Verdict of the “Star Chamber” under Sir William Cash MP on the UK-EU Political Agreement

13 March 2019

  • Sir William Cash MP
  • Suella Braverman MP
  • Robert Courts MP
  • Nigel Dodds MP
  • David Jones MP
  • Dominic Raab MP
  • Michael Tomlinson MP
  • Dr TD Grant
  • Martin Howe QC
  • Barnabas Reynolds

Summary of our Conclusions

1. Yesterday’s documents considered individually and collectively do not deliver “legally binding changes” to the Withdrawal Agreement (“WA”) or to the Northern Ireland/Ireland backstop Protocol (“the Protocol”). They fail to fulfil the commitment made by government to the House in response to the Brady amendment “to obtain legally binding changes to the withdrawal agreement”.

2. They do not provide any exit mechanism from the Protocol which is under the UK’s control. Any exit by the UK from the Protocol cannot take place without the agreement of the EU and therefore the position remains as set out in paras 14-16 of the Attorney General’s advice dated 13 November 2018 that “the Protocol will endure indefinitely until a superseding agreement takes its place”, and that the WA “cannot provide a legal means of compelling the EU to conclude such an agreement”.

3. The suggestion that “bad faith” by the EU could provide a legal route for the UK out of the Protocol is not credible in practice within any determinate or reasonable timeframe. The AG’s advice at para 29 was that demonstrably bad-faith conduct on the part of the EU “would be highly unlikely; all they would have to do to show good faith would be to consider the UK’s proposals, even if they ultimately rejected them.” The threshold for demonstrating bad faith before an international tribunal is very high, and nothing in the documents make this a credible possibility.

4. The UK could not unilaterally disapply the Protocol by alleging bad faith, but would be bound to submit the dispute to arbitration under Part 6 of the WA, and would need a prior finding by the panel of breach on the part of the EU in order to invoke the right under Art.178(2) of the WA to suspend (not terminate) provisions of the WA or Protocol. Any arbitration would be at best a lengthy and uncertain procedure which under Art.174 requires a reference to the ECJ of any questions of EU law involved. Even if the arbitration panel found in favour of the UK, para 14 of the Joint Instrument confirms that it would not enable the UK to exit the backstop.

5. The Attorney General’s further advice today (12 March 2019) indicates at para 17 that there is a “reduced risk” of the UK being trapped in the Protocol but this is caveated by the words “at least in so far as that situation had been brought about by the bad faith or want of best endeavours of the EU.” We consider that the prospects of such findings against the EU are remote, and note that at para 10 the AG only goes so far as to say that “it is arguable” that the UK could secure termination of relevant obligations under the Protocol. Such faint and remote prospects of escaping from the Protocol do not materially change the position the UK would find itself in if it were to ratify the WA. We agree with the AG’s final para 19 that “the legal risk remains unchanged that if through no such demonstrable failure of either party, but simply because of intractable differences, that situation does arise, the United Kingdom would have, at least while the fundamental circumstances remained the same, no internationally lawful means of exiting the Protocol’s arrangements, save by agreement.”

Our supporting reasoning

We have been asked to address issues arising from the latest combined UK-EU withdrawal documentation, comprising the 26 November 2018 draft Withdrawal Agreement with its “backstop” Protocol and the Political Declaration, and to assess the impact (if any) of the three subsequent documents produced on 11 March 2019.

We shall address the implications of this combined package of documents in the context of the “Brady” amendment to a motion of the House of Commons on 29 January 2019 that “requires the Northern Ireland backstop to be replaced with alternative arrangements to avoid a hard border; supports leaving the European Union with a deal and would therefore support the Withdrawal Agreement subject to this change”.

In particular we shall consider the degree of certainty we believe the latest package of documents achieves in ensuring the Northern Ireland backstop will be replaced with alternative arrangements, either before or after it comes into force.

The problem to be addressed

1. Introduction

On 29 January 2019, the House of Commons passed the ‘Brady’ amendment. It inserted the following text into the House’s motion: “and requires the Northern Ireland backstop to be replaced with alternative arrangements to avoid a hard border; supports leaving the European Union with a deal
and would therefore support the Withdrawal Agreement subject to this change.”

In responding to the House’s approval of the Brady amendment, the Prime Minister said: “We will now take this mandate forward and seek to obtain legally binding changes to the withdrawal agreement that deal with concerns on the backstop while guaranteeing no return to a hard border between Northern Ireland and Ireland.”

Our task is therefore to analyse the subsequent arrangements negotiated between the UK and the EU, and (1) to assess whether and if so to what extent they amount to “legally binding changes”, and (2) to consider whether they deal with the House’s concerns about the backstop as expressed in the Brady amendment.

2. Replace, or change, the backstop?

The Brady amendment refers to the backstop being “replaced”, while the Prime Minister’s response refers to unspecified “changes to” the withdrawal agreement. The backstop (formal title, the Protocol on Ireland/Northern Ireland) is physically attached to the text of the Withdrawal Agreement (“WA”) and, by virtue of Art.182 WA, “shall form an integral part of this Agreement”. It follows that on the wording as it stands, the Protocol stands equally with the main WA text as part of the legally binding international treaty which would be concluded if the WA is approved and ratified.

The terms of the Brady amendment would be satisfied by a change to the WA which consists of removing the current Protocol and replacing it with text setting out alternative arrangements. Since this is not being done, it needs to be considered whether any changes to the WA will achieve by alternative means the effect intended by the House in the Brady amendment.

A change to the WA which gives the UK the legal right to prevent the backstop Protocol coming into effect would achieve the same effect as the Brady amendment, since the UK could then effectively insist on alternative arrangements regarding the NI border.

