Tuesday, 4 April 2017

Guide to the Brexit Negotiations

Professor Steve Peers

Last week the Brexit process formally got underway, as the UK formally sent the ‘Article 50’ withdrawal letter to the EU on March 29, and the EU in turn drew up a draft of its Brexit negotiating guidelines.

The following is a detailed annotation of the text of the EU’s draft guidelines, which I compare throughout to the UK position – which is most fully set out in Theresa May’s Chatham House speech in January (discussed here). The draft guidelines might be changed before final adoption (due for April 29), but at present it seems unlikely there will be radical changes (if any), so my analysis is based on the text as it now stands. I will update this blog post if there are significant amendments.

I also make reference to the draft position of the European Parliament (EP). This is less important than the draft EU position, since the European Parliament is not (even partly) the negotiator in the talks – no matter how much it might claim to be, or how much some in the UK media believe that it is.  However, the EP position is still worth considering because the EP has a veto over the final Article 50 deal, and this power could influence the actual negotiators’ position. (There’s a full discussion of the EP’s role here). My comments are both legal and political – although we should keep in mind that most legal analysis about Article 50 TEU (which sets out the process of leaving the EU) is purely speculative in the absence of any ECJ case law.

While the initial attention in the UK has focussed on a misunderstanding of what the draft EU guidelines say about Gibraltar, the most significant issue is actually that the EU and UK in principle have many negotiating objectives in common. Most notably, the EU has accepted the UK’s objective of aiming towards a comprehensive EU/UK free trade agreement (FTA) without provisions on the free movement of persons or contributions to the EU budget.   

The devil will therefore be in the considerable details. For example, the EU and UK still disagree on the timing of Brexit talks, (possibly) the role of the ECJ, financial issues (the ‘divorce bill’), the Gibraltar issue (although this will be an issue for bilateral talks with Spain), whether the UK should comply with environmental and other standards as part of a deal, and whether ‘sectoral’ deals are possible. Further points of detailed disagreement will surely emerge as the talks get underway.

Is an EU/UK relationship based on a free trade deal the right way forward? Frankly, in my view, it’s not the first or even the second best option. But it is still a vastly better option than reversion to trade with the EU on a ‘WTO-only’ basis, as some are contemplating. This would lead to significant tariff and non-tariff barriers on many goods and services traded, and therefore hurt both sides.

There’s a democratic dimension to this too. Many of those voting ‘Leave’ were particularly concerned about EU budget contributions and the free movement of people, and an UK/EU FTA would resolve both concerns. But equally the ‘Leave’ side argued repeatedly that any concerns about a drop in UK trade with the EU were misplaced, because the UK and the EU would sign a free trade agreement without commitments relating to the free movement of persons or EU budget contributions.

They also claimed that such an agreement could be reached quickly and easily, on the UK’s terms. This was an obvious falsehood, as the EU’s draft negotiating guidelines have confirmed.  But rather than focus on this untruth, I believe we should move forward. It’s up to the government to move on with the negotiations, never forgetting that a reversion to a WTO-only trading relationship with the UK’s largest trading partner would both make a nonsense of the government’s aim of a free-trading ‘Global Britain’ and lack the slightest shred of democratic legitimacy. Moreover, it would damage both sides, and in particular put many people in the UK out of a job. These negotiations won’t be easy, but they are necessary. The time for fantasies, flag-waving and faffing about is over.

Annex I

On 29 March 2017, the European Council has received the notification by the United Kingdom of its intention to withdraw from the European Union and Euratom. This allows for the opening of negotiations as foreseen by the Treaty.

Note that the guidelines refer throughout to ‘negotiations’. Some have erroneously asserted that the EU will draft a text and present to the UK on a ‘take-it-or-leave-it’ basis. This is incorrect: Article 50(2) TEU refers several times to ‘negotiations’, and the draft guidelines reflect this.

The guidelines do not refer to any possible withdrawal of the UK notification, although the EP resolution (point L) assumes that a withdrawal of the notification is possible with consent. There is also an argument that the time period for withdrawal in Article 50 could be extended indefinitely. I will return to this issue another time.

Some have speculated that the EP might insist that the UK hold a referendum result on the outcome of the talks, as a condition for its consent to the deal. This would almost certainly backfire spectacularly, and in any event the draft EP resolution does not address this.

European integration has brought peace and prosperity to Europe and allowed for an unprecedented level and scope of cooperation on matters of common interest in a rapidly changing world. Therefore, the Union's overall objective in these negotiations will be to preserve its interests, those of its Member States, its citizens and its businesses.

The United Kingdom's decision to leave the Union creates significant uncertainties that have the potential to cause disruption, in particular in the UK but also in other Member States. Citizens who have built their lives on the basis of rights flowing from the British membership of the EU face the prospect of losing those rights. Businesses and other stakeholders will lose the predictability and certainty that come with EU law. With this in mind, we must proceed according to a phased approach giving priority to an orderly withdrawal.

The EU expressly asserts here its goal of an ‘orderly withdrawal’. Indeed Article 50(2) creates an obligation to negotiate with the withdrawing Member State (implicitly in good faith, as with any other negotiation under international law). Legally the wording of Article 50 equally suggests it aims to ensure an orderly withdrawal, and the ECJ is very likely to interpret it the same way. This necessarily implies that the EU cannot simply walk away from negotiations. However, in the event of a legal challenge it is unlikely (based on its prior case law on the EU’s international relations, which emphasises political discretion for the EU institutions) that the ECJ would rule that the general obligation to negotiate requires the EU to adopt (or refrain from adopting) specific negotiation objectives.

Politically this paragraph (and other similar references) confirms that the EU is aiming to reach a deal with the UK. So it would be false for anyone to assert that it is not.

In these negotiations the Union will act as one. It will be constructive throughout and will strive to find an agreement. This is in the best interest of both sides. The Union will work hard to achieve that outcome, but it will prepare itself to be able to handle the situation also if the negotiations were to fail.

The first sentence confirms that the EU will negotiate as a bloc, as Article 50 provides for (‘the Union shall negotiate and conclude an agreement with that State’). This does not preclude informal bilateral talks between the UK and key Member States’ governments, which is common when the EU discusses trade deals with non-EU states.

The overall tone here is diplomatic and asserts an intention to reach a deal, although the final sentence also asserts that the EU will prepare itself for a ‘no deal’ scenario as an alternative. This would likely entail getting draft EU laws ready for approval and gearing up national customs administrations et al for a UK departure presumably in one of two scenarios: a) on the date set out pursuant to Article 50; b) if the UK purports to leave the EU in breach of the terms of Article 50. In the latter scenario the EU Member States might not wait for the end of the Article 50 period, but declare that the UK has committed a material breach of the EU Treaties within the meaning of Article 60 of the Vienna Convention on the Law of Treaties, and suspend the UK from the Treaties.

The guidelines make no mention of amending the Treaties to remove reference to the UK, but one would not expect them to, since logically that is not an issue which the EU would negotiate with the UK.  It is not yet clear whether the EU will embark upon such a course; legally it might be argued that Article 50 is ‘self-executing’ in the sense that references to the UK automatically become legally irrelevant on Brexit Day. One issue might be whether other changes to the Treaties might be made at the same time; if so, this would become difficult to negotiate. Even a purely technical Treaty amendment which only removes references to the UK would take some time to ratify, so one would expect work to get underway soon, if it is going to happen – although there could be a legal problem in that in principle the UK ought to participate in the Treaty amendment process as long as it is a member. One solution might be to draw up such a treaty after Brexit Day, and backdate its application date.

Paras 30 and 32 of the draft EP resolution refer to possible changes to EU law as a result of the UK’s departure as an issue for the EU-27, although they make no specific reference to Treaty amendment.

These guidelines define the framework for negotiations under Article 50 TEU and set out the overall positions and principles that the Union will pursue throughout the negotiation. The European Council will remain permanently seized of the matter, and will update these guidelines in the course of the negotiations as necessary.

Article 50(2) states that the negotiation must take place ‘in light of’ these guidelines. It is unclear precisely what legal effect this has. For instance, could an outvoted Member State challenge the Article 50 deal for breaching the guidelines? As it happens, there is a pending ECJ case which raises (among other things) the legal effect of European Council guidelines in the asylum context, discussed here.

The predominant effect of the guidelines will of course be political; note that the European Council is prepared to ‘update’ them (ie amend or elaborate upon them further). They will necessarily be fleshed out in detail in a rather longer treaty text, on the basis of drafts tabled and negotiated by each side.

I. Core principles

1. The European Council will continue to base itself on the principles set out in the statement of Heads of State or Government and of the Presidents of the European Council and the European Commission on 29 June 2016. It reiterates its wish to have the United Kingdom as a close partner in the future. It further reiterates that any agreement with the United Kingdom will have to be based on a balance of rights and obligations, and ensure a level-playing field. Preserving the integrity of the Single Market excludes participation based on a sector-by-sector approach. A non-member of the Union, that does not live up to the same obligations as a member, cannot have the same rights and enjoy the same benefits as a member. In this context, the European Council welcomes the recognition by the British Government that the four freedoms of the Single Market are indivisible and that there can be no "cherry picking".

