Part III - Interpreting the Agreements

6 Interpreting the Protocol

Stephen Weatherill

6.1 Introduction

This chapter provides an overview of what the Protocol does and why it does it. This is a more challenging task than one might initially imagine because the Protocol is written with what one might generously describe as calculated ambiguity, or, less charitably, outright evasion. What the Protocol does is not what it says.

Elucidating its true meaning requires a journey through four particular matters: (i) identifying the customs territory to which Northern Ireland belongs, (ii) the effect on West to East (Northern Ireland to GB) trade, (iii) the effect on East to West (GB to Northern Ireland) trade, and (iv) state aid. Each of these issues is considered in greater detail in subsequent chapters. The argument of this chapter is that the approach to drafting the provisions of the Protocol is the same: in each case the Protocol is written in a way that understates the nature and the extent of the commitments made on the UK side.

The Protocol’s resolution of all four issues has come under pressure as particular affected parties resist its full implications. Helpful adjustments have been made. The Protocol has been supplemented, first, by the Trade and Cooperation Agreement (TCA) which, by providing for tariff-free trade between the UK and the EU, reduces, but does not eliminate, the need for obstacles to intra-UK trade caused by divergent tariffs; second, by decisions taken by the Joint Committee (JC); and third, by adjustments in the form of unilateral declarations, some temporary, some permanent.¹ Life under the Protocol has, however, been regrettably infected by a persisting disregard for and misrepresentation of its terms by members of the UK government.

6.2 Content of the Protocol

The Protocol locks Northern Ireland (but not GB) into regulatory alignment with an extensive body of EU rules governing manufactured and agricultural goods. The detail is found in Annex 2 to the Protocol, to which deceptively brief reference is made in Protocol Article 5(4). Two hundred and eighty-seven EU legislative instruments are listed in Annex 2. The list is not static² and it may be amended by the JC.³ Northern Ireland–EU alignment is extended by the Protocol also to cover trade rules concerning the EU’s customs regime,⁴ value added tax (VAT) and excise rules,⁵ the single electricity market,⁶ and state aid rules in respect of measures which affect the trade between Northern Ireland and the EU which is subject to the Protocol.⁷ The Protocol also touches on particular aspects of individual rights to equal treatment,⁸ and the preservation of the Common Travel Area (CTA) covering the UK and Ireland.⁹ A series of Annexes contain intricate detail on exactly which EU measures are to be applied by the UK in Northern Ireland. Protocol Article 19 stipulates that ‘Annexes 1 to 7 shall form an integral part of this Protocol’.

The application of these rules distinguishes Northern Ireland sharply from GB. This entails both changes in patterns of trade (because Northern Ireland products will be different from products made in GB and production and supply chains are likely to be disrupted) and the emergence of a customs and regulatory border between Northern Ireland and GB (because there has to be some kind of border between territories with divergent regulatory regimes). In so far as the UK chooses to diverge in GB from the EU model, those divisive issues will become ever more prominent over time. In short, then, in the areas covered by the Protocol, Northern Ireland is much more a part of the EU’s internal market than it is part of the UK’s internal market.

6.3 Why the Protocol Exists

Both sides, the EU and the UK, have made significant compromises in order to address what are recognized in the Protocol as ‘the unique circumstances on the island of Ireland’.¹⁰ The aim is to keep the border between Northern Ireland and Ireland as soft or invisible as it was immediately before the UK left the EU. This is a matter of economic significance, but much more so of political importance.

The EU has agreed three significant adjustments to its orthodox approach. It has agreed that: (i) its de jure external border, between Northern Ireland and Ireland, shall remain soft or invisible; (ii) its de facto external border, between Northern Ireland and GB, shall be located within the territory of, and policed by, a non-member state; and (iii) its economic freedoms in particular and its internal market legislative acquis in general shall be divided. In Northern Ireland an important package of EU rules shall be applied but not the whole of internal market law.¹¹ The Treaty rules governing free movement of persons and services are not engaged; also excluded from the Protocol are legislative provisions on consumer and environmental protection which, though tied to the internal market by their legal base, do not directly address the composition of products.¹²

For its part, the UK, in order to maintain a soft or invisible border between Northern Ireland and Ireland and also to provide room for GB to pursue a different regulatory course from the EU, has accepted that the rules governing the matters covered by the Protocol will be different in Northern Ireland and in GB. This entails that the border between Northern Ireland and GB acquires a higher legal, economic and political significance than in the past. It is not an international border, but it now counts as a border between a jurisdiction, Northern Ireland, which shares the regulatory features mandated by the Protocol with the EU, and a jurisdiction, GB, which has largely (subject to discussion of state aid; see Section 6.7) separated itself from EU rules. That border, though in formal terms internal to the UK, must become harder than before. The Protocol thus represents a delicate balance which is the result of departures from orthodox approaches made on both sides.

Another way to understand this is to grasp that although the UK might have wanted three things – a soft border between Northern Ireland and Ireland; a soft border between Northern Ireland and GB; and the freedom to depart from the EU’s regulatory model – it is not possible for the EU to accept all three, given the need to preserve the integrity of its internal market, which must be protected by an external border (somewhere). The UK can attain any two of those three objectives – but not all three. Prime Minister May’s deal gave up the freedom to depart from the EU’s regulatory model; Prime Minister Johnson’s deal gave up the soft border between Northern Ireland and GB (albeit with the consequence that the regulatory freedom thereby released attaches to GB, not the whole of the UK). Looming in the future is the acute anxiety that, if the UK persists with its refusal to comply with the Protocol, the soft border between Northern Ireland and Ireland may have to be given up.

6.4 Customs Territory

The Protocol provides that de jure Northern Ireland is part of the UK customs territory.¹³ But the effect of the Protocol is that Northern Ireland is de facto part of the EU’s customs territory for the purposes which are covered by the Protocol. This is the consequence of Article 5(3), which provides¹⁴ that Northern Ireland is locked into the entirety of the EU’s Customs Code, the Common Customs Tariff, legislation setting up a Union system of relief from customs duty, and international agreements containing customs provisions in so far as they are applicable in the EU (subject only to a reservation to the JC of the task of establishing the conditions applicable to certain fishery and aquaculture products), and several other customs-related measures,¹⁵ among them the EU’s trade defence instruments covering, inter alia, anti-dumping and anti-subsidy measures.

Several articles distant from the claim that Northern Ireland is part of the UK customs territory is Article 13(1). This confirms that this claim is in effect untrue, but does so in the spectacularly evasive terms which are the Protocol’s hallmark. Article 13(1) declares that

any reference to the territory defined in Article 4 of Regulation (EU) No 952/2013 in the applicable provisions of the Withdrawal Agreement and of this Protocol, as well as in the provisions of Union law made applicable to and in the United Kingdom in respect of Northern Ireland by this Protocol, shall be read as including the part of the territory of the United Kingdom to which Regulation (EU) No 952/2013 applies by virtue of Article 5(3) of this Protocol.

