On Monday 15 October 2012, the Prime Minister of the UK government and the First Minister of the Scottish Parliament, publicly signed in a formal ceremony a document entitled simply ‘Agreement between the United Kingdom Government and the Scottish Government on an Independence Referendum for Scotland, now known colloquially (in Scotland at least) as the ‘Edinburgh Agreement’ (for discussion of its content see Aileen McHarg’s earlier post). The signing ceremony placed emphasis on the agreement between the Scottish and UK governments as resting not in the meeting of minds of David Cameron and Alex Salmond, but on the existence of a text defining in legal terms their respective commitments (the Scottish government website refers to its signing as ‘the ratification’ of the agreement). But did this text have any legal status, and if not does it matter?
Why should anyone care?
It is unlikely that either side will want to breach the agreement, so why should we care whether it is legally binding or not? Well it is interesting, is it not, to know? Is the Edinburgh Agreement a legal agreement? Or just a political agreement? If a legal agreement, what sort of agreement is it, and how might it be enforced? If a political agreement, what are the social and political mechanisms for enforcement and do any legal methods of enforcing it remain? If it was not a legal agreement why did the parties almost seem to present it as such? Similar questions have been important in similar agreements elsewhere – the legal status of agreements between sovereign governments and sub-state actors present legal conundrums the world over, and for this reason alone the matter merits some thought.
From a more esoteric point of view, the legal status of the Edinburgh Agreement is important to questioning the predisposition of lawyers to view law as always available to those who wish to create binding obligations, and to view legal agreements as always more binding than non-legal alternatives. As I will argue, the Edinburgh Agreement challenges this legal perspective.
Finally, and more provocatively, if Scotland were to vote yes and separation of Scotland from the UK were negotiated, a whole host of agreements between the UK and Scottish government would require to reached, and the difficulty of finding a binding legal form for those agreements would be likely to come to the fore.
The structure of the agreement
The Edinburgh Agreement comprises a main agreement signed by David Cameron Prime Minister, and also Michael Moore, Secretary of State for Scotland, on behalf of the UK Government, and Alex Salmond, First Minister, and Nicola Sturgeon, Deputy First Minister, on behalf of the Scottish Government. This main agreement sets out in broad terms the principles to which the two governments have committed – it reads more like a ‘Declaration of Principles’ from which more detailed commitments flow (see the similar structure of Middle East Peace Accords!). To this front-piece agreement, a memorandum of agreement and a draft section 30 Order are attached, which the front-piece agreement states ‘form part of this agreement’. Technically speaking therefore, the ‘Edinburgh Agreement’ comprises all three documents.
The second part of the document, the Memorandum of Agreement, has clauses and tight drafting and looks much more ‘legal’ than the opening agreement. It sets out ‘elements of the agreement that require legislative provision in the section 30 Order’, but also ‘the elements that have been agreed between the governments on a non-statutory basis.’ The third document, the draft Section 30 Order is a draft piece of legislation (but of course must still pass through processes of affirmation by the UK and Scottish Parliaments to become law).
With this attention to detail, the pervasive legal language, a set of clear commitments, and the signature of government representatives, surely this is a legal document? It does not look like a mere ‘political pact’, and indeed one need only compare with the famous political pact – the Coalition Agreement - to see the differences. The Coalition Agreement was between two political parties rather than government representatives, and although it set out commitments to policies affecting the constitution, the content of the agreement was not itself a constitutional matter. The nature of the commitments to broad areas of common policy meant that, while coherently drafted, they were not written in the precise terms of legal obligation found in the Edinburgh Agreement – this was not the nature of the thing.
So if not a political pact, what is it? The difficulty of legal status is the difficulty in deciding what sort of legal document it is.
Is it a bird (that is: a treaty)?
The word ‘ratification’ and the ceremony attaching to the signing of the Edinburgh Agreement indicated a treaty-like process. This was an agreement between two heads of two governments, and their senior Ministers. However it was not of course a treaty in law because a treaty, according to the Vienna Convention on the Law of Treaties, ‘is an international agreement concluded by states’. International law adds that the parties must intend to create legal obligations on the international plane. In the case of the Edinburgh Agreement only one party was a sovereign state.
