Official Report: Minutes of Evidence

The Chairperson (Mr Buckley): I welcome our guests. Thank you very much for taking the time to come to the Committee. I remind you all that the format of the session is that each witness will present their paper for up to eight minutes to be followed by questions from members. The intention is to invite those witnesses whose papers are wider in scope to speak first, followed by those whose papers focus on particular areas. I understand that Dr Katie Johnston will introduce the panel before we go to each witness individually. I will hand over to you, Katie.

Dr Katie Johnston (University of Liverpool): Thank you very much and good morning. Thank you to the Committee for the opportunity to contribute to its inquiry. I will give a brief introduction. In April 2024, the University of Liverpool convened a workshop on legal aspects of the reform of the 1998 agreement. All today's panellists participated in that workshop, as well as Alan Whysall from the University College London (UCL) constitution unit and Daniel Holder from the Committee on the Administration of Justice (CAJ), who, I believe, have already given evidence to the Committee.

The papers from that conference were published last December, and, as the Committee can see, they covered a range of perspectives on reform, from domestic law to international law. While some of those papers consider the merits of particular reforms, others are more focused on the process of reform.

I will leave it there, for now.

The Chairperson (Mr Buckley): Thank you very much. I invite Professor Ciarán Burke to speak. He is joining the Committee remotely. I refer Committee members to Ciarán's speaking notes in their packs. Ciarán, you have eight minutes in which to brief the Committee.

Professor Ciarán Burke (Friedrich Schiller University Jena): Chairperson and members of the Committee, thank you for the invitation to appear before the Committee

My remarks will draw on my recent research concerning the Belfast/Good Friday Agreement, Brexit and the continuing evolution of peace settlements in international law.

I will begin with what I regard as an important premise: the Belfast/Good Friday Agreement has been an extraordinary success. It transformed a violent constitutional conflict into a political and democratic process and provided institutions through which profound disagreement could be managed peacefully. Any discussion of reform must begin with the acknowledgement of that achievement, but success and permanence are not the same thing.

One of the most striking lessons learned from international law is that agreements that are intended to endure are not interpreted as though they were frozen in the circumstances of their creation. That has been repeatedly recognised by international courts. In both the Namibia advisory opinion and, later, the navigational rights case, the International Court of Justice in The Hague emphasised the fact that long-term agreements frequently contain concepts that must be understood in an evolving manner if they are to continue to fulfil their intended purpose. That observation is certainly relevant to Northern Ireland.

The question before us is not whether the agreement was appropriate in 1998 — it plainly was — but whether the institutional arrangements continue to fulfil their intended function under the conditions of 2025. That is a different question and, indeed, it may be the central constitutional question facing Northern Ireland today.

The agreement was negotiated in a very particular environment when the UK and Ireland were both members of the European Union. The border had become less significant, practically and psychologically, political relationships between London, Dublin and Belfast were developing in a wider European framework, and the institutions reflected those realities. Brexit will undoubtedly be discussed. It did not invalidate the agreement, but it undeniably altered many of the background conditions upon which the agreement depended. The challenge, therefore, is not whether the agreement remains important — clearly it does — but whether the institutions remain sufficiently adapted to contemporary realities.

From the perspective of international law, that should not be regarded as a radical proposition. In fact, there is a useful analogy from an unexpected source. In the whaling in the Antarctic case, the International Court of Justice considered whether states could continue to rely on methods that reflected the state of international legal knowledge at the time that a treaty was concluded, despite the existence of newer and more effective approaches. The court's answer was, in essence, "No": a treaty intended to endure could not simply ignore advances in knowledge, in international law or, indeed, in technology pertaining to whaling, that affected its operation.

Peace agreements are, obviously, different from environmental treaties, but the underlying principle is strikingly similar. If, in 2026, we know more about institutional resilience, implementation, deadlock prevention and peacebuilding than we knew in 1998, it would be surprising if we were not at least willing to consider that learning. That means not abandoning the agreement; it means taking it seriously enough to ensure that it continues to function.

That brings me to what I regard as the most important lesson from almost three decades of experience: the principal challenge may be not constitutional design but implementation. Twenty years ago, Kofi Annan observed that one of the most underestimated threats to peace settlements was not disagreement at the negotiating table but the gradual erosion of political will after an agreement had been reached. In many respects, that insight remains highly relevant today.

Across the world, peace agreements rarely fail because their objectives are unclear; rather, they struggle because implementation becomes uneven. Attention moves elsewhere, and institutions are left without effective mechanisms for review, monitoring and adaptation to circumstances that will undoubtedly have changed. Viewed from that perspective, some of the recurring difficulties experienced in Northern Ireland in the past quarter century are not unusual. They are challenges that many post-conflict societies encounter. The question is how to address them.

For that reason, if I were to suggest one area that deserves the Committee's particular attention, it would be institutional learning. How do institutions identify implementation failures? How do they recognise emerging pressures before those become crises? How do they adapt to changing circumstances while maintaining legitimacy across the community? Those questions are, ultimately, more important than any single reform proposal. Reform should be understood not as a departure from the Belfast/Good Friday Agreement but, rather, as part of the continuing responsibility to preserve it. Perhaps the most useful way of thinking about the agreement today is not as a completed settlement but as a constitutional peace process. The objective is not to revisit the peace settlement but to ensure that the settlement remains capable of producing peace.