The June 2016 statement can be found here. The second sentence matches the UK government’s objective of a future close partnership. Any ‘sectoral’ deals are excluded; this contradicts the UK government position, which had the objective of obtaining such deals. In any event sectoral deals would be potentially vulnerable to challenge under WTO rules, which require any free trade agreement to cover ‘substantially’ all trade in goods (Article XXIV GATT) or services (Article V GATS). It would however be possible for mutual recognition agreements to cover certain sectors of trade. There is no reference in the guidelines to the UK government’s objective to achieve some form of special customs agreement, except implicitly as regards Northern Ireland.

2. Negotiations under Article 50 TEU will be conducted as a single package. In accordance with the principle that nothing is agreed until everything is agreed, individual items cannot be settled separately. The Union will approach the negotiations with unified positions, and will engage with the United Kingdom exclusively through the channels set out in these guidelines and in the negotiating directives. So as not to undercut the position of the Union, there will be no separate negotiations between individual Member States and the United Kingdom on matters pertaining to the withdrawal of the United Kingdom from the Union.

It’s not clear whether the ‘single package’ refers more narrowly to the specific issues which the EU wants to negotiate first, or the talks on later status of UK/EU relations. This is important because there is a strong case for concluding a first treaty on withdrawal issues, including in particular the status of UK and EU citizens who have moved between Member States, separately from any subsequent treaties.

This paragraph reconfirms the ‘no separate negotiations’ point already made in the guidelines. As noted above, this does not prevent some informal bilateral discussions on specific issues. For instance, UK/Ireland discussions on border issues could be significant, and the guidelines in effect require separate talks between the UK and Spain regarding Gibraltar. They also refer to bilateral UK/Cyprus issues.

Article 50 refers to the EU acting by qualified majority vote (without the UK voting), so the common assumption that any Article 50 deal must be agreed unanimously and ratified by all national parliaments is incorrect. However, some aspects of the overall Brexit deal (ie on future relationship) might entail unanimous voting and national ratification; and the EU might be unwilling to proceed with the Article 50 deal against the opposition of one or more Member States.

II. A phased approach to negotiations

3. On the date of withdrawal, the Treaties will cease to apply to the United Kingdom, to those of its overseas countries and territories currently associated to the Union, and to territories for whose external relations the United Kingdom is responsible. The main purpose of the negotiations will be to ensure the United Kingdom's orderly withdrawal so as to reduce uncertainty and, to the extent possible, minimise disruption caused by this abrupt change.

This refers to the entities referred to in Article 353 TFEU, ie Gibraltar, the Channel Islands, the Isle of Man, and some small remaining semi-colonies. For a discussion of their status, see this House of Lords report. The reference to minimising disruption is important, although qualified (‘to the extent possible’).

To that effect, the first phase of negotiations will aim to:
settle the disentanglement of the United Kingdom from the Union and from all the rights and obligations the United Kingdom derives from commitments undertaken as Member State;
provide as much clarity and legal certainty as possible to citizens, businesses, stakeholders and international partners on the immediate effects of the United Kingdom's withdrawal from the Union.
The European Council will monitor progress closely and determine when sufficient progress has been achieved to allow negotiations to proceed to the next phase.

This asserts the EU’s determination to discuss withdrawal issues before trade issues, whereas the UK has wanted to discuss them in tandem. However, the EU guidelines do not require the withdrawal issues to be fully agreed before talks on post-Brexit issues get underway. The ‘sufficient progress’ test is deliberately flexible; it can be argued that agreement in principle on the main issues will be sufficient, even if some details have to be worked out.  Para 14 of the EP draft resolution largely matches this, although the EP will not have any formal role in determining whether ‘substantial progress’ (the slightly stricter test which the EP proposes) has been met.

4. While an agreement on a future relationship between the Union and the United Kingdom as such can only be concluded once the United Kingdom has become a third country, Article 50 TEU requires to take account of the framework for its future relationship with the Union in the arrangements for withdrawal. To this end, an overall understanding on the framework for the future relationship could be identified during a second phase of the negotiations under Article 50. The Union and its Member States stand ready to engage in preliminary and preparatory discussions to this end in the context of negotiations under Article 50 TEU, as soon as sufficient progress has been made in the first phase towards reaching a satisfactory agreement on the arrangements for an orderly withdrawal.

The EP draft resolution (point 15) also rules out conclusion of a ‘future relationship’ treaty before Brexit, so it seems unlikely to happen.

This discussion of a ‘future relationship’ between the UK and the EU (concerning trade and other issues) is separate from a possible ‘transitional arrangement’ referred to in the next paragraph (‘may also’). So it is unclear whether a transitional deal could also only be concluded after the UK leaves.

In any event, it is arguable that a longer-term deal could be concluded (or at least signed and applied provisionally) on the date of withdrawal, or failing that, it could be backdated to that date. Such an approach would be consistent with the EU’s objectives of an orderly withdrawal and minimising disruption. However, the references to ‘preliminary and preparatory’ and ‘overall understanding’ seem to rule that out.

In any event, it would be hard in practical terms to negotiate all the details of the long-term framework within the time available, even though this is Theresa May’s objective. This is particularly true given that some of the available time will be taken up negotiating withdrawal issues before moving on to the ‘future relationship’ talks.

5. To the extent necessary and legally possible, the negotiations may also seek to determine transitional arrangements which are in the interest of the Union and, as appropriate, to provide for bridges towards the foreseeable framework for the future relationship. Any such transitional arrangements must be clearly defined, limited in time, and subject to effective enforcement mechanisms. Should a time-limited prolongation of Union acquis be considered, this would require existing Union regulatory, budgetary, supervisory and enforcement instruments and structures to apply.

It is certainly in the EU’s interest to maintain trade flows and other relationships with the UK, and as noted above this would be consistent with the EU’s objectives of an orderly withdrawal and minimising disruption. They would obviously be ‘necessary’ to the end of avoiding a switch to a WTO-only framework for trade between the UK and the EU. Again as noted already, this paragraph is unclear about when a transitional deal might be concluded, although as compared to the previous paragraph it does not explicitly state that talks could not be completed or that an agreement could not be concluded before Brexit Day. It is also vague about the relevant legal base: is it Article 50 (qualified majority vote) or other Treaty provisions (more likely to entail unanimity and national ratification)? The exact scope of Article 50 is unclear legally; the wording here suggests some uncertainty about what is ‘legally possible’, which it may be necessary to ask the ECJ to clarify (see Annex II).

The transitional deal would have to be limited in time, which could reassure those who do not want the UK to remain in such an arrangement indefinitely. It is not expressly clear that all aspects of EU membership would be carried over. In particular, it is not clear that the only transition which the EU would consider would be continuing in force the EU acquis. The distinction is important because a) flexibility on this issue could involve ending the free movement of people, or continuing it with an ‘emergency brake’, from Brexit Day; and b) an ‘enforcement mechanism’ might fall short of continuing EU ‘existing regulatory, budgetary, supervisory and enforcement’ measures. The latter phrase would logically entail keeping in force the ECJ’s current jurisdiction, while a mere ‘enforcement mechanism’ could entail a more limited role for the ECJ (no references from UK courts or direct impact on UK law) or a different dispute mechanism system entirely, such as using the EFTA Court which already exists and rules on certain EU law issues as regards Norway, Iceland and Liechtenstein. However, the more different a transitional system would be from current EU membership, the longer it would take to negotiate, and the bigger the risk of running out of time.

The UK government’s position seems to contemplate some form of interim transitional period, without calling it such. The Chatham House speech refers instead to ‘phased implementation’, during which aspects of EU law will still apply. But despite the difference in wording, fundamentally there are strong similarities between the UK and EU positions here.

The EP draft resolution is similar (para 28). It sets a maximum time limit of three years for the transitional deal, although otherwise it is as vague as the European Council draft guidelines.

6. The core principles set out above should apply equally to the negotiations on an orderly withdrawal, to any preliminary and preparatory discussions on the framework for a future relationship, and to any form of transitional arrangements.

7. The two year timeframe set out in Article 50 TEU ends on 29 March 2019.

It should be noted that Article 50(3) provides that the withdrawal agreement may set a different date (which could implicitly be later or earlier), and that the UK and the remaining EU, voting unanimously, may postpone the overall deadline.

III. Agreement on arrangements for an orderly withdrawal

8. The right for every EU citizen, and of his or her family members, to live, to work or to study in any EU Member State is a fundamental aspect of the European Union. Along with other rights provided under EU law, it has shaped the lives and choices of millions of people. Agreeing reciprocal guarantees to settle the status and situations at the date of withdrawal of EU and UK citizens, and their families, affected by the United Kingdom's withdrawal from the Union will be a matter of priority for the negotiations. Such guarantees must be enforceable and non-discriminatory.

The EU, like the UK, places priority on an early deal on the rights of each other’s citizens. While some had claimed that the EU could not or would not address this issue in the Article 50 talks, it clearly intends to do so. The reference to ‘non-discriminatory’ arrangements confirms that there will not be bilateral talks on this issue, as some people had expected. While some have claimed that the EU lacks legal power to regulate the position of non-EU citizens, Articles 77-79 TFEU confer such powers and the EU has adopted a number of laws in this area. In any event, it is strongly arguable that Article 50 confers power upon the EU to negotiate any issue which arose pursuant to the withdrawing Member State’s membership of the European Union, and the status of UK and EU citizens is one obvious example of such an issue.

