EU (Withdrawal) Bill and State aid: a note

The EU (Withdrawal) Bill, published last week, has already attracted much comment.

One feature of it that has not been mentioned, however, is a State aid aspect that emerges from the Explanatory Notes to the Bill.

Clause 4 of the Bill provides for EU rights and obligations which are currently recognised in domestic law as a result of section 2(1) of the European Communities Act 1972 to continue being recognised in domestic law after exit.  As paragraph 87 of the Explanatory Notes explains, that provision will mean that directly effective rights contained within EU treaties continue to be law (unless something is done to stop them being law).

Paragraph 89 goes on to list the provisions of the Treaties that the Government considers will be preserved as a result of clause 4.

That list of “preserved Articles” includes Article 108(3) TFEU: the “standstill” provision that precludes a Member State from implementing State aid unless and until it is notified to an approved by the Commission.  That provision has direct effect: so, for example, a third party affected by the implementation of a non-notified State aid has the right to obtain relief from the UK courts.

In other words, the Government’s position is that (absent some other measure), Article 108(3) will, under clause 4, continue to be law the day after Brexit.

That would, however, be a very odd result.  Article 108(3) makes no sense in a context where, out of the EU, the UK is unable to notify a State aid to the Commission and the Commission has no power to approve it.

It is therefore practically certain that the Government will need to address the matter by statutory instrument: depending on the decisions it takes about a State aid regime after Brexit it will either need to modify Article 108(3) to make it operable or provide that it lapses on Brexit day.  So far, the Government has said nothing about its intentions in that regard.


George Peretz QC


Posted in Brexit issues | Comments Off on EU (Withdrawal) Bill and State aid: a note

House of Lords EU Market Sub-Committee calls for evidence on impact of Brexit on State aid

The House of Lords EU Internal Market Sub-Committee, chaired by Lord Whitty, has decided to launch an inquiry into the impact of Brexit on UK competition policy. The inquiry will explore the opportunities and challenges of leaving the EU for antitrust rules, merger control and state aid, as well as considering the potential future relationship between UK and EU competition authorities. See here.

On State aid, the Committee has said it would like to hear evidence on the following issues: –

 (1) Are state aid provisions likely to form an essential component of any future trade agreement between the UK and EU? Do any existing trade agreements between the EU and third countries provide a useful precedent for future UK-EU state aid arrangements?

 (2) Will the UK require a domestic state aid authority after Brexit?

 (3) What would be the opportunities and challenges for state aid or subsidy controls in the UK if no trade agreement were to be reached with the EU? Would WTO anti-subsidy rules restrict the UK’s ability to support industries, or individual companies, through favourable tax arrangements?

 (4) How will the Government’s industrial strategy shape its approach to state aid after Brexit? To what extent has the European Commission’s state aid policy limited interventions that the UK Government may have otherwise pursued?

 (5) What, if any role, might the devolved institutions play in UK state aid control post-Brexit? Are there any potential implications for the UK internal market?

 (6) Will it be necessary for the UK and EU to agree a transitional arrangement for state aid matters after the UK’s withdrawal from the EU? If so, what transitional issues would such an arrangement need to address?


The UK State Aid Law Association has already prepared papers (available on this website here and here) on possible State aid regimes post-Brexit that address many of these issues.  But we shall be submitting further evidence to the inquiry.

Any members who would like to contribute to that process should either post a blog in response or send me comments by e-mail (  As the submission deadline is 15 September, any thoughts by 6 September, please.


Posted in Brexit issues | Comments Off on House of Lords EU Market Sub-Committee calls for evidence on impact of Brexit on State aid









  1. In an earlier paper published at the end of 2016, two of us (George Peretz QC and Kelyn Bacon QC) suggested that the EU was likely to make control of State aid a condition of any type of trade arrangement with the UK after Brexit.  The paper also argued that some form of anti-subsidy control was likely in any event to be needed in order to secure UK compliance with anti-subsidy provisions in the WTO agreement and other free trade agreements.  It also pointed out that there was a strong policy interest, in the era of devolution, in preventing “subsidy races” and other distortions of competition between the various nations of the United Kingdom – an aim currently secured, without any need to deal with the point in the various devolution Acts, by the fact that EU State aid rules apply to the devolved governments.

2. The paper contemplated both an “EEA model” (under which, at least for this purpose, the UK would accept the institutions of the EEA, namely the EFTA Surveillance Authority (“ESA”) and the EFTA Court) and a “domestic” model under which State aid would be regulated purely by domestic bodies and under domestic law.

3. Since that paper, the Government has made it clear that it does not intend to join the EEA.  Although we do not understand that the option of limited participation in EEA institutions has been ruled out, and we refer to it below, the purpose of the present paper is to look in more detail at the “domestic option”.

4. There have also been two further important developments.

5. First, the view taken in the paper that the EU is likely to insist on State aid control as the condition of any comprehensive fair trading agreement is confirmed by the statement in §20 of the European Council’s Guidelines that such an agreement “must ensure a level playing field in terms of competition and state aid”.

6. Second, the Government’s White Paper “Legislating for the United Kingdom’s withdrawal from the European Union”[1] makes it clear that the Government intends to ensure that “wherever practical and sensible, the same laws and regulations will apply immediately before and immediately after our departure”[2]. The White Paper does not mention the State aid rules: but there is nothing in it to suggest that those rules are an exception to the declared aim of converting all the acquis communautaire into UK law, at least for a transitional period.

7. Indeed, §10.11 of the White Paper states that “there are rights in the EU Treaties that can be relied on directly in court by an individual, and the Great Repeal Bill will incorporate those rights into UK law.” Given that declared aim, it is significant that Article 108(3) TFEU provides directly effective rights that can be relied on by those affected by an unlawful aid measure. (Article 108(3) is the “standstill provision” that precludes the implementation of an aid measure not covered by one of the block exemptions unless and until the measure is declared compatible by the Commission, and it is well-established that affected third parties have EU law rights to obtain an effective remedy in relation to infringements of that provision in the form of injunctions, declarations, and damages).  So failure to convert that provision into UK law would appear to be inconsistent with the policy set out in the White Paper.  Further, Article 108(3) effectively imports the definition of State aid in Article 107(1) TFEU: as the Court of Appeal put in in R v CEC ex parte Lunn Poly[3] at §22 “It is accepted that the last sentence of Article [108](3) is of direct effect. Whether or not there has been a contravention of the final sentence of Article [108](3) can only fall to be determined if the Court decides for itself whether or not the matters complained of constitute an aid which should have been notified to the Commission.”  Therefore, carrying over Article 108(3) into UK law requires the carrying over of Article 107(1), whether or not Article 107(1) itself is regarded as being “directly effective”[4].

8. Moreover, failure to convert the State aid rules into UK law would have uncertain repercussions into other areas of the law (such as, in particular, tax) where the content of rights and obligations is affected by the existence of the State aid regime[5], thereby generating the uncertainty as to the law in those other areas which the White Paper seeks to avoid.

9. In theory it might be possible to invent a different but equivalent form of anti-subsidy regime which would meet the policy objectives set out above and satisfy the EU’s requirements for a “level playing field” in this area.  But retaining (as far as possible) the current State aid regime seems to us to be greatly preferable to attempting to create some new system of anti-subsidy control.  The State aid rules are familiar in the UK, and there is considerable legal and policy-making experience in their application.  They are bound to be acceptable to the EU as meeting the “level playing field” requirement. And any other form of prohibition on subsidies would inevitably generate considerable legal uncertainty: in particular, there is comparatively minimal case-law relating to the WTO provisions on subsidies that would be the only obvious alternative model, and the decisions that have been taken under the WTO Agreement on Subsidies and Countervailing Measures indicate that in some respects the WTO definition of a subsidy differs from the EU definition of a State aid (for example there is no equivalent of the requirement of a grant through “State resources”).

10. We therefore take it for the purposes of this paper that the United Kingdom will decide to continue to be bound by the State aid rules, at least in sectors where a comprehensive trade agreement is negotiated with the EU.

11. Further, as was explained in the earlier paper, one of the main difficulties, if not the principal difficulty, in operating the State aid rules is the time taken to obtain decisions from the Commission and rulings from the European Courts – delays that may frustrate sensible measures that fall outside the various block exemptions, and which would be likely to be cleared, but which cannot wait for notification and clearance.   A regime outside the EU might well be able to operate more swiftly than the EU institutions have been able to do, not least because it would be able to deal directly with individual aid givers rather than with the relevant Member State’s central Government. Moreover, the UK would doubtless want to “carry over” all the existing block exemptions (and would face no resistance from the EU in doing so): the effect of that would be that the vast majority of State aid would remain exempt (since the vast majority of State aid granted is covered by the existing block exemptions).

Issues in carrying over the State aid rules to UK law: substantive law

12. One issue that arises if a decision is made to preserve the State aid rules post-Brexit is what to do about the point that EU State aid rules catch only aids that have a potential effect on trade between EU Member States. In our view, that “jurisdictional” aspect of the rules would have to be preserved in some form, despite the oddity of referring to the EU post-Brexit: the absence of any such jurisdictional condition in a domestic regime (or its amendment to, for example, require an effect on trade within the United Kingdom) would bring into the domestic State aid regime measures of purely local interest (such as grants to local leisure centres and so on) that the Commission has made clear in its decisions are not caught by the EU regime.  That issue could possibly be solved by widening the scope of the current EU de minimis Regulation[6], but that might create other difficulties, particularly if commitments had been made to the EU as to the content of the domestic regime.  The cleanest solution to us would seem to be to adopt the approach used in the Stabilisation and Association Agreements concluded between the EU and the countries of the former Yugoslavia, namely replacing “in so far as it affects trade between Member States” with “in so far as it affects trade between the EU and the UK”.

