George Peretz QC of Monckton Chambers looks at the subsidy control provisions of the UK-EU Trade and Cooperation Agreement and what they mean for the United Kingdom’s future state aid regime.
We now know – which we did not know at the time of the BEG online conference – that the UK and EU have agreed a Trade and Cooperation Agreement (the TCA) to govern their relationship as from 1 January 2021. The TCA contains at Article 3 extensive provisions on subsidy control. Those provisions reflect a compromise between the EU and UK positions discussed below. A further critical part of the picture is the effect of Article 10 of the Ireland/Northern Ireland Protocol (the Protocol) which, under the Withdrawal Agreement, is now in force as directly effective law in the UK (the UK government having abandoned the provisions of the Internal Market Bill that would have given it powers to renege on its promises under the Withdrawal Agreement by removing or limiting the effect of Article 10 in UK law).
The FTA negotiations: the EU position
The EU’s opening position was that the UK should continue to follow EU State aid rules, albeit applied by a domestic State aid authority – essentially the regime agreed in the “backstop” by the May Government in 2018. To that extent, the EU was seeking arrangements similar to those that it has with Turkey and Ukraine, which both agree to apply EU State aid rules domestically. The EU was also proposing, however, that UK courts would continue to be able to – and at the supreme appellate level required to – refer questions of interpretation to the ECJ, a court on which the United Kingdom is no longer represented. That latter proposal had no equivalent in the EU’s agreements with Turkey or Ukraine: and it may be noted that even Norway, Iceland, and Liechtenstein are not subject to such a provision (the EFTA Court being independent of the ECJ and composed of judges from those countries). And it was correspondingly far removed from the provisions in CETA (which do not provide for dispute resolution in the area of subsidies at all).
In order to understand that opening position, the key thing to bear in mind is that there is a dramatic asymmetry between the position of the United Kingdom outside (but next to and trading extensively with) the EU, and that of EU Member States. State aid control is in the structure of the EU treaties: it is a fundamental aspect of the regime binding all Member States. The State aid regime limits Member States’ ability to (for example) offer large grants to footloose multinationals in return for locating in their territory, or to subsidise their car exports so as to undercut other Member States’ car industries.
But – critically – compliance with those regimes does not just benefit other Member States: it also (incidentally but inevitably) helps the UK. If Germany cannot freely subsidise its production of manufactured goods, that benefits UK manufacturers selling goods into the EU as well as French ones. And if France cannot grant large incentives to banks that decide to re-locate in Paris, that benefits London as well as Frankfurt. In contrast, after the end of transition, the UK Government will—subject to Article 10 of the Protocol—be able to subsidise manufacturing production, or grant incentives to banks to locate in the United Kingdom, as it likes.
Put another way, the reason why the United Kingdom never had as a major negotiating priority the demand that the EU avoid subsidising industries that compete with UK industry – something that would be a key UK concern given the extent to which its imports come from the EU and compete with its domestic industry – is that the United Kingdom can, in effect, free-ride on the EU State aid regime: and that ability to free-ride while refusing to accept any of its obligations is obviously intolerable for the EU (that is to say, it is an impossible sell to its voters).
One response to that point made by some UK commentators was to observe that, under WTO rules, the EU can impose countervailing duties if it finds that UK subsidies are damaging its domestic industries. However, such duties take considerable time and effort to implement and monitor, and are inevitably reactive (putting the fire out rather than stopping it from starting). Moreover, countervailing duties cannot deal with services – such as the bank location example I gave above – at all: which is a critical point given the very large volume of UK services exports to the EU – much of which volume is likely to remain, for example through EU subsidiaries, even after the end of transition. It should also be noted that the United Kingdom sought agreement from the EU in the area of services access beyond anything previously agreed, for example in relation to freight haulage access, business travel to provide services, mutual recognition of professional qualifications, and equivalence in the area of financial services (though neither of the latter two requests succeeded in finding their way into the TCA, and the business travel provisions are little better than those granted to Canada or Japan). Nor can countervailing duties deal with concerns such as the effect of subsidies in giving a leg-up to UK businesses competing for public procurement business in the EU, or seeking to acquire assets or companies in the EU. Given the deep integration between the UK and EU economies – an integration that will take time to unwind – EU Member States cannot be expected to sell to their voters a situation in which UK businesses get such leg-ups while their own businesses do not and cannot.