The reference to an ‘enforceable’ guarantee does not necessarily entail using the ECJ. There could be some other form of dispute settlement, or a commitment to make the agreement binding in national law and to hold discussions about any issues which arise might suffice.

It should be noted that the guidelines make no reference to the idea, promoted by some in the EP, of an optional right for UK citizens to retain EU citizenship. The EP draft resolution does refer to this issue, but states that such arrangements should be reciprocal (para 27). It is hard to imagine the UK government agreeing to such a deal.

9. Also, the United Kingdom leaving the Union will impact EU businesses trading with and operating in the United Kingdom and UK businesses trading with and operating in the Union. Similarly, it may affect those who have entered into contracts and business arrangements or take part in EU-funded programmes based on the assumption of continued British EU membership. Negotiations should seek to prevent a legal vacuum once the Treaties cease to apply to the United Kingdom and, to the extent possible, address uncertainties.

This indicates a general intention to retain contracts and legal arrangements in force, if they have been concluded before Brexit Day. This could be relevant to research funding, regional funding or farm subsidies, for instance. It could also be the basis for arguing that UK banks who already have a licence to sell financial services to the EU market can retain it.

10. A single financial settlement should ensure that the Union and the United Kingdom both respect the obligations undertaken before the date of withdrawal. The settlement should cover all legal and budgetary commitments as well as liabilities, including contingent liabilities.

This is likely to be one of the most difficult issues to negotiate. The draft guidelines do not put a specific figure on the ‘bill’, although press reports note amounts such as €50 billion. There is no reason why the amount of any bill should be paid upfront, as some of it relates to funding over the next few years and to pensions payable over the longer term. For detailed discussions of this issue see this House of Lords report and this Brueghel report.

11. The Union has consistently supported the goal of peace and reconciliation enshrined in the Good Friday Agreement, and continuing to support and protect the achievements, benefits and commitments of the Peace Process will remain of paramount importance. In view of the unique circumstances on the island of Ireland, flexible and imaginative solutions will be required, including with the aim of avoiding a hard border, while respecting the integrity of the Union legal order. In this context, the Union should also recognise existing bilateral agreements and arrangements between the United Kingdom and Ireland which are compatible with EU law.

The EU and UK both make a priority of addressing the border issue between Ireland and Northern Ireland. (It should be noted that there is no specific reference to Scotland in the EU guidelines; the EP resolution merely notes that Scotland voted to Remain). There are indications of great willingness to compromise here (‘flexible and imaginative solutions’) and it should be noted that Protocols to the Treaties require the EU to facilitate the operation of the Common Travel Area between the UK and Ireland (the words ‘should also recognise’ these arrangements do not adequately take account of this legal obligation). However, the Protocols do not lay down in detail what happens in the event of UK withdrawal, which therefore has to be the subject of negotiations. Equally, arguments about the history of the UK and Ireland are irrelevant, as there has never been a previous situation when one of those countries was in the EEC/EC/EU and the other was not.

12. The Union should agree with the United Kingdom on arrangements as regards the Sovereign Base Areas of the United Kingdom in Cyprus and recognise in that respect bilateral agreements and arrangements between the Republic of Cyprus and the United Kingdom which are compatible with EU law, in particular as regards the situation of those EU citizens resident or working in the Sovereign Base Areas.

There is a specific Protocol to the 2003 Accession Treaty on the position of the UK sovereign base in Cyprus, which will presumably have to be amended in order to take account of any such agreement. Here, despite the overall EU-wide approach to talks, there is recognition that there will be a bilateral agreement between the UK and Cyprus alongside an EU-UK agreement.

13. Following the withdrawal, the Union with 27 Member States will continue to have the rights and obligations of the Union with 28 Member States in relation to international agreements. The United Kingdom will no longer be covered by agreements concluded by the Union or by Member States acting on its behalf or by both acting jointly. The European Council expects the United Kingdom to honour its share of international commitments contracted in the context of its EU membership. In such instances, a constructive dialogue with the United Kingdom on a possible common approach towards third country partners and international organisations concerned should be engaged.

The EU position is that the UK ceases to be part of any treaty with non-EU countries within the scope of EU law – whether concluded by the EU alone, by the Member States alone (but affecting EU law) or by both together. This is particularly relevant to trade agreements (see discussion by Markus Gehring here) but affects other agreements as well (for example, environmental deals).

The UK will likely seek to secure replacement agreements with the countries concerned. This is even encouraged by the EU, which seeks a ‘constructive dialogue’ on these issues. The reference to ‘international organisations’ most obviously refers to the WTO.

In practical terms, the issue which particularly arises is ‘tariff rate quotas’, ie allowing in a certain amount of products at a low tariff. Say the EU allows 100,000 tons of olives imported from Morocco at a low tariff: the obvious solution is to split that between the UK and the remaining EU based on recent trade flows (ie how much of those olives were imported into the UK in the last three years, as compared to the rest of the EU?).

Non-EU countries will obviously have to agree to this process, and the UK might want to focus particularly on replicating those agreements with a major impact on UK exports: for instance, the EU agreements on the protection of the name ‘Scotch Whisky’.

14. While the future location of the seats of EU agencies and facilities located in the United Kingdom is a matter for the 27 Member States, arrangements should be found to facilitate their transfer.

The intention is obviously to move these bodies as soon as possible after Brexit Day, if not before. The UK will not be involved in decisions on where they move to, but will be involved in the logistics of moving them.

15. Arrangements ensuring legal certainty and equal treatment should be found for all court procedures pending before the Court of Justice of the European Union upon the date of withdrawal that involve the United Kingdom or natural or legal persons in the United Kingdom. The Court of Justice of the European Union should remain competent to adjudicate in these procedures. Similarly, arrangements should be found for administrative procedures pending before the European Commission and Union agencies upon the date of the withdrawal that involve the United Kingdom or natural or legal persons in the United Kingdom. In addition, arrangements should be foreseen for the possibility of administrative or court proceedings to be initiated post-exit for facts that have occurred before the withdrawal date.

The EU assumes that EU court cases pending on Brexit Day should remain within the competence of the Court to decide. This is a classic transitional issue and the EU suggestion makes sense, since the cases concern the pre-Brexit legal position. (Compare to the planned Great Repeal Bill, which will provide that pre-Brexit ECJ case law will still be binding). It is slightly more ambiguous as regards pending administrative procedures like competition law and state aid (‘arrangements should be found’). The same applies to cases with EU law issues pending in the UK courts on Brexit Day, or which are brought in the UK courts in summer 2019 relating to a 2018 tax bill, for instance.

This paragraph is too limited, as the transitional deal needs to take explicit account of all legal proceedings pending on Brexit Day pursuant to EU law, not just those which involve or may involve the EU institutions: for instance a pending claim to recognise a German court judgment in the UK, or the French authorities’ obligation to execute a European Arrest Warrant issued by the UK before that date. Most implementation of EU law is carried out by national courts and administrations, not EU bodies, and the transitional rules should take account of this.

16. The withdrawal agreement should include appropriate dispute settlement mechanisms regarding the application and interpretation of the withdrawal agreement, as well as duly circumscribed institutional arrangements allowing for the adoption of measures necessary to deal with situations not foreseen in the withdrawal agreement. This should be done bearing in mind the Union's interest to effectively protect its autonomy and its legal order, including the role of the Court of Justice of the European Union.

The institutional arrangements will presumably entail an EU/UK Joint Committee with the power to take decisions by joint agreement. There are similar bodies in other EU treaties with non-EU states. The first and third sentences, read together, do not explicitly insist that the ECJ have jurisdiction over the withdrawal agreement (although it will inevitably have jurisdiction as regards the EU side). The reference to ensuring EU autonomy reflects ECJ case law which states that treaties with non-EU countries cannot affect the separate development of EU law or the essential features of the Court’s powers: see Opinion 1/91 and Opinion 1/00. Compare with point 17 of the EP resolution, which explicitly calls for the ECJ to have jurisdiction over the withdrawal agreement.

IV. Preliminary and preparatory discussions on a framework for the Union - United Kingdom future relationship

17. The European Council welcomes and shares the United Kingdom's desire to establish a close partnership between the Union and the United Kingdom after its departure. While a relationship between the Union and a non Member State cannot offer the same benefits as Union membership, strong and constructive ties will remain in both sides' interest and should encompass more than just trade.

The EU accepts in general the UK position of having a close partnership, including but going beyond trade. There is no explicit reference to the form of the relationship, which is relevant given that it could affect whether the EU side has to vote unanimously and ask national parliaments to ratify any treaty. (EU treaties with non-EU states can be partly applied provisionally pending such ratification).Point 22 of the EP draft resolution hints at a possible association agreement: this entails unanimous voting and usually also national ratification.

18. The British government has indicated that it will not seek to remain in the single market, but would like to pursue an ambitious free trade agreement with the European Union. Based on the Union's interests, the European Council stands ready to initiate work towards such an agreement, to be finalised and concluded once the United Kingdom is no longer a Member State.

The EU accepts the UK position of seeking a far-reaching free trade deal, rather than continued single market participation. Note that there is no reference to continuing with the free movement of persons or contributions to the EU budget – two key objectives of the UK side. The timing is an issue, as noted already: no finalisation or conclusion until after Brexit Day, which means that a transitional deal will be important in the meantime. Although in theory this could be on or soon after Brexit Day, there is unlikely to be enough time for that.