Thank you, Chair. I look forward to your questions.

The Chairperson (Mr Buckley): Thank you very much, Ciarán; that is much appreciated.

Dr David Torrance (House of Commons): Thank you, Chair and members, for hearing from all of us. I work at the House of Commons Library, so my expertise is, generally speaking, in the devolved institutions, not just in Northern Ireland but in Scotland and Wales, and, to some degree, those of the United Kingdom Parliament. Therein, there are some interesting comparisons that I will come back to.

My paper is very much focused on the process around agreeing to institutional change in Northern Ireland. My analysis was based on three previous examples of that occurring: the St Andrews Agreement of 2006, which was the first substantial change to the institutions set up in the late 1990s; the New Decade, New Approach agreement of 2020; and, most recently, the 'Safeguarding the Union' Command Paper process of just a couple of years ago.

Basically, my argument is that there is a convention that governs the process of institutional change. To backtrack slightly, the Belfast/Good Friday agreement comprises two components: the multi-party agreement (MPA) between the UK Government, the Irish Government and the political parties in Northern Ireland at the time; and the British-Irish Agreement, which is the treaty aspect. My focus is much more on the MPA. Interestingly, that envisaged changes, as you can see if you go back to the original document, although it observed that any changes to the institutions would be for the two Governments to implement merely "in consultation" with political parties in Northern Ireland. The practice since that point has diverged from that and goes much beyond mere consultation.

What do I mean by a "convention"? In the 1950s, I think, Ivor Jennings identified how conventions could be identified by three limbs. First, do precedents exist? Secondly, do the actors in the process consider themselves bound by the convention? Thirdly, is there a reason for the rule? As I argue in the paper, all three limbs are met in the case of an institutional reform convention in Northern Ireland. There are obvious precedents, of which I have mentioned three; there is considerable evidence that the main actors in the process consider themselves bound by that convention; and the reason for the convention is fairly clear, which is that any sustainable institutional reform needs a large degree of consensus.

The Northern Ireland convention has three limbs. Chiefly, and most importantly, it needs agreement from the major political parties in the Northern Ireland Assembly — not all of them, but certainly the two largest as determined by the most recent election. Secondly, there needs to be a degree of buy-in, however tacit, from the Irish Government in Dublin. Again, looking back at the past three processes, it is clear that Dublin is kept very much in the loop and offers a degree of commentary through the process. Thirdly and finally, it needs agreement at Westminster.

There is a separate convention, which has slipped off the radar in the House of Commons, whereby, certainly since the 1970s, Northern Ireland matters have been treated in a non-partisan way, which means that there is generally agreement between the two main political parties in the House of Commons. Of course, generally speaking, institutional changes require an amendment to the Northern Ireland Act 1998, so that is important.

The broader point is that all institutional reform operates in a particular context. I looked at another paper that I produced for the House of Commons on the original Parliament in Northern Ireland in the early 1970s. The Government produced a Green Paper on Northern Ireland that, for the first time in half a century, mooted major institutional changes to the then Government and Parliament in Northern Ireland. Obviously, back then, the impetus for that came from outside. At that point, there was the looming threat of direct rule, and Brian Faulkner, the then Prime Minister of Northern Ireland, felt compelled to meet some of London's concerns in that respect.

In the House of Commons, the Modernisation Committee is looking at making changes to the House of Commons for the first time since the late 1990s. A recent analysis of previous attempts to reform that body suggested that reforms require a degree of consensus but stand the best chance of implementation when they are led by the governing party, which in this case is Labour. However, even there, consensus is important.

The Senedd — the Welsh Parliament — recently adapted to a new electoral system. That, too, followed a consensus from a cross-party independent commission on electoral reform. In a more minor way, and closer to home for me, in the Scottish Parliament, after decades of debate about electing Committee Convenors in that legislature, the change was finally implemented just a couple of weeks ago following elections to that body. Again, there had been a degree of consensus on that but also some reluctance from the main political parties to engage.

The meta-point, if you like, is that consensus is crucial. However, in the case of Northern Ireland, given the broad architecture of the Belfast/Good Friday Agreement, consensus between the major political parties represented in the Northern Ireland Assembly is the most crucial aspect of the convention.

The Chairperson (Mr Buckley): Thank you very much, David. I now invite Dr Patrick Graham, from the University of New England, who intends to speak remotely. I refer members to his paper in their information packs. I hand over to you, Patrick.

Dr Patrick Graham (University of New England): Thank you, Chair, and thank you very much to the entire Committee for the invitation to speak to you remotely today — very remotely, in my case, here in Sydney. Thanks very much as well to Laura Gourley for setting up the logistics for me. It is a privilege to speak to you all.

There is a growing eagerness to reform the way that power is administered in Northern Ireland. It is clear why that is the case. The institutions have spent a substantial amount of time — the better part of the past five years — either collapsed or, essentially, not functioning. I also support institutional reform, and some easy wins could be introduced. For example, having joint First Ministers; having them elected on a joint ticket by the Assembly, going back to the pre-St Andrew position; and having a supermajority procedure for electing the Speaker, which could, potentially, even be blind to designations.

