64.Any new UK-EU relationship will need to be embedded in an institutional and legal structure. Hitherto much debate has focused on ‘off the shelf’ models, ranging from the ‘Norway model’ to a ‘Canada-style’ Free Trade Agreement. In this chapter we briefly review the available models, in light of the benefits outlined in the previous chapter.
65.The European Free Trade Association (EFTA) is an intergovernmental organisation that promotes free trade and economic integration. It was founded in 1960 by seven states (including the United Kingdom), which were at that time either unable or unwilling to join the European Economic Community, the precursor to the EU. Five states, including the UK, have left EFTA at the same time as joining the EEC/EU, while two others have joined, so EFTA now comprises four members: Norway, Switzerland (both founder members), Iceland and Liechtenstein. Of these, all but Switzerland have subsequently joined the European Economic Area, so there are currently two variants of the EFTA model.
66.Switzerland is the only EFTA state not to be party to the EEA Agreement. Its relationship with the EU is governed by over 100 bilateral agreements, many of which are linked, with the result that, in return for preferential market access for air transport, carriage of goods by rail and road, trade in agricultural products, mutual recognition, government procurement and scientific co-operation, Switzerland is also required to accept the principle of freedom of movement. Its attempts to restrict such free movement since a 2014 referendum have led the EU in return to limit Swiss access to Erasmus and Horizon funding. Switzerland does not enjoy general access to the Single Market in financial services, and its per capita contributions to the EU budget are considerably lower than those of an EU Member State.
67.Disputes between the EU and Switzerland are handled at a political level by a Joint Committee—there is no judicial oversight. Dr Markus Gehring, of Cambridge University, giving evidence to our inquiry into the options for trade in late 2016, argued that the absence of independent judicial oversight had in reality only increased the indirect authority of the CJEU: “It is basically impossible for the Swiss side to get any change negotiated in the joint committee, because the Commission officials feel legally bound by the definitive judgment of the Court of Justice.”
68.The European Economic Area (EEA) Agreement dates from 1994. It brought together all the EU Member States and Norway, Iceland and Liechtenstein. EEA membership enables these three states to participate fully in the EU Single Market, including the Single Market in services. In return, they comply with the ‘four freedoms’, including free movement of persons. Norway makes substantial contributions to the EU budget, as part of the Multiannual Financial Framework; Liechtenstein’s per capita contributions are lower, and Iceland is a net recipient of EU funds.
69.The EEA states are required to implement into national law all EU Single Market legislation, which includes legislation on consumer protection, company law, environmental protection and social policy, and the EEA Agreement itself is constantly updated with the introduction of new EU legislation. Since 1994, more than 5,000 new legal acts have been incorporated into the Agreement either as Annexes or Protocols. Nevertheless, the principles of direct effect and primacy of EU over national law are not part of EEA law, and the EFTA Court, which oversees the Agreement, has refused to include them in the EEA legal order.
70.Although the three non-EU EEA countries are a part of the Single Market in services, including financial services, they are not part of the EU’s customs union, nor is EFTA itself a customs union. EFTA countries thus have the autonomy to negotiate Free Trade Agreements with third countries, either independently or through EFTA. This also means that there is a customs border, albeit one that is typically administered with a light touch, between EU and EEA states (notably between Sweden and Norway).
71.The EEA Agreement excludes the common agriculture and fisheries policies; the Common Foreign and Security Policy; Justice and Home Affairs; and Economic and Monetary Union.
72.This short summary demonstrates that EFTA/EEA membership would deliver some of the benefits sought by the UK Government, albeit at a cost. It would enable full participation in the Single Market for goods and services, though this would require a significant sacrifice of regulatory autonomy. Substantial financial contributions to the EU could be required, though the exact level would be dependent on the UK’s participation in EU programmes. The UK would also be required to accept free movement of people, although Article 112 of the EEA Agreement allows EEA states to apply an ‘emergency brake’ in certain circumstances. EEA membership would not entail accepting the direct effect of EU law, or the direct jurisdiction of the CJEU, though in practice Single Market legislation would have to be implemented, and the UK would have to accept the jurisdiction of the EFTA Court. In return, the UK would be able to nominate a judge (or perhaps judges) to sit on the EFTA Court. EEA/EFTA membership would not entail membership of the customs union, allowing the UK to pursue an independent trade policy, and independent policies on farming and fisheries. This would also entail the creation of a customs border, not least to ensure compliance with ‘rules of origin’, which specify the levels of non-domestic content that may be incorporated into manufactured goods, so that the correct tariff can be levied. Trade in food and fish would have to be regulated by separate bilateral treaties.