19. Any free trade agreement should be balanced, ambitious and wide-ranging. It cannot, however, amount to participation in the Single Market or parts thereof, as this would undermine its integrity and proper functioning. It must ensure a level playing field in terms of competition and state aid, and must encompass safeguards against unfair competitive advantages through, inter alia, fiscal, social and environmental dumping.

It is common for EU trade agreements to make some reference to state aid and competition law, although only some of them require the non-EU state to apply EU competition and state aid law as such. The guidelines leave it open as to what exactly the EU side will aim for. In any event WTO rules also contain some restrictions on granting subsidies and set out the possibility for trading partners to retaliate if subsidies are granted, although these rules are less far-reaching and enforceable than those applying to EU Member States. (Note that the EU does not ban state aids entirely, but restricts them to certain cases).

The ‘safeguards against unfair competitive advantages’ the EU side wants refers to – among other things – ‘fiscal, social and environmental dumping’. This does not explicitly refer to the adoption of EU law by the UK, and in any event there is no EU law on minimum corporate tax rates. (Compare to para 24 of the EP resolution, which refers more explicitly to EU laws, although to ‘tax evasion and avoidance’ as distinct from tax rates). The UK might argue that it would be sufficient to remain party to international treaties on environmental law and social protection (in the ILO, the Council of Europe and the UN Covenant on Economic, Social and Cultural Rights), and to hold regular discussions on corporate tax rates – which could cut both ways. In any event, many in the UK would welcome safeguards against cuts in environmental and social standards, and would be concerned about how public services could be funded in the event of large cuts in corporate tax.

It remains to be seen what exactly the EU side would accept as safeguards, but the EU’s position should be seen in its overall context: as noted already, the EU is not insisting on free movement of people or financial contributions. It is striking that the EU side makes no explicit references to services, where the UK has a trade surplus. If the UK wants to maintain that strong surplus by having an advanced trade relationship that offers more free trade in services than the EU usually agrees (but still falling short of single market participation) this is the condition which the EU wants to set.

20. Beyond trade, the EU stands ready to consider establishing a partnership in other areas, in particular the fight against terrorism and international crime as well as security and defence.

This matches the UK’s position. Neither side sets out any real details here.

21. The future partnership must include appropriate enforcement and dispute settlement mechanisms that do not affect the Union's autonomy, in particular its decision-making procedures.

There is no specific mention of the ECJ here, and the autonomy issue was discussed above. It’s not usual for the Court to have jurisdiction as regards non-EU states, bar a few exceptions like the European Aviation Area treaty, which facilitates aviation between EU and other European countries. But it is common for the EU to agree dispute settlement similar to the WTO dispute settlement system in agreements with non-EU countries. Interestingly, the EU does not use such systems in its trade agreements in practice, although it does often use the WTO.

In short, the WTO system provides for panels of experts to decide on whether there is a breach of WTO law; their decisions can be appealed to an Appellate Body. If a WTO party which was found to have breached WTO law does not comply with these rulings, the victorious party which brought the complaint can be authorised to retaliate against it with proportionate trade sanctions. Dispute settlement bodies are not unique to the EU and WTO – there is an active system in the NAFTA agreement between the US, Canada and Mexico, for instance. Such systems fall short of the legal effect of EU law in national legal systems, but still place some constraints upon the parties to trade treaties.

22. After the United Kingdom leaves the Union, no agreement between the EU and the United Kingdom may apply to the territory of Gibraltar without the agreement between the Kingdom of Spain and the United Kingdom.

This clause has attracted much overreaction. It is not a claim to Gibraltar territory or joint sovereignty, so there is no need to respond to it with talk of military action. It merely acknowledges that the EU will not apply post-Brexit treaties to the UK unless the UK and Spain have separately agreed to this. It will be up to the UK and Spain to find agreement for each treaty, or failing that to accept that the treaty in question will either not be concluded or not apply to Gibraltar.

V. Principle of sincere cooperation

23. Until it leaves the Union, the United Kingdom remains a full Member of the European Union, subject to all rights and obligations set out in the Treaties and under EU law, including the principle of sincere cooperation.

This reflects the UK’s continued position that it will apply EU law until Brexit Day. Point 5 of the EP draft resolution matches it.

24. The European Council recognises the need, in the international context, to take into account the specificities of the United Kingdom as a withdrawing Member State, provided it remains loyal to the Union's interests while still a Member. Similarly the Union expects the United Kingdom to recognise the need of the 27 Member States to meet and discuss matters related to the situation after the withdrawal of the United Kingdom.

The first sentence is ambiguous about a key issue: can the UK already discuss trade deals with non-EU countries? It can’t conclude them without violating EU law (para 23); normally it could not negotiate them either, but does this sentence accept the argument that discussions on a post-Brexit deal would be acceptable, as a consequence of the UK’s position as a withdrawing Member State? The second sentence asserts the remaining EU’s right to meet without the UK, presumably going beyond the talks relating to Brexit without the UK present as referred to in Article 50. However, such meetings must remain informal, as the next paragraph confirms.  

Compare to point 6 of the draft EP resolution, which more explicitly argues that the UK cannot negotiate with non-EU countries before Brexit, and argues that the UK should be excluded from EU trade talks with non-EU countries if it does. Such an exclusion would not be legal; the remedy in such cases of alleged breach of EU law is for the Commission or another Member State to bring the UK to the ECJ.  

25. While the United Kingdom is still a member, all ongoing EU business must continue to proceed as smoothly as possible at 28. The European Council remains committed to drive forward with ambition the priorities the Union has set itself. Negotiations with the United Kingdom will be kept separate from ongoing Union business, and shall not interfere with its progress.

The UK retains its formal position as a Member State until Brexit Day, although obviously it will have declining influence as there will be little interest in addressing its concerns and the other Member States merely have to wait out any veto or participation in a blocking minority vote by the UK.

VI. Procedural arrangements for negotiations under Article 50

The European Council endorses the arrangements set out in the statement of 27 Heads of State or Government on 15 December 2016.

These procedural arrangements are discussed in the next annex.

Annex II – Procedural rules

Here are the December 2016 procedural rules, with annotations.

1. The first step following the notification by the United Kingdom will be the adoption by the European Council of guidelines that will define the framework for negotiations under Article 50 TEU and set out the overall positions and principles that the EU will pursue throughout the negotiation. The European Council will remain permanently seized of the matter, and will update these guidelines in the course of the negotiations as necessary.

This refers to the guidelines of the European Council (the Member States’ Heads of State and Government), the draft of which is now available and was discussed above. Article 50 requires the guidelines to be adopted at the start of the process.

2. After the adoption of the guidelines, the European Council will invite the General Affairs Council to proceed swiftly with the adoption of the decision authorising the opening of the negotiations, following a recommendation by the European Commission, and to deal with the subsequent steps in the process. The Council will also adopt negotiating directives on substance as well as on the detailed arrangements governing the relationship between the Council and its preparatory bodies on the one hand and the Union negotiator on the other. These negotiating directives may be amended and supplemented as necessary throughout the negotiations, to reflect the European Council guidelines as they evolve.

This follows the normal process of EU treaty negotiation with non-EU countries, as set out in Article 218 TFEU: a Commission recommendation to start talks, and a Council mandate with detailed negotiation objectives, which may be amendment throughout the talks. That mandate will likely be more detailed than these guidelines. The General Affairs Council is made up of Member States’ foreign or Europe ministers, and meets monthly. The Council’s ‘preparatory bodies’ consist of working groups and Member States’ permanent representatives (the latter meets weekly, and is known as ‘Coreper’).

3. The Council will be invited to nominate the European Commission as the Union negotiator. The Commission's nomination of Michel Barnier as chief negotiator is welcome. To ensure transparency and build trust, the Union negotiator's team will be ready to integrate a representative of the rotating Presidency of the Council. Representatives of the President of the European Council will be present and participate, in a supporting role, in all negotiation sessions, alongside the European Commission representatives. The Union negotiator will systematically report to the European Council, the Council and its preparatory bodies.

Article 50 leaves open who the negotiators will be, and some had assumed it would be the Council. Nominating the Commission follows the usual approach under Article 218 TFEU. However, including a representative of the Council Presidency (which rotates every six months) and the European Council President (Donald Tusk) in the talks is an innovation, which reflects their importance. The obligation to report back to EU bodies reflects Article 218 TFEU.

4. Between the meetings of the European Council, the Council and Coreper, assisted by a dedicated Working Party with a permanent chair, will ensure that the negotiations are conducted in line with the European Council guidelines and the Council negotiating directives, and provide guidance to the Union negotiator.

Article 218 TFEU (and Article 207 TFEU, regarding trade) refer to Council working parties supervising Commission negotiators, so this is nothing new. As regards trade, the Commission refers to the committee which supervises it as the ‘mothers-in-law’ (don’t shoot the messenger!).

5. The members of the European Council, the Council and its preparatory bodies representing the United Kingdom will not participate in the discussions or in the decisions concerning it.

This simply respects the wording of Article 50, extended (logically enough) to the Council’s working parties, not just to the official meetings of ministers or heads of state.

6. Representatives of the 27 Heads of State or Government (Sherpas/Permanent Representatives) will be involved in the preparation of the European Council as necessary. Representatives of the European Parliament will be invited at such preparatory meetings.