A change to the WA which gives the UK a legal right to exit the Protocol after a period of time is not equivalent to the Brady amendment, in view of the negative effects on the UK’s constitutional integrity and on its international trade policy of being trapped in the backstop for that period of time. On the other hand, a lock-in to the backstop for a limited period is clearly better than the indefinite lock-in which could happen under the current text. However, a limited period is not
being stipulated or even attempted by the latest arrangements.

3. The central problem: indefinite lock-in if there is a stalemate in negotiations

According to Art.184 of the WA and Art.2(1) of the Protocol, the EU and the UK will use their “best endeavours” to conclude an agreement on the future relationship between the UK and the EU before the end of the transition period. If such an agreement is concluded by then, it is envisaged that it would supersede the Protocol and prevent its “backstop” provisions coming into effect. Further, if the Protocol does come into effect, there is a review mechanism under Art.20 of the Protocol by which the UK/EU Joint Committee can decide that the Protocol is no longer necessary if alternative arrangements are put in place.

However, each of these routes out of the backstop Protocol requires the agreement of the EU. A superseding agreement obviously requires the EU to agree it. The review mechanism requires that the EU and UK should “decide jointly” that the Protocol should no longer apply. This means that the EU member of the Joint Committee has a veto.

The Protocol itself in Art.1(4) and its recitals says that it is “intended to apply only temporarily”. But the problem is, despite this expression of intention and the expression of the parties’ “best endeavours” to reach a replacement agreement, if the parties fail to agree then the backstop Protocol would come into force and will endure indefinitely.

The Attorney-General’s letter of advice to the Prime Minister dated 13 November 2018 is attached for reference, as well as his latest letter of 12 March 2019. In paras 12-16, headed “The indefinite nature of the Protocol”, he explains with clarity why, despite the references within the text to its intended temporary nature and to the intention of the parties to supersede it with another agreement, it will kick in and then stay in force unless and until there is a joint agreement to supersede it. He concludes at 16:

“16. It is difficult to conclude otherwise than that the Protocol is intended to subsist even when negotiations have clearly broken down. The ordinary meaning of the provisions set out above and considered in their context allows no obvious room for the termination of the Protocol, save by the achievement of an agreement fulfilling the same objectives. Therefore, despite statements in the Protocol that it is not intended to be permanent, and the clear intention of the parties that it should be replaced by alternative, permanent arrangements, in international law the Protocol would endure indefinitely until a superseding agreement took its place, in whole or in part, as set out therein. Further, the Withdrawal Agreement cannot provide a legal means of compelling the EU to conclude such an agreement.” [emphasis in original text]

After considering the review mechanism in the Protocol (now in Art.20 but in Art.19 of the draft at the date of his letter), the Attorney-General reaches the following conclusion:

“30. In conclusion, the current drafting of the Protocol, including Article 19 [now 20], does not provide for a mechanism that is likely to enable the UK lawfully to exit the UK wide customs union without a subsequent agreement. This remains the case even if parties are still negotiating many years later, and even if the parties believe that talks have clearly broken down and there is no prospect of a future relationship agreement. The resolution of such a stalemate would have to be political.”

4. Consequences of “lock in”

The existence of the Protocol and the fact that it will automatically come into force in default of an alternative agreement being reached with the EU will profoundly affect the negotiating position of the UK. This is because the prospect of being locked in to the Protocol if agreement is not reached will impel the UK to make concessions in its negotiations with the EU in order to avert the adverse consequences of failing to reach agreement of the Protocol then coming into force at the end of the transition period. Thus the Protocol with its lock-in risk will gravely undermine the negotiating position of the UK as compared with a scenario where the Protocol were replaced in accordance with the Brady amendment or the UK had a unilateral right to withdraw from it. This is because the consequences of the Protocol coming into force are asymmetric in the extreme.

For the UK, a failure to conclude an alternative agreement would result in the whole UK being locked into a customs union with the EU with no say on external tariff policy or on the EU’s external trade agreements which the UK would be obliged to implement. In addition, the whole would be subject to a range of “level playing field” obligations. A particularly damaging obligation is to have to apply EU State aid law as it dynamically evolves, while having no political representation on the EU Commission and without the political influence to prevent these wide-ranging and highly discretionary rules being deployed in a way which damages the UK economy.

(The State aid rules would be applied directly by the Commission in Northern Ireland and under Commission supervision in the rest of the UK.).

In addition, the application of the Protocol regime to Northern Ireland would violate the principle of democratic consent (which is enshrined in the Belfast Agreement) and have damaging constitutional and political consequences, as we outline in a later section.

From the EU27 perspective, the consequences of the Protocol coming into force are very different. The Protocol would guarantee continued tariff-free access for EU27 producers of goods into the UK market. In addition, and by contrast with the scenario if the UK were to conclude a Free Trade Agreement with the EU and operate an independent global trade policy, the Protocol guarantees that the UK must maintain the tariff barriers mandated by the EU’s Common External
Tariff against competing third country imports.

Therefore the Protocol locks in the EU27’s large surplus in trade in goods with the UK, which has grown to £95bn in 2017 from rough balance in 1998. In addition, since EU27 goods exports to the UK are heavily concentrated in high tariff sectors, the Protocol guarantees that EU27 producers will continue to be able to export their goods to the UK at above prevailing world prices behind the shield of the Common External Tariff.

Given this apparently strong economic incentive for the EU to push the UK into the backstop Protocol and keep it there, the adequacy of the legal arrangements has to be assessed on the assumption that the EU may use all the levers that are legally open to them (and without needing to engage in bad faith) either to push the UK into the Protocol, or to force the UK in lieu into an agreement which is similarly beneficial to the EU and as detrimental to the UK as the Protocol. Unless the UK has a legally secure route out of the deadlock problem, its negotiating position will be fatally undermined.