73.Catherine McGuinness, Chair of the Policy and Resources Committee of the City of London Corporation, while indicating that the Corporation was “trying to operate within the red lines that have been put down” by the Government, said that Single Market membership via the EEA would “give us what we are looking for”, in particular access to talent and free trade in services. Frances O’Grady, of the TUC, also supported Single Market membership, while Dr Sylvia de Mars noted that EEA/EFTA membership could go some way to meeting the Government’s objections in respect of the direct jurisdiction of the CJEU, in that “there could be a judicial tribunal independent from the Court of Justice that, generally speaking, made its own decisions, purely to do with EFTA, and might occasionally ask a question of the Court of Justice”.
74.As for EFTA membership outside the EEA, Daniel Hannan MEP was “in favour of our rejoining EFTA”. He acknowledged the Government’s ‘red line’ of taking back control of the UK’s borders and immigration policy, but saw “absolutely no reason why you, as a sovereign Parliament, should not agree an immigration policy based on the principle that EU nationals … have a presumption that they can take up jobs”. He has developed a similar argument in published articles, including a paper published by the Bruges Group, in which he outlined the benefits enjoyed by both EFTA and EFTA/EEA states.
75.In 2010, however, the Council of the European Union described the model of EU-Swiss relations as “complex”, “unwieldy to manage”, and as having “clearly reached its limits”. In the absence of any appetite on the EU side to enter into such an arrangement, and given the tensions and difficulties already alluded to, the ‘Swiss model’, or a variant thereof, may not be feasible.
76.The EU’s customs union has a Common External Tariff, which is imposed on all goods imported from third countries. It is enforced through the Customs Union Code, and almost 80% of the revenue generated by tariffs go directly to the EU’s budget. Once goods have been imported into the customs union, they may move freely, without further controls, across the EU.
77.There appears to be no prospect of a non-EU member state being part of the EU’s customs union, which, dating back to the Treaty of Rome, is entwined within the EU’s institutional and legal structures. A more feasible alternative would be to form a separate customs union with the EU customs union.
78.The only precedent for such an arrangement is the EU-Turkey customs union, which dates from 1995. This covers industrial goods, but excludes agriculture and services. It is generally acknowledged to be unsatisfactory and one-sided, in that Turkey is required to apply the EU’s external tariff, but does not benefit from the EU’s free trade agreements. This means that Turkey has to waive tariffs on goods imported from, say South Korea, but South Korea is not in return required to waive tariffs on Turkish goods. Nor does the EU-Turkey customs union deliver one of the essential benefits of a customs union, the avoidance of border controls—as Daniel Hannan pointed out, the EU-Turkey border remains “heavily policed and in some places militarised”. In December 2016 the Commission called for modernisation of the EU-Turkey customs union, describing it as “less and less equipped to deal with the modern day challenges of trade integration”.
79.The Turkey precedent does not preclude a more comprehensive, bespoke customs union. As we have noted, Carolyn Fairbairn told us of the CBI’s support for a customs union (preferably alongside Single Market membership) to facilitate free trade in goods, not least for the “150,000 businesses in the UK that only export to the EU”.
80.Catherine McGuinness, on the other hand, noted that the Corporation, with its strong interest in the financial services sector, “do not take a position on the customs union, because, frankly, it is not directly applicable to us”. Frances O’Grady also acknowledged that a customs union “does not deal with the 80% of the economy in the services sector”.
81.What is clear, however, is that a customs union would prevent the UK from pursuing an independent trade policy—its trade policy would be tied to that of the EU. But there is no agreement on whether that would mean subservience to EU trade policy or something closer to a partnership of equals. For Daniel Hannan it meant “giving the EU 100% control of our trade policy, with zero input into what that trade policy should be”. Jude Kirton-Darling, on the other hand, believed that an offer to enter into a customs union would unlock concessions from the EU side: “If the UK decided to be part of a full customs union with the EU, there would be an arrangement to ensure that the UK was part of the decision-making as well, because we are such a big economy.”
82.The Prime Minister, in her Mansion House speech, proposed a ‘customs partnership’ with the EU as one of two potential options for the future UK-EU customs arrangement. Although the Government has yet to publish a detailed proposal, she indicated that this would see the UK “mirror the EU’s requirements for imports from the rest of the world, applying the same tariffs and the same rules of origin as the EU for those goods arriving in the UK and intended for the EU”. At the same time, she said, “the UK would also be able to apply its own tariffs and trade policy for goods intended for the UK market”.