This gives a slightly greater role than usual to the European Parliament.

7. The Union negotiator will be invited to keep the European Parliament closely and regularly informed throughout the negotiation. The Presidency of the Council will be prepared to inform and exchange views with the European Parliament before and after each meeting of the General Affairs Council. The President of the European Parliament will be invited to be heard at the beginning of meetings of the European Council.

The first sentence reflects Article 218 TFEU, which calls for the EP to be fully and immediately informed throughout negotiations. The second and third sentences go into more detail, and give a bigger role to the EP than usual during EU talks with non-EU countries. As noted already, however, the EP is not a negotiator as such, although its power to give its consent to the final deal (not referred to explicitly here) nevertheless means it is a significant actor. The additional meetings referred to here will be a forum for the EP to influence the negotiations.

There is no reference to the ECJ, which can rule on legal issues concerning draft treaties with non-EU countries pursuant to Article 218 TFEU, at the request of the Commission, Council, EP or a Member State. It is not clear whether the whole of Article 218 necessarily applies to the Brexit treaty process, since Article 50 only refers to some of it. There are also other ways that Brexit legal questions can reach the ECJ: national courts can ask the Court about them, the Commission can sue a Member State (or Member States can sue each other), or Member States or the EU institutions can sue the [other] EU institutions for various aspects of their conduct of the negotiation.

Barnard & Peers: chapter 27

Photo credit: the Register

Friday, 31 March 2017

The White Paper on the Great Repeal Bill: Invasion of the Parliamentary Control Snatchers

Professor Steve Peers

Yesterday’s White Paper for the forthcoming Great Repeal Bill outlines the key elements of the domestic law aspects of the process of the UK leaving the EU. It indicates broadly how the future proposal for a Bill will do two main things: (a) convert existing EU law applied in the UK to UK law including, in some cases, law of the devolved bodies in Scotland, Wales and Northern Ireland); and (b) provide for a process of amending ex-EU law in future.

While the White Paper has thought some issues through in detail, there are some key points on which it is either vague or unconvincing (or both). In particular, it contains no real detail or substantiated argument on the most important issue: the power of the executive to amend laws without an Act of Parliament.

Converting EU law

The White Paper confirms that the UK will continue to apply EU law until Brexit Day – which will be March 29, 2019, unless (a) the EU/UK withdrawal agreement specifies otherwise, or (b) the EU and UK agree to extend this date, or (c) the UK rescinds its notification to withdraw from the EU (if that is legally possible). It also confirms the intention to remove the European Communities Act – the main Act of Parliament that gives effect to EU membership in UK law – as of that date.

But it seeks to retain in force the EU rules which apply in the UK as of that date, subject to amendments (as discussed below).  The EU rules in question are directly-applicable Regulations and EU Treaty provisions, as well as Directives which were already transposed into UK law by using the European Communities Act. Also, the Treaties will remain relevant for the interpretation of ex-EU legislation which was initially based on them (para 2.10).

What about the EU courts? Here the White Paper draws a distinction. On the one hand, post-Brexit laws will not be subject to the jurisdiction of the EU Court, and the UK’s courts will not be obliged to take ECJ rulings into account (presumably they could choose to do so, however). On the other hand, ex-EU laws will still have to be interpreted by reference to pre-Brexit case law of the EU courts, to ensure certainty. There’s no reference to post-Brexit case law, but again presumably the UK courts could choose to consider it.

The latter rule will not be absolute, however: pre-Brexit ECJ case law will have the same binding effect as UK Supreme Court judgments, and so the Supreme Court could overrule those judgments in the same limited circumstances in which it can (and rarely does) overrule itself. It will of course be possible for Parliament to overturn ECJ case law by amending the legislation, and the government hints that it might want to consider clarifying the circumstances in which the Supreme Court could overturn prior ECJ rulings.

A similar distinction arises as regards the supremacy of EU law: the EU court rule that EU law takes precedence over any conflicting national law, so the latter has to be disapplied by national courts in the event of any conflict. The supremacy rule will not apply to post-Brexit UK legislation, but it will still apply to pre-Brexit UK laws that conflict with ex-EU law.

Despite the general conversion of EU law, the White Paper insists on an exception for the EU Charter of Rights.  In that case, the ex-EU legislation will be interpreted after Brexit by the ‘rights underlying’ the Charter (ie the ECHR and other international treaties which the UK has signed), rather than the Charter itself.

Amending ex-EU law

Obviously the body of ‘ex-EU’ law will not remain unchanged forever. The White Paper discusses the process by which it can be changed in future. There are two key issues here. First, will any changes be the responsibility of the Westminster Parliament or government, or the parliaments and executives of the devolved bodies in Scotland, Wales and Northern Ireland? Secondly, within Westminster, what changes will be subject to Acts of Parliament, and what will be subject to delegated powers conferred upon the executive?

On the first point, the White Paper is vague about exactly what powers will be exercised by the devolved administrations, and this is likely to be a highly contested issue in practice.

On the second point, the distinction between Acts of Parliament and executive powers is important because an Act of Parliament allows for extensive public discussion and parliamentary scrutiny, whereas an act of the executive (usually in the form of 'Statutory Instruments') is subject to limited public or parliamentary involvement. For instance, there is far less time for discussion, and no prospect of tabling amendments.

The White Paper starts by mentioning some possible delegated powers for the Great Repeal Bill that will surely be uncontroversial. For instance, it refers to the references in ex-EU law to 'EU law', powers of the EU institutions, or information sharing with the EU. Since most of these references will be redundant (except possibly as regards continued sharing of police information, as the White Paper notes), there are limited policy choices to be made when replacing them. So there can be little objection in principle to the executive using such powers.

However, there is a further category of changes to EU law which will be more substantive. The White Paper gives a non-exhaustive list of cases where the government thinks it should have delegated powers: where a policy might change in light of the Brexit talks with the EU; where a policy changes as a direct consequence of leaving the EU; and where the level of detail is 'not appropriate' for an Act of Parliament. 

More generally, the government argues that these powers must be widely defined and must apply not only to ex-EU law converted into EU law, but also Acts of Parliament linked to EU membership. The power for the executive to amend Acts of Parliament is controversial, and such clauses are widely known as 'Henry VIII clauses'. Although the House of Lords Constitution Committee had previously argued that there should be extra powers of parliamentary scrutiny in this context, the government implicitly rejects this view. 


The overall objective of ensuring legal continuity by retaining pre-Brexit EU law in force is logical, and the White Paper has thought many of the details through. It makes sense to ensure that legal continuity as much as possible by creating a distinct body of ex-EU law, where a form of the principle of supremacy still applies and ECJ judgments remain binding, subject to the rare case that the UK Supreme Court might want to overturn.

However, some of the detail has not been considered: what about future ECJ case law? What about cases concerning EU law in the UK pending in the UK courts or the EU courts on Brexit Day? What if a condition of ensuring market access to the EU after Brexit is to take relevant EU court case law ‘into account’? (Note that this falls short of making those judgments binding, and is even an even longer way from retaining the supremacy of EU law over national law). What about the domestic legal impact of any alternative court or dispute settlement system that might rule on UK/EU disputes after Brexit?

The hostility to the EU Charter forms an exception to the rule that prior EU law continues to apply, and raises much legal uncertainty. Does it also mean that ECJ rulings referring to the Charter should be ignored, at least to the extent that they refer to the Charter? Since many such rulings refer to other EU laws and interpret them in light of the Charter, there will in effect be an odd requirement to keep following part of a ruling but not all of it. But this will be like trying to remove an egg from an omelette, because the judicial reasoning on the Charter and the EU legislation is intertwined.  

To some extent, this effect will be limited by the requirement to interpret the ex-EU law in light of the ‘underlying rights’ instead of the Charter. But what does this mean in light of the government’s intention to repeal the Human Rights Act, and replace it with a British Bill of Rights? What if the ECJ’s interpretation of the Charter was arguably more ambitious in a particular case than the relevant ‘underlying rights’ in the Charter? What if the relevant ‘underlying rights’ are set out in a human rights treaty which the UK has not ratified, or not made part of its domestic law? And there is no mention of the pre-Charter case law of the ECJ on human rights as ‘general principles of law’; what happens then?

Moving on to the amendment process, the White Paper’s initial examples of very technical changes to ex-EU laws that the government might wish to make are frankly misleading. For the White Paper then goes on to refer to a non-exhaustive list of broad discretionary powers which the government wants to make changes to the statute book in light of talks with the EU, to make consequential changes to policies or to fill in details of laws.

As drafted, these powers are potentially nearly limitless. They could, for instance, be used to adopt every detail of future policies on agriculture, fisheries, trade with non-EU countries or extradition to the EU without full parliamentary scrutiny or public discussion, because each of these are areas where the new laws could be regarded as changes consequential to leaving the EU. 

So how should Parliament limit government powers? It will be hard to avoid conferring some substantive delegated powers on the government, as the time frame to implement a Brexit agreement with the EU (or the absence of one) before the likely Brexit Day of March 29, 2019 might be tight. On the other hand, there might be several months to spare, or there might be a transitional agreement keeping EU law in force for some time, so giving Parliament more time to act.