13. We also note that the grounds on which compatibility can be declared (set out in Articles 107(2) and (3)) would in the long run need tweaking in the UK context, removing plainly irrelevant grounds (such as the division of Germany) and amending the references to regional development and the internal market.

14. There would also be a risk that the interpretation of key concepts in the definition of State aid might differ over time as between the EU and the UK.  That risk will be minimised if UK courts are permitted to have regard to decisions of the ECJ, and the EFTA Court, on State aid issues (a provision that may well also be appropriate for other areas of competition law, on the likely assumption that the UK wishes to retain the model based on Articles 101 and 102 TFEU).  It is also likely that, as in the EU/Ukraine Agreement, there will be a provision in any EU/UK free trade agreement that provides for a joint committee to discuss and to seek to resolve any differences in interpretation that may arise.  Neither of those suggestions involves, in our view, a concession that the United Kingdom will become a passive “rule taker” in this area: rather, we would hope and expect that there would be a flow of ideas in both directions, and that as the United Kingdom develops its own approach to difficult and controversial aspects of the State aid rules, that approach might well over time have influence on the EU’s approach to those concepts.

Procedural and enforcement aspects

15. However, there are some very substantial problems that need to be addressed in any “carry over”.

16. It may assist at this point if we briefly summarise some relevant characteristics of the EU State aid regime.

17. First, the definition of State aid applies to a very wide range of public measures, ranging from grants to tax advantages to oral guarantees.  State aid can take the form of (for example) a contract, a tax waiver, or legislation.  A grant by a local council to a local football team can be a State aid; so can a difference of tax treatment created by legislation. A domestic State aid regime will need to be able to deal with all those types of measure, implemented by bodies ranging from a small local authority or public body right up to legislation and large national infrastructure projects of high public and political significance.

18. Second, the EU regime accords a wide power to the Commission to declare State aid compatible with the common market on a number of very broadly defined policy grounds. That power involves policy decisions (in particular, balancing the public policy benefits of a measure with anti-competitive distortions) which are unlikely to be suitable for judicial resolution.  Given the wide definition of State aid, and the myriad reasons why State aid may be necessary in order to achieve important public policy goals, this power to (in effect) approve the grant of aid is a critical element of the regime, and is inherent in the structure of Article 107 and 108 TFEU.

19. Third, because State aid may well not be transparent or obvious, there needs to be a body able to investigate possible breaches and to take measures designed to bring those breaches to an end.

20. That summary shows that any domestic carry over of the State aid rules needs to address the following issues: –

  • State aid can have positive and negative effects, and deciding how the trade-off is made and what to approve is a matter of judgment and discretion.
  • There needs to be a body charged with applying and enforcing the State aid rules that is capable of taking often complex decisions of economic policy.
  • That body needs to be politically independent: and since it will be considering key political decisions taken by central government, it needs to be able to resist, and be seen to be able to resist, high-level political pressure.
  • The domestic mechanism needs to be able to resolve the situation where a State aid takes the form of legislation (for example, legislation that confers a selected tax advantage).

The enforcement authority

21. We start by addressing a preliminary question: is it necessary for there to be an enforcement authority at all?  Could the question of whether a measure infringed the State aid rules be left to the ordinary courts to resolve?

22. The key problem with “leaving it to the courts” would, in our view, be that, while the courts are well-placed to rule on the legal question of whether a measure is State aid, they are ill-equipped to perform the key function of assessing whether the aid should be approved (i.e., in EU State aid law, the assessment of compatibility – an assessment carried out by the Commission, exclusively and subject only to judicial review leaving it a wide margin of appreciation).  The key reason why they are ill-equipped is that the question at the heart of any such assessment is whether a public policy objective is such as to justify the distortion of competition inherent in State support: an assessment that involves value judgments of a kind that any court will be uncomfortable in making (particularly in cases where the measure is politically controversial, as many large infrastructure projects are), and which differ in order of magnitude from the relatively narrow value judgments that courts now make in the course of economic assessments under Article 101(3) TFEU (the provision that allows agreements that appreciably affect competition to be exempted if they satisfy certain conditions).

23. A further, practical, reason for not “leaving it to the courts” is that there will be many cases where public authorities and private bodies involved in complex projects where there is clearly State aid will wish to ensure that there is a binding ruling that the aid is approved.  If there is no public enforcement authority empowered to give binding rulings, then Ministers and others involved in key infrastructure would have to apply to court for an approval order (with all the expense and delay that flow from complex litigation) in order to achieve certainty: and anyone seeking to oppose such a measure could threaten delay and costly litigation even in a case where the policy case for the measure was strong.  Further, it is possible, with a public enforcement authority, to have preliminary discussions before any formal notification is made, and a process of iterative negotiation as changes to the project are made in order to satisfy the enforcement authority that it is compatible.  It is hard to see how any court procedure could replicate those features.  Further, we do not see how any court, comprised of whichever judge happens to be available at the time, could ever make the necessary assessment in the speed with which a regulator, familiar with the policy issues, is able to act in highly urgent cases: see the Commission’s decision-making in the bank rescue cases during the financial crisis (and we observe that the balance between the need for such extraordinary measures and the distortions of competition they entailed, and the negotiation of the detailed provisions in those measures, were classic examples of assessments that are in our view wholly unsuitable for litigation or court decision).

24. We also do not see how a court-based system would be able to create and develop consistent policy in response to technological or market developments: see, for example, the way in which the Commission has been able to develop policy in areas such as broadband and transport infrastructure.  That would lead to considerable uncertainty.  Nor, in a court-based system, would there be a regulator able to build up a constructive dialogue with the Commission, ESA, and potentially other anti-subsidy regulators, and to contribute to the development of anti-subsidy policy as a key element of future trade agreements: given the Government’s objective that the UK become a leader in global free trade, that would, in our view, be a significant missed opportunity.  Indeed, without an independent regulator, we suspect that any court faced with taking a decision as to whether to approve aid would tend to reach for, and stick rigidly to, any available Commission decision or policy statement, thereby losing any chance to develop a distinctive UK policy approach in this area.

25. We therefore strongly recommend that the prime responsibility for enforcement of the rules be conferred on a public authority, and not the courts.  The obvious candidate for the role of enforcement authority is the Competition and Markets Authority (“CMA”).  It currently has jurisdiction across the United Kingdom in relation to (amongst other things) competition law enforcement and mergers.  It has the necessary combination of legal, economic and policy expertise.  It has experience of analysing the effect on competition of government policies and of conducting complex investigations involving detailed factual inquiry and economic assessments. Its independence is widely recognised. Finally, it already has experience of giving advice to public bodies on the competition implications of their policies or on proposals for legislation (a function it exercises under section 7(1) and (1A) of the Enterprise Act 2002): see, for example, its Guidelines on Competition Impact Assessment (CMA50) published in September 2015.

26. The CMA performs its functions on behalf of the Crown[7].  It might be regarded as somewhat anomalous for one part of the Crown to regulate other parts of the Crown, especially in a situation where litigation is likely.  However, it should be noted that that situation already potentially arises under the existing UK competition rules: it is entirely possible for bodies that are part of the Crown to be “undertakings” subject to the prohibitions in the Competition Act 1998 and therefore to be (a) subject to the powers of the CMA to make directions and impose penalties and (b) potential appellants to the Competition Appeal Tribunal (“CAT”) against such decisions.

27. Although we have not yet seen the relevant provisions of the Great Repeal Bill, we assume that the powers given to the Secretary of State to adapt EU law to make it suitable for a post-Brexit context will be wide enough to allow a substitution of the CMA for the Commission in the domestic provision that replaces Articles 107 and 108.

29. It has to be recognised, however, that exercising a State aid function would, for reasons that will by now be apparent, place the CMA in a far more politically-exposed position than its current remit generally involves: essentially, a State aid function would give the CMA what would in effect be a “veto” power over potentially very important and controversial decisions taken by UK, and devolved government, Ministers.

30. It therefore seems to us that, if the CMA is to be given this role, its independence and its authority is likely to need strengthening.  It might well be appropriate, for example, for there to be a formal role for Parliament (perhaps via the BEIS Select Committee) and the devolved Parliaments in appointing its key officials.

31. The CMA would also need to be adequately resourced (taking into account that there are expected to be other significant increases in demands on its resources as a result of Brexit): though its expertise in competition law and economics give it a good base for discharging a State aid role, it would need to recruit and train up specialists in the distinct area of State aid.

32.  It seems to us that the functions of the CMA would need to mirror the functions of the Commission under the EU State aid rules (subject to the issue of legislation which we deal with below): that is to say, while the courts could have the power to determine that a measure is not an aid, the CMA would need to have the sole power to declare a State aid to be compatible, and would need to to be given powers to investigate and remedy infringements of the standstill obligation (including the power to block a State aid measure and to require a State aid to be repaid).

