In Shakespeare’s Macbeth, the three witches greet Banquo with the observation that “Thou shalt get kings, though thou be none.” And so it proves. Not everyone, of course, can predict the future with such unerring accuracy. But businesses and governments alike have to be pretty good at predicting what might happen in the future. If they have concerns about the future, they are likely to take steps to deal with the eventualities. That may be by signing contracts or agreeing to binding international treaties which make provision for contingencies.
On one view, the EU-UK Withdrawal Agreement has done just that with its provisions on the Irish backstop issue. Article 1(4) of the Protocol confirms that the backstop arrangements are intended to be temporary. However it also contains terms which make tolerably clear that the Protocol shall continue to apply unless and until it is superseded by a subsequent agreement. Giving those words a natural and ordinary meaning, the backstop sets out a default position which shall continue to apply until such time as a successor agreement on those arrangements is reached.
What happens, however, if the parties are unable to reach a future agreement? Applying the wording of the Protocol, the backstop provisions would then continue for an indefinite period of time. It is therefore correct that this gives rise to a potential conflict between the intended temporary nature of the arrangements and what happens as a matter of practical reality. There is a degree of judgment as to when the arrangements might be considered to stop being “temporary”. “Indefinite” is a long time and, as John Maynard Keynes observed, in the long-run we are all dead. But repeat, failed negotiations to agree a successor agreement might suggest, at the very least, that a replacement agreement was not “on the cards.”
It is in those circumstances that some people, including Lord Pannick in his letter to The Times of 15 March 2019, have suggested that Article 62 of the Vienna Convention of 23 May 1969 on the Law of Treaties (‘VCLT’) might be invoked by the UK. It is being touted as a possible legal route which would permit unilateral withdrawal by the UK from the backstop arrangements. Article 62(1) VCLT provides:
“A fundamental change of circumstances which has occurred with regard to those existing at the time of the conclusion of a treaty, and which was not foreseen by the parties, may not be invoked as a ground for terminating or withdrawing from the treaty unless: (a) The existence of those circumstances constituted an essential basis of the consent of the parties to be bound by the treaty; and (b) The effect of the change is radically to transform the extent of obligations still to be performed under the treaty.”
Clearly, therefore, circumstances may so radically change that the underlying premise for the States parties’ agreement to an international treaty is nullified. The rug is pulled from beneath their feet.
Does this provision apply to the UK in its dealings with the European Union (‘EU’)? The VCLT makes clear that it does not apply to treaties between States (such as the UK) and international organisations (such as the EU), except where dealing with treaties which are the constituent instrument of that organisation or treaties which are adopted within the international organisation itself. See Articles 1, 3 and 5 of the VCLT. There is a separate convention dealing with relations with International Organisations: the Vienna Convention on the Law of Treaties between States and International Organizations or between International Organizations, 1986. But an insufficient number of States parties have ratified it for it to have been brought into force. The UK has ratified it, but the EU has neither ratified it nor formally confirmed it as an international organisation. Only 32 States have ratified it and it is a condition of its entry into force that 35 do so.
However, the Court of Justice of the European Union (‘CJEU’) in a series of cases has been prepared to apply some provisions of public international law as part of the EU’s internal legal order. In Case C-386/08 Brita  ECR I-1289, the CJEU at  held that “the rules laid down in the Vienna Convention apply to an agreement concluded between a State and an international organisation, such as the EC-Israel Association Agreement, in so far as the rules are an expression of general international customary law. Consequently, the EC-Israel Association Agreement must be interpreted in accordance with those rules.” At , the CJEU stated that:
“In addition, the Court has held that, even though the Vienna Convention does not bind either the Community or all its Member States, a series of provisions in that convention reflect the rules of customary international law which, as such, are binding upon the Community institutions and form part of the Community legal order (see, to that effect, Racke, paragraphs 24, 45 and 46; see, also, as regards the reference to the Vienna Convention for the purposes of the interpretation of association agreements concluded by the European Communities, Case C‑416/96 El-Yassini  ECR I‑1209, paragraph 47, and Case C‑268/99 Jany and Others  ECR I‑8615, paragraph 35 and the case-law cited).”