However, I want to inject a little bit of caution into that discussion, particularly where it concerns the petition of concern and the process for nominating the First Minister and the deputy First Minister. In my paper, I focused on the principle of cross-community consent. That means, of course, that the big decisions in Northern Ireland cannot be carried by a bare majority alone. Change must happen together and collaboratively; it should be agreed between the communities and not be done over anyone's head. Those decisions need agreement across the political divide. The UK Supreme Court calls that a fundamental constitutional principle. It is a principle that separates the Good Friday/Belfast Agreement structures from the 50 years of majority rule that came before them in the first half century of Northern Ireland's existence. It also separates them from the noble, but ultimately doomed, Sunningdale experiment in 1974. The principle of cross-community consent is why our institutions exist in the first place.

There are two key reasons for caution. First, it is not entirely clear, to me at least, that politics has moved as far as some of the calls for reform might assume. We are well over a quarter of a century on from the Good Friday Agreement's signing, and approximately five in six voters still give their first preference to an avowedly nationalist or unionist candidate. The middle ground has grown significantly, no doubt, and the 2022 election could prove to have been a pivotal moment, but that middle ground is still a long way from being close in size to the two main political traditions.

Much of what is proposed may not necessarily be workable — at least as it is currently imagined. For example, the idea that the smaller party in the nationalist or unionist bloc might step in to take the First Minister or the deputy First Minister position over their larger rival may not fully capture the political cost of doing so. Similarly, the idea that one community might be locked out in favour of the centre ground — the so-called "others" — would raise immediate questions of legitimacy, at least as things currently stand. Even a two-thirds threshold might not prevent the largest parties from holding the veto in practice.

My second concern relates to the fact that significant changes to the principle in recent years have been delivered from well beyond the Northern Irish political parties. By that I mean, for example, the protocol's democratic consent vote, meaning that the petition of concern — the central cross-community safeguard — was simply switched off. That was done by ministerial regulation from London without the express agreement of the main Northern Irish parties. The courts upheld that when it was challenged but ultimately did so on quite a narrow technical reading of the words on the page, treating the statutory text as decisive and setting to one side the deeper constitutional values underpinning the Good Friday structures.

I fully understand why that route was chosen; it kept the show on the road, as it were. That precedent should cause us at least some pause for thought, however, because it means that a principle that we call "fundamental" can be reshaped by ministerial action at Westminster without consultation, or certainly without the agreement of and consensus among the Northern Irish parties, and, rather, validated after the fact by whichever interpretive method the courts may reach for on any given day.

Over the past 25 years, in interpreting disputes arising from the power-sharing structures, the courts have, using that interpretive method, swung between a strict reading of the text on the one hand and a broad reading of its purpose on the other. Both approaches have been used to narrow the principle of cross-community consent. I do not think that the meaning of the Good Friday settlement should depend on the changing whims of judicial method; were it so, nobody could be quite sure what the constitution requires from one case to the next.

I also want to throw it out quickly that we might rethink how we frame the principle of cross-community consent. Understandably, it is mostly thought of or treated as a brake, a blocking device or an obstacle to good government, and, undoubtedly, sometimes it is. However, it can be more than that; properly understood, it enlarges the Northern Irish people's power and their ability to govern themselves. It forces that breadth of agreement that supports and sustains progress and makes that sustainable. When, for example, policing and justice powers were devolved in 2010, which were some of the most sensitive powers imaginable, given Irish history, it was the requirement of cross-community consent that made that transfer durable.

It keeps those big, contested political questions where, in my view, they belong: with the electorate, not with the courts. Even collapses of government, uncomfortable and messy though they were to watch, have acted as a kind of safety valve, jolting otherwise impassive Governments into addressing real failures. For example, the 2017 impasse produced 'New Decade, New Approach', and the 2022 impasse produced the Windsor framework and 'Safeguarding the Union'. It is fair to say that, on each occasion, the parties that forced the issue were rewarded by their own voters. That is the democratic process working. It is slow and awkward and can be extremely frustrating, but progress was ultimately achieved and the electorate responded.

In summary, I accept entirely that significant change is needed to avoid further paralysis and breakdowns. However, as David said a moment ago, the reforms must achieve consensus. They must achieve and carry substantial support from our traditions, and they must be designed in a way that keeps further disputes outside of the courtroom as much as possible. If that happens, they can succeed as St Andrews and New Decade, New Approach did. They are imperfect, but they have great value. If they do not, they are unlikely to hold.

The Chairperson (Mr Buckley): Thank you very much, Dr Graham. I will now ask Dr Maria Xiouri from the University of Lincoln to speak. I refer members to her paper.

Dr Maria Xiouri (University of Lincoln): Good morning. First, I thank the members of the Committee for the invitation. My expertise lies in public international law and particularly the law of treaties. Therefore, the evidence that I can provide is on those aspects of the possible reform of the institutions that were created by the Good Friday Agreement and which relate to the law of treaties. I have written a relevant article, as has been mentioned, of which Dr Johnston was the guest editor.

The reason why the reform might have aspects related to the law of treaties is that, as was mentioned, the Good Friday Agreement consists of the bilateral treaty concluded between the UK and Ireland — the British-Irish Agreement — and a non-legally binding agreement between the UK Government, the Irish Government, political parties in Northern Ireland — the so-called multi-party agreement. Each of them is annexed to the other. Non-legally binding agreements, also called, among other things, "political agreements" or "gentlemen's agreements", are often used in relations between states and established commitments of a political or moral nature, but not legally binding obligations.