That said, it was fair to point out that (as observed above) the EU’s opening position went beyond the position it has adopted in relation to Turkey and Ukraine and would be more appropriate for a State that is seeking to align with State aid rules than one which wishes to move away from them. By the closing stages of the negotiations the EU appeared to have modified its position so that it was in principle prepared to accept a UK subsidy regime that fell short of full alignment.
The UK position
The UK government’s opening position was, essentially, that commitments on subsidies should be confined to those found in CETA (the FTA between Canada and the EU). Those amount to little more than an agreement to keep each other informed of subsidies every two years. Oddly, reference was also made to the FTA between Japan and the EU even though that FTA – as I note below – contains somewhat stronger provisions than the UK was then offering.
That position was not easy to reconcile with the Political Declaration agreed with the EU in October 2019 before the UK general election, where at §77 it agreed to “maintain a robust and comprehensive framework for … state aid control”. But that promise still appeared to have some life: in March 2020, Minister for the Cabinet Office, Michael Gove told the Commons Select Committee on the Future Relationship that the government planned to produce a “robust” subsidy regime that would satisfy the EU. That was consistent with Conservative Party statements during the 2019 general election that a Conservative government would set up a UK subsidy regime, although those proposals were unclear in a number of respects. It appeared at that stage that the current government was listening to a number of experts in the area – several of whom were on record as supporters of Brexit – who believed that some form of UK subsidy regime would be needed for domestic purposes, not least to prevent harmful subsidy races between different UK home nations (the devolved administrations being largely fiscally autonomous from the UK government).
However, over the summer it became increasingly clear that nothing was being done to develop these ideas into concrete proposals: certainly, there was no attempt (as one would have expected were proposals being seriously developed) to consult with experts, businesses and others affected about the design of such a regime. And, on 9 September, BEIS announced that any such regime would not be announced until well into 2021, and that as of 1 January 2021 the United Kingdom would have no domestic regime, apart from “guidance” to public authorities on the (largely unenforced) WTO subsidy regime. It therefore appeared that the UK government was not prepared to make any commitments on a domestic subsidy regime to the EU, beyond a vague commitment to develop proposals at some point. That impression was confirmed when the UK government announced that it planned to remove the State aid provisions of the EU Treaty from domestic law after the end of transition without replacing them (and, somewhat dubiously, to do so by statutory instrument under the EU Withdrawal Act, notwithstanding the statements made at the time of that legislation that statutory instruments would not be used to effect major changes in policy, which were to be reserved for primary legislation – a position which, it now appears, is likely to be challenged by the Good Law Project in litigation).
But even at this later stage the UK government’s position was, to use a technical term, all over the place. First, at the same time as announcing that there would be no UK subsidy for the indefinite future, the government also asked Parliament, in the Internal Market Bill, to give it powers to impose a UK subsidy control regime on the devolved parliaments by making subsidy control a “reserved matter” under the devolution acts: a provision now enacted as section 52 of the Internal Market Act 2020. So the government appeared to be laying the foundations for a house it disavowed any clear plan to build. Second, at around the same time the UK government agreed an FTA with Japan that contained substantive commitments that the UK had conspicuously not been prepared to make to the EU: in particular, a commitment to prohibit subsidies in the form of unlimited guarantees and subsidies to failing firms without a realistic restructuring plan. It may well be that little thought was given to the EU dimension before those commitments were made – the commitments were rolled-over from the EU/Japan FTA that served as a baseline for the UK/Japan negotiations. But whether accidental or not, the commitments to Japan forced the UK to offer at least those commitments to the EU. And implementing those commitments into domestic law would require legislation (the Japan FTA requires that “reasonable measures” be taken to extend those prohibitions to sub-national governments – and it is hard to see that, given the powers sought in the Internal Market Bill, it would be reasonable for the UK government not to use those powers to legislate across the UK).
A further indication of some softening of the UK position came in evidence given by Lord Frost – the Prime Minister’s Sherpa for the EU trade negotiations – to the Future Relationship Select Committee on 7 October, in which he indicated that there “probably” would be an “administrative” UK subsidy regime “further down the line”.