83.No clear precedent for such a customs partnership exists, and media reports suggested that the EU quickly dismissed the UK’s proposals—the Secretary of State, Rt Hon David Davis MP, giving evidence on 1 May 2018, confirmed that the Commission had “pushed back”. There has also been protracted discussion within the Cabinet on how such a partnership would work. In the absence of detailed proposals, Joe Owen was unsure whether “we would ask every country with an external EU border also to run our tariff regimes”, while Dr Meredith Crowley warned us that it “would impose very significant costs on smaller businesses”, and that the administrative costs might “exceed any benefits of having the flexibility to negotiate lower tariffs … with third-country partners”. The complexity might also make it harder to negotiate free trade deals with third countries: “If Mexico knows that my free trade access to the UK is only through the sort of complex bureaucratic system … it might make the UK less desirable”. Finally, she said, “the real concern about the customs partnership is long-term regulatory divergence”, which could reopen “the problem of whether you want to institute border checks”.
84.The second option proposed by the Prime Minister was a streamlined customs arrangement known as ‘maximum facilitation’. Such an arrangement appears to be premised upon the existence of a customs border, alongside administrative and technological measures to make trade across that border as frictionless as possible. At this stage it is difficult to say with certainty what benefits or costs it would entail, though on 23 May 2018 the Chief Executive of HMRC, Mr Jon Thompson, giving evidence to the House of Commons Treasury Select Committee, estimated the cost to businesses as “between £17 billion and £20 billion annually”. Mr Davis told us that the EU also “pushed back” on this proposal.
85.A Free Trade Agreement (FTA) is the Government’s preferred model, and is also envisaged by the EU, in light of the UK Government’s ‘red lines’. The difference between the two sides is over the scope and depth of such a FTA, with the Prime Minister calling for “the broadest and deepest possible partnership—covering more sectors and co-operating more fully than any Free Trade Agreement anywhere in the world today”. While the detail of the UK’s proposals has yet to be published, the Prime Minister envisaged “reciprocal binding commitments to ensure fair and open competition”; a “completely independent” arbitration mechanism; ongoing dialogue and regulatory cooperation; an arrangement for data protection going beyond the EU’s existing system of adequacy decisions; and an arrangement which, while bringing free movement to an end, would “facilitate” the “valuable links” between UK and EU citizens, also allowing “businesses across the EU and the UK … to attract and employ the people they need”.
86.There can be little doubt that a FTA of the kind proposed by the Prime Minister, were it to be feasible, would deliver substantial benefits to the UK, while in large part protecting the Government’s ‘red lines’. Underpinning it, and supported by ongoing regulatory dialogue, would be “a comprehensive system of mutual recognition”. As we have noted, Daniel Hannan pointed to the Australia-New Zealand FTA as a precedent for such a system, and Catherine McGuinness, of the City of London Corporation, supported the Government’s objective of securing “mutual market access on the basis of mutual recognition of regulatory outcomes”.
87.But so far, as Joe Owen noted, the EU “has shown little interest in that kind of position”. Sylvia de Mars developed this point, noting that mutual recognition as it exists in the EU’s agreements with third countries is “a very different animal” from mutual recognition within the Single Market: the former typically involves “the mutual recognition of conformity assessment, whereby [the EU] will say, ‘We trust that when you, a Canadian laboratory, evaluate the safety of this children’s toy, it meets the EU standards, so, fine, it’s welcome to come over’”. Mutual recognition within the EU Single Market, on the other hand, “is significantly more to do with having identical regulation that is identically enforced than it is with understanding the differences in each other’s regulations. There is so much more harmonised legislation than there is acknowledgement of concrete differences.” Carolyn Fairbairn, of the CBI, also suggested that the “fear among the EU 27 that we may be tempted to tear up some of the regulation that has been developed during the past 10 years” could undermine hopes for a system of outcome-based mutual recognition.
