Vdarious changes to the institutions established by the Belfast (Good Friday) Agreement have been proposed. Does the law require a political consensus in favour of reforming them?
The political merits and demerits of reforming the institutions established by the Belfast (Good Friday) Agreement (“BGFA”) lie outside the scope of this article. Instead, it focuses on the legal frameworks which would govern any process of reform. In the January 2023 issue of this magazine, Rory Montgomery expressed some concerns about suggestions that the British-Irish Agreement (“BIA”), as “the only legally binding text”, “could be changed between the [UK and Irish] Governments alone”. More recently, Sir Tony Blair noted in his evidence to the Northern Ireland Affairs Committee on 16 March 2023 that “the Agreement itself provides for the possibility of review … if the parties agree among themselves that it is going to make for an effective form of Government”. The aim of this article is to clarify the extent to which international and/or domestic law requires any such political consensus. To repeat, the prudence or imprudence of giving lawful effect to politically contentious reforms plays no part in this analysis.
The BIA is the name for the treaty which establishes a legal basis for the BGFA under international law. The text of the BIA consists of a preamble, four articles, and two annexes; with Annex 1 being the text of the Agreement Reached in the Multi-Party Talks (“ARMPT”). The text of the ARMPT likewise includes a copy of the BIA in one of its two annexes, and thus there is an upside-down, mirror-like relationship between the political agreement and its international law twin. Therefore, when viewed in the round it does not matter whether one thinks of the BIA or the ARMPT as the text of the BGFA. It is more important to emphasise that the specific terms of the BGFA are only binding on the international law plane. While there is no international court or tribunal with jurisdiction to enforce those terms (Ireland recognises the jurisdiction of the International Court of Justice, but “with the exception of any legal dispute with the United Kingdom of Great Britain and Northern Ireland in regard to Northern Ireland”, which excludes any litigation about the BIA there), its status as an international law treaty demands compliance in accordance with the rule of law, which is to say nothing of its political and diplomatic status more broadly.
Under domestic law, which for the purposes of this article means UK law, the BGFA has legal effects (i) via the principle of legality, which holds that the UK Parliament will be presumed by the courts to legislate compatibly with its international law obligations unless it clearly indicates it intends not to, and (ii) via UK legislation which has either directly or indirectly incorporated its terms. In these ways, some of the substance of the BGFA has been made justiciable in the UK (i.e. judicially enforceable), unlike the legally binding but judicially unenforceable character of the BGFA under international law.
The main statute which gives domestic effect to the BGFA is the Northern Ireland Act 1998 (“NIA”), but the relationship between the BGFA and the NIA is not a particularly mirror-like one. Leaving to one side that the prisoners section of the BGFA was given expedited effect by the Northern Ireland (Sentences) Act 1998, because it contained a time-sensitive commitment, there are other differences with notable legal effects. The Court of Appeal in Northern Ireland observed in the Buick case, for instance, that section 23(1) of the NIA, which provides that executive power in Northern Ireland shall continue to be vested in His Majesty, is “a variation on the terms of the Agreement”, which envisaged that the Northern Ireland Assembly would be “the prime source of authority in respect of all devolved responsibilities” and would exercise “full legislative and executive authority”.
On the other hand, the NIA contains references to the ARMPT throughout it, which complicates the task of legal interpretation for domestic courts to a certain extent. In section 20, for example, the NIA provides that the Executive Committee “shall have the functions set out in paragraphs 19 and 20 of Strand One of the Belfast Agreement”. With this said, however much such references to the ARMPT in the NIA enable the courts to draw on the ARMPT as an aid to statutory interpretation, the specific terms of the statute book as a whole (including other legislation giving partial effect to the ARMPT) usually take precedence from the domestic judiciary’s point of view. The Robinson case is of course a famous and noteworthy exception, but its details need not be rehearsed here. The UK Supreme Court’s more recent judgment in the Allister case should be highlighted, however, as it confirms even statutes with a constitutional character, like the NIA, can be impliedly repealed by later legislation with unambiguous terms.
The Legal Requisites for Amendment
The substantive sections of the BGFA relating to review procedures should be read with the foregoing explanation of the legal status enjoyed by the BGFA in mind. Paragraphs 6 and 7 in the “Review procedures following implementation” sub-section at the end of the ARMPT provide as follows:
- If there are difficulties in the operation of a particular institution, which have implications for another institution, they may review their operations separately and jointly and agree on remedial action to be taken under their respective authorities.
- If difficulties arise which require remedial action across the range of institutions, or otherwise require amendment of the British-Irish Agreement or relevant legislation, the process of review will fall to the two Governments in consultation with the parties in the Assembly. Each Government will be responsible for action in its own jurisdiction.
Article 2 of the BIA binds the two Governments to support and where appropriate implement the provisions of the ARMPT, which includes the review procedures above. Thus, by complying with the appropriate review procedures, it is possible for the two Governments to amend the Agreement “in consultation with the parties in the Assembly” where difficulties have arisen and there are reform proposals involving a range of the institutions.
Indeed, it is paragraph 7, quoted above, which provides the legal basis for amendments to the original features of the BGFA which have been implemented following the St Andrews Agreement and the New Decade, New Approach deal. They did not come with international treaty twins, as the ARMPT did, as none were required to confer international legal effect upon them. Paragraph 7 served this function.
The most significant aspect of the applicable review procedures at present is the requirement for “consultation with the parties in the Assembly”. This is obviously important as it requires consultation not just with the parties to the ARMPT, but also anti-Agreement unionists who now participate in the Assembly, like the Democratic Unionist Party. However, consultation is different from consent and, as such, it seems clear that the two Governments can lawfully act in the absence of a political consensus in favour of reforming the institutions established by the Agreement. Furthermore, it is unlikely that any unambiguous amendments to the NIA (or any other ARMPT-implementing legislation) could be successfully challenged in UK courts for failing to adhere to the original terms of the ARMPT in a context where such amendments were giving effect to reforms lawfully agreed in accordance with the “Review procedures following implementation” sub-section of the ARMPT.