What is ‘the territory defined in Article 4 of Regulation (EU) No 952/2013’? It is the EU customs territory. What is ‘the part of the territory of the United Kingdom to which Regulation (EU) No 952/2013 applies by virtue of Article 5(3) of this Protocol’? It is Northern Ireland. So although the Protocol says that Northern Ireland is part of the customs territory of the UK, that is not what it does. The Protocol treats Northern Ireland for most purposes as part of the EU’s customs territory. Once it is appreciated that the Protocol contains misleading advertising, it is easier to understand.

6.5 West–East Trade: Northern Ireland to GB

Article 6 of the Protocol asserts that trade from Northern Ireland to GB shall be unfettered, and more generally it asserts an intent to protect the UK’s internal market. But, again, the reality is different.

Article 6(2) of the Protocol directs the EU and the UK to ‘use their best endeavours to facilitate the trade between Northern Ireland and other parts of the United Kingdom’, but this is explicitly stated to occur ‘in accordance with applicable legislation and taking into account their respective regulatory regimes as well as the implementation thereof’. Article 6(2) directs that the JC shall adopt appropriate recommendations with a view to avoiding controls at the ports and airports of Northern Ireland – but only ‘to the extent possible’. Neither, then, offers a basis for setting aside the obligations arising under the Protocol with regard to trade in goods from Northern Ireland to GB. And there are such obligations.

Article 5(3) of the Protocol requires that the normal formalities applicable to goods leaving the EU’s customs territory shall apply to goods leaving Northern Ireland for GB. This entails that a pre-departure declaration be lodged, which shall take the form of a customs declaration, a re-export declaration or an exit summary declaration.¹⁶ Mitigation has occurred. In January 2021, the UK, in a unilateral declaration, announced that it would not require export and exit summary declarations.¹⁷ The EU took note and has accepted this. That acceptance is doubtless conditional on the provision of equivalent information through other means.

Article 6 also envisages impediments to trade between Northern Ireland and GB ‘to the extent strictly required by any international obligations of the Union’ and requires that ‘[t]he United Kingdom shall ensure full protection under international requirements and commitments that are relevant to the prohibitions and restrictions on the exportation of goods from the Union to third countries’. This entails checks to comply with the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). There are also special rules on movement of some cultural goods.

If the UK chooses to carry out no further checks on goods moving from Northern Ireland to GB, this would prevent it making a distinction between goods made in Northern Ireland and goods made in the EU which have been moved to Northern Ireland. Though not precluded by the Protocol, this would generate a risk of trade diversion in so far as the EU–Northern Ireland–GB route allows evasion of the customs and regulatory checks applied at other crossings from the EU into the UK. This does not seem sustainable.

Article 6 of the Protocol says that it is dedicated to protection of the UK internal market and that nothing shall prevent the UK from ensuring unfettered market access for goods moving from Northern Ireland to other parts of the UK’s internal market. But that is not what the Protocol does. The Protocol entails that the border between Northern Ireland and GB shall acquire a higher legal, economic and political significance than it held in the past.

6.6 East–West Trade: GB to Northern Ireland

The Protocol does not suggest that trade in goods moving from GB to Northern Ireland will be unfettered. And it will not be.

Protocol Article 5(1) provides that no customs duties shall be payable for a good brought into Northern Ireland from another part of the UK by direct transport unless that good is at risk of subsequently being moved into the Union, whether by itself or forming part of another good following processing. So it seems that the norm is that no duties on GB to Northern Ireland trade are payable, while there is an exception where the good is at risk of onward movement to the EU when payment of duties is required. But this is not the case. Protocol Article 5(2) reverses the presumption by providing that a good brought into Northern Ireland from GB is considered to be at risk of subsequently being moved into the Union unless it is established that that good will not be subject to commercial processing in Northern Ireland and fulfils criteria to be established by the JC. The key point is that, under the Protocol, goods are deemed to be at risk of onward movement and so attract an obligation to pay duties, unless it is shown that they are not.

The starting point, then, is that duties are payable: in the matter of duties, exports from GB to Northern Ireland are treated in the same way as exports from GB to the EU generally. Exceptions foreseen by Article 5 Protocol are few and limited – for UK residents’ personal property; for goods shown to be not at risk of onward movement in rebuttal of the presumption explained in the previous paragraph; for consignments of negligible value; for consignments sent by one individual to another; and for goods contained in travellers’ personal baggage. There is also tightly drawn scope for reimbursing duties paid. This confirms that what the Protocol says is not what it does – what it does is not treat Northern Ireland as part of the customs territory of the UK. It treats Northern Ireland as part of the EU’s customs territory. Mitigation has occurred. The TCA has reduced the scale of the problem by securing tariff-free trade between the UK and the EU, with the consequence that tariffs are payable only on goods which do not qualify for such treatment under the Agreement, most obviously goods imported into GB from third countries. Moreover, in December 2020, the JC agreed criteria to determine when goods are not at risk of onward movement.¹⁸ The sharp edges of the Protocol have in this way been successfully softened, but not eliminated.

Obstacles to trade in goods between GB and Northern Ireland are not limited to payment of duties. The Protocol requires compliance with the obligations imposed by the EU Customs Code in matters such as entry summary declarations and customs declarations. It also requires that checks on goods be carried out in order to ensure that they comply with the EU rules which are applicable in Northern Ireland but not in GB. The full range of checks is not spelled out in the Protocol, in line with its thematic concern to avoid telling the full story about its extent. But compliance with EU rules on the many matters covered by the Protocol such as product composition, safety, technical standards and SPS requirements applicable to agricultural products and food will need to be checked according to the normal rules and procedures governing entry to the EU’s territory, because GB will no longer be bound by these rules. Grasping the required intensity, location and nature of such checks, left undefined by the Protocol, will need reference sector-by-sector to applicable EU procedures in order to identify precisely what is at stake.¹⁹

The TCA does not change this because it does not include the type of regulatory alignment between the UK and the EU which would have permitted the EU to assume the compliance of GB goods with its rules. ‘Grace’ periods agreed by the two parties in 2021 were of only temporary effect. Looking forward, the greater the regulatory divergence that develops between GB and Northern Ireland, the greater the incentives to use the GB–Northern Ireland–Ireland trade route, and the more significant will the checks on GB to Northern Ireland trade in goods need to be.

In sum, the Protocol says that it protects the UK’s internal market, but that is not what it does. It makes significant changes to it, especially, but not only, for GB to Northern Ireland trade in goods, and it establishes the Irish Sea as a politically, legally and economically significant frontier within it.

