Bill – that sought to give domestic effect to one of those treaties, often in ways that bore striking resemblance to techniques employed by the Human Rights Act 1998 (“HRA”) to give domestic effect to the European Convention on Human Rights.

Section 19 of the Bill, which resembled section 3 of the HRA, directed that legislation, including UK legislation within Scottish competence, be read and given effect compatibly with relevant provisions of the United Nations Convention on the Rights of the Child (“UNCRC”) so far as possible; section 21, borrowing from section 4 of the HRA, authorised Scottish courts to issue “incompatibility declarators” in respect of legislation enacted after the Bill's entry into force; and (going further than the HRA), section 20 allowed courts to strike down legislation enacted prior to the Bill's entry into force. Two key factors underpinned the Court's conclusion that sections 19–21 were outwith the Scottish Parliament's competence. First, the Court adopted a broad reading of the prohibition, set out in the Scotland Act 1998, on the modification by Holyrood of certain “protected” legislative provisions. Second, the Court took a very wide view of one such provision: namely, section 28(7) of the Scotland Act. This provision states that section 28, which invests Holyrood with legislative authority, “does not affect the power of the Parliament of the United Kingdom to make laws for Scotland”. Section 28(7) had generally been understood simply as affirming that the devolution scheme set out in the Scotland Act did not displace the doctrine of parliamentary sovereignty. However, the Court read section 28(7) as vouchsafing Westminster's “unqualified” power to legislate for Scotland – a concept that, as we will see, imposes restrictions on Holyrood that transcend those flowing from parliamentary sovereignty itself.

Perhaps the Court's least surprising conclusion was that section 20, authorising “strike-down declarators” in respect of legislation enacted prior to the Bill's entry into force, was outwith devolved competence. But even this did not straightforwardly follow. The Scottish Parliament can repeal Westminster legislation insofar as it applies to Scotland and relates to matters within devolved competence. Why could Holyrood not authorise courts, within corresponding parameters, to strike down UK legislation? The Court said that this would make the continuation in force in Scotland of relevant UK legislation conditional on the courts’ willingness to find it compatible with the UNCRC, which, in turn, would qualify Westminster's legislative authority in breach of section 28(7) of the Scotland Act. Of course, the continuation in force in Scotland of relevant UK legislation is already conditional on Holyrood not repealing it, but, on the Supreme Court's analysis, that form of conditionality is acceptable whereas the arrangement envisioned by this Bill was not. The decisive factor, it seems, is that Westminster has itself directly authorised the Scottish Parliament to repeal UK legislation but has not similarly authorised the Scottish courts to strike down such legislation. By the same token, the court took the contestable (and arguably paradoxical) view that the European Communities Act 1972 limited parliamentary sovereignty but that this was acceptable on the ground that “Parliament itself effected that qualification of its sovereignty”.

More remarkable were the Court's conclusions about sections 19 and 21, not least because of its associated pronouncements upon corresponding provisions of the HRA. Section 19 was outside competence because (in the Court's view) provisions such as section 3 of the HRA and section 19 of the Bill require courts “to modify the meaning and effect of legislation enacted by Parliament” thereby “plainly impos[ing] a qualification upon its legislative power”. This simultaneously contradicts established understandings of section 3 of the HRA, which is not generally understood to limit Westminster's legislative power, and exploits the distinction drawn in the judgment between the sovereignty of Parliament (which apparently endures despite such provisions) and the more demanding prohibition (imposed by section 28(7) of the Scotland Act) upon Scottish legislation that disturbs the UK Parliament's “unqualified” power. If anything, the Court's analysis of section 21 of the Bill is even more arresting: it was held to be outwith competence because section 4 of the HRA, on which section 21 of the Bill is modelled, was said to qualify sovereignty itself thanks, inter alia, to the political pressure unleashed by declarations of incompatibility. This remarkable analysis at once departs from the well-founded view that such declaratory powers are compatible with sovereignty and inverts orthodoxy by proceeding again on the basis that Parliament “can qualify its own sovereignty”.

