What are the domestic law constraints on triggering Article 16 of the Protocol?

In this post, George Peretz Q.C. analyses the domestic law constraints on triggering Article 16 of the Ireland/Northern Ireland Protocol.

The UK government has, very publicly, been considering the use of the “safeguards provision” in Article 16 of the Ireland/Northern Ireland Protocol (the “Protocol”) as a response to perceived difficulties in operating the Protocol.  The use of Article 16 might well generate a response by the EU under the provisions of Article 16 itself, the remainder of the Withdrawal Agreement (the “WA”) of which it forms part, or under the Trade and Cooperation Agreement (the “TCA”).  But a further factor is the possibility that UK courts might be asked to review the lawfulness in domestic law of any measures made on the basis of Article 16.  How realistic is that possibility?

To answer that question, we need to look, first, at the ability of the UK government, as a matter of domestic law, to act inconsistently with the Protocol.  After that, we will look at the question of how that analysis is affected if the UK government is purporting to act in a way permitted by Article 16.

Supremacy of the WA (and of the Protocol)

The starting point is that the WA, of which the Protocol forms an integral part (see Article 182 of the WA), is incorporated into domestic law.  That is achieved by section 7A(1)-(3) of the EU Withdrawal Act 2018 (the “2018 Act”), a section inserted by the EU Withdrawal Agreement Act 2020 (the “2020 Act”).  That section provides that: –

“(1) Subsection (2) applies to—

(a) all such rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the withdrawal agreement, and

(b) all such remedies and procedures from time to time provided for by or under the withdrawal agreement, as in accordance with the withdrawal agreement are without further enactment to be given legal effect or used in the United Kingdom.

(2) The rights, powers, liabilities, obligations, restrictions, remedies and procedures concerned are to be—

(a) recognised and available in domestic law, and

(b) enforced, allowed and followed accordingly.

(3) Every enactment (including an enactment contained in this Act) is to be read and has effect subject to subsection (2).”

That language bears a very substantial, and evidently deliberate, resemblance to section 2(1) of the European Communities Act 1972, the provision that provided the conduit through which EU law flowed into and took effect in the UK domestic legal order during the UK’s membership of the EU.  In Thoburn v Sunderland CC [2002] EWHC 195 (Admin), Laws LJ (in a passage which has frequently been cited and approved since) said that the European Communities Act 1972 was a “constitutional statute” and that: –

“63. Ordinary statutes may be impliedly repealed. Constitutional statutes may not. For the repeal of a constitutional Act or the abrogation of a fundamental right to be effected by statute, the court would apply this test: is it shown that the legislature’s actual – not imputed, constructive or presumed – intention was to effect the repeal or abrogation? I think the test could only be met by express words in the later statute, or by words so specific that the inference of an actual determination to effect the result contended for was irresistible. The ordinary rule of implied repeal does not satisfy this test. Accordingly, it has no application to constitutional statutes. I should add that in my judgment general words could not be supplemented, so as to effect a repeal or significant amendment to a constitutional statute, by reference to what was said in Parliament by the minister promoting the Bill pursuant to Pepper v Hart [1993] AC 593. A constitutional statute can only be repealed, or amended in a way which significantly affects its provisions touching fundamental rights or otherwise the relation between citizen and State, by unambiguous words on the face of the later statute.“

The conclusion that section 7A has that constitutional status is strengthened by the fact that it is intended to implement the UK’s obligation under Article 4 of the WA, namely that: –

“1.   The provisions of this Agreement and the provisions of Union law made applicable by this Agreement shall produce in respect of and in the United Kingdom the same legal effects as those which they produce within the Union and its Member States.

Accordingly, legal or natural persons shall in particular be able to rely directly on the provisions contained or referred to in this Agreement which meet the conditions for direct effect under Union law.

2.   The United Kingdom shall ensure compliance with paragraph 1, including as regards the required powers of its judicial and administrative authorities to disapply inconsistent or incompatible domestic provisions, through domestic primary legislation.“

Nor is that analysis affected by section 38 of the 2020 Act, which “recognises” the sovereignty of Parliament and states that its sovereignty subsists notwithstanding section 7A.  That is because the analysis in Thoburn already fully takes account of the principle of Parliamentary sovereignty: in Laws LJ’s words at [70], its analysis “gives full weight both to the proper supremacy of [EU] law and to the proper supremacy of the United Kingdom Parliament.”