The Court went on to find that Articles 31 and 34 of the VCLT did codify customary principles of public international law and could be applied when reviewing the legality of the EC-Israel Association Agreement. The CJEU has conducted a similar exercise when construing the Fisheries Partnership Agreement between the EU and Morocco when finding that it could not lawfully be applied to the territory of the Western Sahara in Case C-266/16 Western Sahara Campaign UK  ECLI:EU:C:2018:118, CJEU.
The next question is accordingly whether or not Article 62 VCLT codifies a pre-existing peremptory norm of public international law (jus cogens). There is a good case for saying that it does. The International Court of Justice (‘ICJ’) in its Gabcikovo/Nagymaros Project judgment  ICJ Reports 7 (Judgment of 25 September 1997) cited with approval its judgment in the Fisheries Jurisdiction case,  ICJ Reports 63, , in support of the proposition that “Article 62 of the Vienna Convention on the Law of Treaties . . . may in many respects be considered as a codification of existing customary law on the subject of the termination of a treaty relationship on account of change of circumstances.” The learning suggests that the principle of rebus sic stantibus operates as an exception, in limited and exceptional circumstances, to the general principle of pacta sunt servanda.
However, the ICJ also went on in the same paragraph to state that:
“A fundamental change of circumstances must have been unforeseen; the existence of the circumstances at the time of the Treaty’s conclusion must have constituted an essential basis of the consent of the parties to be bound by the Treaty. The negative and conditional wording of Article 62 of the Vienna Convention on the Law of Treaties is a clear indication moreover that the stability of treaty relations requires that the plea of fundamental change of circumstances be applied only in exceptional cases.”
How are those principles to be applied in the context of the Irish backstop? The terms of Article 62 VCLT will have to be interpreted in good faith in accordance with their ordinary meaning, as well as in their context and in the light of their object and purpose: see Article 31 VCLT and the CJEU judgment in Brita (supra) at .
Can it really be suggested that the UK and the EU have not foreseen the possibility of not being able to reach agreement on the seemingly intractable issue of a soft Irish border outside the confines of a customs union? It is what they have each spent months discussing. A number of different solutions have been proposed during the course of negotiations. But no concrete and practicable solution has been endorsed by both sides as part of the Withdrawal Agreement. Instead, they have put the backstop arrangements in place until a better solution can be found.
True it is that the UK sought to negotiate for a defined time limit for the ongoing operation of the backstop arrangements. But the EU, reflecting no doubt the wishes of Ireland as a continuing Member State of the EU, declined to agree to make the backstop arrangements subject to any defined cut-off point. In those circumstances, can it realistically be said that the parties have not foreseen the possibility of no better solution being found? It seems to me to be clear that they have foreseen that risk and addressed it by the provisions which require the backstop arrangements to be maintained in force until a better solution has been found. They have foreseen the risk, but put the default provisions in place which will continue to operate unless replaced. In the same way that a bank may charge a default rate of interest on borrowings, unless and until a fixed rate is agreed.
In those circumstances, it seems difficult to suggest that the UK could, in good faith, choose to invoke the terms of Article 62 and unilaterally withdraw from the backstop arrangements. If, as is the case, the problem has been foreseen, good faith negotiation would require the UK and the EU as parties to the Withdrawal Agreement to negotiate for an amendment of that treaty to deal with the anticipated impasse. It is not presently known, of course, what the particular sticking point may be. But that is very far from concluding that negotiating deadlock would be an unforeseen change in circumstances. It must surely be incumbent on the UK to seek to amend the default positions which apply in the event of no further agreement, rather than to exercise a unilateral right of withdrawal which could not be justified under Article 62 VCLT.