That statement was full of ambiguities. But perhaps the biggest ambiguity was what he meant by an “administrative” regime. A regime that simply operates as an administrative principle – with no more legal effect than general administrative guidance such as the Treasury’s “Managing Public Money” – was not going to be enough for the EU, which was looking for a binding and enforceable legal regime. (Such a regime would also have no effect on public authorities fiscally independent of the UK government, such as the devolved governments.) But a regime that is “administrative” in the sense that most decisions were taken in an administrative procedure by an independent regulator, with some reserve enforcement power in the hands of the courts to deal with cases where correct procedures were not followed, would be enough (indeed, the EU State aid regime would qualify as “administrative” in that sense).
In the end, as we shall see, the UK government conceded a regime that was “administrative” in that latter sense – even though the way in which it has so far chosen to implement means that the system that we now have, at least for the moment, is better described as “judicial” than “administrative”. But before looking at what has been agreed, I should say something about Article 10.
Article 10
The effect of Article 10 (and Article 12) of the Protocol is that, as from the end of transition, the whole of EU State aid law, including current mechanisms of enforcement by the Commission, national courts, and the Court of Justice of the EU (CJEU) applies to the whole United Kingdom after transition, to the extent that UK measures have a potential effect on trade in goods between Northern Ireland and the EU (in particular, of course, Ireland). Article 4 of the Withdrawal Agreement requires Article 10 of the Protocol (along with the rest of the Agreement) to have direct effect in UK law, and section 7A of the EU Withdrawal Act 2018 currently provides for it to have such effect. In consequence, as matters stand, Article 10 will be directly enforceable against all UK public authorities in the UK courts – and its extent (and in particular the application of the “potential effect on trade in goods between Ireland and Northern Ireland” test) is a matter for the Commission or, if action is brought before the UK courts, the UK courts, subject in each case to the supervision of the CJEU. Private parties will be able to take alleged breaches of Article 10 to the UK courts, which will be obliged to grant injunctions and award damages in accordance with the principles laid down by EU law. They will also be able, as now, to complain to the Commission, which will retain its current powers in relation to such measures (to approve tor not approve them if notified, and if they are not notified to order their withdrawal and the recovery of aid already granted, in each case subject to appeals by those affected to the General Court and ultimately the CJEU).
Essentially, Article 10 leaves the United Kingdom significantly bound by EU State aid rules after transition, subject only to a threshold (‘effect on trade”) which is notoriously low and imprecise, and which falls to be interpreted and applied by the Commission and, ultimately, the CJEU. I have no doubt, for example, that the vast amount of State aid being granted by the government during the current Covid-19 crisis would have fallen within Article 10 – and have been subject to Commission clearance – had the current crisis fallen after transition. Similarly, a UK-wide tax measure such as the rules on the taxation of controlled foreign companies that were found to be a State aid last year would also, in my view, obviously be caught by Article 10 (and would therefore have to be disapplied as soon as found to be so by either the Commission or a national court).
After a period of conspicuous refusal to engage with these issues, it seems that the government at some point during 2020 realised what the consequences were of what it had agreed. However, its reaction was both unexpected and unconscionable, namely to insert clauses into the Internal Market Bill, that give Ministers a free hand to remove the direct effect of Article 10 (and to “modify” it, “interpret” it and restrict remedies available for breach of it) despite the admitted fact that such action would be flatly contrary to Article 4 of the Withdrawal Agreement it signed in January 2020. As many have pointed out, there is no UK precedent for avowedly breaching an agreement binding on the United Kingdom. And the government’s attempts to defend its action – to protect the peace settlement in Northern Ireland and to respond to alleged “bad faith” by the EU – were hard to accept as any more than irrelevant piffle: there was no plausible link between the current government’s concerns about the effect of Article 10 and peace in Northern Ireland, and there has been no substantiated or relevant allegation of bad faith.
Fortunately, the government decided in early December to withdraw those provisions from the Bill. That concession was probably inevitable given the overwhelming and determined opposition to those clauses in the House of Lords, but was covered by the fig-leaf of a declaration by the Joint Committee on the Protocol, which the government presented as having achieved its objectives. That claim was, however, unconvincing given both that the declaration was not binding on any court and that it was entirely anodyne in terms of its content.
The TCA
The subsidy control provisions of the TCA represent a compromise between the EU and UK positions.