The best way forward is to rule certain issues off-limits entirely, particularly issues where EU law provides for essentially domestic legal harmonisation: for instance employment law, environmental law, consumer law, discrimination law. In those areas there is generally no reason why the law necessarily has to change if the UK leaves the EU, because the relevant laws are not usually about cross-border matters. (There are exceptions, like European Works Councils or carbon trading laws).

In other substantive areas, where there is a direct link with leaving the EU, Parliament should be given a right to rule on whether the conditions for conferring delegated powers on the government are met. Effectively it could decide whether the trigger for those powers had to be pulled due to lack of time or not. (Thanks to Professor Tammy Hervey for a version of this idea). There could be particular limits on the power of the executive to amend Acts of Parliament.

Also, the government could be pressed to make more effort to table Acts of Parliament well in advance of Brexit Day on the planned changes to some key areas, for instance agriculture and fisheries, to enable full parliamentary scrutiny. In particular, the planned customs bill could include rules setting out the domestic legal framework for UK’s post-Brexit international trade law (more on that specific issue another time).  

Finally, on the issue of devolved assemblies, it is striking that the Brexit Minister’s foreword to the White Paper talks generally about a “significant increase in the decision-making power” of the devolved bodies, but the actual White Paper then does not give any detail of this. On the other hand it does go into some detail about the powers which can’t be conferred upon devolved assemblies, due to an intention to ensure a UK-wide single market.

Maybe we need a short, simple phrase to refer to the promise of devolved powers which is not then substantiated by any detail. Might I suggest…“the Vow”.

Barnard & Peers: chapter 27

Photo credit: BBC

Wednesday, 29 March 2017

Judicial control of EU foreign policy: the ECJ judgment in Rosneft

Stian Øby Johansen, PhD fellow at the University of Oslo Faculty of Law*

Yesterday the Court of Justice of the European Union (the CJEU) delivered its judgment in the long-awaited Rosneft case (C-72/15, ECLI:EU:C:2017:236). The judgment clarifies some aspects of the CJEU's jurisdiction over the Common Foreign and Security Policy (CFSP). Moreover, it is an important precedent in the field of EU sanctions law generally, and also resolve some questions of interpretation that are particular to the Russian sanctions.

In this blog post I will focus on what the judgment in Rosneft adds to the existing case-law on the review of CFSP decisions. Thus, I will not be discussing any of the more specific questions of EU sanctions law nor summarize the full 197 paragraph judgment. For those looking for a quick summary of the case, I refer to the succinct post by Maya Lester QC at the Sanctions Law blog.

CFSP sanctions: decisions and implementing regulations

I have already sketched out the background of the case in considerable detail in a blog post I wrote here following the oral hearing, and I co-wrote a blog post (with Alexander Arnesen) on Verfassungsblog on the Opinion of Advocate General Wathelet.  I will not repeat all this background here. But there is a couple of details that are essential to understanding the issue at hand here and the Rosneft judgment more broadly.

Recall that EU law sanctions (also known as "restrictive measures") are enacted through a two-step process. First, the Council adopts a decision under TEU Title V, Chapter 2. This decision is then implemented in Union law (and thus domestically in the EU member states) by virtue of a regulation adopted under TFEU article 215. In the case of the Russia sanctions, as in most sanctions regimes, the wording of the respective decisions and regulations are virtually identical.

The use of two legal instruments with different legal bases -- one decision with a CFSP legal basis, and one regulation with a legal basis in the TFEU -- complicates matters when it comes to the jurisdiction of the CJEU. With regard to the regulation, the judgment in Rosneft confirms the obvious: any regulation adopted on the basis of the TFEU article 215 is within the jurisdiction of the CJEU (Rosneft paras 105-106). That is so irrespective of whether the regulation merely restates the decision.

On the other hand, the decision is a act adopted under the CFSP. The CFSP treaty provisions and acts adopted under them are carved out of the CJEU's otherwise general jurisdiction over Union law (see TEU article 24 and TFEU article 275). To this carve-out there are two exceptions, a.k.a  claw-backs: the CJEU has jurisdiction to monitor compliance with TEU article 40 and to decide on the legality of decisions concerning "restrictive measures against natural or legal persons" (emphasis added). As I will come back to, both claw-backs were at play in Rosneft.

The remainder of this blog post focuses on the CJEU's jurisdiction over such CFSP decisions, notably on the clarifications and contributions the Rosneft judgment offers to the ever-expanding case law in this field. (Key judgments in the previous years that have discussed these issues include Case C-155/14 P H v. Council et al  [2016], Case C‑439/13 P Elitaliana SpA v. EULEX Kosovo [2015], and Opinion 2/13 EU Accession to the ECHR [2014].)

The general scope of the CJEU's jurisdiction over CFSP decisions

With regard to the general scope of the CJEU's jurisdiction over CFSP decisions, the Rosneft judgment further cements the approach that has emerged in the case-law over the last couple of years.
First, the limitations on the CJEU's jurisdiction and the two claw-backs explicitly provided for in TEU article 24(1) and TFEU article 275(2) have to be taken seriously. In relation to CFSP acts, the CJEU only has jurisdiction to (1) monitor compliance with TEU article 40, and (2) decide on the legality of restrictive measures against natural or legal persons. (Rosneft para 60.)

Second, while recognizing the explicit limitations on its jurisdiction in the treaties, the CJEU reiterates that those limitations must be interpreted narrowly (Rosneft para 74-75). Or, put differently, the provisions clawing back jurisdiction must be interpreted expansively.

What is new in Rosneft is the application of these starting points to a new issue: do the claw-backs apply in the context of preliminary rulings? This question must be answered separately for the each claw-back provision (Rosneft para 61).

Preliminary rulings and the legality of CFSP decisions in light of TEU article 40

One of the easier questions before the CJEU in Rosneft was whether the validity of CFSP decisions in light of TEU Article 40 could be determined in a preliminary ruling. The identical claw-back provisions in TEU Article 24(1) and TFEU Article 275(2) simply provide that the CJEU has jurisdiction to "monitor compliance with Article 40 [TEU]".

There is nothing to suggest that such monitoring may only happen e.g  in actions for annulment. In other fields of EU law, it has long been clear that the CJEU has jurisdiction to declare Union acts invalid in preliminary rulings (see e.g. Case 314/85 Foto-Frost [1987]). Thus, due to the lack of an express and specific limitation, the CJEU concluded that its jurisdiction extended also to monitoring compliance of CFSP decisions with TEU article 40 in preliminary rulings (Rosneft paras 62-63).

Preliminary rulings and the legality of CFSP decisions concerning restrictive measures

The key jurisdictional question that the Grand Chamber had to deal with in Rosneft was whether it had jurisdiction to decide on the validity of a CFSP targeted sanctions decision in a preliminary ruling procedure. Due to the CFSP carve-out and the peculiar wording of the claw-back provision for restrictive measures in TEU article 24(1) and TFEU article 275(2), there has been significant uncertainty as to the correct answer to this question.

The relevant part of TEU article 24(1) reads as follows (emphasis added):

"The [CJEU] shall not have jurisdiction [over the CFSP], with the exception of its jurisdiction [...] to review the legality of certain decisions as provided for by the second paragraph of Article 275 of the [TFEU]".

The relevant part of TFEU article 275(2) reads as follows (emphasis added):
"the Court shall have jurisdiction [...] to rule on proceedings, brought in accordance with the conditions laid down in the fourth paragraph of Article 263 of this Treaty, reviewing the legality of decisions providing for restrictive measures against natural or legal persons [adopted under the CFSP]"

In other words: TEU article 24(1) limits the jurisdiction to "certain decisions as provided for" by TFEU article 275(2), which in turn refers to "proceedings" that are "brought in accordance with the conditions laid down" in TFEU article 263(4). The latter provision provides that actions for annulment can be brought against acts of the EU institutions before the CJEU:

"Any natural or legal person may [...] institute proceedings against an act addressed to that person or which is of direct and individual concern to them and does not entail implementing measures."

The combined text of these provisions arguably suggests that the jurisdiction to determine the validity of CFSP targeted sanctions decisions only extends to actions for annulment brought by individuals. TFEU Article 263(4) ostensibly provides for the institution of proceedings for annulment; the types of acts that may be annulled are listed in TFEU Article 263(1)-(2). Conversely, then, the CJEU would lack jurisdiction to determine the validity of such a decision in a preliminary ruling. Against this, one may argue that jurisdiction to rule on the validity of Union acts is inherent to the complete system of legal remedies that the Union treaties establish.

How to solve this conundrum? AG Wathelet essentially suggested that the Court should interpret TFEU article 275(2) in the context of TEU article 24(1). In particular, he emphasized the use of the term "certain decisions" in TEU article 24(1), which suggest that the reference in TFEU article 275(2) to TFEU article 263(4) concerns the type of act (“restrictive measures”), and not the type of proceedings (i.e. actions for annulment). See AG Wathelet in Rosneft, para 61 et seq. Consequently, the validity of CFSP decisions can be determined in any kind of proceedings -- also in preliminary rulings.

A further point is that the French language version of TFEU article 275(2) differs from the English in a crucial respect. It provides that the CJEU has jurisdiction "pour controller la légalité de certaines decisions visées à l’article 275, second alinéa  [TFUE]". This seems to suggest an interpretation in line with that of AG Wathelet. Although  AG Wathelet does not discuss the language discrepancy directly, French is the "langue de travail" at the CJEU, and one may speculate that AG Wathelet relied more heavily on the French version of the treaty text than the English.