33. We would though advocate that the CMA’s State aid procedures should, at least in the long run, provide rather better protection for beneficiaries of State aid than EU procedures currently do, in order to resolve what is at present one of the most frequent criticisms of the EU State aid regime. The domestic procedures should start from the proposition that, given the profound impact of State aid decisions on the beneficiary of aid, the beneficiary should be treated as a full party to the proceedings and not just as an “interested party”.  In particular, it might well be appropriate for the beneficiary of a possible aid measure to be able to notify that measure to the CMA in order to ensure legal certainty, rather than to reserve notification to public authorities (and we note in that respect that the “Member State only” notification model would in any event be inappropriate for a domestic regime).

34. The CMA’s decisions should be subject to a right of appeal to a court (as is the case now for decisions of the Commission) – most obviously (not least because of its expertise in competition law and economics, as well as its UK-wide jurisdiction) to the CAT.  We see no difficulty in principle in there being a full right of appeal on the question of whether a measure is State aid at all (as pointed out in Lunn Poly,that is a role that the courts already discharge, and do so on the merits in even on a judicial review). But we do not consider that decisions as to compatibility of aid measures should be subject to a full merits appeal to a court given the non-judicial nature of the assessment being made: it seems to us that such decisions should be challengeable on only on a judicial review basis.  It is for discussion whether that difference would need to be expressly set out in legislation: our view is that that is the approach the courts would be likely to adopt in any event (as has the General Court, which, despite the single standard of review laid down in the EU Treaties, subjects Commission decisions on whether there is aid to a more intrusive standard of review than decisions on compatibility).

35. We would also suggest that, if the CAT is given an appellate role in relation to State aid decisions by the CMA, any other court faced with a State aid issue should have power to transfer that issue to the CAT for determination: that power would be analogous to the (now implemented) power in in relation to competition law issues in section 16 of the Enterprise Act 2002.

36. We also consider that, apart from primary legislation (discussed in the next section) the general rule should be that a State aid measure implemented without CMA approval should be void and any aid granted recoverable (subject to a limitation period).  In the long run, consideration might be given to permitting the CMA to declare compatibility retrospectively. The advantage of that would be to avoid the problem for beneficiaries of unnotified aid that even if the aid is declared compatible they may still be required to repay the value of interest on the period between the grant of the aid and the declaration of compatibility[8], but the disadvantage might well be to remove a strong incentive to notify potential State aid and hence damage the enforceability of the State aid rules, and the existence of such a material procedural difference between the domestic regime and the EU regime might be controversial at EU level.

37. We have considered whether the Secretary of State should have a “last resort” power to overrule the CMA’s assessment of compatibility on specified public interest grounds, akin to the provisions that currently exist in the mergers regime.  Such an arrangement may of course not be possible under the treaties under which the UK agrees to continue to apply the State aid rules: but even if it is, the risk would be that the Secretary of State would be asked to “second guess” every decision taken by the CMA, which in turn would undermine the independence of that decision making.  We therefore would not recommend that such a power be taken, even if permitted by the EU/UK agreement: but any such power should be very strictly confined to specified matters of major significance.

37. A further key issue is the real risk of differences emerging as between the UK and EU approach to compatibility. We note that the risk of different approaches to compatibility already arises as between the Commission and the ESA, but in practice the ESA appears to have been content to follow the approach of the Commission.  We suspect that, given the size of the UK economy compared to those of the EEA/EFTA Member States, the risk of divergent approaches emerging in the case of the United Kingdom is rather greater than has been the case for the ESA. As we have already observed, if the current approach of the European Courts to the judicial review of compatibility is followed in a domestic regime, the scope for judicial resolution of a divergent approach to compatibility may be limited.  That leaves only a joint committee mechanism by which both the EU and UK agree to seek to resolve any differences by negotiation, with a reserve power on both sides to take protective trade measures in case of irreconcilable differences: there might also need to be a domestic power to amend the legislation to reflect any agreement reached by the UK at that level.  We would, however, hope that both the EU and UK would be prepared to live with some differences in approach: and, indeed, we would hope and expect that both sides would be able to learn from differences of approach by the other, to the mutual benefit of both.


38. As pointed out above, legislation that establishes a favourable treatment of a person or class of person may well constitute a State aid measure: the complex litigation surrounding the aggregates levy is a clear example, as is the Lunn Poly case referred to above (although the Court of Appeal’s decision in that case is now accepted to have been incorrect in the light of subsequent ECJ case-law).

39. We do not think that it is constitutionally feasible, or politically realistic, for the CMA – or even a court – to be given power to prevent the passage of primary legislation that constitutes a State aid measure.  Nor do we think that it would be acceptable for the CMA or a court to make an order that effectively renders a State aid measure contained in primary legislation ineffective (for example, by ordering recovery of a tax advantage given by primary legislation to a favoured class).

40. Nonetheless, any State aid provision in a UK/EU free trade agreement is bound to require action to be taken to deal with State aids granted by primary legislation.  It seems to us that the way forward in the case of State aid measure contained in primary legislation would be (a) for the CMA to have power to declare such a measure compatible (in which case no further issues arises) but (b) in a case where it considered that there was an incompatible State aid, for it to be required to apply to a court (probably the CAT, though it might be considered that the High Court – or its Scottish or Northern Ireland equivalents – would be more appropriate) for a declaration of incompatibility.  In such an application, the issues would be (a) whether the measure was in fact State aid (on which the Court would reach its own view) and (b) whether the CMA had acted lawfully in reaching the view that the measure was incompatible.  If the court agreed on both points, it could issue a declaration of incompatibility along the lines of section 4 of the Human Rights Act 1998. As with that provision, the effect of that declaration would, we suggest, be to give the appropriate Minister the power to amend the legislation in question so as to remove the State aid aspect (or so as to satisfy the CMA that the remaining State aid aspect of the measure was compatible).

41. We have considered whether third parties should have the right to apply to the court (or to the CAT) for a declaration that a measure in primary legislation was State aid.  However, we see problems with this given that the Court could not form a view on compatibility (an issue that we consider should be reserved to the CMA).  We therefore suggest that the power to apply for a declaration that a provision of primary legislation amounted to State aid should be reserved to the CMA: however, in order to protect third party rights, a refusal by it to do so would be challengeable by way of judicial review.

42. In the case of primary legislation we do not think that it would be acceptable, or consistent with fundamental UK constitutional principle, for the measure to be void unless and until declared compatible by the CMA.

43. A decision would need to be taken on whether primary legislation passed by the devolved Parliaments should for these purposes be treated in the same way as legislation passed by the UK Parliament.  The devolution Acts already make provision for primary legislation passed by the devolved legislatures to be held to be invalid if that legislation exceeds the scope of those legislatures’ authority (e.g. by legislating for a reserved matter or in a way that contravenes the ECHR): so there is no fundamental constitutional objection to their primary legislation being declared invalid on State aid grounds.  But we can well see that, politically, it might be considered inappropriate for the CMA to be given power, in effect, to veto primary legislation passed by the devolved legislatures: so the correct option there might well be for the CMA to have to apply to the appropriate court for an order declaring the legislation to be invalid rather than (as in the case of other measures) having the power to do that itself (albeit subject to an appeal).

44. As far as secondary legislation is concerned, we see no fundamental objection to such legislation being declared void on State aid grounds: however, again, it might be thought to be appropriate for the CMA (or third parties) to have to apply to the appropriate court for an order declaring the legislation to be incompatible.


45. As we have indicated, a UK domestic State aid law would require a limit on the powers of the devolved legislatures to adopt legislation amounting to an incompatible State aid measure.

46. That constraint would replicate the current constraint imposed on the devolved legislatures by EU law.  However, the constraint would be imposed by UK legislation.  Since such legislation would significantly affect devolved competence in non-reserved matters (for example, by limiting the ability of the devolved legislatures to create tax distinctions amounting to State aid in matters falling within their tax competence, or in providing for grants in non-reserved fields), it would appear to engage the Sewel Convention (now set out in statute[9]) by which the UK Parliament will not “normally” legislate in matters falling within devolved competence without the consent of the devolved legislatures.

47. As is clear from the Supreme Court’s judgment in R(Miller) v Secretary of State for Exiting the EU[10], the Sewel Convention imposes a political, rather than a legal, constraint on the ability of the UK Parliament, and the Courts will decline to rule on whether the convention has been breached in any particular case.

48. We do not comment on the political aspects, save to note that the question of State aid is only one of the areas where this problem will arise[11], so that the politics of the approach to the issue are likely to be conditioned by wider issues than State aid.

49. To the extent that the United Kingdom accepts commitments in its hoped-for trade agreement with the EU not to implement incompatible State aid, we note that any devolved legislation that did so could be prevented by Ministerial order under section 114(1)(d) of the Government of Wales Act 2006 and its equivalents in the other devolution Acts on the basis that it contravened the United Kingdom’s international obligations: but we do not consider that that power on its own would be adequate to deal with the problem.

50. We also note that the area of “Regulation of anti-competitive practices and agreements; abuse of dominant position; monopolies and mergers” is reserved to the UK Parliament[12].  But it seems to us to be clear that the effect of that provision is to prevent the devolved legislatures from legislating in the area of competition law, and does not limit the competence of the devolved legislatures to pass a law that constitutes a State aid measure (in, for example, the exercise of a non-reserved tax competence).  Nor would we regard a State aid law as naturally falling within the description “regulation of anti-competitive practices and agreements”: it might just fit the description “anti-competitive practice” but we consider that it would be hard to persuade a court that it was Parliament’s intent in this provision to reserve to Westminster a general power to impose State aid control on the devolved legislatures.

The Great Repeal Bill and transitional measures

51. As will be clear from the above, any domestic implementation of the State aid rules will require considerable adaptation.

52. It is not clear to us whether it will be possible to set up a workable State aid scheme under the powers to be contained in the Great Repeal Bill (which has not yet been published).  It is therefore not possible to express any view as to whether the powers to make amendments to EU law as it is converted into domestic law will be wide enough to enable  the creation of such a scheme on the back of converting Article 108(3) into domestic law. We note, however, that the Bill is also likely to take powers to implement the United Kingdom’s obligations under the withdrawal agreement and any UK/EU Treaty, and therefore suspect that that route is likely to be provide a sounder legal basis for a new regime.


53. On the assumption that the United Kingdom continues to accept a State aid regime, it seems to us (subject to any transitional provisions that may be made under the withdrawal agreement) that Commission decisions under EU rules authorising State aid measures that continue to be implemented after Brexit should continue to apply after Brexit as if they were decisions of the CMA under the new UK rules: the CMA would have power to amend those decisions as necessary on the same basis that the Commission has power to amend its compatibility decisions.


54. We do not underestimate the challenges of creating a domestic State aid regime.  But we do not think that those challenges can be avoided if any satisfactory trade arrangement is to be negotiated with the EU and if the Government does not accept the option of using the EEA institutions for these purposes (or if an agreement on use of the EEA institutions cannot be reached).

55. We do not, however, think that the creation of a domestic State aid regime should be seen as an unwelcome imposition forced on the United Kingdom in order to obtain benefits elsewhere.  We consider that, in fact, State aid control will be important aspect of protecting the integrity of the United Kingdom’s own internal market.  We would also argue that State aid law does not generally prevent well-focused subsidies granted in pursuance of a wider industrial strategy, and that many of the problems attributed to the State aid rules arise because of the delays associated with the requirement for notification and approval of aid falling outside the block exemptions, a problem that is likely to be much less severe under a domestic regime.  Implementing a domestic State aid regime will also put the United Kingdom in a strong position to negotiate detailed anti-subsidy arrangements that are likely to form a key aspect of the other free trade agreements which it hopes to conclude port-Brexit (and will ensure its compliance with existing provisions in existing EU free trade agreements with third countries which the Government hopes to extend to the United Kingdom after Brexit).  State aid control should also be regarded, as it has been since 1973, as part of the armoury of measures designed to ensure the efficient spending of public money by public bodies and to provide assurance to business and investors that the level competitive playing field will not be distorted by unwarranted government subsidies to favoured players.  Its continuance will, in our view, bring many benefits to the UK economy and taxpayers.


GEORGE PERETZ QC, Monckton Chambers

KELYN BACON QC, Brick Court Chambers

ISABEL TAYLOR, Partner, Slaughter & May




[1] Cm 9446.

[2] Introduction by the Secretary of State for Exiting the EU, fourth paragraph.

[3] [1999] 1 CMLR 1357. See also Case C-368/04 Transalpine Ölleitung [2006] ECR I-9957, §§38–39.

[4] For this reason it seems to us that it is not necessary to determine whether Article 107(1) is itself directly effective. If a decision on this point were required, however, for the purposes of considering the powers under the Great Repeal Bill, we would be inclined to consider that Article 107(1) is indeed directly effective; the compatibility provisions in Articles 107(2) and (3) are, however, clearly not directly effective.

[5] For an example, see the decision of the First-tier Tribunal (Tax Chamber) in Western Ferries (Clyde) v HMRC [2011] UKFTT 243 (TC) at §163 (need to construe tonnage tax rules so as to avoid State aid).

[6] Commission Regulation 1407/2013/EU

[7] Para. 8 of Sched. 4 to the Enterprise and Regulatory Reform Act 2013

[8] See Case C-199/06 CELF [2008] ECR I-469, §52.

[9] See section 28(8) of the Scotland Act 1998 and section 107(6) of the Government of Wales Act 2006.

[10] [2017] UKSC 5, at §151.

[11] Agriculture is another area where similar issues arise: see G. Peretz, ‘Storm Clouds over the Welsh Mountains: Agriculture and Devolution’ U.K. Const. L. Blog (30th Mar 2017) (available at

[12] See paragraph 69 of Schedule 7A to the Government of Wales Act 2006 (when it comes into force) and paragraph C3 of Schedule 5 to the Scotland Act 1998.



The Court of Justice gives Ms Vestager two Christmas presents: the Commission wins its appeals in Aer Lingus/Ryanair and in Banco De Santander/Autogrill España

In the midst of a flood of significant judgments on a number of areas of EU law handed down today, the CJEU has given judgment on two important State aid cases. In each case, the Commission won its appeal against an adverse finding of the General Court.

In Joined Cases C‑164/15 P and C‑165/15 P Commission v Aer Lingus and Ryanair, ( the Court of Justice restored the orthodox position that the amount of aid to be recovered is the amount of aid given (plus interest), not the amount by which the beneficiary has economically benefited from the aid.

Those amounts can be very different. For example, if a company is given €10m to build a factory in a remote location rather than in a convenient location, the grant may do no more than overcome the additional costs of locating in that area. If that grant is found to be unlawful State aid and ordered to be recovered, the company can be left well out of pocket: it does not have the €10m, and has a factory in a remote location rather than a convenient one.

In the Aer Lingus case, the aid came in the form of a lower rate of Irish air transport tax for very short flights. The tax was a fixed amount per passenger, and was passed onto the passenger in the form of higher ticket prices. The Commission’s decision ordered Ireland to recover from the beneficiary airlines the difference between the lower and higher rates. The airlines argued that that failed to reflect the fact that the benefit of the lower rate was effectively passed onto the passengers and could not be recovered from the passengers: so the effect of that recovery was to put the beneficiary airlines in a worse position than the airlines who paid higher rate tax (because those airlines, too, had simply passed the duty on).

The General Court agreed with the airlines, and limited recovery to the amount by which the airlines could be said to have gained business as a result of paying a lesser rate.

The CJEU overturned the General Court and agreed with the Commission. At §92, the CJEU summarises the position:

recovery of [unlawful] aid entails the restitution of the advantage procured by the aid for the recipient, not the restitution of any economic benefit the recipient may have enjoyed as a result of exploiting the advantage. That benefit may not be the same as the advantage constituting the aid and there may indeed be no such benefit, but that cannot justify any failure to recover that aid or the recovery of a different sum from that constituting the advantage procured by the unlawful aid in question.

That is a clear restoration of the orthodox dogma.

Those who adopt a questioning attitude to the dogmas of State aid may, however, find that statement unsatisfying. What principle underlies the distinction between the “advantage” (the crude “cash value” of the aid) and the “benefit” (the amount by which the beneficiary actually benefited)? Without a satisfactory justification in principle, the fact that a beneficiary can end up having to “repay” far more than it ever actually benefited begins to look penal (particularly as it is not the beneficiary that has breached the rules, but the Member State).

It is, in the present writer’s view, hard to discern a coherent principle behind the dogma. The CJEU notes (at §89) that the aim of a recovery order is “to restore the situation as it was before the aid was granted”: but that ignores the point that (in the present case as well as in the case of the factory example above) the effect of the recovery order is to leave the beneficiary very significantly worse off than it was before the aid was given. More telling, perhaps, is the Court’s observation at §91 that “the recovery of unlawful aid with a view to re-establishing the status quo ante does not imply reconstructing past events differently on the basis of hypothetical elements such as the choices, often numerous, which could have been made by the operators concerned, since the choices actually made with the aid might prove to be irreversible”: that is, in the present writer’s view, a fancy way of saying that, because it can be difficult to establish economic benefit (since it will to some extent involve establishing a counterfactual) the effort should not even be made. That is not principle: it is expediency.

Another interesting aspect of the case – although one on which the CJEU upheld the General Court – is that the difference in treatment was also attacked as contrary to Article 56 TFEU (freedom to provide services) on the basis that it subjected higher rate flights – more likely to be between Ireland and other Member States – to a higher tax than that applied to flights likely to be within Ireland. The beneficiary airlines pointed out that, if that was right, the airlines that paid higher rate would be entitled to a refund of unlawfully levied tax under the well-known San Giorgio principle. If that happened, the tax difference that was the substance of the State aid would vanish.

The CJEU dismissed that argument. It pointed out that it was for the national courts to enforce san Giorgio rights and that, at the time of the decision, there had been no successful claim. The Commission had to proceed on the basis of matters as they stood.

Again, that approach appears artificial: it appears to leave open the very unattractive prospect that the beneficiary airlines will find themselves having to bear the burden of the higher rate duty when those airlines on which it was actually imposed obtain San Giorgio refunds. Matters may, though, not be quite that simple: the Irish tax authorities may well (given the fact that the tax was passed on) have an “unjust enrichment” defence to the San Giorgio claims: and to the extent that those claims succeed, there would seem to be no reason in principle why the beneficiary airlines (who are now, in effect, required to pay the higher rate tax) should not also be able to make such claims. It will be for the Irish courts to deal with those matters.

Ms Vestager’s other victory today is in Joined Cases C‑20/15 P and C‑21/15 P Commission v World Duty Free (formerly Autogrill España) and Banco de Santander ( That case concerns a Spanish tax rule that provides that, in the event that an undertaking taxable in Spain acquires a shareholding in a “foreign company” equal to at least 5% of that company’s capital and retains that shareholding for an uninterrupted period of at least one year, the goodwill resulting from that shareholding, as recorded in the undertaking’s accounts as a separate intangible asset, may be deducted, in the form of an amortisation, from the basis of assessment for the corporation tax for which the undertaking is liable. The measure at issue states that, to be classified as a ‘foreign company’, a company must be liable to pay a tax that is identical to the tax applicable in Spain and its income must derive mainly from business activities carried out abroad. The Commission found that that rule gave a selective advantage to Spanish companies that met the “foreign company” condition, and hence was unlawful State aid to such companies.

The General Court annulled that decision, essentially on the basis that the advantage created by the measure was accessible to any undertaking and was directed not to a particular category of undertakings, which would have been the only undertakings favoured by that measure, but to a category of economic transactions. Put another way, because, in principle, any Spanish company could acquire a qualifying interest in a foreign company the measure was, like the Ritz hotel, open to all and not selective.

The CJEU disagreed with the General Court. It held that the selectivity condition was met as long as the Commission could show that some undertakings qualified for the advantage and that others, in a similar legal and factual situation, did not. In the present case, the mere holding of a qualifying interest in a foreign company did not place an undertaking in a different factual and legal situation from one that did not (that assessment being made having regard to the objectives of Spanish corporation tax): and it was irrelevant that one could not pin down a specific category of undertakings (eg undertakings supplying particular goods or services) that did not satisfy the condition over and above the mere fact of failing to satisfy it.

The Banco de Santander judgment is particularly significant because of its overlap with the controversial Commission decisions in the “tax ruling” cases (Apple, Starbucks etc.). The controversy in those cases centres on the Commission’s approach to selectivity. There are differences between the cases: but the CJEU’s wide approach to the concept of selectivity is likely to be seized on by the Commission as it seeks to defend its decisions in the ongoing appeals.



Posted in European Court judgment | Comments Off on The Court of Justice gives Ms Vestager two Christmas presents: the Commission wins its appeals in Aer Lingus/Ryanair and in Banco De Santander/Autogrill España

Paper on post-Brexit options for State aid

 Note: the following paper is a draft paper prepared by us (George Peretz QC and Kelyn Bacon QC) as a contribution to the Commercial Bar Association’s consideration of post-Brexit issues.  The views expressed are however ours alone.  We are grateful to Robin Griffith of Clifford Chance and Isabel Taylor of Slaughter and May for their comments on an earlier draft: any errors are ours.

In this paper, we first provide a brief summary of the EU State rules and concerns that have been expressed about them, and then consider various options that might be considered for maintaining or replacing the State aid rules, or aspects of them, after Brexit.  We conclude by looking at transitional issues that will need to be dealt with, in the agreement with the EU under Article 50 TEU and/or in the Great Repeal Bill, if it is decided not to retain the State aid rules, or to maintain them in a substantially different form.


In our view, it is likely that retention of State aid control will form an essential component of any comprehensive trade deal between the United Kingdom and the EU (whether in or outside the single market).  But we also see considerable advantages to the United Kingdom in agreeing to retain such a regime: and such a regime, out of the EU, is likely substantially to reduce a number of the disadvantages of the EU regime.

In any event, a number of transitional issues will need to be dealt with in domestic legislation and/or the Article 50 agreement with the EU.


The essential thinking behind the EU State aid rules (and other international prohibitions on subsidisation, such as the current WTO rules, which we discuss below) is that the grant of subsidies to firms of one State, in a single market or free trade area, will often distort competition to the detriment of competing firms from other participating States.  Put shortly, it is one thing to open up your domestic markets to foreign competition, but quite another thing to open your domestic markets up to subsidised competition.  And the freedom to export to another country without restriction is of little value if the government of that country can freely subsidise its domestic producers so as to defeat competition from imports.

On the other hand, there may well be powerful arguments for subsidies in order to achieve important domestic (or indeed pan-European) policy aims, such as regional development, promoting R&D, encouraging training, dealing with natural disasters, and supporting important fundamentally viable businesses over short-term market turbulence. These are recognised in the range of justifications that permit the Commission to authorise State aid.

The EU State aid rules (now in Articles 107 and 108 TFEU) date from the earliest days of what is now the European Union.  State aid provisions formed part of the European Coal and Steel Community Treaty in 1952, and the current provisions in the TFEU are in essential respects the same as those in the 1957 Treaty of Rome.  But it may be noted that the historical origins of the State aid rules go even further back, to the 1947 GATT: much of the wording of GATT Article 16 on subsidies found its way into the drafting of the Treaty provisions on State aid.

Brief summary of the State aid rules

In a nutshell, the State aid rules prevent Member States from granting State aid save where the Commission has approved that aid as being justified.

The essential definition of a State aid is that it is an economic advantage, granted to an undertaking out of State resources, which favours certain undertakings over others (i.e. is selective), and which potentially distorts competition and trade between Member States.

Unpacking that definition, the following important points follow in terms of the scope of the EU State aid rules.

First, the State aid rules have a wide scope.  They apply to all sectors of the economy.  They also apply to a wide variety of State measures: not just straight subsidies, but also to measures that are economically equivalent (such as access to government assets on favourable terms, favourable tax treatment, guarantees and so on).  That extensive scope means that the rules will usually catch any attempt the dress up in some other legal form what is in economic terms a subsidy.  But it also leads to criticism that that the Commission, supported by the Court of Justice, has a tendency unduly to expand the scope of the rules.  The recent controversy over the Commission’s decisions finding that tax rulings given to certain multinational companies amounted to State aid (decisions currently on appeal to the EU General Court) are a topical example.

Second, and very importantly in practice, the State aid rules do not apply to measures taken by the State that are equivalent to those that a rational private operator in the market would take (the “market economy operator principle” or “MEOP”)).  Since it is accepted that, in most cases, rational private investors might take a range of views, the MEOP in practice allows a range of measures to be taken by Member States provided that they ensure that they have sufficient evidence that the measure is one that a rational private operator could realistically have taken.

Third, they do not apply to measures that do not (even potentially) affect competition or trade between States.  It is generally accepted that the case-law of the CJEU and the practice of the Commission has set that hurdle quite low, although the Commission has recently taken a number of decisions that indicate that it is trying to raise that hurdle.  Moreover, a de minimis regulation has created a safe harbour for many small aid measures.

As noted above, the State aid rules acknowledge the existence of a range of powerful policy justifications for subsidies.  So the Commission is given wide power under Article 107(2) and (3) TFEU to approve State aids (“declare them compatible with the common market”) on a range of public policy grounds.  That approval mechanism has the following key features.

First, it is unlawful (under Article 108(3) TFEU) for a Member State to implement an aid measure before obtaining approval from the Commission (known as the “standstill obligation”).  National courts are required to enforce that rule if the matter comes before then, although there is flexibility in how that is to be done in individual cases.

Second, however, the effects of that rule have been significantly reduced in recent years by a series of block exemptions (including, most importantly, the so-called General Block Exemption Regulation, applying across a whole range of sectors to numerous types of aid measures) which automatically clear the bulk of Member States’ aid measures without any need to obtain the approval of the Commission.

Third, where individual clearance is required from the Commission, that can be obtained very rapidly in an emergency (there are some examples of banking aid being given within 24 hours); but in general it is a process that takes months or even years.

Fourth, though the EU Courts will carefully scrutinise Commission decisions as to whether a measure is or is not State aid (mainly a question of law), they allow the Commission a wide latitude in terms of its aid approval policy, intervening only in the case of legal or procedural errors, or serious flaws in reasoning or fact-finding.


We suspect that there is wide support for the aim of the State aid rules: ensuring that State subsidies that have an economically distortive effect should be given only where they are appropriate and proportionate to deal with market failures.  It may be noted that the State aid rules do not prevent either nationalisation or privatisation: and many EU countries manage a range of industrial strategies while fully complying with the State aid rules (Germany, for example).

Policy concerns about the State aid rules have generally focused on three areas.

First, as noted above, there is concern that the Commission and the CJEU have tended to widen the scope of the State aid rules to catch measures that should not be the concern of a regime whose principal purpose (at least historically) was to protect competition in the internal market against distortions caused by unjustified subsidies.  Those concerns centre on both the definition of State aid and on the approach taken to the requirements that a State distort competition and affect trade between Member States.

The second set of concerns focuses on the policy of the Commission in deciding whether to approve aid notified to it. Concerns have centred on lack of transparency, lack of economic rigour and, partly as a result of those failings, a concern that the Commission’s approach is sometimes too “political”.  In general, however, it is fair to say that the Commission (encouraged in particular by successive UK Governments as well as academic commentators) has tended over recent years both to set out its approach in different sectors in considerable detail, and to adopt a more rigorous economic approach to identifying the market failure sought to be addressed by the aid measure and to evaluating whether the measure is the most appropriate means of addressing that failure.

The third set of concerns relates to procedural issues: for present purposes the most important of these is the delay caused by the time taken by the Commission to deal with individually notified measures, given the unlawfulness of proceeding with those measures before the Commission’s approval has been obtained.  Those delays are aggravated by the delays caused by appeals to the EU Courts against Commission decisions.  There is no doubt that those delays can prove frustrating to policy-makers and to businesses whose projects depend on State support, and in some cases those delays can stop a desirable project or make it more expensive.  On the other hand, the increasing scope of block exemptions has (as noted above) substantially reduced the number of projects that have to be notified to the Commission for approval, with the consequent reduction in the risk of projects being delayed by appeals against decisions to the EU Courts.  Another area of concern is the limited procedural rights given to aid recipients in the process, even though (in the case of investigations for unlawful aid) the consequence of a finding of aid is often an order for recovery with very serious adverse effects on the aid recipient.

The various concerns summarised above were essentially those expressed by respondents to the Coalition Government’s Review of the Balance of Competences.  §3.7 of the section dealing with Competition and Consumer Policy reported that “there was broad agreement in principle on the current balance of competence on State aid, but some expressed concern about its limits, about real or apparent extension of EU competence into areas of domestic policy, and about the way State aid controls are exercised”.State aid rules in comprehensive trade agreements with the EU


We assume here that the United Kingdom will seek to negotiate a comprehensive trade agreement with the EU.

We note in that respect that, with the exception of Switzerland, every other European country with which the EU has entered into comprehensive trade agreements has accepted that it will comply with State aid rules.  There are, in essence, two models.

The first is the EEA model.  The EEA Agreement effectively replicates the EU State aid rules[1], with the EFTA Court playing the same role as the EU Courts and the EFTA Surveillance Authority (“ESA”) playing a role equivalent to that of the Commission.  The significant differences of substance are that: –

  • the EEA Agreement does not have direct effect[2].  However, EEA States are required to (and have) incorporated into their domestic law the obligation not to implement aid unless and until approved[3] and to implement, for example, prohibition and recovery decisions by the ESA.  Moreover, EEA States must also, under that Agreement, pay damages to any third party harmed by a manifest and serious breach of the standstill obligation[4] – and the ESA has indicated that almost any breach of that obligation would trigger a duty to pay damages[5];
  • the EEA agreement does not apply to agricultural products falling outside the scope of Article 8(3) EEA, or to the fisheries sector[6]; and
  • EFTA Court opinions given in response to references from EFTA States are “advisory”, rather than binding on the courts of those Member States – see Article 34 of the Surveillance and Court Agreement.

The second is what might be called the “domestic implementation” model, and is found in agreements with European countries outside the single market.  Perhaps the most pertinent example (since it is with a large State that will not be applying for EU membership for the foreseeable future) is the Association Agreement between the EU and Ukraine (“the Ukraine Agreement”)[7].  Article 262 of the Ukraine Agreement sets out the State aid rules; Article 264 provides that they are to be applied “using as sources of interpretation the criteria arising from the application of [the EU State aid rules] including the relevant jurisprudence of the [CJEU], as well as [Commission frameworks and guidance].” Article 263 requires each of the EU and Ukraine annually to report to each other on the State aid granted on each side.  Most interestingly for present purposes, Article 267 requires Ukraine to implement a domestic system of State aid control, with “an operationally independent authority … entrusted with the powers necessary for the full application of [the State aid rules][8].

As far as Switzerland is concerned, it has a series of bilateral agreements with the EU. Of those, the ones that mention State aid are the 1972 Free Trade Agreement and the 1999 Agreement on Air Transport.

The 1972 FTA contains, at Article 23(1)(iii), a general prohibition on “any public aid which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods.” We understand that this provision has not been applied in Switzerland and that as a matter of Swiss law it is of limited application.  However, the Commission has on at least one occasion raised what it regards as infringements of Article 23 with the Swiss Government[9].

The 1999 Air Transport Agreement is more thorough in its reference to familiar concepts of EU State aid law, containing at Article 13 a provision that closely reflects Article 107 TFEU. However, although specific provision is made in relation to enforcement of the Articles reflecting Articles 101 and 102 TFEU (the general prohibitions on anti-competitive agreements and abuse of dominant position) by the European Commission and the Swiss authorities, no enforcement mechanism for the prohibition is provided other than, at Article 14, a general requirement to keep measures falling within Article 13 under review.  No specific authority has been given power on the Swiss side to enforce this rule.

We also understand that Swiss law contains general prohibitions on public subsidies that fail to meet conditions of economic efficiency and a general requirement that Swiss government bodies respect competitive neutrality, but we also understand that these are not often invoked before the Swiss courts.

The EU has been prepared to negotiate agreements with countries outside Europe – notably CETA (Canada) and the ongoing negotiations on TTIP (United States) – that do not contain prohibitions on the grant of subsidies.  However, Article 7 of CETA reflects and reinforces WTO anti-subsidy obligations by providing for notification to each other of subsidies granted and for a consultation procedure between Canada and the EU if either considers that the other is harming it by granting subsidies.

What can be concluded from that brief survey is that the EU, as far as Europe is concerned, has generally insisted on compliance with State aid rules as a condition of a comprehensive trade arrangement.  (We suspect that, as in other areas, the case of Switzerland is not a reliable precedent.)  It is ultimately a question for diplomats, rather than us, whether and to what extent the EU would so insist in the case of the United Kingdom. But it is at least possible that State aid compliance will be a “red line” condition on the EU side for any comprehensive trade deal anywhere in the range between CETA and the EEA Agreement, not least because it will be hard to explain to EU voters why their employers should potentially face competition from subsidised UK businesses when their employers are unable to receive equivalent subsidies.  Moreover, the EU will bear in mind that (unlike the US or Canada) compliance with State aid rules is not a novelty as far as the United Kingdom is concerned and that the United Kingdom has considerable experience and expertise in applying the State aid rules over the last four decades.

WTO Rules

It is also important to be aware that, even outside any trade agreement with the EU containing State aid rules, the United Kingdom will, in relation to goods, still be bound by WTO anti-subsidy rules.

A good account of those rules can be found in a paper by David Unterhalter SC and Thomas Sebastian[10], as well as in Bacon “EU Law of State Aid” Ch.4[11], which we simply summarise.

As they point out, there is considerable overlap between the WTO concept of “subsidy” in Article 1 of the Agreement on Subsidies and Countervailing measures (the “SCM Agreement”) and the concept of “state aid” under Article 107(1) TFEU. Both concepts involve: (1) measures which are taken by governments or which are imputable to governments; (2) the grant of benefits (to use WTO terminology) or advantages (to use EU terminology) which are assessed using market-based tests; and (3) measures which are not generally applied but which are specific (to use WTO terminology) or selective (to use EU terminology). Moreover, measures which are purely regulatory in nature, for instance exemptions from labour or environmental standards, would fall outside the scope of both sets of rules as WTO law requires the presence of a “financial contribution”/ “income or price support” while EU law requires the involvement of “state resources”.

However, as they also point out, there are considerable differences between the concepts. First, the WTO regime does not apply to services, but only to goods.  Second,  measures which do not involve any cost to the government, such as a price control measure, would clearly be outside the scope of Article 107(1) TFEU but may fall within the scope of the SCM Agreement (although the extent to which the SCM Agreement applies to such measures remains somewhat unclear, as there are few decisions on this point). Likewise, the complex EU law approach to the assessment of selectivity in cases involving tax exemption measures has no direct analogue in WTO law.

Moreover, the enforcement mechanisms for the WTO rules are (i) either state-to-state dispute resolution (there being no mechanism for private enforcement, injunctions or damages, or for actions to be brought in ordinary courts) or (ii) the imposition by the adversely-affected state of countervailing duties on products from the infringing state.  (Calculation of the appropriate rate of countervailing duties is generally more complex than calculating the amount of unlawful State aid that has to be repaid.).  There is therefore no scope for WTO enforcement of the rules by private operators: they have no right of action in national courts and no independent body to which they can complain, and their only option is to persuade their own government to invoke the WTO procedure.

Finally, there is no procedure in the WTO rules for the approval of justified subsidies on public interest grounds, as is possible under Article 107(2) and (3) TFEU.


We now turn to the policy considerations which, in our view, the UK Government should bear in mind in deciding, in the context of negotiations with the EU, what, if any, State aid regime should be retained post-Brexit.

We appreciate that the State aid question will be but one of numerous issues on which the Government will need to negotiate.  However, in deciding its wider negotiating position, the Government will need to form a view on the extent to which a State aid regime imposes burdens on, or benefits, the United Kingdom, and it is this question that we address.

We start by acknowledging that any constraint on the ability of public bodies to act as they see fit in relation to the expenditure of public money requires careful justification.  That is particularly because the United Kingdom has a number of well-established means of ensuring that public money is well-spent (an advantage not enjoyed to the same degree by all EU Member States).

That said, however, we see the following advantages in retaining some form of domestic State aid or anti-subsidy control.