88.The EU’s proposal, in contrast, is more limited, drawing heavily on the EU’s existing FTAs with Canada and South Korea. The European Council in March proposed “a balanced, ambitious and wide-ranging free trade agreement (FTA) insofar as there are sufficient guarantees for a level playing field”. This would include tariff-free trade in goods, “appropriate customs cooperation”, and a framework for “voluntary regulatory cooperation”. There was no acknowledgement of the possibility of a wider agreement on mutual recognition, and the offer in respect of trade in services was limited to “allowing market access to provide services under host state rules, including as regards right of establishment for providers, to an extent consistent with the fact that the UK will become a third country and the Union and the UK will no longer share a common regulatory, supervisory, enforcement and judiciary framework”.
89.Just as the Irish land border brings into sharp focus the trade-offs between the various benefits sought by the Government and its ‘red lines’, so none of the existing models that we have described is sufficient in itself to meet the desire of both sides to avoid the creation of a ‘hard border’. A genuinely open border, without physical infrastructure, requires the free, uncontrolled movement of goods. That in turn requires the avoidance not only of tariffs, but of non-tariff barriers. Within the EU, the former is addressed by means of the customs union; the latter by means of the harmonised rules of the Single Market.
90.The March 2018 European Council guidelines stated: “Divergence in external tariffs and internal rules as well as absence of common institutions and a shared legal system, necessitates checks and controls to uphold the integrity of the EU Single Market as well as of the UK market.” This is borne out by the key sentence in paragraph 49 of the December 2017 Joint Report, which we have already quoted, according to which the United Kingdom undertook, in the absence of agreed solutions, to “maintain full alignment with those rules of the Internal Market and the Customs Union which, now or in the future, support North-South cooperation, the all-island economy and the protection of the 1998 Agreement” (our emphasis). In other words, neither customs union nor Single Market membership is sufficient in isolation to allow an open border.
91.As we noted above (see paragraphs 58–59), the geographical scope of the reference to ‘full alignment’ in paragraph 49 of the Joint Report remains contested. In Jude Kirton-Darling’s view, “The agreement is that if no alternative is put forward by the UK there will be no hard border on the island of Ireland and no hard border in the Irish Sea, and there will be regulatory alignment, which basically covers the single market in a significant number of areas, and the customs union.”
92.Others offered different emphases. Daniel Hannan told us that “the customs union is a complete red herring here. As the EU itself has accepted in its internal memos, British membership of the customs union does not obviate the need for border checks in Ireland.” Carolyn Fairbairn, on the other hand, supported customs union membership specifically to address the Irish border: “If there is a better solution to the Northern Ireland border, we would like to hear it.”
93.The models outlined in this chapter all relate to the future economic partnership. Separate agreements will be needed to achieve the wider benefits of the future UK-EU relationship. While we have not sought detailed evidence on these other agreements in this inquiry, their existence adds substantially to the scale and complexity of the forthcoming negotiations, and we briefly list them here:
94.The number of sectors in which agreements will be needed means that the final UK-EU relationship is likely to be composite: the economic relationship, however framed, will be just one of many elements. The EU’s chief negotiator, Michel Barnier, has proposed grouping the various agreements under four ‘pillars’: trade, socio-economic cooperation (including a range of sectoral agreements), internal security and law enforcement, and foreign and defence policy. The European Council has also identified “socio-economic cooperation”, “law enforcement and judicial cooperation in criminal matters”, and “cooperation in the fields of foreign, security and defence policy” as distinct priorities. The Government has divided its proposals for a “deep and special partnership” into three broad headings: economic partnership, security partnership, and cross-cutting cooperation.
95.Only the European Parliament, however, has proposed a specific model that would bring the various elements of the UK-EU relationship together. The Parliament resolution in March 2018 states that an Association Agreement “provides a flexible framework allowing for varying degrees of cooperation across a wide variety of policy areas”. The resolution continues:
“An association agreement negotiated and agreed between the EU and the UK following the latter’s withdrawal pursuant to Article 8 TEU and Article 217 TFEU could provide an appropriate framework for the future relationship, and secure a consistent governance framework, which should include a robust dispute resolution mechanism, thus avoiding a proliferation of bilateral agreements and the shortcomings which characterise the EU’s relationship with Switzerland.”
96.The key features of Association Agreements are described in Box 1.
Since the 1960s, the (now) EU has negotiated Association Agreements with non-Member States (third countries). There is no template, but these agreements are adopted under Article 217 (TFEU) (or earlier variants), which defines such agreements as “establishing an association involving reciprocal rights and obligations, common action and special procedure”.
These agreements are so-called ‘mixed agreements’, where the EU and the individual Member States share competence. This means that they must be ratified by the European Parliament, the Council of the EU, and all individual member states under their national procedures for the ratification of international agreements. In some Member States this process includes agreement by regional parliaments.