6.7 State Aid

The Protocol’s treatment of state aid shares the deceptive character of several other provisions, but its geographical reach is considerably wider. Article 10 of the Protocol, combined with Annex 5, locks control of aid covered by the Protocol into the wider network of control exercised across the territory of the twenty-seven member states. Article 10 catches measures where they ‘affect that trade between Northern Ireland and the Union which is subject to this Protocol’. This is plainly inspired by Article 107 TFEU, the key EU Treaty provision, which subjects to control any aid granted by a member state which distorts or threatens to distort competition ‘in so far as it affects trade between Member States’. It creates a jurisdictionally significant threshold. Aid which does not affect trade between member states is not subject to EU law and is instead a matter of purely national concern: in the same way, aid which does not affect trade between Northern Ireland and the EU is not subject to the Protocol and is instead purely a matter for the UK. However, in EU law the Court interprets this jurisdictionally significant threshold as low.²⁰

The same applies to the Protocol. Article 10 Protocol is not limited to aid granted directly to firms based in Northern Ireland. It is much wider than that. Again, there is far more to it than the Protocol admits. Probably, the grant of aid to small manufacturers of goods in, say, Kent would not affect the trade between Northern Ireland and the EU which is subject to the Protocol, but the grant of aid to firms which are based in GB but have a presence in Northern Ireland is likely to fall within Article 10. Once aid is being used in the Northern Ireland market (even if mainly being used in GB), that may help the recipient to expand its activities into Ireland (and beyond), and it may deter Irish (or other EU) firms from entering the Northern Ireland market because any competitive advantage they enjoy is eroded by the aid provided by the UK. The effect on interstate trade which is jurisdictionally necessary to trigger the Protocol is then present. The larger the aid, the more likely it is to fall within the Protocol’s scope. Such aid is not automatically unlawful, but it must be compatible with EU law. This entails supervision by the Commission according to the terms set by the Treaty and secondary legislation,²¹ and it engages also the role of the Court of Justice of the European Union (CJEU) and of national courts in the same way as under EU law generally.

The only exception is in Protocol Article 10(2), which carves out a limited exception applicable to the agricultural sector. This concession is stated to apply up to a maximum overall annual level of support to be determined by the JC, which was duly set in December 2020.²²

6.8 Enforcement

Some parts of the Protocol are subject to the arbitration procedures established by the Withdrawal Agreement (WA), including the exceptional derogation in Protocol Article 16, but Protocol Articles 5 and 7 to 10, which contain the trade rules considered earlier, are enforced by methods which are closely aligned to those which prevail under orthodox EU law and which applied generally throughout the period of UK membership of the EU.²³ That means that those provisions of the Protocol are enforced through two distinct routes: through the supervisory jurisdiction conferred on the Commission backed up by the role of the CJEU; and through enforcement by private parties before national courts relying on (inter alia) the direct effect and primacy of EU law, which extends to the Protocol, embracing also the preliminary reference procedure.

Article 13(2) adds that ‘the provisions of this Protocol referring to Union law or to concepts or provisions thereof shall in their implementation and application be interpreted in conformity with the relevant case law of the Court of Justice of the European Union’. This Protocol-specific provision asserts a longer-term and firmer obligation of compliance with the case law of the CJEU than is imposed by Article 4(4) and (5) WA. Here, there is nothing deceptive about the Protocol. The split within the UK internal market is plain. EU law orthodoxy continues to prevail in Northern Ireland, but not (state aid aside) in GB.

6.9 Conclusion

Ever since the UK’s 2016 vote to leave the EU was won on a promise of an undeliverable confection of benefits, it has been plain that the form that the withdrawal would ultimately take would always disappoint some, many or even all of those involved. So too with the Protocol. Just as nobody wanted this Brexit, nobody wanted this Protocol. The Protocol is the product of reluctant compromise on both sides.

The Protocol entails the creation of a customs and regulatory border between GB and Northern Ireland and, in the areas it covers, its effect is to locate Northern Ireland within the EU’s rather than the UK’s internal market. The EU has nervously agreed to outsource policing of its de facto external border to a third country, the UK, and has accepted the divisibility of its economic freedoms. This cannot fully satisfy all involved. Yet the problems are caused not by the Protocol but by Brexit itself. The Protocol represents the unavoidable post-Brexit choice among three objectives: no hard border between Northern Ireland and GB; no hard border between Northern Ireland and Ireland; and regulatory autonomy for the UK (or GB). Only two of these can be achieved.

The Protocol is, as explained, drafted in evasive terms, and the areas where clarity is lacking all underplay the extent of the UK’s readiness to accept fragmentation of its own internal market. The Protocol belongs alongside the 1998 Agreement as a subtly written offer to all involved to accentuate the positive and live with the negative. It is a genuine attempt to address the unique circumstances arising on the island of Ireland. With goodwill and honesty, the disappointments could be minimized. Those conditions seem increasingly elusive.

In late 2020, the UK Internal Market Bill included provisions that would have directly contradicted the obligations accepted by the UK under the WA, including the Protocol. The UK government eventually backed down. But, in March 2021, it once again set a course in conflict with the terms of the Protocol by announcing unilateral derogations from its requirements. This provoked the EU to begin legal proceedings.²⁴ British politicians, including the Prime Minister, continue routinely to deny the impact of the Protocol.²⁵ In July 2021, the UK government published a Command Paper which showed shameless disdain for the terms of the deal carefully negotiated and agreed in late 2019 and which included several proposals to amend the Protocol which the UK government is fully aware are unacceptable to the EU.²⁶ It seems painfully clear that the Protocol was a device to ‘get Brexit done’ as part of a strategy that secured a General Election victory in December 2019 and that the UK government sees no further value in it. The same people who campaigned for Brexit with no regard for its consequences on the island of Ireland are now, having agreed the Protocol, showing disregard for its terms. Can this end well? Only if London is somehow forced to take seriously its binding international obligations. That one needs to make such an observation emphasizes the scale of the reputational damage which this government is prepared to inflict on the UK.

7 International Law Rules on Treaty Interpretation

Steven R Ratner

7.1 Introduction

International law is central to the interpretation of both of the Brexit-related treaties. The Trade and Cooperation Agreement (TCA) explicitly requires the parties and any dispute settlement body to interpret it according to the rules of interpretation of public international law, notably the 1969 Vienna Convention on the Law of Treaties (VCLT). The Withdrawal Agreement (WA), and thus the Protocol, by specifying that any of its provisions concerning Union law or concepts must be interpreted in accordance with EU law (including the case law of the Court of Justice of the European Union (CJEU)), implies that its many provisions not concerning EU law will need to be interpreted by the default rules of treaty interpretation, namely those of the VCLT. This chapter provides a brief overview of those rules and some of their implications for these two instruments. I focus on Articles 31 and 32 of the VCLT that concern the interpretation of treaties.

7.2 Parsing the TCA and WA Provisions on Interpretation

TCA Article 4(1) states:

The provisions of this Agreement and any supplementing agreement shall be interpreted in good faith in accordance with their ordinary meaning in their context and in light of the object and purpose of the agreement in accordance with customary rules of interpretation of public international law, including those codified in the Vienna Convention on the Law of Treaties, done at Vienna on 23 May 1969.

The following two paragraphs add, ‘for greater certainty’, that parties are not obliged to interpret the agreement in accordance with their domestic law; and that an interpretation by the courts of either party shall not bind the courts of the other party.