The multi-party agreement is structured around a three-stranded approach, comprising strand one, to do with the political institutions in Northern Ireland; strand two, which is the relationship between Northern Ireland and Ireland; and, as already known, strand three, which is the relationship between Ireland and Britain. Of course, the multi-party agreement also contains other important provisions. By contrast, the British-Irish Agreement contains four articles.

Therefore, it seems to constitute, to some extent, an umbrella agreement that includes fundamental points of the Belfast/Good Friday Agreement and is meant to be complemented by other instruments, particularly the multi-party agreement. Through interpretation of an article of the British-Irish Agreement — article 2 — I argue that, contrary to strand one institutions, the establishment of strand two and strand three institutions in accordance with the provisions of the multi-party agreement is a legally binding obligation. The Good Friday Agreement is therefore an important example of a non-legally binding agreement partly being given legal effect as a result of a reference in a treaty.

It is important to point out that the significant advantage of non-legally binding agreements, such as the multi-party agreement, is that they can be amended more easily than a treaty, the amendment of which may be subject to formalities. For that reason, the use of non-legally binding agreements may be preferred in cases where states will not be able to amend an agreement easily in response to new developments. In any event, an important question that arises about the relationship between the British-Irish agreement and the multi-party agreement is reform. What are the consequences of possible reform of the institutions created by the Good Friday Agreement through the British-Irish Agreement? Will an amendment of the British-Irish Agreement be required?

On that question, it could be argued, on the basis of interpretation of article 2 of the British-Irish Agreement, that the two states commit to implementing the multi-party agreement, as it may be reviewed after 1998. The British-Irish Agreement is quite broad — as I mentioned, it is an umbrella agreement — and is therefore flexible enough to accommodate changes to the multi-party agreement without the need for the British-Irish Agreement itself to be formally amended.

A related issue in that regard is whether there are limits to the extent to which the multi-party agreement can be reviewed. Is it the case that, if certain reforms of the multi-party agreement were contrary to the British-Irish Agreement, they could not be effected unless the British-Irish Agreement was amended? There is a relevant section in the multi-party agreement entitled, "Review procedures following implementation".

Paragraph 5 provides:

Paragraph 7 states:

An interpretation of that section reveals that certain changes to the multi-party agreement may require amendment of the British-Irish Agreement without, however, clarifying what those changes may be. It may be argued that an amendment of the British-Irish Agreement will be required if a change to the multi-party agreement is contrary to the provisions of the British-Irish Agreement. That would happen in the case of change to the institutions listed in article 2 of the British-Irish Agreement, namely the strand one institutions, which, as I mentioned, the UK and Ireland have a legally binding obligation to establish. Dr Johnston made that argument in her article.

Nonetheless, at the same time, the interpretation of that section does not seem to exclude the possibility that, under certain circumstances, an amendment of the British-Irish Agreement may also be required for the reform of the institutions under strand one, if such a reform is contrary to the British-Irish Agreement, which Dr Johnston also mentioned. However, it is possible that minor changes, particularly to single institutions, will not require an amendment of the British-Irish Agreement. Of course, determining what may be contrary to the provisions of the British-Irish Agreement will require an interpretation of its terms, in the context of "good faith" and in light of the objective purpose of the British-Irish Agreement.

In the case of the British-Irish Agreement, it could be argued that, apart from its preamble, which, among other things, confirms the commitment of the UK and Ireland to the principles of democracy, non-violence, partnership, equality, mutual respect and the protection of human rights, of particular importance is article 1, which, among other things, draws attention to the right of self-determination of the people of Ireland and affirms that:

Moreover, the "Declaration of Support", which constitutes some form of preamble, namely an introduction to the multi-party agreement, should, arguably, also be taken into account, in particular this provision:

Therefore, to give some examples, renaming the First Minister and the deputy First Minister as "joint First Ministers" would not need an amendment of the British-Irish Agreement. Enabling the Speaker of the Northern Ireland Assembly to be elected by a two-thirds supermajority would, arguably, also not need an amendment of the British-Irish Agreement.

In most cases, changes to the institutions provided under the multi-party agreement have been effective, although they are not legally binding. There have been other agreements between the two states, such as the 2014 Stormont House Agreement, the 2015 Fresh Start Agreement, and the New Decade, New Approach agreement in 2020. Therefore, it might be observed that, in those cases, reforms to the strand one institutions, provided for under the multi-party agreement, took place through changes to non-legally binding agreements, which meant that no amendment of the British-Irish Agreement was considered necessary. By contrast, the 2006 St Andrews Agreement, which is also a non-legally binding agreement, brought about reforms across all the strands and was incorporated in 2007 into the treaty between Ireland and the UK. Therefore, it was considered that in such a case an amendment of the British-Irish Agreement was required.

In conclusion, it seems that so far, reforms of certain significance to the institutions realised in article 2 of the British-Irish Agreement, namely strand one, have been affected by amendment of the British-Irish Agreement, while reforms to the strand one institutions have been effected mainly through other non-legally binding agreements of the participants in the multi-party agreement. It could be argued, however, that even the reforms to the strand one institutions might require an amendment to the British-Irish Agreement if it could be considered as contrary to its provisions as interpreted in light of its object and purpose and according to the principles of the [Inaudible.]