Finally, it is far from clear that the UK would escape from the backstop Protocol even by concluding a replacement long term trade agreement with the EU. Assuming that such an agreement would have a conventional termination clause of say 12 months’ notice, we see nothing that prevents the EU from requiring that the UK agree that the Protocol or something like it should come back into effect in the event of the long term relationship agreement being terminated. Therefore if the Protocol is once agreed without a legally enforceable exit route for the UK, there is a very high risk that it will become a permanent liability of the UK.

5. Binding nature of the Protocol under international law

A number of suggestions have been in circulation that the UK might be able to escape from the Protocol in the future despite the lock-in problem identified in the Attorney-General’s advice. These suggestions are (1) that the UK could leave the Protocol in the future “because one Parliament cannot bind its successor”, (2) that the UK could simply breach and repudiate the Protocol and take the consequences, and (3) the UK might be able to escape from the Protocol because of various arguments arising under the Vienna Convention on the Law of Treaties (VCLT).

We note that none of these points appears in the Attorney-General’s letter of advice, which concludes that the only way out of the Protocol is political – i.e. ultimately by agreeing terms which are acceptable to the EU. We infer that the Attorney and the Government’s other legal advisers did not (rightly in our view) regard any of these points as providing a secure or credible route of escape from the backstop, in the face of the EU’s refusal to agree terms within the treaty which would give the UK that right.

(1) “One Parliament cannot bind its successor”

This suggestion is based on a misunderstanding of the relationship between domestic law and international law. The doctrine that one Parliament cannot bind its successor is a principle of the UK’s internal constitutional law. As a matter of internal law, an Act passed by one Parliament cannot prevent the next Parliament from repealing it.

However this doctrine has no application to international law obligations under treaties. Under international law, States are bound by treaties which they conclude regardless of changes of government or legislature. Even if there is a revolution and the whole constitution is overthrown, the State continues to be bound by treaties validly concluded by the previous regime.

Although the backstop Protocol is very unusual or probably unique in treaties of this kind in having no conventional termination clause, if it is concluded under the authority of the present Parliament it will continue to bind the UK under future Parliaments.

(2) “Just breach the Protocol and face the consequences”

The second suggestion simply to breach the treaty faces severe problems and should be regarded as unrealistic in view of the legal, constitutional and practical problems that would arise. The UK
has a good reputation for adherence to treaty obligations. That reputation is important for our credibility, particularly in entering into new trade agreements, and it would be gravely damaged by repudiating a lawful treaty. In addition, a treaty breach could give rise to retaliatory action by the EU which would be rendered lawful by a UK treaty breach.

Title III of Part 8 of the WA empowers an arbitration panel to rule whether a party has breached the WA and, if so, what reparation is due. Unlike more limited international dispute settlement procedures seen in other treaties, it gives an arbitration panel on-going oversight of its rulings, including the power to levy lump sums or penalty payments if the UK failed to comply with a ruling (Art.178(1)), and it would give the other party (the EU if the UK were to repudiate the Protocol) the right to suspend the observance of its own obligations—that is to say, effectively to offset the injury by means of self-help (Art.178(2)).

The exercise of a perfectly lawful right to terminate the EU treaties under Art.50 has engendered great opposition. A proposal to repudiate a legally binding treaty in breach of international law would be likely to generate opposition which is even greater, and those who think that such a course of action is a credible option need to consider how realistic is it that they would overcome such opposition.

The difficulties of breaching the Protocol at the international level are increased by Art.4 of the WA, which requires that the provisions of the WA (including the Protocol) and of EU law which are applied to the UK under the WA must be given “direct effect” and must have supremacy over conflicting UK law. It requires in terms that UK courts must have the power “to disapply inconsistent or incompatible domestic provisions”.

The Bill which implements the WA will, in order to implement Art.4, have to contain clauses which produce the same effect as subsections 2(1) and 2(4) of the European Communities Act 1972. It has been well known from the Factortame case onwards that those provisions empower and require UK domestic courts to disapply even Acts of Parliament if they are incompatible with EU law.

In addition, preliminary references to the ECJ will continue across the board under the WA in respect of the transition period and will continue indefinitely in respect of Northern Ireland under the backstop Protocol. Therefore those who contemplate a treaty breach as a way out the Protocol need to factor in a likely scenario in which any legislation purporting to authorise such a breach would be challenged in the courts and possibly referred to the ECJ for a binding ruling, as well as or in the alternative to being submitted to an arbitration panel under Part 8 of the WA. Both the ECJ and the arbitration panel have similar powers to order reparation and lump sums or penalty payments for breach of their rulings. Payment of such sums would be a legal obligation under international law.

(3) Vienna Convention* arguments

Art.60(3)(b) VCLT recognises the right of one party to a bilateral treaty to terminate or suspend its operation if there is a breach by the other party of a provision of the treaty which is a repudiation of the treaty or a “violation of a provision essential to the accomplishment of the object or purpose of the treaty”. It has been suggested that if a replacement agreement for the backstop Protocol cannot be negotiated in a reasonable period, then that would be evidence of a breach by the EU of its obligation to use best endeavours in good faith to negotiate such an agreement.

This argument faces a number of severe problems. First, it would not be enough for the UK to assert that there was such a breach – the UK would have to persuade an arbitration panel to make a finding that there was a breach by the EU, or risk a finding by the panel that the UK was in breach if the UK chose to act unilaterally. In coming to its decision, that panel would be obliged under Art.274 of the WA to refer any questions of EU law to the ECJ for decision and would be bound by the ECJ’s ruling. So, for example, if the EU were to say that alternative arrangements suggested by the UK were unsuitable because they would conflict with EU law or were not compatible with EU single market rules, it would be the ECJ rather than the independent panel which would be the effective decision maker. As we explain below, the obligation of good faith itself could be seen to be a matter of EU law whose meaning requires to be determined by the ECJ.

(The VCLT is probably, as a strictly formal matter, inapplicable to the WA, which is an agreement between the UK and the EU, a non-State entity. However, the VCLT is generally regarded as codifying pre-existing customary international law and hence the same result is likely whether or not the VCLT as such applies.)