The first sentence of Article 4(1) WA begins with language nearly identical to the key sentence in Article 31 VCLT: ‘A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in light of its object and purpose.’ But Article 4(1) then clarifies (or perhaps conditions) this phrase by adding that such interpretation must be ‘in accordance with’ customary international law, including the VCLT. The choice of words is odd, but not unique, as it appears in a number of EU treaties with non-EU states.¹ It refers to customary international law directly, and the VCLT secondarily, as (a) the EU itself is not a party to the VCLT; (b) the VCLT does not govern treaties between international organizations and states; and (c) customary international law includes other rules of interpretation (discussed in Sections 7.5 and 7.7) not mentioned in the VCLT. Nevertheless, the TCA accepts – as is universally accepted – that the VCLT’s rules of interpretation are customary international law.²

The WA, for its part, contains the following two paragraphs in Article 4:

1. 3. The provisions of this Agreement referring to Union law or to concepts or provisions thereof shall be interpreted and applied in accordance with the methods and general principles of Union law.

2. 4. The provisions of this Agreement referring to Union law or to concepts or provisions thereof shall in their implementation and application be interpreted in conformity with the relevant case law of the Court of Justice of the European Union handed down before the end of the transition period.

From the language of these paragraphs, provisions of the WA that do not refer to EU law, concepts or provisions will not be governed by EU law and CJEU jurisprudence. What, then, will be the framework for their interpretation? The WA cannot mean that non-EU law provisions would be interpreted solely in accordance with the law of one or the other party. The only alternative principles for interpreting any treaty – whether between states or between the EU and a state – are those of public international law. The public international law that applies is essentially that set forth in the TCA, namely customary international law, including the VCLT.

7.3 The VCLT’s Rules of Interpretation: General Themes

Article 31 is a sort of bible when it comes to treaty interpretation. Global, regional and domestic courts routinely cite it as providing the framework for the treaties they interpret. Before delving into the key provisions and their implications for the Brexit agreements, a few key points should be made:

1. (i) The goal of treaty interpretation under the VCLT is to determine the meaning of the treaty viewed from the perspective of the contemporary shared understanding of the parties to the treaties. As James Crawford has pointed out, from the perspective of international law, ‘the parties … own the treaty’,³ even if it is certainly the case that their citizens, animals and plants (and those of other states) may well be affected by how that treaty is interpreted. As a result, the key evidence for interpretation will be the interactions of the parties insofar as it demonstrates that shared understanding.

2. (ii) Treaty interpretation is not just a task for tribunals. Parties to treaties are constantly interpreting them, making claims against the other party in diplomatic settings, bilaterally and multilaterally, confidentially and publicly. Only a small handful of these disputes will make it to tribunals due to jurisdictional obstacles and incentives that states have to avoid formal dispute settlement. Yet the provisions remain relevant nonetheless.

3. (iii) Although the VCLT’s rules of interpretation are laid out in two articles, treaty interpretation is not a formulaic exercise, where boxes are checked and then a decision reached. Moreover, although various indicia of the meaning of the text appear in Articles 31 and 32, the VCLT does not offer a ranking of evidence, insofar as a meaning flowing from one set of evidence will simply override that flowing from recourse to other evidence – with the one obvious exception that the text obviously ‘counts’ more than anything else. At the same time, there is a general understanding among tribunals that some evidence will be more probative of the parties’ contemporary understanding than others.⁴

4. (iv) Treaty interpretation, like statutory interpretation, can easily yield more than one convincing interpretation of a treaty. Just as the states parties to a treaty may have bona fide differences of opinion regarding its meaning, so may judges. At the same time, the tools of treaty interpretation generally succeed in narrowing the range of plausible interpretations, sometimes to one.

5. (v) The VCLT’s rules for treaty interpretation are default rules. The parties to a treaty are free to include clauses in the treaty, or in subsequent instruments, that offer (a) a particular interpretation of a clause, (b) a particular set of indicia to consider or exclude in future interpretations and (c) a specific process for interpretation that may differ from that in the VCLT.

6. (vi) Relatedly, different tribunals have developed interpretive methodologies that are distinct from those in the VCLT, or at least put a particular emphasis on one factor more than others. Notably, both the European Court of Human Rights (ECtHR) and the CJEU have endorsed a broadly teleological approach to treaty interpretation that emphasizes the overall purposes of the European human rights regime or the EU, respectively. In addition, the ECtHR has developed at least one doctrine, the margin of appreciation, that is certainly not part of the VCLT (though it is not precluded by the VCLT either).

7. (vii) While the VCLT does not include an explicit obligation on domestic courts to employ it to interpret treaties, it is, of course, a treaty itself, binding on the states parties (which include the UK). So, depending on each state party’s approach to the direct application of treaties (eg, monist versus dualist), domestic courts will have a domestic law obligation to interpret the treaty in accordance with the VCLT.⁵

7.4 The Starting Point: Text, Object and Purpose, and Context

The VCLT’s first command to a treaty interpreter appears in Article 31(1) and (2):

1. 1. A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.

2. 2. The context for the purpose of the interpretation of a treaty shall comprise, in addition to the text, including its preamble and annexes: (a) any agreement relating to the treaty which was made between all the parties in connexion with the conclusion of the treaty; (b) any instrument which was made by one or more parties in connexion with the conclusion of the treaty and accepted by the other parties as an instrument related to the treaty.

These two articles provide a text-based starting point for treaty interpretation. In the case of the TCA, the text includes all its protocols, annexes, appendices and footnotes.⁶ In the case of the WA, the three Protocols and the nine Annexes are all part of the text.⁷ While the VCLT refers only to the ‘ordinary meaning [of] the terms of the treaty’, tribunals have relied on additional doctrines, which can be said to constitute customary international law, to engage more meaningfully with the text. Various principles from domestic law, such as ejusdem generis and expressio unius est exclusio alterius, are frequently borrowed. The principle of effectiveness, or effet utile, is particularly worth mentioning. This principle requires that a treaty is to be interpreted to give it, as a whole, and the individual provisions within it meaningful effect. The approaches of international courts reveal that the principle of effet utile means that treaty clauses must be interpreted to avoid either rendering them superfluous or depriving them of significance for the relationship between the parties.⁸

As for the ‘object and purpose’, tribunals have significant discretion in their mode of determining it, as well as its impact on the meaning of the text. Often the text will specify the agreement’s purpose, either in the preamble or in an early article. The TCA does so in its very lengthy preamble as well as its rather anodyne Article 1; the WA also has a lengthy preamble and similar Article 1. However, the Protocol has not only a lengthy preamble but a much more specific Article 1: it clarifies that the Protocol is ‘without prejudice’ to the 1998 Agreement; that it respects UK ‘essential State functions and territorial integrity’; and that its arrangements address ‘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’. These clauses clearly inform the interpretation of the Protocol and the WA generally.⁹

The ‘context’ of the treaty is, in a sense, more concrete, as it consists of other written instruments concluded in connection with the treaty. They must be written, agreed upon, ‘relating to the treaty’ and ‘in connection with the conclusion of the treaty’. Typical examples include exchanges of notes between the parties that define terms or include additional commitments. For the TCA and the WA, no simultaneous agreements or exchanges of notes were agreed at the time of their conclusion. At the same time, the term ‘in connection’ does not have a fixed time-limited meaning, so even though more than a year separated the conclusion of the TCA and that of the WA, courts may need to decide whether the WA is part of the context of the TCA as well as whether the TCA is part of the context of the WA. The ‘Political Declaration’ agreed in November 2018 might also be such an instrument¹⁰ (though it might also be part of the negotiating history for purposes of Article 32 of the VCLT).

