The Protocol does its best to refute such claims. Article 1(1) of the Protocol states that it is “without prejudice to the provisions of the [GFA],” and Article 1(3) tells us that the Protocol “sets out arrangements necessary… to protect the [GFA] in all its dimensions.”
But the Protocol says a number of things, not all of which are consistent with what it actually does. For example, it also says that Northern Ireland remains part of the UK customs territory, a claim which is not easy to reconcile with what it does (which is, with limited exceptions, place Northern Ireland under the EU customs codes and its rules on tariffs and checks). So, to evaluate the very serious charge that the Protocol is contrary to the GFA—the agreement that is fundamental to the peace process in Northern Ireland—we have to look at what the GFA says, and at what the Protocol does.
The GFA provides for several things, including commitments on various kinds of cooperation across the island of Ireland and between the UK and Ireland, and commitments on human rights and security. In a situation familiar to lawyers, it silently assumed that both the UK and Ireland would remain in the EU (it refers to cooperation on EU matters), but made no provision for the exit of either Ireland or the UK from the EU or for the Ireland/Northern Ireland border in such a case.
The argument that tends to be made by those claiming the Protocol is inconsistent with the GFA is that, by providing that large parts of EU single market law apply in Northern Ireland when they do not apply elsewhere in the UK, the Protocol infringes section 1(1) of the Northern Ireland Act 1998 (a provision which is set out in the GFA and which the UK government agreed to incorporate in UK legislation). Indeed, the argument of inconsistency with section 1 was a major element of the recent challenge to the Protocol in the Northern Ireland High Court (Allister v Secretary of State for Northern Ireland).
However, what section 1(1) actually says is that “it is hereby declared that Northern Ireland in its entirety remains part of the United Kingdom and shall not cease to be so without the consent of a majority of the people of Northern Ireland voting in a poll held for the purposes of this section.” So the obvious difficulty with the argument of inconsistency with section 1 is that, even though it (unlike Great Britain) remains subject to significant parts of EU law, Northern Ireland nonetheless “remains part of the United Kingdom.”
In response to that difficulty, it was argued that the section had to be read so as to apply to a fundamental change in the constitutional status of Northern Ireland involving “surrender of powers to the EU.”
The High Court convincingly rejected that response. First, it pointed out that in the first Gina Miller case (about the government’s power to serve notice under Article 50 and thus commence the Brexit process) the Supreme Court had previously decided that section 1 gave the people of Northern Ireland the right to determine whether to remain part of the UK or to become part of a united Ireland, but did not regulate any other change in the constitutional status of Northern Ireland. Second, the High Court went through the GFA itself, noting that when the GFA referred to the need for the people of Northern Ireland to agree a change in its “constitutional status,” it did so in the context of the choice between being part of the UK and being part of a united Ireland: for example, it is said that the parties to the GFA “recognise the legitimacy of whatever choice is freely exercised by a majority of the people of Northern Ireland with regard to its status, whether they prefer to continue to support the Union with Great Britain or a sovereign united Ireland” (emphasis added).
A more persuasive argument—which is connected to the GFA, given the UK’s commitment in that agreement to respect obligations under the European Convention on Human Rights—is the claim that the Protocol infringes the right to vote for your legislature set out in Article 3 of Protocol 1 to the ECHR (A3P1). A couple of years ago, I argued here that the transition period in the Brexit Withdrawal Agreement ran the strong risk of being inconsistent with A3P1: and, though the Protocol places Northern Ireland under only certain aspects of EU law (albeit important ones that affect daily life in all sorts of ways), it does so permanently (subject to provisions for periodical votes in the Assembly as to whether to consent to the Protocol’s central provisions). Yet the people of Northern Ireland have no democratic voice in the EU institutions that will be making some of their laws.
The claim of breach of A3P1 was made in Allister, but was also rejected. The High Court pointed out that crass comparisons with colonial rule or Vichy France were wide of the mark. It also relied on the periodical consent mechanism and on the fact that Northern Ireland sends MPs to Westminster which (ultimately) has power to disapply the Protocol as a matter of domestic law (though not, of course, in international law). The court also noted that the Protocol was a complex solution to a complex problem, and that its provisions had been considered and approved by parliament. In the end, it decided that the Protocol lay within the UK’s “margin of appreciation” under A3P1. (It might also be noted that, even if it were established that the Protocol infringed A3P1, a possible remedy would be to allow Northern Ireland residents to vote for Irish MEPs—which might not be the result that those opposing the Protocol actually want.)
In the end, therefore, the claims by hard Brexiters that the Protocol contravened the GFA all failed on legal grounds. What one now hears are claims that the Protocol violates the “spirit” of the GFA. But such claims are supremely paradoxical: the inevitable effect of any successful attempt to overturn the Protocol on grounds of incompatibility with the GFA would be the imposition of a hard border between Ireland and Northern Ireland, with consequences that would strike at the heart of everything that the GFA was seeking to achieve.