Secondly, the obligation to negotiate in good faith rests not just on the EU, but on the EU and UK together. If negotiations break down, it cannot be assumed that one party, the EU, is the one at fault. It could equally be the fault of the UK. An obligation to negotiate in good faith is not the same as an obligation to reach an agreement, and does not require the negotiating party to set aside its own fundamental interests in order to do so. Most likely, if an agreement is not reached, it would be classed as a case where, as contemplated by the Attorney-General in paragraph 13 of his letter of 13 November 2018, “parties, pursuing their best endeavours in good faith, are simply unable to agree a superseding agreement within a reasonable time, or indeed at all”.

Thirdly, the argument that a State or international body has not acted in good faith is inherently a difficult one, “which would require clear and convincing evidence of improper motive and wilful intransigence” (again quoting the Attorney-General at paragraph 29).

A second argument is that the UK has an implied right of withdrawal from the backstop Protocol under Art.56(1)(b) VCLT. This Article however only applies to a treaty which “contains no provision regarding its termination” and does not provide for withdrawal. However, Art.20 of the backstop Protocol could be said to contain an express provision for its termination via the bilateral review mechanism (which gives the EU a veto), and the EU would certainly argue that Art.54 VCLT rather than Art.56 applies to it. Under Art.54, withdrawal may take place only with the consent of the parties, or in accordance with the provisions of the treaty – in this case, via the review mechanism. The EU would also argue that in any event the backstop Protocol is not an instrument which by its nature implies a right of denunciation or withdrawal, in view of their understanding of its relevance to the Irish border and the Belfast Agreement. So the argument that Art.56 VCLT is applicable cannot, in the face of the EU’s likely counter arguments, provide to Parliament any assurance that this could provide a unilateral route out of the backstop.

Thirdly, suggestions have been made that it might be possible for the UK to escape from the backstop in the event of a negotiating deadlock because that would be a “fundamental change of circumstances” under Art.62 VCLT. However, for Art.62 to apply, the change of circumstances has to be “not foreseen by the parties” when the treaty is concluded, and plainly the deadlock problem has been foreseen. Further, it is doubtful that the deadlock problem is a “change of circumstances” at all, since these are normally taken to be changes of circumstances external to the treaty and not just matters which arise from the operation of the agreed treaty terms. Moreover, the EU might well say that the Protocol is an agreement relevant to an international boundary such that, as reflected in Art.62(2)(a) VCLT, the UK may not invoke fundamental change of circumstances as a ground for termination or withdrawal.

In conclusion, none of the Vienna Convention arguments provide a robust or probably even credible alternative to securing changes to the treaty which would secure the UK’s right to escape from the Protocol.

6. Constitutional impact of the Protocol within Northern Ireland

The December 2017 Joint Report of the negotiators included the following paragraph, which complemented the preceding para 49 on alignment of customs and single market rules:

“50. In the absence of agreed solutions, as set out in the previous paragraph, the United Kingdom will ensure that no new regulatory barriers develop between Northern Ireland and the rest of the United Kingdom, unless, consistent with the 1998 Agreement, the Northern Ireland Executive and Assembly agree that distinct arrangements are appropriate for Northern Ireland. In all circumstances, the United Kingdom will continue to ensure the same unfettered access for Northern Ireland’s businesses to the whole of the United Kingdom internal market.”

Under the backstop Protocol, regulatory barriers will be created between Great Britain and Northern Ireland in important respects. Although GB’s external tariffs will be aligned with EU tariffs and there should no tariffs on goods passing between GB and NI, GB and NI will be under different regimes for customs administrative purposes and this means (as pointed out in the Attorney-General’s letter at para 7), customs declarations will be required for goods passing from GB to NI.

Furthermore, the rules of the EU single market on goods and related areas will continue to apply within Northern Ireland. This means the goods from GB which do not comply with EU single market rules, as they stand now or as they are altered by the EU in future, will be prohibited from entering NI and administrative controls will be needed to enforce these restrictions.

We note that if the WA is concluded, both the above restrictions will be imposed by the UK Act of Parliament which implements the WA without the consent of the Northern Ireland Executive or Assembly, either now or in the future.

In addition, the Protocol applies a huge range of EU single market regulations and directives to Northern Ireland which are listed out over 68 pages in Annex 5 (the 68 pages are the list of the titles of these rules, not the rules themselves). These rules are applied to Northern Ireland (but not Great Britain) by Arts.6(2) and 10 of the Protocol.

Broadly speaking these are all the rules of the single market relating to the placing of goods on the market and the processes and procedures they must undergo before being placed on the market, and rules relating to movement of goods (such as rules on live animal exports), but also include connected matters such as EU legislation on intellectual property (Annex 5, para 45), and an eclectic collection of additional legislation such as, for example, Council Regulation (EC) No 2182/2004 of 6 December 2004 concerning medals and tokens similar to euro coins (para 47).

In addition, the EU’s legislation relating to VAT and excise duties will continue to apply in Northern Ireland but not in the rest of the UK (Art.9 and Annex 6). This means for example that Northern Ireland would be restricted from introducing a lower VAT rate to encourage tourism, even if such a rate were brought in elsewhere in the UK.

Under the Protocol, the single market and tax rules must apply in Northern Ireland very much in the same way as EU law now applies in the UK as a Member State. As explained above, Art.4 of the WA requires the UK to legislate to make these provisions of EU law supreme over UK law (including Acts of Parliament) in UK courts. The supervisory powers of the EU Commission and other institutions and the jurisdiction of the ECJ (both in direct actions and in preliminary references) will continue to apply in Northern Ireland, just as if it were still part of an EU Member State: see Protocol Art.14(4).

These provisions mean that the UK Supreme Court will plainly not be supreme when it comes to dealing with these areas of law within Northern Ireland. It will be required to refer any issues of EU law to the ECJ and will be bound by the ECJ’s ruling.