The Scotland Act also contains, in section 101, a statutory requirement that any provision of (inter alia) a Bill of the Scottish Parliament is to be read “as narrowly as is required for it to be within competence, if such a reading is possible”. The Court was asked how far “recourse to [this] interpretative obligation” could cure any competence issues arising in this case. The Court, in answering this question, gave guidance on both the burden this obligation is capable of bearing, and what factors determine its limits. A parallel was drawn between section 101 and section 3 of the HRA. Despite both provisions relying on the notion of what is interpretatively “possible”, the Court considered that curial treatment of section 101 had not given it (or its analogues in other devolution legislation) “as far-reaching an effect” as section 3 of the HRA. Unlike the latter, the Court (citing with affirmation Lord Neuberger's judgment in Attorney General v National Assembly for Wales Commission and others [2012] UKSC 53, [2013] 1 A.C. 792) held that section 101 of the Scotland Act cannot be invoked if it is “inconsistent with the plain words of [the provision in question]”. Although the Court confirmed that section 101 is not “confined to the resolution of linguistic ambiguities”, it also took the view – applying Lord Hope's judgment in DS v HM Advocate [2007] UKPC D1, 2007 SC (P.C.) 1 – that section 101 does not authorise courts to have recourse to the wide range of interpretative techniques that section 3 of the HRA facilitates. The two interpretative obligations are thus “fundamentally different”.

The Court articulated two further principles which, it held, sculpt the boundaries of the section 101 obligation. First, section 101(2) “cannot have been intended to be construed as having the effect of rendering nugatory the pre-enactment safeguards provided by the Scotland Act”. In the Court's view, the existence of pre-enactment safeguards depends “on legislative provisions being drafted with sufficient clarity to enable the requisite assessments to be made”. The legal clarity required by these safeguards-including the process utilised in this case, whereby questions can be referred to the Supreme Court about whether provisions in a Bill would be within competence – would be undermined if section 101 could “usually, if not invariably, solve the problem” with any question of competence. The second principle also revolved heavily around legal certainty: a broad approach to section 101 would allow the promulgation of statutes which are not “a clear and accessible statement of the law”. In this case, “there has been a decision to draft and enact a provision whose plain meaning does not accurately represent the law, and to rely on the courts … to give it a different effect”. However, the limits on legislative competence, the Court held, should be discernible on the face of Bills and not only after “sufficient cases are decided”. The reason for this principle is simple: such a degree of uncertainty is offensive to the rule of law and the use of section 101 to facilitate that is not contemplated, let alone expressly permitted, by the text of the Scotland Act.

The Supreme Court's judgment in this case represents a milestone both for devolution and for the constitution more widely. On devolution, the Court confirms that section 28(7) of the Scotland Act, far from merely reaffirming parliamentary sovereignty, represents a broad, powerful restriction on devolved legislative freedom. Furthermore, the judgment confirms that provisions of the Scotland Act – particularly section 101 – which are intended to secure a degree of devolved legislative flexibility are not capable of evading limitations that are explicitly part of the devolution settlement (such as the pre-enactment safeguards), or limitations (such as those that derive from the rule of law) that form an implicit part of that settlement. As a result, the hierarchical distance between the Scottish Parliament and Westminster appears to be far greater than had previously been thought. On the constitution more widely, this judgment represents an interesting, and perhaps unexpected, departure from orthodoxy. First, it views the HRA as extraordinarily far-reaching in its impact, and as a piece of legislation that compromises parliamentary sovereignty – a characterisation of the HRA that may impact on the debate about its future. Second, it views parliamentary sovereignty as a concept sufficiently flexible as to be capable of tolerating such compromise. The HRA is widely understood to have been designed to preserve a more rigid, less forgiving conception of sovereignty. The Supreme Court, at least implicitly, suggests that such an attempt was neither successful nor necessary.