The Chairperson (Mr Buckley): Thank you very much, Maria. I invite Dr Katie Johnston to be the final speaker and close the opening remarks.

Dr Johnston: Thank you very much, Chair. I thank the Committee and the Clerks for the opportunity to be here this morning. My remarks will focus solely on the international legal obligations created by the 1998 agreement and its implications for the reform of the Assembly and the Executive. I stress that my evidence does not address any political or constitutional requirements for reform or those that may be imposed by domestic laws, as David discussed, nor does my evidence take any view on the desirability of any particular reforms. I will also build on and, to some extent, overlap with what both Ciarán and Maria have said, with which I largely agree.

In a nutshell, the international legal obligations created by the 1998 agreement do not prevent reform of the Assembly and the Executive, but they may impose requirements on the process that needs to be followed in some cases. In certain situations, the requirements may act as a limit on the changes that can be made by one state that is acting unilaterally. The 1998 Belfast/Good Friday Agreement is made up of two interconnected instruments: the multi-party political agreement and the British-Irish Agreement. The British-Irish Agreement is a bilateral treaty between the UK and the Republic of Ireland, and it was also concluded on 10 April 1998. It is registered with the United Nations and creates binding international legal obligations for both states.

On the question of reform, article 2 of the British-Irish Agreement is the most significant because it creates an obligation for the two states to support and, where appropriate, implement the provisions of the multi-party agreement. Essentially, and as Maria said, given that the international treaty obligations can only be modified in certain ways, in some cases requiring a formal amendment process, the question is this: if reforms are made to the Assembly and the Executive that depart from the 1998 text of the multi-party agreement, would they require the UK and Ireland to agree to modify the British-Irish Agreement to avoid the reform being contrary to the article 2 obligation?

The international law and treaty interpretation is highly technical, and I do not propose to go into detail on it now. Maria has already laid out some of those principles, with which I agree. I will, of course, be happy to follow up in writing if the Committee would like more detail on any points.

When applying the rules of treaty interpretation, and in particular taking into account the subsequent practice and agreements since 1998, my research reached two main conclusions on the reform of the Assembly and the Executive. First, some changes can be made to strand one without requiring an amendment to the British-Irish Agreement, or otherwise modified or reinterpreted. That is because, in effect, the article 2 obligation to support and implement the multi-party agreement is best interpreted as referring to the multi-party agreement as it is modified from time to time — there is an evolving meaning to the term.

In practical terms, it means that certain reforms to strand one, such as renaming the Offices of First Minister and deputy First Minister as "joint First Ministers", for example, could be made through UK domestic law alone, without the UK and Ireland needing to agree to amend the British-Irish Agreement.

The second conclusion of my research, however, is that there are situations in which change to the strand one institutions could require amendment of article 2 of the British-Irish Agreement. Making such changes unilaterally and without amendment to the British-Irish Agreement would put the UK in breach of its obligations under that treaty. Much would depend on an analysis of the particular reform proposal, but, broadly speaking, changes involving complete replacement or abandonment of the multi-party agreement or those that would disregard the object and purpose of the British-Irish Agreement or the context of the multi-party agreement as a whole would require the UK and Ireland to amend the treaty. To give an extreme example, were the UK to unilaterally abolish the devolved institutions in Northern Ireland, that would clearly breach article 2 of the British-Irish Agreement, and so breach international law. A more realistic but also more complex scenario could involve significant reform of the cross-community safeguards in the Assembly or Executive. Again, much would depend on the details of the particular reform proposals, but it is likely that a move to a completely voluntary coalition without any accompanying safeguards, such as cross-community or weighted majority voting, would require amendment of the British-Irish Agreement, and so could not be accomplished by the UK unilaterally without breaching the treaty.

Finally, it is important to note that amending or reinterpreting the British-Irish Agreement is not necessarily an onerous or time-consuming process. If that were required, it could consist of a simple agreement between the two Governments, for example, in the form of an exchange of notes or even a more informal interpretive agreement, so, even where treaty action is required, as I have argued, it certainly is not a serious obstacle to reform.

To conclude, my research shows that there is scope for certain reforms to be made to the Assembly and Executive without any action needing to be taken to amend or reinterpret the obligations under the British-Irish Agreement. However, other more significant changes might require action by the UK and Ireland to modify their obligations and could not be enacted unilaterally. As I said, those international legal obligations certainly do not act as an obstacle to reform, but they may impact the process by which such changes can be made.

The Chairperson (Mr Buckley): Thank you very much, Katie. I thank everybody who has taken part this morning. It is fascinating to hear all of your different perspectives and expertise on different aspects of how the agreement intersects with both law and convention when it comes to agreements. I start, Katie, with your interesting analysis and how it intersects with the law in general, particularly the multi-party agreement aspect, combined with the British-Irish Agreement and the obligations under article 2. Most of the speakers have emphasised the need for multi-party agreement in order for the British-Irish Agreement to be interpreted in such a way to make change in line with those agreements as they have evolved since the Belfast/Good Friday Agreement through to St Andrews and, indeed, other agreements. How could the British-Irish Agreement be amended if there was not consensus contained within — if there was no multi-party agreement as such, but there was a desire with no cross-party consent. How would that intersect with the particular aspects around the British-Irish Agreement?