And the Protocol obliges the UK to apply all these legal rules not as they stand today but as they are amended or replaced from time to time by the EU (Art.15(3)). The same applies as regards judgments of the ECJ: the UK will be bound by future judgments of the ECJ when interpreting these EU rules, as well as by judgments given before we leave (Art.15(4)).

These provisions mean that the people of Northern Ireland will be subjected to binding changes in the laws which apply to them, with no possibility of democratic consent either via the Westminster Parliament or via the Northern Ireland Assembly. This imposition of laws on Northern Ireland without the consent of the people of Northern Ireland either to the laws themselves or to the process by which they will be imposed on Northern Ireland represents a clear breach of the principle of consent enshrined in the Belfast Agreement.

Where the EU introduces a new directive or regulation in one of the fields where EU laws apply within NI, the EU will notify the Joint Committee under Art.15(5) of the Protocol. The Joint Committee can then take a decision to add the new piece of legislation to the Protocol and it will then take effect within Northern Ireland.

In theory, the UK could veto the adoption of the new act because decisions of the Joint Committee have to be taken by consensus of both the UK and EU representatives. However, if this is done, the EU then becomes entitled to “take appropriate remedial measures”. This is very similar to the mechanism which applies to the EEA states under the EEA Agreement, and the threat of “remedial measures” by the EU has in practice induced the EEA States to adopt all new EU legislative acts: after a period of resistance, Norway was compelled to abandon its opposition to the Postal Services Directive in 2014.

So, although there is a theoretical right for new EU laws not to apply within Northern Ireland under the Protocol, in practice both Northern Ireland and potentially the UK as a whole would be subject to the threat of retaliatory action by the EU.

An important aspect of our consideration of the proposals to be negotiated with the EU is how far they go to remedy the non-compliance with paragraph 50 of the December 2017 Joint Report and the breaches of the principles of democratic consent under the Protocol. It seems to us that the documents negotiated on 11 March 2019 do nothing to address the issue of the need for democratic consent of the people of Northern Ireland to the EU laws that will apply to them under the Protocol.

Our assessment of the 11 March 2019 documents

On 11th March the Government published a “Political Agreement” with the EU comprising three new documents additional to the draft Withdrawal Agreement and Political Declaration of 26 November 2018: (a) a joint instrument relating to the draft Withdrawal Agreement; (b) a unilateral declaration by the UK in relation to the operation of the Northern Ireland Protocol; and (c) a joint statement supplementing the Political Declaration.

Each is claimed to be intended to provide greater legal certainty that the UK is not bound permanently by the Northern Ireland backstop, nor trapped in a customs union with the EU. None of these documents seeks to address any other concerns with the Withdrawal Agreement and Political Declaration.

Basic problem with the New Documents

We consider that the New Documents are largely focussed on dealing with the wrong alleged problem, which is that of the EU employing ‘bad faith’ in order to trap the UK in the backstop. In seeking to address this alleged problem, the New Documents entirely fail to deal with the real problem, which is that of the UK being indefinitely trapped in the backstop because the EU acting in good faith – or at least not in a way which can credibly be proved to amount to bad faith – fails to offer terms which are acceptable to the UK and the negotiations then break down.

This problem – “good faith deadlock” – is the real risk both in our own view as set out in section 5(3) above and is also the risk identified in the Attorney-General’s letter of 13 November 2018. His letter of 12 March 2019 reaffirms the existence of that same risk in his concluding paragraph 19:

“19. However, the legal risk remains unchanged that if through no such demonstrable failure of either party, but simply because of intractable differences, that situation does arise, the United Kingdom would have, at least while the fundamental circumstances remained the same, no internationally lawful means of exiting the Protocol’s arrangements, save by agreement.”

Again as we point out in section 5(3) above, the EU is entitled to use the negotiating leverage available to it to advance its own interests without that being in bad faith or in breach of best endeavours. The criteria for a review under Art.20 of the Protocol on superseding the backstop are extremely broad and vague, being that any replacement arrangements must “address the unique circumstances on the island of Ireland, maintain the necessary conditions for continued North-South cooperation, avoid a hard border and protect the 1998 Agreement in all its dimensions.” (Art.1(3) of Protocol).

These criteria involve wide questions of political judgement, quite apart from and on top of operational judgements of, for example, whether or not proposed replacement customs procedures would be effective in safeguarding the EU single market from entry of non-conforming or non-tariff bearing goods. This means that a refusal by the EU to agree replacing the backstop could be based on wide grounds involving subjective judgement which it would be quite impossible to demonstrate were being advanced by the EU in bad faith.

The New Documents do not even address this real problem. Instead, they seek to divert attention from that problem by elaborately purporting to strengthen a potential “bad faith” argument.

Legal Status and Effect of the New Documents

(a) Joint instrument relating to the draft Withdrawal Agreement (the “Joint Instrument”)

Purpose and Extent of Joint Instrument

This instrument is a joint document which is intended to interpret the WA and Protocol. It is possible for parties to treaties to enter into supplemental documents which interpret treaties in particular by providing a clearer meaning for treaty provisions which are vague or unclear. However, such interpretative instruments cannot alter or contradict the underlying treaty, or create new substantive provisions . The document explains its interpretative purpose in the last paragraph on page 1 which concludes that “it has legal force and a binding character.” However, it is vital to appreciate that this phrase is preceded by the words “To this effect”, which refer back to the interpretative nature of the document.

Accordingly, it has limited binding character, only to the extent of interpreting and not changing or contradicting the WA or Protocol. The statement made in the Prime Minister’s letter to Members of Parliament of 12 March 2018 (“The Choice Today”) that “We have secured legally binding changes, with comparable legal weight to the Withdrawal Agreement” is manifestly incorrect in two regards. First, neither this instrument nor the other New Documents produce “legally binding changes” to the WA or Protocol; secondly, an interpretative instrument is clearly not “of comparable weight” to the treaty which it interprets.