In Rosneft the CJEU reaches the same conclusion as AG Wathelet; the CJEU has jurisdiction to determine the validity of CFSP decisions in preliminary rulings. In doing so, the CJEU recognizes the textual discrepancy between the otherwise identical claw-back provisions in TFEU article 275(2) and TEU article 24(1), but not (explicitly) the linguistic discrepancy  However, it is slightly more careful than AG Wathelet in grounding its argument in other sources than (con)text -- perhaps because the Court is indeed aware of the discrepancy between the language versions.

There are in particular two supporting arguments that the CJEU relies on. First, it puts forward a systemic argument. According to the CJEU, it is "inherent" in the Union's "complete system of legal remedies or procedures that persons bringing proceedings must, when an action is brought before a national court or tribunal, have the right to challenge the legality of provisions contained in European Union acts" (Rosneft paras 67-68).

Second, the CJEU  emphasizes the fundamental rights dimension of judicial protection (Rosneft paras 69-75). As usual in cases on jurisdiction on CFSP acts, it refers in passing to the concept of the "rule of law" (Rosneft para 72). But it also refers extensively to article 47 of the Charter of Fundamental Rights (Rosneft paras 73-74). While the Charter has been mentioned in passing before in cases concerning CFSP decisions (notably in Case C-455/14 P H v. Council et al. [2016]), the emphasis has usually been on the nebulous concept of the rule of law. In Rosneft we see the reverse: an emphasis on the fundamental right of effective judicial protection, which is laid down in positive primary law in CFR article 47.

The conclusion that the CJEU draws is thus built on a principled and even more solid ground than that of AG Wathelet. The question of whether the validity of CFSP decisions can be determined in preliminary ruling proceedings must therefore be regarded as settled following the Rosneft judgment.

Graham Butler, Assistant Professor of Law, Aarhus University, Denmark

Can the Court of Justice of European Union (‘the Court’) assert jurisdiction and provide a national court with an interpretation of Union law in a case referred to it from a national court under an Article 267 TFEU preliminary reference, when the subject matter is in regard to the Common Foreign and Security Policy (CFSP)? This was one of a number of questions referred to the Court of Justice from the High Court of England and Wales in Rosneft (C-72/15). This week, the Court meeting in a Grand Chamber formation, answered this jurisdictional question in the affirmative. Given the significance of this judgment for the law of the Common Foreign and Security Policy (CFSP), and previous discussion of the Opinion of the Advocate General in 2016, yesterday’s judgment was hotly anticipated given its implications for the ‘specific rules and procedures’ that are applicable to the law of CFSP. As the Court continues in a line of case law to clarify its jurisdiction in CFSP, it is ultimately a question of constitutional importance for the Union’s external relations.

Rosneft concerns the EU’s restrictive measure regime, more popularly known as sanctions. The governance scheme surrounding sanctions is a developed body of case law, in which individuals subject to them have the possibility to challenge them directly before the EU’s General Court, the administrative court of the Union. Given that the locus standi (standing) of taking actions to the Court is a narrow right, the use of preliminary references, otherwise known as referrals from national courts, also functions as an indirect means for legal entities to access the Court for adjudication on matters of Union law. What makes the Rosneft case noteworthy, in comparison to other aspects of CFSP and sanctions case law, is that it is the first case on the Court’s jurisdiction to rule on sanctions not taken directly to the EU General Court. Rather, the Rosneft case arrived at the Court of Justice through the preliminary reference procedure from a national court, in this case, the High Court of Justice (England and Wales) in the United Kingdom, upon the basis of Article 267 TFEU.

Sanctions have a peculiarity in their procedural sense. Firstly, it requires a CFSP Decision, done on an Article 29 TEU legal basis. Secondly, a subsequent Regulation is decided upon an Article 215 TFEU legal basis, which allows sanctions to be implemented throughout the Union. Accordingly, in Rosneft, on the table was Council Decision 2014/512/CFSP, Council Decision 2014/659/CFSP, and Council Decision 2014/872/CFSP (collectively, ‘the Decision’). Furthermore, there was Regulation 833/2014, Regulation 960/2014, and Regulation 129/2014 (collectively, ‘the Regulation’). The Decision taken by the Council, where Member States as a general rule act unanimously, were directly in response to the alleged actions of Russia in Ukraine. Substantively, the applicant contested the implementation measures by way of Regulation taken by the British Government as a result of the CFSP Decision, of which it too was part of, on the grounds that it contained ambiguities. Accordingly, the substantive question was whether the CFSP Decision was one the one hand sufficiently clear, or on the other, imprecise?

In Rosneft, both the Decision and accompanying Regulation were challenged. Yet, it is unclear whether the Court has the jurisdiction to fully answer the questions asked of it, given the first legal act is adopted on a CFSP legal basis (the Decision), and the second legal act on a non-CFSP legal basis (the Regulation). The Court’s jurisdiction in the latter is undisputed given its adoption on Article 215 TFEU, however, much more speculative and up for question is the Court’s jurisdiction on the Decision, given its adoption on a CFSP legal basis. Prior to recent treaty revision, questions surrounding the Court’s jurisdiction rumbled for decades. However, the Treaty of Lisbon, saw a flipping effect, in that jurisdiction of the Court was to be assumed, unless specifically derogated from by the Treaties. One of these derogations was acts adopted upon a CFSP legal basis, which is elaborated in Article 24(1) TEU and Article 275 TFEU.

Firstly, Article 24(1) TEU, inter alia, states that, ‘The Court of Justice of the European Union shall not have jurisdiction with respect to these provisions [CFSP], with the exception of its jurisdiction to monitor compliance with Article 40 of this Treaty and to review the legality of certain decisions as provided for by the second paragraph of Article 275 of the Treaty on the Functioning of the European Union.’ Secondly, Article 275 TFEU states that the Court has the jurisdiction to, ‘...rule on proceedings, brought in accordance with the conditions laid down in the fourth paragraph of Article 263 of this Treaty, reviewing the legality of decisions providing for restrictive measures against natural or legal persons adopted by the Council on the basis of Chapter 2 of Title V of the Treaty on European Union.’

This consequently points to Article 263 TFEU and its fourth paragraph stating, ‘Any natural or legal person may, under the conditions laid down in the first and second paragraphs, institute proceedings against an act addressed to that person or which is of direct and individual concern to them, and against a regulatory act which is of direct concern to them and does not entail implementing measures.’ The first and second paragraphs in Article 263 TFEU do not appear to envisage the possibility for the Court to have the ability to answer questions on preliminary references from national courts. The leading academic material of EU procedural law previously acknowledged that the Court ‘may afford possibilities’ in this area (Lenaerts et al. 2014:458), recognising that it is by no means a settled question. This is, until the right opportunity arose to address it, which was Rosneft.
So what did the Advocate-General say firstly? AG Wathelet said the Court did have the jurisdiction to answer the substantive questions of it by the national court. Yet how did he reach this view in light of the treaties, and their apparent formulation to exclude the Court in such matters? Whilst acknowledging the Court’s jurisdiction in CFSP matters appears to be limited by Article 24(1) TEU and Article 275 TFEU ‘at first sight’ (para. 39), he skirted a narrow interpretation of Article 263 TFEU and its apparent lack of foresight for seeing preliminary references in the equation. For the aforementioned Article 24(1) TEU and Article 275 TFEU, it can be assumed there was there a need for them to have the intended same effect. However, they are worded differently, and thus, the Advocate-General said, might put out the ‘false impression’, that the Court had no jurisdiction. Thus, he said, the two articles enable the Court ‘to review the compliance with Article 40 TEU of all CFSP acts’, (para. 65), regardless of what way the question ends up at the Court, that is, through a direct action, or a preliminary reference.

The Opinion of the Advocate-General is a demonstrative example of how the restatement of certain constitutional provisions within primary law have the ability to be read differently, despite the intentions of the drafters may have been for such restatements to have the same meaning. Given this part of the Opinion of the Advocate-General on jurisdiction, which was non-binding, what did the Court say, and did it reach the same conclusion?


In the judgment issued on 28 March 2017, the Grand Chamber, before going onto matters of substance, had to handle the important question of jurisdiction, and furthermore grapple with the admissibility of the question of jurisdiction. The Council had queried whether the questions referred by the national court could have been answered in respected of the Regulation alone (non-CFSP), rather than contesting the validity of Decision (CFSP) (para. 48). Thus, along this line of thinking, the Court would then not have to assert any jurisdiction on the CFSP legal basis, for which the Council has always viciously defended against any judicial incursion by the Court (Cases C-455/14 P, H v. Council; C-439/13 P Eulex Kosovo; C-263/13 TanzaniaOpinion 2/13; and C-658/11, Mauritius). The Court rejected this Council viewpoint, stating that it is up to national courts alone to ask questions of the Court on the interpretation of Union law (para. 49). The Court was therefore only in a position to not answer a reference when it fails to have a legal question in need of answering, or is only a hypothetical question (para. 50).