Dr Johnston: Thank you for your question. As I said at the outset, my analysis is really confined to international law, and in particular the obligations under the British-Irish Agreement, so it does not say anything about the political or constitutional prior requirements that might be needed for a reform process that is both lawful and legitimate under domestic law and constitutional law and, indeed, politically. Under the British-Irish Agreement, that is purely a matter between the states. It does not foresee any roles for those political parties. At the same time, that is confined to that bilateral treaty. Some of the international instruments that Ciarán talked about may impose requirements for how inclusive the process of reform is, so there are wider international standards within which the treaty is located that may impose those requirements. The British-Irish Agreement is really just a matter between the two Governments.

The Chairperson (Mr Buckley): So, if there was no cross-party agreement in multi-party talks for significant change, would that make it more difficult for the British-Irish Agreement to be amended? My interpretation of what you said at the beginning was that it is, essentially, it is amended in line with the will of the multi-party arrangements and their desire for change in different areas.

Dr Johnston: If there were to be agreement in Northern Ireland among the parties, whatever form that took, according to the requirements, that would be automatically reflected in the treaty if the change relates to strand 1 and if it were within the limits that I set out.

The Chairperson (Mr Buckley): And if there is not agreement? Essentially, my question is whether there can be unilateral change via the British-Irish Agreement without that initial stage, which is multi-party agreement on strand one.

Dr Johnston: If we look at the practice since 1998, we see that all changes to strand one have been made, as David said, through cross-community consensus.

The Chairperson (Mr Buckley): David, you talked a lot about the process for agreement and the convention that exists. My interpretation of what you said — you can correct me — is that, essentially, the convention has been that we have multi-party agreement, followed by buy-in from Dublin and agreement by Westminster. Is that the process that was intended? If not, why has that become the mode of travel that is so often adopted when we come to different agreements?

Dr Torrance: That is an interesting question. Political conventions often just sort of arise circumstantially. In this case, it flowed from the run-up to the Belfast/Good Friday Agreement. In the 1990s, there was much talk, usually from British political actors such as Patrick Mayhew and John Major, of a triple-lock process for reaching agreement, which became the Belfast/Good Friday Agreement: you needed buy-in from the two Governments, the parties at Westminster and, crucially, the people of Northern Ireland, as broadly defined. Perhaps the convention since then, which I would argue exists, has reflected those components. It is recognised that any reform proposal that does not have major inter-party agreement and Dublin buy-in will not work. For example, if the Irish Government strongly attacked a reform proposal or indicated dissent — not that they have ever done so — that would make proceeding quite difficult.

The Chairperson (Mr Buckley): Obviously, that would make things difficult to proceed with. You talked about consensus being crucial and the need for agreement, particularly within the two largest political parties at present. Is it your view that it would not be feasible to implement significant reform in line with the process that you have outlined without agreement from the two largest parties?

Dr Torrance: The previous three examples that my analysis is based on indicate that that would not work. However, a slight check on that was the process that led to the Safeguarding the Union agreement, which strayed a bit from the convention in that it was majorly a conversation between the DUP and the UK Government. Before that agreement, there was tacit buy-in from the other parties that allowed the process to proceed, but, at first sight, it departs slightly from New Decade, New Approach and the St Andrews process. Again, however, that simply reflects the modus operandi of the whole Belfast/Good Friday Agreement: unless you have cross-community consent, important matters such as institutional reform are not going to proceed.

The Chairperson (Mr Buckley): That leads me to my final question, which is to Patrick. You majored on the the principle of cross-community consent. You talked about the democratic consent vote, which brought that particular piece of the agreement to the fore of debate again in the Northern Ireland Assembly. You also talked about concerns that you have in that respect. If that particular issue were to be unravelled, as such, in political calls for reform, is it your opinion that those calls for reform have probably underplayed just how crucial that particular aspect of the agreement is to the survival of agreement and government in Northern Ireland?

Dr Graham: Broadly, it was regrettable, if understandable. We got to a stage in the Brexit withdrawal process where the British Government ultimately chose a type of withdrawal from 2019 onwards where there would have to be —. It got to a stage where one community was going to be very unhappy with what happened, and we reached an impasse. That impasse was then thrown into the judicial process, which, again, in my view, was regrettable but understandable.

Chair, will you expand on or reiterate what you meant when you asked about the process unravelling?

The Chairperson (Mr Buckley): Pardon, Patrick?

Dr Graham: Your question was about what would happen if the process were to unravel, if I recall correctly.

The Chairperson (Mr Buckley): My question is this: do you believe that calls for reform that do not take into account the principle of cross-community consent are ultimately doomed to failure if they do not have buy-in from all major political parties?

Dr Graham: Yes, absolutely and unequivocally. Notwithstanding what Katie and Maria might say about the international aspects, when it comes to the strand one implications, the bounds of what is possible through reform are endless as long as there is consensus and cross-community support. Without that, reform is unsustainable and unviable. In response to your question, my answer is yes.

The Chairperson (Mr Buckley): Thank you. I will now go to questions from members.