Provisions of the Joint Instrument

Para 4 records that the parties “consider that, for example, a systematic refusal to take into consideration adverse proposals or interests, would be incompatible with their obligations” of good faith to replace the Protocol. This merely reiterates the position under international law and was fully taken into account in the AG’s advice of 13 November 2018 when he concluded that it was very unlikely that the EU would put itself in breach of its good faith obligations. We consider this point further in the next section.

Para 7 records an agreement to establish a negotiating track “for replacing the customs and regulatory alignment in goods elements of the Protocol with alternative
arrangements”. This does not expand the UK’s rights beyond the existing Political Declaration, which states that such alternative arrangements will be “considered”. This means that the EU can quite lawfully and in good faith “consider” such arrangements and find them wanting. It does not even contain an “in principle” acceptance of such alternative arrangements by the EU. Once we had ratified the WA and Protocol, the legal position would be that the UK would be entirely dependent on the goodwill of the EU to approve any such arrangements and their refusal to approve them would in practice be likely to be unchallengeable by any legal process. It should also be noted from the footnote that such arrangements would not replace Articles 11 onwards of the Protocol, which include State aid control by the Commission (Art.12).

None of the other provisions of the Joint Instrument in our view produces any material change in the impact of the WA or the Protocol. It largely consists of selectively reiterating points which are within the text or are obvious from the text, rather than resolving any ambiguities.

(b) Unilateral declaration by the UK in relation to the operation of the Northern Ireland Protocol (the “Unilateral Declaration”)

Role and Application of Unilateral Declaration (“UD”)

We consider this “Unilateral Declaration” to be entirely worthless. There are, putting it at the lowest, considerable doubts as to whether and when a unilateral statement by one party to a treaty can affect its meaning if that statement is not acknowledged or adopted by the other treaty party: it could be regarded as just “the UK talking to itself”. But even assuming in its favour that this document could in principle produce some legal effect, analysis of its content shows that it does nothing but reiterate the existing position under the WA.

The second paragraph states that “the objective of the Withdrawal Agreement is not to establish a permanent relationship”, but notably precedes that with the words “subject to Article 1(4) of the Protocol”. Article 1(4) of the Protocol contains the following last sentence: “The provisions of this Protocol shall apply unless and until they are superseded, in whole or in part, by a subsequent agreement.” This paragraph of the UD therefore simply re-iterates what is already in Art.1(4) of the Protocol but in a misleading way in which the last sentence, which formed the bedrock of the Attorney General’s advice, is omitted but is replaced with a reference which produces the same legal effect as if that sentence had been starkly re-iterated in the text of the UD. We cannot understand what diplomatic purpose is served by this paragraph, and must regrettably conclude that it has been couched in a way which will mislead an unwary reader.

The third paragraph of the UD purports to record the UK government’s understanding of the position if it is not possible for the parties to conclude an agreement which replaces  the Protocol in whole or in part, “due to a breach of Article 5 by the Union”. Article 5 WA is the obligation of good faith. Therefore this paragraph is limited to circumstances where the EU has acted in bad faith, and has no application to the far more likely circumstance in which the parties simply cannot reach agreement and the UK is locked in indefinitely. As regards the “bad faith” scenario, the paragraph then merely reiterates the remedies available to the UK in such circumstances without expanding them. As we have already pointed out, the UK would need to establish bad faith in front of an arbitral tribunal, which would be a Herculean task rendered more difficult and longer by the probable need for a reference by the arbitrators of points of EU law to the ECJ under Art.174 of the WA. It is strongly arguable that the meaning of “good faith” and “best endeavours” are points of EU law in the context of the WA and Protocol, in which case the ECJ would need to rule on the extent to which those concepts under EU law might differ from their meaning under international law. Even if the UK were successful in establishing bad faith, remedies under the WA disputes process would be limited to partial and “proportionate” suspension of parts of the treaty obligations, and would not amount to an exit from the backstop Protocol. We expand on this point below.

We can see no legal or diplomatic purpose that is served by the inclusion of this Unilateral Declaration. Regrettably we conclude that it has been generated for home consumption and its purpose is to obscure rather than elucidate the true position under the WA and Protocol.

(c) Joint statement supplementing the Political Declaration (the “Joint Statement”)

The Joint Statement is a non-binding statement of intent in connection with the non-
binding Political Declaration.

Possible Registration with UN

It has been suggested that the Joint Statement be registered with the United Nations. The implication appears to be that such a registration will establish that the Joint Statement is a legal binding instrument and not a mere political declaration. Article 102(1) of the UN Charter provides that “[e]very treaty and every international agreement entered into by any Member of the United Nations… shall as soon as possible be registered with the Secretariat and published by it.” However, neither an omission to register an instrument nor the fulfilment of the registration requirement is conclusive as to the character of the instrument as legally binding or otherwise. To determine the character of an instrument, one must refer to its terms. It is the terms, not the formalities of registration (or their
omission), that matter. This point is reflected in international practice, for example in the 1994 judgment of the International Court of Justice in Qatar v. Bahrain, ICJ Rep. 1994 at pp. 121-22 (paras. 27-29); and in UK practice under the Budapest Memorandum concerning Ukraine in 2014.

Ability to Withdraw from the Northern Ireland Backstop

Meaning of Best Endeavours

In the Joint Instrument the parties seek to clarify the meaning of “best endeavours” in their obligation to conclude, by 31 December 2020, an agreement which supersedes the Protocol in whole or part. As noted above, it is unclear whether the term “best endeavours” is already or may become a matter of Union law whose meaning would be determined under the Withdrawal Agreement by the ECJ. If it were so, the effect of the clarifications in the Joint Instrument would need to be seen in that context. Whilst they may be given some form of effect by the ECJ that effect could, in the experience of the UK with that court, be highly uncertain and may not be in accordance with the expectations of a UK reader.