Was it nonetheless perhaps an international agreement of some sort? The Vienna Convention does make some provision for ‘international agreements concluded between states and other subjects of international law’. However, who constitutes a ‘subject of international law’ is not defined, and again both parties would need to have intended their agreement to be binding as regards international law, and it seems clear that this is not the case.
It is interesting to note that the question of the legal status of documents signed between the UK government and sub-bits of the jurisdiction is not an entirely new one. Similar issues arose around the project of de-colonization (to which no analogy of politics is being drawn – merely analogy of legal condundrum). In 1921, the document that confirmed the move towards Irish independence (although its direction was ambiguous at the time) was termed an ‘Anglo Irish Treaty’. However, even in pre-Vienna Convention days its status as treaty was not clear. The parties signed not as members of government but ‘on behalf of’ the British and Irish delegation respectively. The legal status issue is dealt with by the final clause which provides:
18. This instrument shall be submitted forthwith by His Majesty's Government for the approval of Parliament and by the Irish signatories to a meeting summoned for the purpose of the members elected to sit in the House of Commons of Southern Ireland, and if approved shall be ratified by the necessary legislation.
Similar issues of legal status arose in the later project of decolonization across the British Empire which saw the British Government signing decolonization agreements with (then) sub-state governments or groups of leaders, which had at the moment of signing no status as a new state, and whose agreements were therefore of unclear legal status. The countries were on track to independence leading to arguments that local signatories were not really sub-state entities with no treaty-making capacity, but signing as the ‘state in waiting’, and the resultant documents had some form of international legal status. This argument does not apply to the Edinburgh Agreement, where there is no UK acceptance of intention to act on the international plane or that relations being conducted in the domestic realm are in reality in transition to international relations. This lack of international intention also rules out theories that the UK has made a unilateral internationally binding agreement with the Scottish Government. Interestingly, however, were the referendum to succeed and separation be negotiated, these arguments might start to apply to agreements reached as to the modalities of separation between the two governments.
Is it a plane (that is: contract, or a domestic binding agreement of some other sort)?
Does the agreement have a contractual dimension? It is clear that it is not a contract between two private parties. But, similar to other concordats between the UK and devolved governments, it arguably has a quasi-contractual nature as setting out common understandings and reciprocal obligations with respect to the conducting of the referendum. Indeed, the Agreement is listed on the Scottish Government website in their ‘concordat’ section. (The final paragraph of the Memorandum of Agreement of the Edinburgh Agreement mentions that the ‘spirit’ is that of the Memorandum of Understanding already agreed between the two governments).
What is the Agreement’s legal status as a ‘concordat’? Well, notoriously unclear. Guidance published in 1998 and then in 2005 has suggested that concordats are not designed to create legal obligations or restrictions, but to act as the ground rules for administrative co-operation and exchange of information. The terms of the concordats often also express themselves to be ‘non-binding’ as statements of intent, ‘binding in honour only’, and are not intended to create contractual arrangements (see Condordats and Devolution Guidance Notes, House of Commons Library, SN/PC/3767, 7 October 2005). Nonetheless, commentators talk about these concordats in ‘twilight law terms’, as an area of ‘para constitutional law’, agreements between the governments which serve almost as a form of ‘quasi-contractual’ or ‘soft law’ for civil servants because they create expectations and guide conduct.
Some have gone further, suggesting that a breach of a concordat could become justiciable because the concordats create ‘legitimate expectations’ as to how governments will act on which one government could judicially review the other (see discussion in Devolution and Concordats, House of Commons Research Paper, 99/84, 19 October 1999, at pp 21 - 22).
It is difficult to imagine the facts in which a doctrine of legitimate expectations could play out with respect to the Edinburgh Agreement – even if it were legally sustainable (and there are differences in the doctrine in Scottish and English courts). If either side were to refuse to affirm the section 30 Order, for example, then the Agreement would effectively be rescinded and everyone would be back to the negotiating table. That would leave to a return to the status quo ante namely, a dispute over the capacity of the Scottish Government to hold a referendum. If outright breach is unlikely, what about, for example, foot dragging: what if implementation of the section 30 Order were just to be delayed? (See the Devolution Matters blog for an interesting discussion of timetables relating to the section 30 Order) It is difficult to see how foot-dragging could be argued to frustrate legitimate expectations, particularly as no fixed timetable is given for the passing of the order in the Agreement, unless this was so extensive that it pushed the commitment to a fixed timetable of the Agreement itself, namely to provide for a poll before ‘the end of 2014’. Again, this would require re-negotiation of the Agreement and a return to the status quo ante.