It follows that any UK measure (including retained EU law) is – as a matter of domestic law – subject to the provisions of the Protocol.  If any such measure is inconsistent with “rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under” the Protocol, then as a matter of domestic law the domestic court must grant effective relief. Jack Williams has discussed the relationship between “relevant separation agreement law” (the law flowing through into domestic law by section 7A of the 2018 Act) and “retained EU law” in a blog post available here.

It should be noted that the above analysis applies not just to legislative measures, but also to any non-legislative act or omission by the executive: the suggestion by the current Prime Minister’s former adviser Dominic Cummings that the UK government could simply refuse to implement Protocol obligations or act inconsistently with them therefore suffers from the rather significant flaw that the domestic courts would have to quash or declare unlawful any such act or refusal.

The only exception to the above is where an Act of Parliament, “by express words or by words so specific that the inference of an actual determination to effect the result contended for was irresistible”, qualified the application of section 7A.  In short, to put it somewhat colloquially, the intention of the legislation to apply notwithstanding section 7A would have to be made clear in words visible from outer space: and since giving effect to that intention would be a breach of Article 4 of the WA, it would have to be made clear that the intention was to authorise an infringement of the WA. 

Given the fate of the current government’s previous attempt, in Part 5 of the Internal Market Bill, to get such a measure through the House of Lords (which rejected it by overwhelming majorities), one can see why, according to various newspaper stories, it wants to avoid that course.  Could it avoid it by using subordinate legislation?

Domestic powers of the UK government in relation to the Protocol

Section 8C of the 2018 Act confers what Colton J described in Allister v Secretary of State [2021] NIQB 64, at [166], as an “extraordinarily broad” power on the UK government in relation to the Protocol: –

“(1) A Minister of the Crown may by regulations make such provision as the Minister considers appropriate—

(a) to implement the Protocol on Ireland/Northern Ireland in the withdrawal agreement,

(b) to supplement the effect of section 7A in relation to the Protocol, or

(c) otherwise for the purposes of dealing with matters arising out of, or related to, the Protocol (including matters arising by virtue of section 7A and the Protocol).

(2) Regulations under subsection (1) may make any provision that could be made by an Act of Parliament (including modifying this Act).”

As can be seen, that power may be used to deal with “matters arising out of, or related to” the Protocol and to make “any provision that could be made by an Act of Parliament”. 

Could section 8C be used to make regulations that were inconsistent with the Protocol?  The wide words in section 8C(1)(c) above would appear on their face to permit that.  However, it follows from Thoburn and the analysis in the previous section that such regulations could not modify the effect of the Protocol, and would have to be struck down to the extent that they were inconsistent with it. 

What if the regulations made it clear that, in words visible from outer space, that they intended to apply notwithstanding section 7A, thus qualifying or modifying that section?  Section 8C(2), after all, states that regulations made under that section can do anything that an Act of Parliament can do, including modifying the 2018 Act itself.  And in Allister, at [192]-[210], Colton J upheld regulations made under section 8C as to the operation of the “consent mechanism” in Article 18 of the Protocol notwithstanding that those regulations disapplied (by express words) a provision of the Northern Ireland Act 1998 (another statute recognised as “constitutional”).

Nonetheless, the answer to the question of whether section 8C regulations could modify, or exclude, section 7A is, to a very high degree of certainty, “no”.  Indeed, it is submitted that the answer “yes” is not arguable:

• The general rule, as explained in the passages from the Supreme Court’s judgment in R(Public Law Project) v Lord Chancellor [2016] UKSC 39  set out at [198] of Allister, is that “it is … legitimate to take account of the fact that a delegation to the Executive of power to modify primary legislation must be an exceptional course and that, if there is any doubt about the scope of the power conferred upon the Executive or upon whether it has been exercised, it should be resolved by a restrictive approach.”

• In Allister, the use of section 8C was to create a consent mechanism that complied with the terms of Article 18 and the UK’s Unilateral Declaration attached to it: since Article 18 and the Unilateral Declaration had been before Parliament when it passed section 8C: in Colton J’s words at [205] the consent mechanism process created by the regulations “was set out in the primary legislation and Parliament was fully sighted of its contents when passing the [2020] Act  and when considering the [regulations in that case].”  Since there is no basis for suggesting that use of section 8C regulations to infringe, rather than implement, the Protocol was in Parliament’s contemplation when passing the 2020 Act, the reasoning in Allister would not apply.  