7.5 Role of Subsequent Agreements, Practice and Other Rules of International Law

Article 31(3) of the VCLT begins an additional list of indicia for interpreters beyond the immediate context of the conclusion of the treaty (a list that continues in Article 32, discussed in Section 7.6 of this chapter):

There shall be taken into account, together with the context: (a) any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions; (b) any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation; (c) any relevant rules of international law applicable in the relations between the parties.

The first two additional elements concern the way that the parties understand the meaning of the treaty since its entry into force. Subsequent agreements (a) refer to new treaties or other agreements (eg, oral agreements) where the parties clarify how they wish to interpret or apply the treaty. Such agreements are given significant, even dispositive, weight because they clearly demonstrate the parties’ contemporaneous understanding of the meaning of a treaty. In the case of the TCA and the WA, such agreements are certainly possible. Indeed, the TCA itself presumes that such agreements will be ‘supplementing agreements’ that will be an ‘integral part of the overall bilateral relations … and shall form part of the overall framework’.¹¹ The TCA does not quite call these supplementary agreements an integral part of the TCA text itself, but it suggests that future interpreters using VCLT Article 31(3)(a) should give them heavy weight.

The second interpretive aid, subsequent practice (b), is also very important but much more difficult to identify because it is typically not reduced to a single document. Interpreters have to verify that the two (or more) parties are truly agreed on what a particular provision of the treaty means, as opposed to simply agreeing on something else.¹² In recent years, the UN’s International Law Commission (ILC) has offered detailed guidance to states and courts about the meaning of subsequent practice. That guidance clarifies that the range of acts that might constitute subsequent practice includes executive, legislative, judicial or other acts of the parties (but not of non-state actors), as a well as a conference, or joint institution, of the parties.¹³ So if, for instance, the highest domestic courts of the two parties to a treaty explicitly shared an understanding of a term in the treaty, and that understanding was not contradicted by other organs of the state, those court judgments would constitute the sort of state practice that is to be taken into account under sub-paragraph (b). If one of the many bilateral institutions established under the WA and the TCA, from the highest political level to the expert/technical level, were to agree on an interpretation of the treaties, that too would be important subsequent practice.¹⁴

The third interpretive aid, other relevant rules of international law applicable in the relations of the parties (c), on its face opens the door to courts interpreting either the WA or the TCA to bring in completely extraneous treaties – whether human rights, environmental law, the law of the sea, international trade law (including intellectual property) – as well as rules of customary international law. Much of this process is not controversial. For example, international courts routinely rely on the customary law rules of state responsibility to determine when an action in violation of a treaty is imputable to one of the parties.¹⁵ So, too, may international courts sometimes rely on an obviously relevant treaty for the purpose of interpreting another treaty, such as interpreting the term ‘territorial sea’ by reference to the UN Convention on the Law of the Sea.

But hard cases abound, where the parties or judges may have to interpret whether a particular norm is a ‘rule’, whether it is ‘relevant’ and whether it is ‘applicable in the relations between the parties’.¹⁶ If one party to a dispute believes that a treaty should be interpreted in light of another treaty to which only one of the states is party, it would be hard to view that second treaty as ‘applicable’ (unless the rule in it has become one of customary law). In the case of the Protocol, two obviously delicate questions will concern whether the 1998 Agreement and the ECHR fit within Article 31(3)(c). The former is mentioned numerous times in the Protocol, with an emphasis on the need to preserve its operation;¹⁷ but the EU is not a party to it. Nor is the EU a party to the ECHR, although all members of the EU are party to that treaty. Courts might nonetheless look past these distinctions and determine that both are indeed ‘relevant’ and ‘applicable’. Regarding the TCA, Part 3 on criminal law enforcement co-operation states that the co-operation is ‘based on’ the ECHR and the importance of giving it domestic effect, and that nothing in the TCA modifies the duty to respect the rights in the ECHR.¹⁸ So clearly the ECHR constitutes a relevant rule for purposes of interpreting Part 3.

7.6 Supplementary Means

The VCLT’s other article on interpretation invites tribunals to consider what are deemed ‘supplementary’ indicia:

Recourse may be had to supplementary means of interpretation, including the preparatory work of the treaty and the circumstances of its conclusion, in order to confirm the meaning resulting from the application of article 31, or to determine the meaning when the interpretation according to article 31: (a) Leaves the meaning ambiguous or obscure; or (b) Leads to a result which is manifestly absurd or unreasonable.

Although Article 32 begins with the word ‘may’, its drafters fully expected tribunals to refer to these materials, and many tribunals routinely have recourse to supplementary materials, in particular the negotiating history of a treaty (travaux préparatoires).¹⁹ As the text makes clear, tribunals are not constrained in when they may turn to these supplementary means, as they can be used to either confirm or rebut what has been found from the deployment of Article 31. Yet the travaux, especially of a particular article in a treaty, may be obscure. The deal on the language may have been reached over drinks or coffee at a bar, or just in a conversation off the record, so that those taking notes of the negotiation never wrote down what the parties meant to agree on – or to avoid agreeing on. The negotiating history could be confidential, requiring an interpreting body to request it from the parties. This uncertainty over the availability and reliability of the travaux préeparatoires, combined with a view that these materials should matter less than those showing the contemporaneous understanding of the parties, may render them ultimately of little use for interpreting some treaties, though tribunals take them into account when they can find them.

Article 32 is not limited to the preparatory work of the treaty. It includes any other material related to the conclusion of the treaty not covered in Article 31(2), including statements by the negotiators to their legislature during debates over the latter’s approval of the treaty. In addition, as the ILC has recently made clear, it includes the conduct of one or all parties to the treaty when that practice does not demonstrate the agreement of the parties required for that practice to be considered under Article 31(3)(b).²⁰ So courts might look at the interpretation of the treaties by the parties to demonstrate that a particular understanding is not shared.

In the case of the WA, some of the negotiating record is easily available – notably the first EU Commission proposed draft, as well as the EU–UK agreed draft rejected by Parliament; other documents are public. The statements of the negotiators to their legislatures are well documented; and both treaties have been the subject of disagreements already. Notably, the EU accused the UK of breaching the WA, including its duty of good faith,²¹ in including a particular clause (later withdrawn) in a bill introduced in Parliament, and in delaying certain customs regulations on goods destined for Northern Ireland.²²

7.7 Conflicts: Peremptory Norms and Other Treaties

While international law gives states virtually unlimited discretion to define, refine, alter and terminate their treaty relations, this discretion is not completely unlimited. Article 53 treats as void ab initio any treaty conflicts with a norm of jus cogens. Certainly, civil society actors like to make claims that various parts of treaties so conflict, rendering the treaties void. Yet the scope of jus cogens norms is extremely limited – the ban on the use of force, a few core human rights norms (the ban on genocide, torture or crimes against humanity), the supremacy of the UN Charter, and perhaps a few others. It seems highly unlikely that any provision of the WA or the TCA would violate jus cogens.