As regards other possible forms of breach, key commitments are locked-down by different forms of legal enforcement, meaning that enforcement of ‘legitimate expectations’ is not necessary. The Memorandum of Agreement commitment, for example, that the Scottish Government not hold another referendum on the same day (presumably designed to ensure that the Scottish government did not unilaterally introduce ‘a third question’ by this mechanism), is dealt with in the frame of the section 30 Order. Article 3 makes it clear that power is only devolved if one poll is held, meaning that introducing a new poll would lead to arguments that the entire process is outside the scope of the Scottish Parliament’s devolved power.
In other areas a deliberate choice seems to have been made to leave enforcement to the sphere of politics. For example, one role of the Electoral Commission is to examine the referendum question and issue a report to be laid before the Scottish Parliament. The Scottish Government ‘will respond to the report, indicating its response to any recommendations that the Electoral Commission may make’ (but note no pre-commitment requiring the Scottish Government to fully implement any recommendations of the Commission). Strictly speaking, therefore, the Agreement does not commit the Scottish Government to heeding the Commission’s advice. In fact, this is the same legal position as that which technically applies to the UK government as regards Electoral Commission advice. What then, is to stop the Scottish Government from ignoring this advice? The answer appears to be, ‘public opinion, and political debate’. The Agreement provides for the Scottish Government to produce a referendum Bill which is to be debated and passed by the Scottish Parliament. In answers to the Scottish Affairs Committee, the Secretary of State for Scotland, Michael Moore, noted that the Scottish Parliament had a key role to play in holding the Scottish government to its Edinburgh Agreement commitments, and in particular to Electoral Commission advice. Clearly, a majority government will be able to carry a vote. However, the Parliamentary process, in providing an opportunity for all parties to engage in public debate, gives it a role in holding the Scottish Government to account in terms of how far the legislation honours the commitments made in the Agreement.
If it is not legally binding, why did the parties therefore present it so formally as a ‘super-agreement’?
The role of ‘political enforcement’ may start to explain why the parties presented an Agreement in legal writing, legal form and with full signing ceremony. Paradoxically, the Agreement may have been presented so formally precisely because it had no binding legal status. The main mechanism for enforcing agreements of this type is through two dynamics: fashioning reciprocal agreements which it is in the parties’ self-interest to implement, and creating costs to a party’s reputation in the event of breach.
As with all agreements, a key mechanism for enforcement is that of reciprocity – I will abide by my side, if you abide by yours. Here the Agreement legalizes itself to the extent that it provides for commitments for most eventualities, with clear timetables, and fashioned in precise language.
As regards reputation, whichever party breaches this agreement will look bad, be seen as unreliable in the future, and will be less likely to have people sign credible agreements with it. These costs of breach are real for the UK Government, which holds itself out as a credible deal-maker on the international plane, but also as regards other devolved regions where stakes may be even higher (Northern Ireland). But they are also real for the Scottish Government, which seeks to present itself as a sovereign government in waiting to people who they seek to persuade to vote for independence, but also to external states and organizations (see EU debates). Reputation costs are heightened by having an agreement the commitments of which are clear and leave as little room as possible for ambiguity, and also by making them publicly to great fanfare. The public signing of the Agreement by two governments increases the reputation costs of breach but also extends those costs to future UK or Scottish governments should either of those governments change before all the commitments are implemented and seek to change tack (something that can never be completely ruled out). This dynamic of reciprocity and reputation perhaps also explains why a UK government that had something to lose from the ‘treaty’ symbolism of the joint governmental signing was happy to go along with it. Arguably it delivers its main side of the bargain – the section 30 Order – first and must then rely on trust for delivery of crucial Scottish Government commitments.