• It is significant that although section 8C refers to “supplementing” section 7A, it nowhere expressly provides that section 8C regulations can modify that section – and it is particularly significant that subsection (1)(c) refers only to dealing with “matters arising by virtue of section 7A and the Protocol” – a phrase that significantly avoids any clear statement that regulations could modify or qualify section 7A.

• Finally, section 7C(1) of the 2018 Act provides that: “Any question as to the validity, meaning or effect of any relevant separation agreement law is to be decided, so far as they are applicable (a) in accordance with the withdrawal agreement ..”.   Since regulations under section 8C are “relevant separation agreement law” (section 7C(3)(a)(i)), it would follow that the validity or meaning of any section 8C regulations have to be decided in accordance with the withdrawal agreement (which includes the Protocol) – and it further follows that any section 8C regulations that are inconsistent with the Protocol are invalid.

It can therefore confidently be concluded that, if as a matter of domestic law the UK government wishes to do anything (whether executive action or omission or subordinate legislation) that is inconsistent with the Protocol, it can do so only after persuading Parliament to pass an Act that, in words visible from outer space, modifies, or expressly allows the government to modify, section 7A.  If it fails to get that cover – and the political difficulties in getting it are obvious – then any such act, omission, or legislation would have to be declared unlawful, or quashed, by a domestic court.

Application to Article 16 measures

How does all that apply to measures purportedly taken under Article 16?

Article 16 is part of the Protocol and hence part of domestic law.  Any measure authorised by it is therefore consistent with the Protocol: section 7A would not apply to invalidate or modify it.

However, the contrary is also true: any measure purportedly made under Article 16 that was not in fact consistent with it would, as a matter of domestic law, be invalid and would have to be set aside by the domestic courts.

As a paper by Robert Howse convincingly argues, the scope for Article 16 measures is in fact limited in a number of very significant ways.  Among other things: it is (to put it at its lowest) not obvious that it offers a means of addressing or suspending aspects of the Protocol that form part of its structure of or which are inherent in its basic obligations: eg the role of the Court of Justice of the EU or the basic obligations in relation to State aid, customs, VAT and goods regulation.  Nor (again putting it at its lowest) is it obvious that Article 16 can properly be used to deal with alleged difficulties, or diversions of trade, that are inevitable or obvious consequences of the Protocol: it is not plausible that either party would have intended to include a provision that allowed the other party to escape its obligations by citing the inevitable consequences of those obligations.  Further, it is hard to see how a plausible case can be made for saying that, for example, the State aid provisions in Article 10 have caused any difficulty – no decision has yet been taken under them, and any UK measure that was needed to avoid serious problems in Northern Ireland would, no doubt, be cleared by the Commission under Article 107(3) TFEU.  .

It is also important that any Article 16 measure be strictly necessary (which would suggest, for example, that where a problem such as disruptive checks on the Irish Sea border could be dealt with by alignment of standards in Great Britain with those of the EU, then alignment should be chosen rather than Article 16 measures). 

It is notable that the current government’s analysis of the case for invoking Article 16, set out at [29] of the July Command Paper, does not address any of the above issues: its assertion in that paragraph that it is “clear” that the circumstances exist for its invocation is therefore highly questionable.

It is no doubt true that no court would lightly dismiss as irrational a government’s view as to whether provisions of the Protocol were giving rise to economic or societal difficulties or diversion of trade.  But the basis of such a view would have to be set out and reach a standard of plausibility, if not a very high one.  Moreover, if the above analysis is right, the government would have to advance a plausible case that those difficulties were both unexpected (or at any rate not obvious) consequences of the Protocol and that solutions such as alignment would not resolve them.  It is not entirely clear that the government is in a position to do that.

As a result, it is not at all clear that the government has a solid legal basis for invoking Article 16, at least in relation to the large majority of concerns set out in the July Command Paper. 

Therefore, if the UK government chooses to implement measures that are otherwise in breach of the Protocol but which are justified solely on the basis of Article 16, it is at real risk of having those measures struck down in the domestic courts, especially if the measures exceed a limited duration or scope.  The only way of immunising itself against such a challenge would be to persuade Parliament to pass an Act authorising it to modify or qualify section 7A: a course which, as noted above, is likely to face very serious difficulties in the House of Lords, if not in the House of Commons.

Share this post on social media: