Wednesday 22 February 2023

Allister: the effect of the EU Withdrawal Act


Anurag Deb, PhD researcher, Queen's University

Photo credit: Dom0803, via Wikimedia Commons 

In the uncertain aftermath of the Brexit referendum and shortly after Theresa May had succeeded David Cameron as Prime Minister, she had announced her intention to move a ‘Great Repeal Bill’ to undo what Lord Denning had once described as the ‘incoming tide’ of EU law. That Great Repeal Bill became the EU Withdrawal Act 2018, amended in 2020 by the EU Withdrawal Agreement Act 2020. But did this stop the incoming tide? Not particularly – there are post-exit arrangements baked into the 2018 Act in order to provide certainty to, for example, people who have cases reliant on EU law progressing through the courts at the point of exit, so that their cases are not extinguished overnight.


However, savings provisions to ensure legal certainty are fairly common in legislation. What is somewhat more uncommon is for the legislation designed to achieve one purpose to be amended almost two years later to partially (and significantly) achieve its opposite purpose. This is a post which looks at the effect of section 7A of the 2018 Act, inserted by the 2020 Act, and what the Supreme Court said about it in Allister and Peeples’ applications for judicial review.


Section 7A: the reopened gateway


Section 7A adopts much of the same language as the now-repealed section 2(1) of the European Communities Act 1972 – the very same provision which Denning MR described as governing the ‘incoming tide’ of EU law. This tide does little for most of UK law, however, given that the Withdrawal Agreement, which section 7A incorporates, generally provides for the severance of the UK from the EU legal order. But this is not true of Northern Ireland, which is addressed by its own Protocol.


The content of the Protocol varies widely, but its major function is the continued application in Northern Ireland of aspects of the EU Single Market in goods as well as its customs code. There are additional guarantees regarding equality and non-discrimination, which link with the Good Friday Agreement. The detail of this is covered by Colin Murray in his post on Allister and Peeples, and I will not be covering it here. Nor will I cover the well-trodden discussion of the constitutional statutes doctrine, in light of the judgment. Some have discussed its demise (Murray, David Allen Green and John Bell) while others have defended its continuing health (Oliver Garner and myself). Instead, I will be covering a single paragraph of the Supreme Court judgment – a single paragraph which contains significant possibilities.


The gateway through which the Withdrawal Agreement (and the Protocol) enters domestic UK law is section 7A. In addition to incorporating the Withdrawal Agreement, section 7A also gives effect specifically to two highly important aspects of the Withdrawal Agreement. The first aspect is direct effect (Withdrawal Agreement, Article 4(1); section 7A(1) and (2) of the 2018 Act), by which people may directly rely on the provisions contained in or referred to in the Withdrawal Agreement. The second aspect is to subject every enactment (including the 2018 Act itself) to the provisions of the Withdrawal Agreement (Withdrawal Agreement, Article 4(2); section 7A(3) of the 2018 Act). This trifecta of consequences was used to great effect by the Supreme Court when faced with a particular conundrum.


The modification of the Northern Ireland Act


The Northern Ireland Act 1998 is the primary legislation governing modern devolution in Northern Ireland. The relevant part of the statute for this post is section 42, which creates a powerful legislative mechanism unique in the UK’s legislatures. This mechanism, the petition of concern, is a petition which can be moved by as few as one-third of MLAs and if so moved, triggers a requirement for cross-community support in respect of the matter against which the petition was moved. Cross-community support is significantly more difficult to achieve than a simple majority, especially in a legislative body which can at times be highly polarised. But section 42 was originally applicable to any matter on which the Assembly was to vote. This was amended in 2022, and I have covered the circumstances of that amendment elsewhere.