The Institute for Government (IfG) says that the rights and obligations that arise from an individual Association Agreement with the EU depend on the “exact content of [the] individual agreement”. But it suggests that Article 217 (TFEU) imposes “three main criteria”:
(1)the agreement must create privileged links between the EU and the third country that foster and encourage wide-ranging co-operation;
(2)both the EU and the third country must have reciprocal rights and obligations in their partnership; and
(3)the agreement must include institutions designed to implement and monitor the agreement, such as an Association Council (a Ministerial-level group) and an Association Committee (see below).
The IfG also says that Association Agreements are “typically characterised by four other features”:
(1)a free trade agreement with the EU, the contents of which vary country by country;
(2)access to the Single Market, though the third country is often required to implement part of the relevant EU regulatory framework;
(3)opportunities for co-operation beyond trade in areas of mutual interest, including defence and security, the environment and energy, science and education; and
(4)the inclusion of a clause dealing with respect for human rights and democratic principles.
The EU has negotiated Association Agreements with a number of countries, including near neighbours such as Ukraine, Moldova, and Georgia; with North African and Middle Eastern countries such as Tunisia, Egypt, Lebanon and Israel; and countries further afield such as Chile. In the 1990s a suite of special Association Agreements called ‘Europe Agreements’ were used to prepare the thirteen states that joined the EU since 2004 for membership.
In 1963 the (then) EEC agreed an association with Turkey (the Ankara Agreement). It addressed agricultural products and the free movement of persons, and included a provision to “examine the possibility” of Turkish accession. It covered eight pages and 33 Articles. To ensure the “implementation and the progressive development” of the relationship it created a “Council of Association”, with membership drawn from the Turkish Government, the Commission and the Council.
After more than ten years of negotiation the EU and Ukraine in 2014 signed an Association Agreement which sought to create a “Deep and Comprehensive Free Trade Area”; it came into force on 1 September 2017. The agreement took account of the “close historical relationship and progressively closer links” between the EU and Ukraine, and sought to “strengthen and widen relations in an ambitious and innovative way”. It acknowledged “the European aspirations” of the Ukraine and welcomed its “commitment to building a deep and sustainable democracy and market economy”.
Among many matters, the agreement covered: foreign and security policy; regional stability; conflict prevention; non-proliferation of nuclear weapons; anti-terrorism measures; cooperation on border management and asylum and migration; criminal measures including drug-trafficking and corruption; trade and customs arrangements; cross-border supply of services, including financial services; public procurement; intellectual property and trademarks; and, competition and state aid. The agreement without annexes and protocols is over 170 pages and includes 428 Articles.
Its institutional framework includes regular high-level summit meetings; regular ministerial meetings within an ‘Association Council’; regular senior civil servant meetings through the ‘Association Committee’; a ‘Parliamentary Association Committee’ within which members of the European Parliament and the Ukrainian Parliament meet; and a framework within which regular meetings between representatives of civil society are promoted.
97.Several witnesses supported the European Parliament in advocating an Association Agreement. Daniel Hannan MEP noted that the term was “fairly broad”, allowing flexibility to both sides, but saw it as “a viable option … It seems to me that a narrow vote is not a mandate to walk away, even if that were desirable, which I do not think it is. So some kind of country membership and some kind of associate status is an option.”
98.Jude Kirton-Darling agreed that the “very broad-brush” concept of an Association Agreement could benefit both sides: “I suspect that we will want pillars of ongoing co-operation post Brexit. Different pillars are under discussion in the final agreement, whether they are security policy, co-operation on trade or cultural co-operation. An association agreement allows an evolutionary relationship, which is crucial.” Sylvia de Mars argued that an Association Agreement “plays to the UK’s interests significantly, because if it is an association agreement, giving the UK a special deal on services becomes a possibility … whereas if it is just a free trade agreement, any deal that the UK gets would have to be extended to all the EU’s other trading partners”.
99.One of the key advantages of embracing the Association Agreement model could be a change in the tone of the negotiations. This was articulated by Carolyn Fairbairn:
“Can we get away from the language of ‘free trade agreement’ on to the area of ‘association agreement’, which is the European Parliament language? I know that there are challenges with that, but if, between now and October, we can start to talk about partnership, rather than a Canadian variant, that would be very helpful.”