The UK has achieved its objective of not being subject to EU State aid rules (if one forgets the continuing operation of Article 10 of the Protocol, about which the TCA says and does nothing). The subsidy control provisions are carefully written to avoid importing State aid law concepts into the UK’s commitments – indeed, with the obsessive determination of the Académie Française faced with a list of English words in common use in French, every concept is given a different name to that which the corresponding concept has in the State aid rules (“economic actor” for “undertaking”; “subsidy” for “aid”, and so on). And although it conceded that there would be an independent authority, it is bound only to give it an “appropriate role” (Article 3.9).
On the other hand, the EU secured UK commitment to an overall structure that looks very like the State aid regime. The scope is limited to subsidies that affect or could affect trade between the UK and EU. But there is a de minimis limit that, at SDR 325,000, is only around twice the (very low) EU limit. The definition of “subsidy” looks similar to the definition of “aid” (and carefully carries over, without naming it as such, the approach of the CJEU to the question of whether tax measures amount to State aid). There is a list of prohibited subsidies that look very like the list of “no chance of clearance” types of aid that one finds in any good State aid law textbook, and the principles to be applied in deciding whether other kinds of subsidy, in Article 3.4.1, are acceptable look very like the approach applied by the Commission under Article 107(3) TFEU. Even the exemption for subsidies to compensate undertakings (sorry: “economic actors”) affected by disasters such as pandemics – trumpeted as a UK victory – looks very like Article 107(2)(b) TFEU. Perhaps most importantly, Article 3.10 and .11 commits the UK to giving its domestic courts, on the application of interested parties, the power to review subsidy decisions of granting authorities for compliance with the prohibitions and principles set out in Article 3, and (save where the measure is an Act of the UK Parliament) the power to order recovery of non-compliant subsidies: and it is that provision that ensures that the new regime is not “administrative” in the first sense described above.
The current position
The current position is characterised by the government’s decision to implement Article 3 not by a specific provision in the EU (Future Relationship) Act 2020, nor by specific regulations made under it, but by section 29 of that Act, which simply imports Article 3 into UK law, at least to the extent that it requires to implemented.
That creates a very peculiar regime, which is, extraordinarily, now the law. There is (as yet) no independent authority. But the courts have – through section 29 – the powers that Article 3.10 and.11 requires them to have. Since no court has been given that role, it must (being a power to review the decisions of public authorities) in England and Wales fall to the Administrative Court. But it is entirely unclear how it is to approach that task. Should it be applying the Article 3.4.1 principles itself (which include concepts of proportionality), or should it be reviewing, and if so to what standard, the granting authority’s views as to its compliance with them? Is a pure rationality approach appropriate here (especially when the provisions refer to proportionality)? Is the Administrative Court equipped to deal with the economic arguments that will inevitably arise, particularly if the standard of review is higher than irrationality?
Lawyers will have profitable fun answering all these questions. But granting authorities, and potential recipients of subsidies, will have less fun, at least in big projects where the grant is, or almost certainly is, one falling within the scope of the regime and where the application of the Article 3.4.1 principles is contestable. The advice to them is, at least until there have been some cases, going to be a variation on “there is a non-zero risk that the Administrative Court will strike this down and order recovery”. And that advice is going to be of particular concern to recipients and funders because there is no way of removing that risk: no equivalent to the ability under the State aid rules to get projects notified to and cleared by the Commission as either acceptable aid or not being aid at all. What may well happen in practice is that practitioners fall back on the State aid rules as a type of comfort blanket – not least because in a case where the subsidy would clearly be exempt or cleared under State aid rules and policy guidance, it is not conceivable that the EU would object to it under the TCA and probably inconceivable that the Administrative Court would find it inconsistent with Article 3.
Meanwhile, Article 10 of the Protocol is now fully in force, as described above. Nothing has been done – as could in principle have been done – to get the EU to agree to amend it so as to limit its scope or even to remove it as no longer needed given Article 3 of the TCA.
The priority now must be to implement Article 3 by a detailed statutory subsidy regime giving a proper role to an independent authority – in particular, giving it power to clear subsidies (both individual subsidies and classes of subsidy), and to issue policy guidance. Until that is done, the current position is that a government that promised us freedom from State aid rules and their alleged uncertainty has delivered a system where Article 10 will still require consideration of State aid rules in many cases and where granting authorities and recipients fear to proceed with major projects involving subsidies that would (at least arguably) not comply with State aid rules because there is too much uncertainty about what a court will do. It is, in its way, quite an achievement.
the European Advocate 