The important point, however, is that section 42 remained applicable to any Assembly vote at the time that the Protocol was incorporated. Almost a year after this incorporation, the Northern Ireland Secretary made new regulations with sweeping consequences. These regulations were made to provide for the so-called ‘Stormont consent vote’, by which the Northern Ireland Assembly votes periodically on the question whether or not to continue the application of EU law on goods and customs. The provision of this vote was required by Article 18 of the Protocol as well as the UK Government’s unilateral declaration on providing the vote, neither of which mentioned section 42 or the petition of concern mechanism. Now, these regulations disapplied section 42 to the consent vote. The appellants in Allister and Peeples challenged this disapplication as being in breach of the Northern Ireland Act, which was prohibited by section 10 of the 2018 Act. Thus, the appellants said, the regulations were unlawful.


The Supreme Court resolved this issue in paragraph 108 of its judgment. While the paragraph is lengthy, it essentially says that there was no breach of section 10 of the 2018 Act because section 7A had already modified section 42 of the Northern Ireland Act. This had occurred because Article 18 had created an obligation on the UK Government to provide for the consent vote ‘strictly in accordance’ with the terms of the unilateral declaration, which, together with Article 18, both provide for a decision based on a majority of MLAs. Section 42 would subject this to the requirement of cross-community consent. Therefore, because of the overriding nature of section 7A, the obligation created by Article 18 and the unilateral declaration modified section 42. The regulations, as a result, did not breach the Northern Ireland Act.


The Supreme Court’s reasoning is considerably different from that of the Court of Appeal and the High Court on this point, both of which zeroed in on the breadth of the powers of delegated legislation conferred on the Northern Ireland Secretary under section 8C of the 2018 as a complete answer to this issue. Both courts reasoned that the regulations, having been made under section 8C and in accordance with the conditions attached to that power under the 2018 Act, were made lawfully – independent of the question whether section 7A had already modified the Northern Ireland Act before the making of the regulations.


The implications


The Supreme Court’s reasoning is far from obvious when scrutinising the content of section 42 (as it originally existed), Article 18 and the unilateral declaration. First, section 42 is a conditional provision – it needs a minimum threshold of MLAs to be triggered. While it was possible to trigger section 42 in any matter on which the Assembly voted, possibility does not equate to a guarantee. In other words, section 42 would not automatically have subjected a consent vote to the requirement of cross-community support unless the 30-MLA threshold was triggered. Moreover, both Article 18 and the unilateral declaration contain the possibility of a decision to continue the application of EU law made with cross-community support (which triggers a longer waiting period before the next consent vote) – a matter the Supreme Court curiously omits to mention.


So, how can paragraph 108 be interpreted? Accepting the Court’s judgment as correct (which should be obvious given that no appeal is possible from a judgment of the Supreme Court) raises an interesting point. Section 7A incorporates some EU law, but through the Withdrawal Agreement (and Protocol). The Withdrawal Agreement is not a passive vehicle in this incorporation, but creates rights and obligations distinct from the EU law in respect of which it also creates obligations. After all, the consent vote is a creation, not of EU law, but of the Protocol. Thus, not only does section 7A modify domestic law in consequence of aspects of EU law, but also in consequence of the Withdrawal Agreement as a distinct source of law by itself. Moreover, as nowhere in either Article 18 or the unilateral declaration was there any explicit obligation to disapply section 42, it follows that section 7A modified domestic law by necessary implication of giving effect to the Withdrawal Agreement and the Protocol.


I return briefly to Denning MR’s analogy, with the incorporated EU law as an incoming tide. Incoming tides raise water levels, but they can also change the appearance of coastal areas over time. This is a possible reading of the Supreme Court’s reasoning – that section 7A brings in both EU law and Withdrawal Agreement law and modifies the domestic legal order around these sources of law – both by giving primacy to the explicit provisions of these sources and as a necessary consequence of their incorporation.


What this means for the future of the Protocol’s operation in Northern Ireland is difficult to tell precisely. But what is reasonably certain is that a single paragraph of the Supreme Court’s judgment contains a disproportionately vast potential.

No comments:

Post a Comment