Introduction
The Climate Change Act (Northern Ireland) 2022 (‘CCA’) is, from a constitutional perspective, a fairly unassuming, if far-reaching, exercise of devolved legislative power. It is an example of a devolved part of the UK seizing the opportunity the devolution framework provides to ‘design policies and take public spending decisions that more closely align with local needs and preferences’, as well as to participate in ‘policy experimentation’ from which other parts of the UK might learn. However, the purpose of this post is to argue that the Supreme Court’s recent jurisprudence on devolution, if it is found to apply in Northern Ireland, threatens to imperil, or to at least drastically hollow, exercises of devolved legislative power – like the CCA – that should be uncontroversial from the perspective of devolved law-making power. The potentially far-reaching – and perhaps unforeseen – practical implications of that jurisprudence, we suggest, illustrates the flaws in its reasoning. In the second part of this post, we consider that jurisprudence’s tension with devolution through a wider lens, demonstrating that an account of history and of the conceptual framework upon which devolved law-making depends provide further reasons to hope that a future (and enlarged) Supreme Court changes course.
In the first section of this post, we set out the problem: the ‘conditioning limitation’ identified in the Supreme Court’s Continuity Bill and Treaty Incorporation judgments. In its second section we consider some important ambiguities surrounding the scope, meaning and application of those cases (particularly in Northern Ireland). Third, we apply these issues to the CCA to demonstrate that this jurisprudence may have destabilising implications for that Act, and for other similar exercises of devolved legislative power.
The Problem: The Conditioning Limitation
In Continuity Bill, the Supreme Court said, at [64], that it would be unlawful for the Scottish Parliament to ‘impose a condition on the effect of certain laws made by Parliament for Scotland, unless and until Parliament exercised its sovereignty so as to disapply or repeal it’. Such a condition, the Court said, would modify the protection – in s 28(7) of the Scotland Act 1998 – of Westminster’s power to make laws for Scotland, something prohibited by Schedule 4 of the Act. There, to ‘impose a condition’ meant to make the ‘effect of laws made by the UK Parliament for Scotland conditional on the consent of the Scottish Ministers’ ([53], emphasis added). Section 28(7), the Court said (at [52]) recognised Westminster’s ‘unqualified legislative power’, which could not lawfully be modified.
In Treaty Incorporation, the Court went further, saying (at [28]) that this limitation would be transgressed where Westminster legislation was ‘given a different meaning and effect from those that Parliament intended’ and that ‘[a] provision which required the courts to modify the meaning and effect of legislation enacted by Parliament would plainly impose a qualification upon its legislative power.’. Westminster’s unqualified legislative power included its ‘power to allow existing legislation to remain in force unamended’ ([42]) and would be transgressed (at [52]) by legislation facilitating ‘judicial condemnation’ in the form of a declaration which would ‘impose pressure on Parliament to avoid the opprobrium which such a finding would entail.’ The Court also identified problems with a declaration of that sort where it would (at [53]) ‘make it difficult if not impossible in practice for public authorities to implement provisions of Acts of Parliament’ and where (at [54]) a Minister would be empowered to amend or repeal declaratorily impugned legislation.
There are two apparent readings of the limitation identified by the Court in these cases (which we describe as the conditioning limitation). One reading is a non-delegation doctrine. The Court’s issue, on this reading, is not with the impugnment of legislation per se. As the protected enactments framework recognises, the devolved legislatures are of course empowered to repeal or amend Westminster legislation so far as it is within their competence. Instead, the Court’s concern was with the allocation of powers to other institutions to impugn Westminster legislation (ministers in Continuity Bill, whose permission was required for secondary legislation to become effective, and courts and ministers in Treaty Incorporation, who were empowered to interfere with Westminster legislation in various ways explored elsewhere). As the Court put it, ‘the Scottish Parliament cannot make the effect of Acts of Parliament conditional on decisions taken by other institutions’ (Treaty Incorporation: [30] (emphasis added); [40]).
A broader reading, as McHarg has outlined, includes ‘indirectly’ amending or repealing UK legislation within devolved areas ‘by subjecting it to general conditions.’ This broader reading, McHarg suggests, could capture devolved legislation which requires that ‘functions’ exercisable by virtue of UK legislation ‘will be exercised differently to the way in which the UK Parliament had intended them to be used.’ However, even understood narrowlyas a rule against delegation, the Supreme Court’s reasoning throws open a number of problems. Such a doctrine is potentially both anomalous and incoherent, with Cowie, for instance, suggesting that it calls into question Henry VIII powers which are prolific in the Scottish Parliament’s legislation. The Court found that the triggering, by an incompatibility declarator, of a Scottish ministerial power to make discretionary remedial regulations altering or repealing Westminster legislation would be outwith competence (Treaty Incorporation,at [54]). The position for the enactment of Henry VIII powers, therefore, remains uncertain, since these are quite clearly instances in which the devolved legislatures ‘authorise another body to alter the meaning and effect of legislation enacted by Parliament’ ([30]).
Some Uncertainties
The conditioning limitation gives rise to several ambiguities, the first of which concerns the precise point at which – if it is correct – the non-delegation doctrine begins to bite. Cowie notes, for instance, that there is a meaningful distinction between altering legislation and enforcing its altered meaning. It is unclear, however, when the Court will be satisfied that the devolved legislature, acting lawfully, is amending Westminster legislation and when it is delegating that amending or repealing power to another institution (such as a court or a minister) and so acting unlawfully. That point, we suggest, might be located by reference to a chain of decision: where an institution that is not the devolved legislature has discretion or decisional power to alter the meaning or effect of an Act of the Westminster Parliament, this rule might be engaged. However, even here there is uncertainty: devolved legislation may itself amend a power sourced in Westminster legislation while empowering a minister to alter the scope of that amendment of the power. It is unclear whether discretion merely as to the scope of an amendment would be captured, but it certainly might where, but for this change, Westminster legislation would not otherwise be modified, or be modified to the same extent.
A second ambiguity relates to the institution to which power is delegated, and whether the same restrictions apply universally. The Court appeared to suggest in Treaty Incorporation (at [50]) that delegating powers to courts to impose political opprobrium was particularly problematic for instant purposes. Questions arise, therefore, as to whether allocating similar powers to non-judicial bodies would have the same implications, or whether different authorities can be lawfully empowered in different ways.
The third ambiguity concerns procedural duties. The Court considered procedural duties in passing in Treaty Incorporation (at [33]) by reference to s 1 of the Rights of Children and Young Persons (Wales) Measure 2011, which provides that ‘Welsh Ministers must, when exercising any of their functions, have due regard to the requirements of’ specified parts of the United Nations Convention on the Rights of the Child (‘UNCRC’). Although the Court said that ‘that provision is much more limited in scope than the Scottish UNCRC Bill’ in dispute before it, the Court did use the language of ‘conditions’ to describe a provision analogous to the relevant duties in the CCA explored below. The provision in that case (s 6 of the UNCRC Bill), the Court said (at [59]), ‘imposes a condition – namely, compliance with the UNCRC requirements – upon the lawfulness of the discharge of functions required or authorised by Parliament’. However, the Court was principally concerned with other aspects of this provision which rendered it outwith competence and did not explore whether that type of condition was itself unlawful.
A procedural duty could engage the conditioning limitation because it would make it unlawful for an authority to exercise a power conferred by an Act of Parliament without (for instance) having regard to a matter which is not anticipated in that Act. A procedural duty therefore restrictsthe lawful exercise of a power and, therefore, on the Court’s reasoning, the practical effect given to a provision in an Act of Parliament. That duty could in practice drastically limit the exercise of that power, but whether this amounts to ‘giv[ing a provision] a different meaning and effect from those that Parliament intended’, and therefore falls foul of the conditioning limitation, is ultimately unclear.