The comfort sought to be provided by the Joint Instrument takes examples often at extremes, which are unlikely to be applicable. For instance, it says “a systematic refusal to take into consideration adverse proposals or interests, would be incompatible with their obligations under Article 2(1) of the Protocol [“best endeavours” to conclude an agreement superseding the Protocol in whole or part] and Article 5 of the Withdrawal Agreement [obligations of “good faith” and to take appropriate steps to fulfil the agreement, without prejudice to the (uncertain) application of Union law.]” (Para 4, Joint Instrument). It is almost inconceivable that the UK would be able to demonstrate a systematic refusal to take into consideration adverse interests incompatible with the parties’ best endeavours obligations, with all the uncertainties that concept imports.

The extreme improbability of an arbitral finding of a breach of good faith is well illustrated by the International Court of Justice case Whaling in the Antarctic (Australia v. Japan). Japan was permitted by the Whaling Convention to hunt whales for “scientific purposes” (Art. VIII), but was hunting them and selling the meat at a fish market. The ICJ would not hold Japan to be acting in bad faith.

There are statements made about the swift commencement of negotiations (para 5), speedy work (para 6), a negotiating track (para 7), a high level conference (para 8) and extraordinary high level conferences (para 9). None of these have much legal import, as paragraph 8 indicates when it states that the UK may request an extension of the transitional period.

The interpretation of “best endeavours” has received little attention from the ECJ, rendering it an uncertain concept that is open to all manner of interpretation.

Protection of UK Territorial Integrity

The Joint Instrument notes it intends to protect the territorial integrity of the UK (para 2), but the provisions of the Protocol entail significant encroachments on UK territorial integrity, with a commercial border across the Irish Sea.

The Protocol covers numerous areas, including the rights of individuals, a single travel area, a single customs territory, technical assessments, registrations,  certificates, approvals and so on, VAT and excise, agriculture and the environment, single electricity market and state aid. The provisions on state aid, for instance, significantly hamper UK competitiveness. They are to be given a dynamic meaning determined from time to time by the ECJ, and state aid is to be controlled either directly by the European Commission (Northern Ireland) or indirectly (Great Britain).

This means the Withdrawal Agreement package comprises a semi-permanent trade agreement, with many concessions already having been made in favour of the EU. Similarly, the EU’s much prized provisions on geographic indicators apply until such time as a free trade agreement is entered into between the UK and EU. Until that time this key point of leverage is conceded to the EU.

The Joint Instrument seeks to provide some comfort that the Protocol cannot be applied indefinitely (para 12). However, this comfort is limited. The assertion is that the good faith etc obligations under Article 5 of the Withdrawal Agreement and the best endeavours obligation under Article 2(1) of the Protocol should be interpreted such that it would be “inconsistent with their obligations” if either party acts “with the objective of applying the Protocol indefinitely”. It is wholly implausible that evidence would come available that the EU would “act” with such a demonstrable objective. If disputed the matter goes to the dispute resolution mechanisms, but as above it is quite possible that the determining law would be said to be Union law and reference is made as to its application to the ECJ. It is also to be noted that the dispute resolution mechanics are to be slowed down by a process of prior consultations with the Joint Committee (para 13). It is only when the arbitration panel makes a ruling and there is “persistent failure by a party to comply” with that ruling, “and thus persistent failure by that party to return to compliance with its obligations under the Withdrawal Agreement” that remedies – and only temporary remedies at that – are to be provided. Ultimately after this process, the UK would have the right to enact a “unilateral, proportionate suspension of its obligations under the Withdrawal Agreement” other than the provisions on the rights of former EU citizens “unless and until the [EU] has taken the necessary measures to comply with the ruling of the arbitration panel.” Such a step may expose the UK to damages claims from commercial parties under international arbitration.

The Unilateral Declaration seeks to add weight to the conclusion that the Withdrawal Agreement as a whole will not be indefinite. However, as set out above, this is a document refers to but does not set out Article 1(4) of the Protocol, which states that “the provisions of this Protocol shall apply unless and until they are superseded, in whole or in part, by a subsequent agreement.” The Unilateral Declaration therefore has no impact in respect of the principal matter at hand, that is
the permanency of application of the Protocol for Northern Ireland.

The Unilateral Declaration then seeks to record the UK’s understanding of the best endeavours obligation in Article 2(1) of the Protocol – if it is “not possible” for the parties to conclude an agreement superseding the Protocol in whole or in part. If but only if that situation was brought about by bad faith on the part of the EU, the UK states its belief that it can instigate measures “that could ultimately lead to disapplication of obligations under the Protocol, in accordance with” Part 6 Title III of the Withdrawal Agreement or Article 20 of the Protocol. However:

  • The subsequent agreement could only replace the Protocol in part and still satisfy Art.2
  • There is no time period for the disapplication
  • The UK only states that “obligations” under the Protocol will be disapplied – not all obligations, appearing to affirm that some of the Protocol could endure indefinitely
  • The UK reaffirms (unnecessarily, if the whole Protocol were to fall away) its new obligations to the EU, entered into pursuant to the Withdrawal Agreement, to “uphold its obligations under the 1998 Agreement in all its dimensions and under all circumstances and to avoid a hard border on the island or Ireland”. The 1998 Agreement is therefore then buttressed, for ever, by obligations to the EU, potentially including obligations to persons across the UK, EU and beyond pursuant to Article 4 of the Withdrawal Agreement. This is inappropriate as the scope of the obligations under the 1998 Agreement are relevant only to the UK and the Republic of Ireland.
  • When the declaration states that Article 1(4) of the Protocol makes clear a permanent relationship under the Protocol is not the parties’ intention, whilst that is true for the first sentence of Article 1(4), as already discussed it is far from true for the third sentence of that same Article. This states “[t]he provisions of this Protocol shall apply unless and until they are superseded, in whole or in part, by subsequent agreement.” Quite apart from the fact that such an inconsistency illustrates the dangerous paucity of drafting of the Withdrawal Agreement, such that the UK is not in a position to be certain what it has agreed to, the Attorney General himself placed weight on this provision in determining that the backstop was permanent in his letter to the Prime Minister of 13 November 2018.