A separate question is whether an obligation under either treaty might conflict with another treaty or rule of customary international law, insofar as carrying it out would breach those other rules. In this scenario, there are various possible responses: the states will amend the treaty to avoid the conflict; one or both will interpret it through recourse to Article 31(c)(3) to avoid the conflict (although conceivably the parties could interpret it differently); or the parties will need to assume the consequences of a breach of the other rule. The VCLT does not set any priorities for treaties nor require that one treaty be interpreted to be consistent with another (although the text of a particular treaty can so provide). So, international law does not require that the WA or the TCA be interpreted to be consistent with something as old as the Act of Union of 1800 unifying Ireland and Great Britain (eg, its provisions on commerce), or something as recent as the WTO Agreements. In addition, Article 30 of the VCLT provides that if the parties to one treaty later conclude another treaty on the same subject matter without terminating the first, the earlier treaty applies only to the extent that it is ‘compatible’ with the later treaty, unless the parties specify otherwise.²³

7.8 Conclusion

Articles 31 and 32 VCLT are central to the interpretation of all treaties and are routinely applied by international, regional and domestic courts and arbitral bodies – and of course by treaty parties themselves in their interactions regarding the implementation of treaties. They are the subject of enormous case law and scholarship; the aim of this chapter has been only to highlight their key features when it comes to the Brexit agreements.²⁴ What remains to be seen is how the different courts and tribunals that engage with the Brexit agreements (domestic courts, the CJEU, the arbitration panels) may use the VCLT, perhaps, in different ways. The CJEU has significant experience using the interpretive methodology of the VCLT in interpreting treaties between the EU and other states (while adopting a sui generis framework for interpreting the EU’s constitutive instruments). Yet, as numerous commentators have noted, even as it applies the VCLT framework, its approach is often quite idiosyncratic.²⁵

8 Good Faith and Sincere Co-operation

Christopher McCrudden

8.1 Introduction

The Withdrawal Agreement (WA) has several common provisions which set out various obligations as to how the WA (and the Protocol as an integral part of the WA) is to be interpreted and applied. Among these are the duty of ‘good faith’ and the duty of ‘sincere cooperation’, which feature in Article 5 WA. This provides, in part, that the EU and the UK ‘shall, in full mutual respect and good faith, assist each other in carrying out tasks which flow from this Agreement’. In addition, Article 5 provides that this obligation of good faith ‘is without prejudice to the application of Union law pursuant to this Agreement, in particular the principle of sincere cooperation’.

8.2 How to Approach the Interpretation of Article 5 WA?

‘Good faith’ is a general principle of international law,¹ including international trade law,² and is explicitly included in Article 26 of the Vienna Convention on the Law of Treaties (VCLT).³ What appear to be somewhat equivalent concepts feature in other specifically European agreements. Article 4(3) of the Treaty on European Union (TEU) applies the duty of ‘sincere cooperation’ (sometimes referred to as the principle of ‘loyalty’) to member states and the EU in the application of the EU Treaties.⁴ Article 3 of the European Economic Area (EEA) Agreement applies what the Court of the European Free Trade Association (EFTA Court) terms the duty of ‘sincere cooperation’ or ‘loyalty’ to the parties to that Agreement.⁵ The relevant provision in these two European agreements is drafted in very similar terms to Article 5 WA,⁶ except that they do not use the term ‘good faith’ and Article 4(3) TEU includes the duty to ‘facilitate the achievement of the Union’s tasks’, which neither Article 3 EEA nor Article 5 WA includes. In light of the fact that Article 5 WA thus contains an amalgam of both international law and EU law terms, how should it be interpreted? As more closely related to the international law usage or to EU legal understanding?

Although the matter is not without doubt, a plausible interpretation of Article 5 WA is that the reference to ‘good faith’ is a reference to the international law concept and should be interpreted as such,⁷ but that this is an additional requirement on the parties, over and above the other obligations in Article 5, which is to be interpreted in light of Union law. Thus, although Article 5 is entitled ‘Good Faith’, the term only appears in the first paragraph of that provision. The temptation to interpret the whole Article in light of this, applying an international law lens to the other provisions as well, should be resisted. This understanding draws strength from Article 4 WA. Article 4(4) provides that ‘[t]he provisions of this Agreement referring to Union law or to concepts or provisions thereof shall in their implementation and application be interpreted in conformity with the relevant case law of the Court of Justice of the European Union handed down before the end of the transition period’. With the exception of the ‘good faith’ provision, Article 5 is, therefore, to be interpreted in accordance with Court of Justice of the European Union (CJEU) case law. This interpretation is strengthened further by Article 4(1) WA which provides that ‘[t]he provisions of this Agreement … shall produce in respect of and in the United Kingdom the same legal effects as those which they produce within the Union and its Member States’. It is clear that, with the exception of the ‘good faith’ provision, the remainder of Article 5 is likely to be interpreted in the Union and the member states along the lines of the case law on Article 4(3) TEU.

The matter is not free from doubt, however, given that Article 3 of the Trade and Cooperation Agreement (TCA),⁸ drafted in almost identical terms to Article 5 WA, is subject to Article 4 TCA, which states that the provisions of the TCA must be interpreted according to international law, and not according to EU law.⁹ This would seem to result in a conclusion that Article 3 TCA may give rise to different obligations on the parties than does Article 5 WA. While this appears to be counter-intuitive, this conclusion would accurately reflect the way in which the interpretation of these provisions is heavily context specific, a point we return to subsequently. In light of this, this chapter considers primarily how the meaning and scope of Article 5 WA is clarified by examining the interpretation of the analogous concepts by the CJEU and the EFTA Court.

8.3 Article 5 WA in Detail

8.3.1 ‘Objectives’ of the WA and the Protocol

Each of the Agreements mentioned so far refers to the ‘objectives’ of these Agreements and relates the concept of ‘good faith’ (or equivalent) to the delivery of these ‘objectives’. The Agreements differ significantly in what they identify as their objectives and these differences need to be given sufficient weight when interpreting the obligations of the parties in Article 5 WA. Some of these objectives are considerably more ambitious than others. For example, Article 3 TEU, setting out the objectives of the Union, goes considerably further than Article 1 of the EEA Agreement, the objectives of which are simply ‘… to promote a continuous and balanced strengthening of trade and economic relations between the Contracting Parties with equal conditions of competition, and the respect of the same rules, with a view to creating a homogeneous European Economic Area…’. In the EU context, but not in the EEA, the principle of ‘sincere cooperation’ has been understood as an expression of EU solidarity and a reflection of the principle of ‘federal good faith’,¹⁰ ‘which is designed to secure mutual respect and readiness to co-operate with respect to the powers exercised by the legislative, executive and judicial bodies of different levels of authority within a “federal” system’.¹¹

So, too, the objective of the WA as a whole is stated in deceptively simple, and relatively restrained, terms in Article 1 WA: ‘This Agreement sets out the arrangements for the withdrawal of the United Kingdom of Great Britain and Northern Ireland … from the European Union … .’ The objective of the Protocol, on the other hand, is stated in more specific and wide-ranging terms: addressing the unique circumstances on the island of Ireland, maintaining the necessary conditions for continued North–South co-operation, avoiding a hard border and protecting the 1998 Agreement in all its dimensions. This means that the Protocol has, to an extent, ‘Europeanised’ the 1998 Agreement, meaning that upholding that Agreement has become an EU objective in the context of the WA, and no longer simply a UK–Irish objective.