The fourth ambiguity turns on the idiosyncrasies of each particular devolution scheme and, therefore, the applicability of the conditioning limitation in Wales and Northern Ireland. The conditioning limitation itself turns on the protection of s 28(7) of the Scotland Act 1998 from modification. It provides that ‘[t]his section does not affect the power of the Parliament of the United Kingdom to make laws for Scotland.’ The Government of Wales Act 2006 contains, in s 107(5), the same language (adapted for Wales), making the application of this reasoning straightforward. Section 5(6) of the Northern Ireland Act 1998 (‘NIA’) is the same (adapted for Northern Ireland) but, crucially, with this addition:
‘… but an Act of the Assembly may modify any provision made by or under an Act of Parliament in so far as it is part of the law of Northern Ireland.’
There are further idiosyncrasies in Northern Ireland which the Continuity Bill and Treaty Incorporation reasoning may not adequately accommodate, such as the heavy reliance upon Orders in Council, the sourcing of contemporarily exercisable executive powers in previous legislative schemes which are generally repealed, but whose effect is to some extent preserved, and the existence of powers to modify those empowering frameworks when making secondary legislation. We explore the implications of these matters for the application of the conditioning limitation in Northern Ireland in more detail in the second part of this post, as well as some further problems to which it gives rise. The final part of this post applies this reasoning, and its ambiguities, to the CCA.
The Climate Change Act (Northern Ireland) 2022
Section 29(4) of the CCA, enacted by the Northern Ireland Assembly, provides that ‘[w]hen developing policies each [NI] department must ensure they are consistent with the targets set out in the carbon budget [required by the Act].’ Under s 52(1) each NI department is under ‘duties’ to, amongst other things, ‘exercise its own functions, so far as is possible to do so, in a manner that is consistent with the achievement of that objective’. Those targets for the years 2030 and 2040 are set by the Department for Agriculture, Environment and Rural Affairs and these ‘functions’ are those conferred by or under an Act of the Northern Ireland Assembly or an Act of the UK Parliament. The CCA potentially engages the conditioning limitation for three reasons.
Firstly, the CCA arguably restricts the lawful exercise of power authorised ultimately by virtue of an Act of Parliament. In Re Hassard and others’ application [2025] NIKB 42 McAlinden J quashed a decision made by the Northern Ireland Department for Infrastructure on the basis that (among other things) it had failed to discharge its obligations under s 52 of the CCA. The Department’s decision was made under various provisions of the Roads (Northern Ireland) Order 1993 (as amended), which was itself made by the Privy Council under the authority of the Northern Ireland Act 1974 (‘1974 Act’). Hassard therefore demonstrates that s 52 of the CCA requires Northern Ireland Departments to exercise their functions, including those derived from Westminster legislation (in that case the 1974 Act), consistently with conditions not envisaged, authorised or set out in that legislation, and that a failure to do so may result in the exercise of those functions being quashed. This engages the conditioning limitation as it was described in Treaty Incorporation because the UK Parliament’s ‘unqualified power’ includes the power (at [52]) ‘to allow existing legislation to continue in force unamended’, including the 1974 Act and legislation made under it. The CCA, therefore, to adapt the language of the Court (Treaty Incorporation, at [59]), ‘imposes a condition – namely, compliance with the [targets] – upon the lawfulness of the discharge of functions required or authorised by Parliament’.
Furthermore, the nature of the source of the powers exercised in Hassard – the 1993 Order made by the Privy Council – engages the Continuity Bill limitation because (at [52]) ‘the power of the UK Parliament to make laws for [the devolved part of the UK] includes the power to make laws authorising the making of subordinate legislation by Ministers and other persons’ such as the Privy Council. Under the CCA, the UK Parliament cannot authorise the making of subordinate legislation which empowers Northern Ireland Departments to exercise their functions inconsistently with the CCA’s conditions ‘unless it disapplied or repealed the provision in question’ (Continuity Bill, [53]).