The Court furthermore in its judgment stated that only focusing on reviewing the legality of the Regulation (non-CFSP), and not the questions asked of it as a whole by the national court, which would not be adequately answering questions asked of it (para. 53). Moreover, despite the sharp distinction between a CFSP act and a non-CFSP act, in order to impose a sanction within the EU legal order, the Court noted that they are inextricably tied. Given how sanctions are imposed in the EU legal order, it is a perfect demonstration of the possibility of close-knit relations between CFSP and non-CFSP legal bases, given the Court in Kadi I said the link occurs when it has been made ‘explicitly’ (Joined Cases C-402/05 P and C-415/05 P, para. 202). The Court in Rosneft however hypothesized that even if the latter Regulation implementing a CFSP Decision was to be declared invalid, that would still mean that a Member State was to conform to a CFSP Decision. Thus, in order to invalidate a Regulation following a CFSP Decision, the Court would have to have jurisdiction to examine that CFSP Decision (para. 56).

 Once the admissibility of the question of jurisdiction was answered, the Court progressed onto answering the jurisdictional questions raised, in which it concluded that, ‘Articles 19, 24 and 40 TEU, Article 275 TFEU, and Article 47 of the Charter of Fundamental Rights of the European Union must be interpreted as meaning that the Court of Justice of the European Union has jurisdiction to give preliminary rulings, under Article 267 TFEU, on the validity of an act adopted on the basis of provisions relating to the Common Foreign and Security Policy (CFSP)...’ (Ruling 1 of 3). Yet, Court’s assertion of its jurisdiction was not completely unqualified. Rather, it must be meet one of two conditions. The first condition that it may meet, is that it must relate to Article 40 TEU on the Court having the jurisdiction to determine the boundary between CFSP and non-CFSP in its border-policing role. The second condition that the Court’s allows for the assertion of its jurisdiction, is when it involves the legality of restrictive measures against natural or legal persons.

 The remark on Article 40 TEU is significant from the Court (paras. 60-63). From some corners, the Court has been subject for some remarks for not properly utilising this Article for elucidating what the precise boundaries for a CFSP and non-CFSP. To date, it has shunned such possibilities provided to it to determine the fine lines of this providing, underling the fact that CFSP is an obscure area of the treaties, legally speaking. Rosneft perhaps elucidates some reasons why Article 40 TEU has not been used by the Court to date, namely that it does ‘not make provision for any particular means by which such judicial monitoring is to be carried out’ (para. 62). Thus, given this lack of guidance, the Court finds itself falling back on Article 19 TEU to, ‘ensure that in the interpretation and application of the Treaties the law is observed.’ (paras. 62 and 75).

It was advocated nearly a decade ago that rule of law concerns could be used to provide justification for the Court’s jurisdiction in CFSP cases upon a preliminary reference (De Baere 2008:186). Whilst this can be a common phrase with large recourse in a number of situations to justify Court actions, the Court instead of utilising this argument alone here (para. 72), went one-step further. alluded to the EU’s Charter on Fundamental Rights (CFR), selected Article 47 CFR, the right to an effective remedy and a fair trial, ensuring who has ‘rights and freedoms guaranteed by the law of the Union…the right to an effective remedy’. (para. 73), as a basis for clarifying this position on its jurisdiction.

From the Court’s perspective in CFSP-related cases, it certainly does not want the possibility for national courts to test the possibility for them to attempt and try invalidate Union legal acts in whatever form (paras. 78 and 79). It it long-standing jurisprudence of the Court stemming from Foto-Frost (314/85), that it alone has the ability to invalidate Union law, which the national courts cannot do. Thus, national courts only have the possibility to invalidate implementing national measures subject to their own constitutional requirements, and not the Union legal acts themselves. The most recent example of the Court clarifying (ie. extending) its jurisdiction into the CFSP arena was H v. Council (C-455/14 P). Unlike H v. Council however (Butler 2016:677 and on this blog), in which the Court asserted jurisdiction, it then proceeded to fling the substantive matter back to the General Court for adjudication. The Court here in Rosneft had to proceed and answer the substantive questions itself, which conclusively, upheld the sanctions in question.


The Court and the Opinion of Advocate-General on its jurisdictional points can be commended for not allowing a legal lacuna to be created by further disenfranchising CFSP as a particular sub-set of Union law, and ensuring it was kept as close of the normal rules surround preliminary references as possible. Such a situation, where jurisdiction were not asserted, could have lead national courts to not send preliminary references to the Court in further questions seeking clarification on points of Union law. This potential chilling effect would most certainly hamper not just the nature of sanctions, but also the coherent interpretation of Union law as a whole, for which the Court is the ultimate adjudicator. By coming to the conclusion that the Court did have the jurisdiction, empowering itself with the ability to answer the substantive questions, AG Wathelet acknowledged he was breaking with the view of his colleague, AG Kokott, from her view provided in Opinion 2/13 on the EU’s accession to the ECHR. AG Wathelet said that without the Court having jurisdiction, it would undermine under Articles in the Treaties, namely, Article 23 TEU, which guarantees access to a Court and effective legal protection (para. 66), which albeit by an alternative method, the Court broadly arrived at the same conclusion.

Jurisdictional questions are not just inconsequential matters in the exercise of EU foreign policy, but have ramifications for EU procedural law, and the constitutional framework in which Union law operates. The Court’s judgment, clarifying jurisdiction for itself, when it was in doubt, further widens the potential for its scope for a role in EU foreign policy. Hence, how broad a deference is there at the Court to questions that ultimately hinge upon ‘sensitive’ areas of policy? Do Member States want the Court to have jurisdiction in CFSP? The Treaties do their best to prevent it, and five of the intervening six Member States and the Council in Rosneft pleaded that the Court did not have the ability to rule on the validity of CFSP acts. Yet the Court is no stranger to such questions, as it has dealt with jurisdictional questions on sensitive areas before, albeit in a slightly different context. The Area of Freedom, Security, and Justice (‘AFSJ’ or ‘Justice and Home Affairs’)). The Gestoras (C-354/04 P) and Segi(C-355/04 P) cases here provide us with suitable examples. In a pre-Lisbon context, the Court said to interpret the cases as falling outside the scope of the then Article 35(1) TEU because they were preliminary references would not be in ‘observance of the law’. Thus, the Court ruled in both Gestoras and Segi that jurisdiction for the Court in that field was permissible.

Given the Court’s judgment here in Rosneft, there is no doubt that it had to be slightly inventive to overall what was clearly a shortcoming in the drafting of the Treaties. For the Court to have not asserted jurisdiction in Rosneft would have seemed contrary to the overall premise upon which the Union is a ‘complete system of legal remedies’, which again it cited in Rosneft (para. 66), stemming from Les Verts (294/93). Do the Treaties allow vacuums to be created where judicial review is excluded, or does it by reasonable means provide for judicial review? The latter was not only an easy choice, but also the more logical one. Article 19(1) TEU states that the Court, ‘ensure that in the interpretation and application of the Treaties the law is observed’, and that, ‘Member States shall provide remedies sufficient to ensure effective legal protection in the fields covered by Union law.’

This, coupled with the Court’s own ‘Declaration by the Court…on the occasion of the Judges’ Forum organised to celebrate the 60th anniversary of the signing of the Treaties of Rome’ made the day before the Rosneft judgment was published, commenced with restating the premise that the EU is, ‘is a union governed by the rule of law’. Yet such spirited measures are always dampened by other events, and it is hardly in fitting with recent developments at the General Court. The NF and Others v. European Council cases, and the Orders by the General Court on 28 February 2017 (T-192/16, T-193/16, and T-257/16), stated that it did not have jurisdiction on the question of the legal basis upon which an ‘EU-Turkey statement’ was reached. The likelihood is therefore that such questions about the scope of the Court’s jurisdiction in non-CFSP matters will rumble on.

Whilst this Rosneft judgment has clarified the scope of the Court’s jurisdiction on preliminary reference cases dealing with CFSP-related matters, one has to ask why the litigant did not instead seek to go straight to the EU’s General Court with an action for annulment claim, seeking the annulment of the sanctions applying Union-wide. The Court said that the basis for actions for annulment through direct actions from the treaties do not constitute the only means for which sanctions are challengeable (para. 70). Thus, from this, we can deduce that Rosneft opens the basis for future forum shopping when legal entities are subjected to the Union’s comprehensive sanctions regime under the auspices of CFSP in the future.

Remaining questions on the legal limits of CFSP as a special area of area are yet to be fully answered in a categorical way. One example of such is the doctrine of primacy, with lingering questions on its applicability to CFSP. Even with this, jurisdictional questions in CFSP remain. In a recent Order of the General Court in Jenkinson v. Council (T-602/15), it found it did have the jurisdiction to deal with a staffing case stemming from a CSDP, under the wing of CFSP. This demonstrates the caution of the General Court on leading the way on jurisdictional matters, preferring to let the Court of Justice lead the way.

Nonetheless, Rosneft clarifies that CFSP is one (small) step towards wider integration with the rest of the EU legal order. Former Judge at the Court, Federico Mancini said once in a speech at the Danish Supreme Court (Højesteret) in Copenhagen that without the system of preliminary references, that the ‘roof would collapse’ (Mancini and Keeling 1991:2). Indeed, this week’s Rosneft judgment, ensuring that Article 267 TFEU preliminary references in cases involving CFSP can be heard, upholds this notion rather tightly.

Barnard & Peers: chapter 10, chapter 24
Photo credit: The Hill