How these differences in ambition regarding objectives should lead to differences in the application of the obligations in Article 5 WA is likely to become a contested issue. All that can be said for the moment is that the greater the depth of integration that an agreement is thought to require, the more far-reaching the principle of good faith and the cognate obligations are likely to be, in two different senses: first, in a more extensive treaty the principle of good faith etc will be more far-reaching simply because it applies to more objects; and, second, the application of the principles will involve differential depth of engagement by the parties with each other. If this approach is adopted, then the principle of ‘good faith’ etc in the WA, as applied in the Protocol, is likely to be considered more far-reaching than in Article 26 of the VCLT, Article 3 EEA or Article 3 TCA, but less far-reaching than Article 4(3) TEU. Much will depend, however, on the particular cases in which the issue arises.

8.3.2 Good Faith etc and Sincere Co-operation as Lex Generalis

There is also a significant difference between these Agreements in the extent to which they include detailed provisions on a range of issues which potentially overlap with the duties of good faith etc and sincere co-operation. So, for example, the fact that there are specific provisions in the Protocol regarding the requirement for domestic courts to refer issues of EU law to the CJEU means that the role that the duty of sincere co-operation plays in the EEA context in generating a duty to refer in certain circumstances is unnecessary. Lex specialis displaces lex generalis. The CJEU has been explicit on this point. TEU Article 4(3)’s predecessor provision, the Court held, ‘is worded so generally that there can be no question of applying it autonomously when the situation concerned is governed by a specific provision of the Treaty …’.¹² Where there are gaps in the Protocol,¹³ however, the role of Article 5 may become critical. Where there are specific, detailed requirements, these will reduce the role that the more general concept of ‘good faith’ will need to play. The duty of ‘good faith’ is ‘subsidiary to more specific Treaty provisions’.¹⁴

8.3.3 Obligations Deriving from Article 5 WA

Although the ends (‘objectives’) that each of the Agreements seeks to achieve differ (in some cases significantly, as we have seen), and the Article 5 WA and Article 3 TCA obligations are also situated in the context of differing substantive obligations in the various Agreements, the means by which these objectives and obligations may and should be met are mostly common to both of these Agreements. There are three basic obligations in Article 5 WA and Article 3 TCA, applying to both the EU and the UK: (i) a positive obligation to assist each other in carrying out tasks which flow from the Agreement; (ii) a positive obligation to take all appropriate measures, whether general or particular, to ensure fulfilment of the obligations arising from the Agreement; and (iii) a negative obligation to refrain from any measures which could jeopardize the attainment of the objective of the Agreement.

With the exception of (i), which is not included in Article 3 EEA, these obligations mirror equivalently drafted obligations in Article 4(3) TEU, Article 3 EEA and Article 3 TCA. These all relate to the means by which the objectives of the relevant Agreements may and should be fulfilled. As a result, the interpretation of these common ‘means’ provisions should be seen as closely aligned.¹⁵ Following the approach adopted by Hreinsson in the EEA context, we may say that Article 5 WA will be construed in the same manner as the corresponding EU and EFTA provisions, ‘so long as this interpretation is not irreconcilable with the provisions and characteristics’ of the relevant Agreement.¹⁶ Unless any of the differences in drafting are material in a particular case, the CJEU’s application of the concept of ‘sincere co-operation’ and the EFTA Court’s interpretation of the ‘loyalty’ provision in Article 3 EFTA will both be highly relevant in understanding the meaning and scope of the obligations in Article 5 WA.

8.3.4 Implications of Article 5 WA Obligations

Several important principles deriving from the CJEU and the EFTA Court jurisprudence on ‘sincere co-operation’ apply to Article 5 WA:

1. (i) The obligations set out in Article 5 are legal obligations, not simply political commitments.¹⁷ ‘Good faith’ is sometimes said to be non-justiciable and simply a declaration of political principle. This is inconsistent with the uninterrupted jurisprudence of analogous concepts by the CJEU and the EFTA Court, and Article 5 WA itself requires that this consistent jurisprudence should be followed in the interests of consistency. Obiter comments to the contrary by UK judges, however illustrious, should not be followed.¹⁸

2. (ii) The other obligations in Article 5 should not be seen through the international law lens of ‘good faith’ but should rather be seen in light of the interpretation of the analogous concepts in the EEA and the EU. This has significant implications, not only in distancing it from general international law understandings of ‘good faith’. Some have interpreted a ‘good faith’ obligation as consisting of little beyond imposing an obligation to behave ‘reasonably’. This understanding is common in diplomatic and political uses of the term and is reflected in some judicial interpretations of the term. This approach gives insufficient importance to the way in which the other obligations in Article 5 WA have come to possess the more technical meaning that this chapter now considers.

3. (iii) ‘Good faith’ and the other obligations in Article 5 WA may also differ significantly from apparently similar domestic law concepts. ‘Good faith’ is not, for example, simply the absence of ‘bad faith’, of a type that is common in UK public law. The positive obligations inherent in a duty of good faith go significantly beyond the concept of the absence of bad faith, as it is hoped this chapter demonstrates. That said, there are, as we have seen, some points of comparison in the use of similar concepts in national federal systems.¹⁹

4. (iv) The obligation of ‘good faith’ and the other obligations in Article 5 WA are addressed to all the organs of the UK, including to UK courts for matters within their jurisdiction.²⁰ The principle also applies to the institutions and bodies established by the WA and the Protocol, as well as to all the organs of the EU. The duty of good faith etc rests on mutual duties, and mutual respect, which bind not only the UK but also the Union institutions. This does not mean either that the ‘good faith’ obligation is symmetrical or that it depends on reciprocity.

Taken as a whole, the duty of ‘good faith’ will be seen to express the ‘gravitational force’ that the Protocol should have on all decisions taken by the UK and the EU that impact those objectives.²¹ The implications this will have for the UK and the EU have yet to be fully identified. That will be a work in progress for many years. However, as a first approximation, it will be useful to consider what the obligations in Article 5 WA are likely to require, using the three basic obligations identified above as the basic structure, considering them in reverse order.

Article 5 lists three obligations. First, the negative obligation to refrain from any measures which could jeopardize the attainment of the objective of the Agreement. As a result, the freedom of action in the future of both the EU and the UK is limited, ‘excluding the right to legislate at will concerning the subject matter of the treaty’.²² At its most basic, Article 5 WA requires that the legal obligations of the Agreement must be upheld, echoing Article 26 VCLT, incorporating the principle of pacta sunt servanda. The obligations in Article 5 WA can be seen in part as the equivalent to or an expression of these international law principles.²³ The decision by the UK government to introduce legislation (the UK Internal Market Bill) that included provisions that explicitly empowered ministers to breach international obligations, including those under the Protocol, if that was necessary in order to protect the UK internal market, was a clear breach of this first element of the Article 5 WA obligations. Whether a matter falls within the sovereign power of the EU or the UK does not alter the obligation, therefore, that, in situations covered by the WA and the Protocol, EU and UK rules must comply with these Agreements.