100.The most detailed account of a possible Association Agreement was provided by former MEP Andrew Duff. He argued that the EU’s Association Agreements with Ukraine (for which see Box 1), Georgia and Moldova disproved the Commission’s assertion that “there is nothing to choose between Canada and Norway”. In contrast, these Agreements, intended to lead progressively towards accession (though there is nothing to suggest that Association Agreements need be confined to potential accession countries), were “designed to be more dynamic than either the EEA or CETA”. Within a single “portmanteau treaty”, there could be agreements on trade, mutual recognition, mobility of workers, and security. The Association Agreement would be supported by “permanent joint institutions established at summit, ministerial, parliamentary and technical levels, a platform for civil society and a tripartite tribunal for the arbitration of disputes”.
101.When the Secretary of State, Mr Davis, gave evidence on 1 May 2018, we asked him about the European Parliament’s proposal for an Association Agreement. While he warned that he would not want such an agreement to “bring in ECJ jurisdiction”, he did not rule it out: “As a concept, I have no intrinsic objection to it.”
102.The various models proposed for the future UK-EU relationship all deliver benefits, but all do so at a cost. Compromises will be needed if the two sides’ respective ‘red lines’ are not to preclude the deep partnership to which both aspire.
104.The existing models, such as EEA/EFTA, or a Canada-style Free Trade Agreement, are helpful in clarifying the options, but they must not be allowed to dictate the terms of the negotiations. Existing models fail to capture the full possibilities of the UK-EU relationship—but a gesture of good will, from one or other side, may be needed to unlock these possibilities.
105.We note the European Parliament’s support for a UK-EU Association Agreement, and applaud the Parliament’s readiness to contemplate innovative approaches to the future UK-EU relationship. We also note that Association Agreements are by their nature dynamic and evolutionary, and that a UK commitment to such a partnership could bring about a positive change in the tone and language of the negotiations.
51 See European Union Committee, (5th Report, Session 2016–17, HL Paper 72), para 144.
52 See Zsolt Darvas, ‘Single market access from outside the EU: three key prerequisites’, 19 July 2016: [accessed 24 May 2018].
53 A fuller discussion of the EFTA Court is contained in European Union Committee, (15th Report, Session 2017–19, HL Paper 130).
54 The European Union Committee explored the mechanics of operating controls at the EU’s external border with EEA/EFTA states in a meeting with Swiss and Norwegian border officials on 6 February 2018: see oral evidence taken on (Session 2017–19).
55 While it is not entirely clear whether the EFTA Court is bound by the post-1992 judgments of the CJEU, the system is designed to foster a homogenous legal order, and thus any divergence from CJEU jurisprudence is likely to be, at the most, at the margins. For more on this, see European Union Committee, (15th Report, Session 2017–19, HL Paper 130), paras 34–49.
59 Daniel Hannan MEP The Bruges Group, The case for EFTA, (no date): [accessed 24 May 2018]
60 The Council of the European Union, ‘Council conclusions on EU relations with EFTA countries’, 14 December 2010, para 6: [accessed 25 May 2018]
62 European Commission, ‘Commission proposes to modernise the Customs Union with Turkey’, 21 December 2016: [accessed 24 May 2018]
67 Rt Hon Theresa May MP, speech on ‘our future economic partnership with the European Union’, 2 March 2018: [accessed 24 May 2018]
68 Oral evidence taken on 1 May 2018 (Session 2017–19), (Rt Hon David Davis MP)
71 House of Commons Treasury Select Committee, oral evidence taken on 23 May 2018 (Session 2017–19), (Mr Jon Thompson)
72 Oral evidence taken on 1 May 2018 (Session 2017–19), (Rt Hon David Davis MP)
73 For a description of the process for granting adequacy decisions, see European Union Committee, (3rd Report, Session 2017–19, HL Paper 7).
81 See for instance HM Government, Framework for the UK-EU partnership: economic partnership, May 2018, p. 4: [accessed 25 May 2018]
82 European Parliament, ‘Guidelines on the framework of future EU-UK relations’, 14 March 2018, Preamble, para J: [accessed 5 June 2018]
83 European Parliament, ‘Guidelines on the framework of future EU-UK relations’, 14 March 2018, para 5: [accessed 5 June 2018]
84 Institute for Government, ‘Association Agreements’, (March 22 2018): [accessed 24 May 2018]
89 Written evidence from Andrew Duff ()
90 Oral evidence taken on 1 May 2018 (Session 2017–19), (Rt Hon David Davis MP)