The Belfast Agreement (similarly lacking in any legal status) got through some of its darkest days because no-one wished to be seen as the party who breached first. Where parties did renege, they always tried to present their rejection of an obligation as an alternative reading of Agreement commitments rather than breach. Public debate and response acted as the court deciding whether they got away with this. Interestingly, however, eventually a court of law did find the Belfast Agreement to have a quasi constitutional status that affected the interpretation of its implementing legislation – the Northern Ireland Act 1998 where that legislation was ambiguous (Robinson v Secretary of State for Northern Ireland, 2002 UKHL 32). There seems little that is ambiguous about the draft section 30 Order, but it would be likely that any ambiguity would be resolved by recourse to the background Agreement text, as definitive of the parties’ intentions.
Dogs that do not bark
Interestingly one additional matter relating to the referendum is not specifically dealt with by the Agreement; namely the Agreement does not tightly specify what will follow from the referendum result. The Agreement talks of the parties agreeing that the referendum, ‘deliver a fair test and a decisive expression of the views of people in Scotland and a result that everyone will respect’. Paragraph 30 of the Memorandum of Agreement provides that ‘The two governments are committed to continue to work together constructively in the light of the outcome, whatever it is, in the best interests of the people of Scotland and of the rest of the United Kingdom.’ But even this does not clearly say what either government must do to respect the result.
What would be the obligations of the UK government in the event of a ‘yes’ vote? Historical statements, precedent, and the statements of the current UK government seem to make it clear that a yes vote will be implemented by the UK government and that this is not contentious. However, the exclusion of any firm commitment to implement the result should perhaps still be noted, particularly given that the only other legal provision for a secession referendum, in the case of Northern Ireland, provided for in the Northern Ireland Act 1998, does make clear legal provision for implementing the result in Westminster legislation. Even if implementation of a ‘yes’ vote is assured, does this mean that the UK government has to agree to independence in whatever shape the Scottish Government wants? In fact, the wording of paragraph 30 may indicate not. This wording reflects reluctance to spell out a pre-commitment to implementation of the referendum that might risk pre-empting a complex negotiation over what form independence would take: the UK government clearly do not want to pre-commit to matters such as the currency, the monarchy, Trident, etc (see further, responses of Michael Moore to the Scottish Affairs Committee). However, the ambiguity of this paragraph also begs the question of what compliance a commitment to ‘work together constructively in the light of the outcome’ imposes on an SNP-led Scottish Government. One can imagine the arguments that could arise with respect to how it articulates its commitment to the current devolution arrangements as a government in relation to its (presumably on-going) aspirations to independence of the SNP as a party.
Who cares anyway: is this not a ridiculously academic discussion?
Well yes: lawyers love to consider hypotheticals and it is 99.99% likely that this discussion is a purely hypothetical one. To be super-clear: I am not suggesting or trying to suggest that the Edinburgh Agreement is flimsy, likely to be dishonoured, or does not settle the questions it purports to settle. But while an academic discussion, perhaps it is not ridiculously so. Hypotheticals are not merely fanciful intellectual exercises; they assist us in understanding what the legal consequences of a factual situation are. And surely it is interesting to consider whether law matters or does not matter in the context of two governments wanting to commit to a common resolution of an issue of immense constitutional significance?
The discussion challenges our legal sense of the availability of law. The Edinburgh Agreement could not be signed in a legally binding form because no easy, appropriate legal form exists for this type of agreement in our legal system. This is not so important in Scotland where violence and instability is not an issue. But it has been very important in other contexts where negotiations between governments and sub-state actors (governments or would-be governments) take place against a backdrop of violence, and where central governments change mid-process between pro- and anti- peace agreement parties and may wish to renege on the commitments of their predecessors without reputational loss.
The lack of legal status also challenges ideas about the usefulness of law. No-one much cares whether the Edinburgh Agreement has legal status or not because it may simply not matter. The Agreement has perhaps all the ingredients for successful auto-enforcement: tight legal drafting, a culture of governments honouring formal commitments, and reciprocal self-interest locked down by the reputational costs of breach. It is difficult to see how the ‘added-value’ of a legal agreement - court enforcement - would help. These ingredients will either remain or will not: if the Agreement self-executes no court is necessary, while if a party for some presently unpredictable reason pushes a ‘destruct’ button, no court is sufficient.
Christine Bell is Professor of Constitutional Law at the University of Edinburgh