The drafting of this supposed clarification hardly gives confidence as to the conviction as to the outcome ostensibly desired.

Political Declaration

The Joint Statement seeks to reassert the link between the Withdrawal Agreement and the Political Declaration (para 2). It states a shared ambition to have a future relationship in place by the end of the transition period (para 3). Notably, it does not indicate what that future relationship might be. Many in the UK regard it as a free trade agreement with collaboration on security and other such matters. However, there are aspects of the Political Declaration which are seriously damaging for future UK interests and where therefore an enhanced linkage between the WA and the PD would damage UK interests. Particular points of concern are:

Para 23 on tariffs. This is incompatible with the future conclusion of a conventional Canada-style Free Trade Agreement between the UK and the EU, because of its reference to a single customs territory and to the absence of origin controls. The EU could use this paragraph as a lever to deny the UK the right to negotiate an FTA to replace the backstop.

Para 124 on disputes and enforcement. This states that the long term agreement should contain arrangements for disputes and enforcement based on the WA; hence the role of the ECJ in Art.174 of the WA would become a permanent feature of the future relationship between the UK and the EU. Because the UK has sold the pass on this issue in the PD, the EU could insist on the UK having to succumb to this perpetual ECJ jurisdiction as part of the price of being let out of the Protocol. Indeed, the EU would undoubtedly argue that the UK is in breach of its own “best endeavours in good faith” obligation if it fails to submit to such a clause.

Para 75 on fishing. This states that the parties “should establish a new fisheries agreement on, inter alia, access to waters and quota shares.” The EU could insist on the UK succumbing to what is in substance the continuation of the Common Fisheries Policy, failing which the EU would be entitled to refuse to let the UK out of the backstop Protocol.

The remainder of the Joint Statement is a statement of warm intent, which is not justiciable or binding. Most notably this includes the self-described commitment to work “at speed” on a subsequent agreement that establishes by 31 December 2020 alternative arrangements to the Northern Ireland backstop. Various further aspirations as to how this work will proceed are then set out in paragraph 6, but these are just that – and non-binding.

Annex: Summary of Key Sovereignty Issues with the Withdrawal Agreement and Political Declaration

Legal Status

The Agreement, once entered into, cannot be exited except by subsequent agreement with the EU. Although Parliament cannot bind itself within the UK’s territory, the UK can bind itself internationally by Treaty, ceding governance to the Treaty’s provisions and those empowered to interpret and apply them – in many instances the ECJ.


The Agreement provides for an uncapped and uncertain payment liability for the UK to contribute to EU liabilities, determined ultimately by the ECJ.

Citizens Rights

Ex-EU citizens in the UK continue to have EU rights, interpreted by the ECJ. These may diverge from the UK’s future choices of entitlements for all UK citizens, which would create two classes of UK citizens.

The Agreement gives numerous direct entitlements to citizens, corporates and others across the EU, which they can claim under the wording of the Agreement, as interpreted by the ECJ through the EU doctrine of direct effect (in Article 4). The precise extent and content of this commitment is extremely unclear and is in the hands of the ECJ. It overrides any conflicting UK legislation and is to be implemented directly into UK law and enforced by UK administrative bodies and Courts.

Transitional Period

There is a transitional period, extendable until an unspecified date, under which the EU’s laws apply but without governance rights for the UK. There is no provision adjusting eg State aid provisions so as clearly to allow the UK to protect its service industries from a potential no-deal outcome after that period. By contrast, some member states appear already to believe they can protect their industries from such a no deal outcome.

The Backstop

The Agreement provides for a rudimentary trade deal on goods, leaving Northern Ireland in elements of the single market and customs union unless agreed otherwise, and applying some of these provisions to mainland Great Britain. There is no binding trade deal on the services sector, which represents 80% of the economy.

Unless a replacement trade deal is agreed:

  1. The Northern Ireland single market and customs union laws in goods, with partial application to mainland Great Britain, will apply indefinitely. Certain checks will be required across the Irish Sea, eg customs declarations for GB goods going to NI. Goods that do not meet EU requirements cannot enter NI. NI will be economically part of the EU’s single market in goods.
  2. EU State aid law will apply in the UK, as interpreted by the ECJ from time to time and administered directly (for NI) or indirectly (for GB) by the European Commission. This already governs UK tax policy and could prevent deregulation. It could be expanded further by the ECJ.
  3. Existing EU laws on social and environmental policy will apply, regardless of any changes in circumstances or subsequent political wishes.
  4. The EU’s much-prized protections on geographical indicators – ie product descriptions for EU goods such as champagne etc – continue to apply.

There is no enforceable requirement for the EU to agree a reasonable trade deal replacing these provisions.

Who Decides What The Agreement Means?

The ECJ is deferred to in interpreting many aspects of the Withdrawal Agreement, directly or indirectly.

The drafting is extremely unclear in many areas, leading to uncertainty and unpredictability as to what the UK is agreeing to, and allowing for significant ECJ discretion. This is inconsistent with normal Treaty practice where an independent arbitral body holds the parties to their word.

The ECJ is a politicised court which interprets and makes law according to perceived EU purpose underpinning the provisions. Its judgments can be highly unpredictable. For instance it interpreted the original “free movement of workers” as free movement of all people. It has also marginalised agreed legal text, introduced in 1993 partly at the behest of the UK, intended to devolve decision-making back to the member states from the EU (through the doctrine of