Second, Article 5 WA imposes the positive obligation to take all appropriate measures, whether general or particular, to ensure fulfilment of the obligations arising from the Agreement. This applies in particular to the implementation and enforcement of these obligations, including regarding the provision of national remedies, and respect for the allocation of competences in the Protocol. Several examples may be identified from other chapters: regarding the scope of ‘direct effect’, as discussed in Chapter 3, in which the duty of good faith tips the balance in favour of a generous interpretation; regarding the availability of damages remedies for failure by the UK to track replacement directives, as discussed in Chapter 12, in which the duty of good faith will be critical in deciding the scope of state liability;²⁴ regarding the debateable role of the Charter of Fundamental Rights (CFR), as discussed in Chapter 13, in which the duty of good faith is relevant in deciding when it should apply. Other examples may also be identified. The UK is required to ensure that directives which have been implemented and satisfy certain conditions prevail over conflicting national law.²⁵ This aspect of the duty of ‘good faith’ requires respect for the principle of consistent interpretation, developed in EU law.²⁶

The third obligation in Article 5 WA consists in a positive obligation on each party to assist the other in carrying out tasks which flow from this Agreement. This includes the duty, in certain circumstances, to negotiate with the other party, to inform the other party of relevant developments, and not to act unilaterally where this would lead to difficulties for the other party in the sphere of the Protocol. An exemplification of these principles can be seen in the inclusion of ‘good faith’ in Article 169(1) WA. (‘The Union and the United Kingdom shall endeavour to resolve any dispute regarding the interpretation and application of the provisions of this Agreement by entering into consultations in the Joint Committee in good faith, with the aim of reaching a mutually agreed solution.’) This third wing of the obligation goes beyond the duty to negotiate, however. In infringement proceedings, the UK would also be required to co-operate in any inquiry of the Commission and to supply it with all the necessary information, and vice versa.²⁷ The EFTA Court has decided that the EFTA Surveillance Authority was under a duty in a state aid case not only to respect the right of the state to be heard but also to co-operate sincerely with the latter in that procedure.²⁸

An application of this third element of the Article 5 WA set of obligations is to be found in the Commission’s letter to the UK objecting to the unilateral extension by the UK of grace periods,²⁹ and supporting the commencement of legal action by the EU. It drew heavily on WA Article 5’s good faith provision. ‘The UK’, the Commission wrote, ‘has resorted to this unilateral action without any discussion or consultation with the EU side in the bodies established by the Agreement. It has therefore acted in breach of the mutual trust and spirit of cooperation ….’ The Commission identified two separate sets of legal complaints: (i) a violation of the substantive provisions of the Protocol, Article 5(3) and (4) of the Protocol on Ireland–Northern Ireland, read in conjunction with relevant Union law listed in Annex 2; and, separately, (ii) a violation of the duty of good faith provided by Article 5 WA. The Commission required the UK to enter into consultations in the Joint Committee under Article 169 WA, with the aim of reaching a mutually agreed solution; failure to do so would lead to the triggering by the EU of the dispute settlement procedure under WA Title III of Part Six.

8.4 ‘Sincere Co-operation’ in Article 5 WA

What are we to make of the proviso in Article 5 WA that the obligations we have just examined are ‘without prejudice to the application of Union law pursuant to this Agreement, in particular the principle of sincere cooperation’? There appear to be three main implications of this provision. First, the duties that arise from Article 5 cannot be used to limit the scope or application of the principle of sincere co-operation in applying Union law (for example vis-à-vis a member state) that continues to apply to the EU under Article 4(3) TEU. This proviso would prevent the UK from arguing that an obligation on the EU under Article 4(3) TEU was trumped by Article 5 WA.

Second, the proviso is not drafted in such a way that it applies only to the EU institutions. At this point, we need to tread with care. Since the UK has left the EU, the Article 4(3) TEU duty of sincere co-operation clearly does not apply to the UK. There is no overarching duty on the UK to fulfil the aims of the Union in the exercise of its sovereign authority. However, there are various places in the WA and the Protocol where the UK has agreed to be bound to follow ‘Union law’, as several chapters of this book describe.³⁰ Where that is the case, we need to bring in another critical common provision of the WA. We have seen that Article 4 WA provides that the ‘… the provisions of Union law made applicable by this Agreement shall produce in respect of and in the United Kingdom the same legal effects as those which they produce within the Union and its Member States’. This means that the general principles that continue to apply to the interpretation and application of Union law in the EU also apply to the interpretation and application of Union law if and when it applies in the UK by virtue of the WA and the Protocol. One of these general principles is the duty of sincere co-operation. If the duty of sincere co-operation in Article 4(3) TEU goes beyond the other obligations of the parties in Article 5 WA, then the interpretation of Union law, based on the duty of sincere co-operation, applies.

Arguably, there is a third implication. Article 4 WA also provides that ‘[t]he provisions of this Agreement … shall produce in respect of and in the United Kingdom the same legal effects as those which they produce within the Union and its Member States’. The implications of this provision, when read with Article 5 WA, are somewhat less clear. The Union must interpret and apply the WA in such a way that the aims of the Union are upheld. A maximalist interpretation would be that the UK is also required by Article 4 WA to act in such a way in implementing and interpreting the WA as to fulfil the aims of the Union, but that seems a step too far. When Article 5 states that the obligations in Article 5 WA are ‘without prejudice to the application of Union law pursuant to this Agreement, in particular the principle of sincere cooperation’, this implies that the intention was that the duty of sincere co-operation under Article 4(3) TEU applies to the application of Union law only, not to the provisions of the WA itself, which do not constitute ‘Union law’.

8.5 Conclusions

Early indications are, then, that Article 5 WA is likely to play a significant role in future disputes between the UK and the EU over the application of the Protocol, not least because it helps to address a feature of EU–UK relations that seems unlikely to disappear quickly, namely the absence of trust. Article 5 WA stresses that, for the parties to the Protocol, it is about following not just the letter of that Agreement but its spirit. It means avoiding weak, implausible interpretations of that Agreement, and working through differences, not looking for reasons to get around the obligations. At the domestic level, however, in practice the primary role of Article 5 WA will be ‘gap filling’: where gaps in coverage are identified, or where anomalous results are likely to arise, or where the issue is simply not otherwise addressed in the text of the Protocol. Where there is no express provision to deal with situations, it is likely to be argued that because the Protocol is intended to meet a specified objective or contains a general obligation, Article 5 WA imposes a specific obligation to do or not to do something. The Northern Ireland legal system, in particular, is so multilayered³¹ that it would be unlikely for gaps not to be identified or for anomalous consequences not to be detected, and so Article 5 WA may become of critical importance.