Mr Sheehan: Thank you all for your presentations. There is a lot to absorb from all of them. The point has been made clearly that, without agreement, it is difficult to move forward towards any reform of the institutions. The work of this Committee is to discuss potential areas of reform that would bring to the institutions greater stability, greater accountability, more effective democracy and so on. The Committee will take evidence, as we are doing today, and, when we have heard all the evidence, we will draw up a report. It is unlikely, however, that we will get consensus on a final report, so the report will probably reflect that. The Committee Clerk can step in and correct me at any time if I am wrong, but the report will reflect the different views of Committee members or parties and so on. Given that that is how the Committee will do its work, the report will be debated in the Assembly and, presumably, forwarded to the two Governments for whatever work they can do in regard to it. It has come across clearly from the presentations that, if there is no consensus, it will be almost impossible to bring forward any reforms. Would any of you like to comment on that?

Dr Torrance: It is not impossible, because, as most of us have covered, it has happened in the past. In each case, there were long-drawn-out processes that involved a degree of inter-party negotiation. It may look impossible at this juncture, but, if you look at the past examples, you see that there have been significant institutional reforms that achieved major inter-party agreement. To a degree, it is more of a political question. It involves the major parties that are represented in the Northern Ireland Assembly reaching agreement. There does appear to be movement. I was talking to Katie beforehand about Sinn Féin's publication yesterday, which covers fairly major points to do with institutional reform. That is a change from what we have had previously, and I detect some signs of a growing appetite for institutional change, though it may not quite be there yet. The word "impossible" is too strong. It is possible based on previous examples, but that can take some time.

Mr Sheehan: Does anyone else want to make a contribution?

Professor Burke: To some extent, the possibility is, of course, a matter of political will and how close people are prepared to get to one another and how important they see reform as being. Taking an external perspective and looking at this comparatively, many modern peace settlements now contain formal review mechanisms that have the capacity to change the workings of those agreements over time. I can cite the example of the Dayton Accords in Bosnia and Herzegovina, South Africa's constitutional tradition or Colombia's peace process. The common lesson from those is that peace agreements increasingly operate as living constitutional frameworks that evolve over time, rather than one-off bargains. It could be said that that is best practice in peacebuilding. One thing that we have learned internationally is that successful peace settlements tend to require ongoing maintenance. The original agreement creates peace, but it is the subsequent decades, often including some reform, that determine whether peace becomes self-sustaining. Perhaps bearing that in mind might have the capacity to bring parties closer to consensus on the potential for reform.

Mr Sheehan: I suppose the evidence that you are giving is from a legal or constitutional basis, and we cannot look at those aspects in isolation, because what we are talking about here is politics and the political. According to many recent opinion polls, people are losing faith in the institutions here. One would think that, without some sort of reform, confidence in the institutions will continue to decrease and that, at some stage, one presumes, they will lose whatever legitimacy they previously had. In that sense, is it not incumbent on the co-guarantors of the agreements to take the lead? We know that, previously, when the two Governments have taken their eyes off the process here, it has gone into decline to a certain extent. There is the view of, "Everything is sorted there, so we do not need to bother ourselves any more", and difficulties have arisen as a result of that. Is it not the case now that there needs to be closer involvement of the two Governments and that, at some stage, if agreement cannot be got here among the parties, the Governments will have to move forward with some sort of institutional reform? Is that scenario even possible?

Dr Johnston: I agree. As a few of us have said, the article 2 obligation in the British-Irish Agreement is on the states, the UK and the Republic of Ireland, to support and, where appropriate, implement the multi-party agreement. I think that you could certainly interpret an obligation to support to include things like maintenance, ensuring it continues to function well. You could certainly get that from the article 2 obligation to support the multi-party agreement.

Dr Xiouri: This political agreement is outside my area of expertise, but, yes, there should be the greatest possible agreement among political parties in Northern Ireland, and obviously the views of the UK and Ireland will also be very important in that regard. Yes, the greatest possible agreement will be required for such changes.

Professor Burke: You have made a few points together. To some extent, this echoes what Katie said and the points that were made earlier. All durable peace settlements require sensitivity to perceptions of exclusion or non-consent across communities. One of the lessons of the agreement is that legitimacy cannot be sustained if significant sections of society believe that their concerns are not being heard. Your initial point, to some extent, outlined a lack of confidence or diminishing confidence in the agreement as it is now, on the basis of the practice of the institutions. In that context, it is worthwhile evaluating whether that is a lack of confidence in practice or in principle. My belief — again, I am not directly involved in the political institutions; you are obviously the expert here — is that there remains a certain degree of confidence in the principles that were agreed to, such as consent, power-sharing, parity of esteem and peaceful dispute resolution, but less confidence in how the institutions have been able to deliver them. To some extent, that very fact militates in favour of reform, so we are faced with the fundamental question as to whether, in 2026, the existing institutions represent the most effective means of delivering those core principles in modern circumstances.

Could the British and Irish Governments act without the buy-in of the political parties? Arguably, that would be possible. They are, perhaps, best placed to deliver a draft of a reform proposal that could be shared with the parties. Often, when an agreement is reached on these things, a lot of the ingredients of the first draft make it into the final draft. Perhaps their perceived neutrality or externality would make them ideally placed to start that process.

The Chairperson (Mr Buckley): Thank you. We have talked about the process of reform, and much of your contributions have focused on the legality of some aspects of reform. For the purposes of the discussion on the Belfast/Good Friday Agreement, is there a legal definition of multi-party agreement, or how is that interpreted?