Secondly, the CCA empowers departments to adjust the targets with which they must act consistently (s 54). Even, therefore, if it were accepted that any amendment of the Westminster legislation-sourced powers is lawfully effected by the Northern Ireland Assembly in the CCA, the departments can alter the intensity of that amendment. This adjustment may have implications not only for the scope of the condition imposed but may also bring newpowers within the conditioned category (whether extant or in future legislation), which the CCA would not otherwise touch.
Thirdly, the way in which s 52 of the CCA conditions departmental functions may go beyond the purview of the CCA itself. In Hassard, McAlinden J set out what s 52(1) required, saying, at [199]:
‘… in order to comply with its section 52 duty when making a decision such as this, the DfI needs to be able to produce cogent evidence that its decision has been made following careful planning, synchronisation and co‑ordination between all Northern Ireland government departments, the result of which demonstrates that the project fits into plans, strategies and policies which map out a realistic and achievable pathway for Northern Ireland to achieve net zero by 2050, meeting the interval targets on the way and staying within the carbon budgets that have now been set. […]’
The Judge’s use of ‘synchronisation’ and ‘co-ordination’ in the context of climate change mitigation mirrors the duties on the Northern Ireland Executive as a whole to make decisions collectively in prescribed circumstances. It is possible to suggest that the CCA’s requirements, pervasive across departmental functions, mean most (if not all) decisions required or authorised by the CCA may fall within the scope of decisions which are ‘significant’ and ‘cross-cutting’ under the NIA. That conclusion would engage collective decision-making duties under Stormont’s justiciable Ministerial Code, depriving Ministers of the authority to act if such decisions are not made collectively. If that is right, the CCA’s impact here is not by explicit amendment to the relevant provisions of the NIA (sections 20 and 28A, which are within the competence of the Assembly to amend, see section 8(b) of the NIA read with Sch 3, para 42(a)) but in obliging Departments (and their Ministers) to act consistently with cooperation requirements in ways not anticipated in the empowering legislation. This may not be the case, but the fact that it is not unthinkable exposes the problems posed by the conditioning limitation and its interaction with other aspects of the devolution framework.
We consider that it is certainly arguable that the CCA falls foul of the conditioning limitation as it is described in Continuity Bill and Treaty Incorporation if it is found to apply in Northern Ireland. Where that limitation is broadly construed, this is so because the CCA restricts the practical effect given to Westminster legislation and its power to authorise power-conferring delegated legislation. Unless it repeals the relevant provisions of the CCA, extant and future legislation enacted by Westminster which empowers Northern Ireland Departments and Ministers will all otherwise be conditioned by the CCA (Continuity Bill, [53]). Further, because of the role played by departments in setting and adjusting the targets, which could bring new Westminster-sourced powers in scope, this might engage the narrower ‘non-delegation’ conception of the conditioning limitation by ‘mak[ing] the effect of Acts of Parliament conditional on decisions taken by other institutions’ (Treaty Incorporation: [30]).
The purpose of this examination is to highlight the depth and breadth of the problems the conditioning limitation poses to otherwise constitutionally innocuous exercises of devolved legislative power. The CCA is precisely the kind of legislation that, under the devolution framework, the devolved legislatures should be able to enact, but the conditioning limitation – at the very least – brings its lawfulness into question. One response to this issue might be for the Supreme Court to confine the conditioning limitation to Scotland (and potentially Wales). However, in our next post we explain why that approach would be unprincipled and ahistoric. Rather, it is our hope that the Supreme Court significantly curtails, or abandons, the conditioning limitation in its entirety, closing off the possibility – which is open at present – that it might apply to the CCA and other similar enactments. This approach is preferable because, as the example of the CCA illustrates, the conditioning limitation is a Pandora’s Box which sits uncomfortably with the devolution framework itself.
Thanks are owed to Aileen McHarg, Mike Gordon and Paul Scott for comments on an earlier draft. Any errors or omissions are our own.
UK Constitutional Law Association 