Domestic considerations

There seem to us to be two principal domestic considerations

The first is that the United Kingdom will want to ensure that it respects its obligations under the WTO SCM Agreement.  The UK Government can of course ensure through administrative means that its own conduct complies with those obligations.  But there are large number of public bodies which have wide powers to make their own spending decisions without reference to Whitehall.  Given the overlap between those obligations and the State aid rules, it has to date been unnecessary to consider the extent to which UK law needs to ensure that public bodies do not take measures that conflict with the SCM Agreement. But in the absence of those rules, it may well be necessary to ensure, by means of domestic law, that support measures adopted by public bodies do not put the United Kingdom in breach of its WTO obligations.

The second, linked to the first, is that increasing devolution (both to Scotland, Wales and Northern Ireland and increasingly within England) means that there are now a large number of public bodies with their own substantial tax and spending powers independent from the financial control of the UK Government.  That strengthens the case for a form of legal control on the ability of those bodies to subsidise favoured firms: legal control that to date has been provided by the State aid rules (and is provided by the State aid rules in EU Member States with fiscally autonomous regional government, such as Germany and Spain[12]).  We recognise that there are devolution issues here (and that, under the Sewel convention, it may well be that the consent of the devolved administrations would be needed before their powers were limited by a form of State aid control).  But there is a powerful policy case for such control, given that it is in no-one’s interests for there to be “subsidy races” between different parts of the United Kingdom to attract investment.  Moreover, any such provision would do no more than re-instate the limitations until now imposed by the State aid rules.

EU considerations

There also seem to us to be advantages of retaining State aid rules in terms of protecting the interests of UK business.

First, if the United Kingdom were to enter into a commitment to comply with State aid rules (whether in the form of the EEA Agreement or the Ukraine Agreement), that would carry with it a corresponding obligation on the EU institutions to prevent State aid that harmed competition in the United Kingdom.

Further, the EU State aid rules only catch measures that (at least potentially) distort competition in the EU/EEA.  If the United Kingdom were to leave not only the EU but also the EEA, a measure that affected competition only in the United Kingdom (for example, potentially, an Irish subsidy aimed at assisting exports to the United Kingdom) would not as such be caught by the EU State aid rules[13].   In practice, the Commission would, if it objected to the measure, often be able to find that even a measure targeted at exports to a non-EU/EEA country has sufficient effects within the EU/EEA to satisfy the “effect on trade between member States” requirement: but if the reality is that the effect of a State aid measure is centred on a State outside the EU/EEA, the Commission is perhaps unlikely to make it an enforcement priority.

Those issues would not arise if the UK were to remain within the EEA or were to negotiate an agreement similar to the Ukraine agreement, since the effect of both the EEA and Ukraine agreements is to give the EU institutions the power (and the duty) to regulate State aid measures by EU Member States that harm competition in (respectively) EEA States and Ukraine.

Second, when an EU Member State takes State aid measures that harm businesses trading in (respectively) an EEA State or one of the States with agreements similar to the Ukraine agreement, the relevant Agreement gives a right of action in the courts of the Member State concerned to obtain damages.  So, for example, if the United Kingdom were party to EEA/Ukraine type State aid provisions, and if the French Government decided to subsidise steel exports to the United Kingdom, UK steel manufacturers would have the right to sue the French Government for breach of the State aid rules for losses suffered by them in the United Kingdom.  Such actions have been rare to date, though there is no doubt that in principle State liability does arise.

Third, in cases where the United Kingdom has granted subsidies to UK companies operating in the EU, the fact that such subsidies have been approved (or block exempted) under provisions analogous to the EU State aid provisions will make it in practical terms difficult for the EU to take retaliatory measures against the United Kingdom under the WTO SCM Agreement.  We should emphasise though that that is a practical rather than a legal point: Article 265 of the Ukraine Agreement makes it clear that the State aid provisions in that Agreement are without prejudice to the right of both parties to invoke the WTO SCM Agreement.

Fourth, under both the EEA and Ukraine-type arrangements, the United Kingdom would retain a role in the development of EU State aid law (which, given the importance of the EU market to the United Kingdom, will remain a matter of important policy concern for the United Kingdom).  In the EEA model, the United Kingdom would have a direct role in influencing the practice and jurisprudence of the ESA and EFTA Court, both of which in turn influence the development of Commission and Court of Justice thinking[14]. The United Kingdom’s role in the ESA and EFTA Court would be considerable, given its size and importance, and would be likely to increase the influence of the EFTA institutions.  And, as an EEA State, the United Kingdom would have the right to intervene, itself, in any EEA-relevant case (including State aid cases) before the Court of Justice[15].  But even in the Ukraine model, the United Kingdom would have a right to be consulted about and to influence any decision or policy development in the State aid field that affected its interests.

Would retaining a State aid regime outside the EU give scope for improvements vis-à-vis the current EU regime?

Compared to the EU regime, we see some advantages for the United Kingdom in moving to a State aid regime along either the EEA or the Ukraine lines.

First, although in both cases the notion of State aid would be the same as the EU concept, day-to-day enforcement would be in the hands either of the ESA (in which the United Kingdom would be a major player) in the case of the EEA agreement, or in the hands of a UK agency in the case of the Ukraine model.  As we have already pointed out, in the area of clearance of State aid on the ground of compatibility there is considerable scope for policy judgment even when detailed guidelines exist; and even in terms of the definition of State aid (which is a question of law) we do not think that either agency would be as tempted as the Commission sometimes is  “push the boundaries” (see, for example, in its controversial decisions in the tax ruling cases, which are widely argued to be an attempt to deal in the State aid field with what in fact are wider policy concerns about tax avoidance by certain multinational companies). That addresses the policy concern we identified at §11 above.

Second, in relation to the policy concern we identified at §12 above (lack of economic rigour and transparency in approval decisions), the United Kingdom would, in relation to either the ESA or a domestic agency, be in a very good position to ensure that the agency took transparent and economically rigorous decisions.  Indeed, a number of State aid law practitioners take the view that ESA decisions are clearer and better reasoned than those of the Commission, and involve greater participation by the beneficiary of aid (though that could be because the ESA takes a small fraction of the State aid decisions taken by the Commission).

Finally, in relation to the policy concern we identified at §13 above (delay) the United Kingdom would, in relation to both the ESA or a domestic agency, be in a good position to ensure both speedy decision-making and speedy appeals (we note, in that respect, that appeals to the EFTA Court typically take between six months and one year – which compares very favourably to the period of five or more years it can take to appeal a decision to the General Court and ultimately to the Court of Justice).  The ability to get substantially swifter decision-making would, in our view, very substantially improve the State aid regime compared to the present situation, and very significantly reduce the constraint and uncertainty the EU regime imposes on public authorities and on business.

Choice of model

If the Government decides that it is right to retain a State aid regime on either the EEA or Ukraine model, which is preferable?

That choice is likely largely to be dictated by the extent to which the arrangement with the EU involves UK participation in the EEA.  Full membership of the EEA would of course entail the “EEA option” in the State aid area.  But if the United Kingdom decides to make some use of the EEA institutions under some arrangement under which those institutions are “borrowed” for certain purposes, then it would seem to us to be sensible to make use of them in the State aid field, given the established expertise and reputation of both the ESA and EFTA Court.  It also avoids the legal issues and likely greater expense of setting up a national State aid regime.  It should though be noted that, due to the absence of the principle of direct effect in the EEA Agreement, the United Kingdom would have to make domestic legislative provision for the State aid rules – something which the EU doctrine of direct effect has made unnecessary to date.

Creating a national State aid regime would raise a number of issues.  The body would have to be demonstrably independent, and (given that much of its work would involve dealing with central and devolved governments) would have to be strongly protected against political pressure.  The obvious body to take that responsibility would be the Competition and Markets Authority (“CMA”), but it would have to be recognised that State aid regulation would be a considerable expansion of its responsibilities into an area that it has not to date had to deal with, and it would have to be resourced accordingly.  It also has to be recognised that conferring State aid control powers on the CMA would put the CMA in a position where it was effectively reviewing important policy decisions by Ministers.  It might be that a more “judicial” model was more appropriate, so that enforcement decisions would be taken by a court, perhaps the Competition Appeal Tribunal, on application by a specialist State aid monitoring body: but although the question of whether a measure is State aid or not is suitable for judicial resolution, the question of whether State aid is justified and should be approved on public interest grounds is not obviously one that should be decided by judges (though judicial review of such decisions would be approporiate).  A further issue is that it would be difficult to see how, in the UK constitutional system, a State aid regulator would deal with cases where the State aid was in the form of primary UK legislation: its powers would, we would have thought, there need to be confined to a declaratory power.  It would also have to be decided what powers the body had to deal with secondary legislation incorporating unlawful State aid.

We therefore think that, other things being equal, an arrangement that brought the United Kingdom into the EEA State aid regime would be the best way forward, if the present State aid regime is to be broadly maintained: and, as we suggested above, it may be that the EEA institutions can be “borrowed” for that purpose even if, in some other respects, the United Kingdom does not wish to take part in the EEA Agreement.


We finally turn to transitional issues that need to be considered.  We start here by observing that uncertainty as to the transitional position – particularly if the United Kingdom does maintain some form of State aid control – could well cause delay in infrastructure projects if it is not clear how any new regime will deal with aid necessary to fund those projects.  It seems to us that the following issues arise.