Dr Johnston: Multi-party agreement is a provision in the terms of the British-Irish Agreement, and it refers to the political agreement that was reached on 10 April, which is then annexed to the British-Irish Agreement. It is referring to that document.

The Chairperson (Mr Buckley): The point is that parties may agree with various forms of reform. However, for the purposes of multi-party agreement, which intersects with the British-Irish Agreement and how it is interpreted, is there a legal definition of what that entails?

Dr Johnston: I am sorry; multi-party agreement as a standard?

The Chairperson (Mr Buckley): Yes.

Dr Johnston: Off the top of my head, I am not sure, but I will be happy to follow up on that in writing.

The Chairperson (Mr Buckley): That would be interesting, thank you.

Mrs Guy: Apologies for being late. When I came in, there was a bit of discussion around cross-community consent. In 1998, that very clearly involved two traditions. I think we would all agree that Northern Ireland has changed and is more diverse now. Certainly, politically, we are more a collection of minorities rather than two main communities. In that context, would it not be appropriate to revisit what cross-community consent looks like? At the moment, we operate cross-community votes etc by parallel consent. There are proposals to look at a weighted majority. Would a weighted majority achieve that? Is that a viable proposition to achieve what would be cross-community consent without using the existing mechanisms?

The Chairperson (Mr Buckley): Any particular member of the panel?

Mrs Guy: I believe that Patrick was speaking to that issue when I came in.

Dr Graham: That is a great question, but I do not have an answer to it. I would defer to Brendan O'Leary's standard, which is to say that once we get to around 20% of members reflecting non-aligned or other traditions — I do not like to use the term "other", but we know what it means when we say it. We are potentially getting to that point, but we are not quite there yet. When it comes to supermajorities voting on certain matters, it is important that a substantial element of the unionist and nationalist communities shows a degree of consensus in that way. As for what percentage or standard is required to progress a measure when the main nationalist or main unionist party is not on board, that is going to depend on the issue. Some elements would not be appropriate in that regard. In other areas, for example electing a Speaker, if there was buy-in from albeit a minority of one of the traditions or designations, that could go ahead. With regard to revisiting the fundamental principle around designation and a more inclusive approach to the other, as it were, that is going to play out over the next five or 10 years. It really depends on electoral success, and what Brendan O'Leary said on that matter is a fair standard.

Dr Torrance: The obvious point is that any change to the nature of cross-community consent would require cross-community consent. Therefore, it comes back to the political point on the degree to which there is willingness from the two communities, as currently defined and understood, to expand the definition of cross-community consent.

Mrs Guy: There is definitely a tension between the idea of cross-community consent and democracy. People like myself are excluded democratically from cross-community votes as a cross-community party. That has got to be a glaring anomaly that needs to be addressed.

Dr Torrance: I am afraid that that is beyond my pay grade.

Mrs Guy: A lot of the framing of these conversations is around the original intent of the Good Friday Agreement and what was drafted at the time. The petition of concern, for example, was intended as a safeguard for minorities but is applied as a veto. We have a petition of concern right now that is intended to override a democratic vote in that context. Again, people have talked about the confidence in our institutions. That erodes confidence like nothing else, when a majority of your democratic institutions are being told that we have a mechanism here that will apply if you do not get what you want. Can somebody comment on the necessity of that or the ability to change it?

Dr Johnston: I should emphasise again that I am confining myself to the limits of international law. Many aspects of the petition of concern, and the circumstances and conditions under which it can be employed, are not set out at all in the multi-party agreement, so this is a clear example of where change could be made to strand one through political agreement or UK domestic legislation. It certainly would not require treaty amendment. That is without prejudice to those other points. What keeps coming up is the question about whether consensus is required, what consensus is required and whether something can be done unilaterally by the UK Government. An implication of my and Maria's position is that, yes, those things could be done without consensus or unilaterally by the UK Government lawfully under the treaty. As others on the panel have shown, however, that is far from the end of the story. International law really sets the outer limits of what is acceptable. It is quite common in peace agreements to have a non-binding political agreement that is underwritten or guaranteed by hard international legal obligations. They are there as outer limits to provide reassurance to the parties.

Dr Torrance: The petition of concern was, of course, reformed as part of the New Decade, New Approach agreement, and that took legislative form at Westminster, so there are precedents for that to be changed, but it has to be through political inter-party agreement.

Dr Xiouri: You mentioned something in your previous question, in case it was addressed to both of us, in relation to the legal definition of non-legally binding agreements. Non-legally binding agreements do not create international legal obligations. Of course, international law deals with them. There are non-binding instruments that define non-political and non-legally binding agreements. There is also ongoing work by the International Law Commission of the United Nations that tries to define non-legally binding agreements and regulate, because questions are raised on the extent to which those types of agreements can be used.

There is also previous non-legally binding work by the Inter-American Juridical Committee that defines political commitments as:

That is a definition that exists in that regard. Of course, I do not know what kind of definition the UN International Law Commission will go for. Strictly speaking, legally, they are outside the area of international law and the law in general. A question arises about the relationship and whether they could have some type of legal effect. As I said, the UN International Law Commission is working on clarifying those issues. In principle, they do not create legally binding obligations.

The Chairperson (Mr Buckley): OK. I have no other indications from members. I thank Katie, Maria, David, Patrick and Ciarán for their presentations and written submissions. They are really helpful to the Committee in its work. Thank you very much.