The Article 50 agreement and any new arrangements with the EU or EEA would have to deal with : –

State aid notified to or being considered by, but not yet decided by, the Commission at the time of Brexit;

  • the status of any State aid cases involving the UK that were pending before the EU Courts at the time of Brexit, whether references to the Court of Justice or direct actions in the General Court (or on appeal to the Court of Justice)[16];
  • the extent of the United Kingdom’s post-Brexit obligation to annul (and usually recover) any unlawful aid implemented before Brexit;
  • the extent to which the Commission (or anyone else) had power post-Brexit to order the United Kingdom to recover unlawful aid granted before Brexit, or would the Commission be confined to opening a WTO dispute; and
  • the extent to which the United Kingdom is required, post-Brexit, to comply with the terms of any Commission decisions addressed to it before Brexit, including in particular the numerous Commission decisions approving aid schemes.

If the United Kingdom were to retain a domestic or EEA-type State aid regime, arrangements would need to be made for existing State aid decisions approving ongoing State aid measures to “carry over” to the new regime.

Provision would also need to be made, whether in any new arrangements with the EU/EEA or as a matter of domestic law, for the status of the residual aid measures currently implemented in the UK that predate the accession of the United Kingdom to the EU and are therefore, under the EU rules, regarded as existing aids that are not subject to the same rules as apply to new aid measures: see, for example, the BBC licence fee arrangements.

State aid damages actions against the authority granting an unlawful State aid are rare and none have been successful so far in the UK. But in principle, in domestic law, in relation to unlawful State aid put into effect before Brexit, it seems to us to be clear, without any legislation, that third parties would be entitled to damages in relation to the pre-Brexit period, and could sue for such damages after Brexit.



[1] See Art. 61 and 62 of the EEA Agreement. Art, 61 essentially repeats Art.107 TFEU.  Art. 62 requires “constant review” of existing and planned measures in the EEA to ensure compatibility with Art.61, a task which in the EEA/EFTA States is allocated to the ESA. The ESA then has, under Art.5 of the Surveillance and Court Agreement (“SCA”), the general duty to ensure the compliance of the EEA/EFTA States with their duties under the EEA Agreement, and Article 24 SCA then enumerates compliance with the State aid rules as an aspect of that duty and points to Protocol 3 SCA. That Protocol effectively incorporates the equivalent provisions to Art.108 TFEU: it provides for the duty to notify new aid (Art.2), and an obligation not to put that aid into effect before approval by the ESA (Art.3).

[2] See e.g. Case E-4/01 Karlsson v Iceland at §28.

[3] See §22 of the ESA’s guidelines on enforcement of the EEA State aid rules by national courts, available at

[4] Case E-4/01 Karlsson v Iceland at §29.

[5] See §§43ff of the ESA’s guidelines on enforcement of the EEA State aid rules by national courts, available at

[6] Art.4 of protocol 9 to the EEA.

[7] [2014] OJ L161/3

[8] There are similar provisions in Accession Agreements with Albania, Bosnia and Herzegovina, Macedonia, Montenegro, Serbia, and Turkey.

[9] the Commission decided that certain company tax regimes in Swiss Cantons in favour of holding, mixed and management companies were a form of State aid incompatible with Art.23, and asked the Council for a mandate to negotiate a satisfactory resolution.


[11] In both the present 2nd and forthcoming 3rd editions.

[12]In Spain, there is a specific provision (art.11 of Ley 15/2007 de Defensa de la Competencia) allowing the Comisión Nacional de Competencia (the national competition authority: “CNC”) to review and report on any State aids when asked by a local or regional Government to do so, and to require local and regional authoirities to supply it with relevant information. The CNC is also provided, by the Spanish Ministry of Foreign Affairs, with a copy of all Spanish State aid notifications.

[13] see e.g. Case T-34/02 EURL Le Levant 001 ECLI:EU:T:2006:59 at §§115-117.

[14] Art.64 EEA gives the ESA a formal right of consultation in relation to the Commission’s development of EU State law and policy, and the ESA may intervene in cases in the EU Courts – indeed it did so recently in a State aid reference from the UK’s Court of Appeal in Case C-518/13 Eventech v Parking Adjudicator ECLI:EU:C:2015:9.

[15] As Norway has done in some EU State aid cases: see e.g. Joined Cases T-371 and 394/94 British Airways et al v Commission [1998] ECR II-2405

[16] An example of a case affecting the UK likely still to be before the EU Courts at that time is Case C-356/15 Austria v Commission (Hinkley Point)

Posted in Brexit issues | Comments Off on Paper on post-Brexit options for State aid

Ombudsman rejects complaint of maladministration in relation to third party rights in Commission investigations

Some of the readers of this blog will be aware of John Temple Lang’s complaint to the Ombudsman, and related article (“The Charter and the EU State Aid Procedure” in de Vries, Bernitz and Weatherill, The EU Charter of Fundamental Rights as a Binding Instrument), in which he argues that the Commission’s practice of refusing to provide aid beneficiaries and other interested parties with access to its State aid file was in breach of Articles 41 and 47 of the Charter of Fundamental Rights of the EU.

In a recent decision (see full text of decision here), the Ombudsman has rejected that complaint. She found that the fact that beneficiaries and other interested parties do not have access to the State aid file reflects their limited role in a State aid investigation, and the fact that the investigation is instigated against a Member State rather than against the aid beneficiary.

Continue reading

Posted in Uncategorized | Comments Off on Ombudsman rejects complaint of maladministration in relation to third party rights in Commission investigations

Book launch on 10 October – “EU Renewable Electricity Law and Policy”

UKSALA members are warmly invited to attend the launch of the book “EU Renewable Electricity Law and Policy” by Tim Maxian Rusche, a member of the legal service of the European Commission. Since a central topic of the book is State aid, this work is likely to be of interest to many of our members.

The event will take place on Monday 10 October at 12.30pm, in the Moot Court at Kings College, Strand Campus. Full details and a registration link are here.

Posted in UKSALA news | Comments Off on Book launch on 10 October – “EU Renewable Electricity Law and Policy”

Ferracci and Montessori: ecclesiastical aid and post-Lisbon admissibility

On 15 September 2015 the General Court handed down two parallel judgments in the Ferracci and Montessori cases, concerning Italian rules granting tax exemptions for various “non-commercial” entities, including ecclesiastical institutions. The judgments are noteworthy for their comments on the impossibility defence to recovery and the application of the State aid rules to entities with predominantly non-commercial activities. Nevertheless, for this blogger, the most unusual aspect of the judgments is that this is the very first occasion where a challenge to a State aid decision was found to be admissible on the grounds that the decision was a “regulatory act which … does not entail implementing measures” for the purposes of the fourth subparagraph of Article 263 TFEU.

Continue reading

Posted in European Court judgment | Comments Off on Ferracci and Montessori: ecclesiastical aid and post-Lisbon admissibility

Berlin seminar slides and handouts

The slides and other materials from the joint seminar with the Berliner Gesprächskreis zum Europäischen Beihilfenrecht, on 24 June 2016, are all available here for those who are interested (click on individual speaker names to obtain their slides or speeches).

We are, once again, enormously grateful to the Berliner Gesprächskreis for organising and hosting such a successful event – and for their solidarity on the day following the result of the EU referendum overnight.


Posted in UKSALA news | Comments Off on Berlin seminar slides and handouts

Apple State aid decision just announced

In a press release today (, the Commission has announced that it has decided that Apple received State aid from Ireland in the form of “undue tax benefits of up to €13 billion”.

According to Ms Vestager, the Commissioner responsible for competition, “Ireland granted illegal tax benefits to Apple, which enabled it to pay substantially less tax than other businesses over many years. In fact, this selective treatment allowed Apple to pay an effective corporate tax rate of 1 per cent on its European profits in 2003 down to 0.005 per cent in 2014.”

The thrust of the decision (the full text of which will not be made public for a while) appears to be as follows.  First, as a matter of background, Apple had set up sales arrangements across the EU which meant that sales were regarded for tax purposes as effected in Ireland.  Second, against that background (which does not appear to be questioned, though see below), Ireland  agreed, in tax rulings in 1991 and 2007, “artificial” arrangements which allowed profits from those sales to be allocated to a head office “not based in any country”.  It is those tax rulings which are attacked as “selective”.

As to recovery, the Commission interestingly notes that, having looked at its decision, other EU Member States may want to open up the conclusion that the sales arrangements entered into by Apple across the EU did result in sales being made (for tax purposes) in Ireland.  If they do that, the Commission says, that will impact on the amount that Ireland needs to recover, since it will reduce the volume of Irish sales.  Further, in a comment that may reflect adverse US comment on its investigation, the Commission states that “the amount of unpaid taxes to be recovered by the Irish authorities would also be reduced if the US authorities were to require Apple to pay larger amounts of money to their US parent company for this period to finance research and development efforts.”

Some commentators have suggested that this decision calls Ireland’s generally low corporation tax rate into question.  That is simply wrong: the decision is not about Ireland’s generally low rate (which is not subject to control in EU law) but rather about what amounts, in effect, to an alleged waiver by Ireland of its own tax rules by incorrectly accepting Apple’s allocation of profits to its head office based in no country, when that allocation had “no factual or economic justification”.

As usual in these cases, the devil is in the detail: it will be important to look carefully at the actual decision (reported to be 130 pages long) before any confident assessment can be made of its implications.




Posted in Commission decision | Comments Off on Apple State aid decision just announced