One concern that has been widely expressed during the passage of the Subsidy Control Bill through Parliament is its relationship with the current government’s “levelling-up” agenda. Although that agenda is somewhat inchoate (although it may become a bit clearer after the White Paper is published), and is likely to involve spending that does not raise any subsidy control issues (such as spending on public services and general infrastructure), subsidies to private business conditional on locating activity in particular areas and justified by the need to address relative disadvantage or deprivation in those areas is likely to play a key part. How does the Bill address subsidies of that kind (which I will refer to as “regional development subsidies”), and does it do so in a satisfactory way?
The ancien regime
Before Brexit, the EU State aid regime dealt with regional development subsidies in two main ways. First, and most importantly in practice, the General Block Exemption Regulation (“GBER”) created a “safe harbour” for regional development subsidies if they (a) provided for investment in particular areas set out in a “regional aid map” approved at EU level and (b) ticked various boxes in terms of the proportion of investment supported by the aid (“aid intensity”) and other criteria, the strictness of which varied as between different areas as set out in the regional aid map. That meant that local authorities could confidently grant regional development subsidies, without fear of objection or challenge under State aid rules, as long as (a) the area concerned was in the right part of the regional aid map and (b) the subsidies ticked the necessary boxes.
The other way in which the EU regime dealt with regional development subsidies was that, where the proposal fell outside GBER, it had to be notified to the Commission for clearance – and in deciding whether to approve it the Commission would take a view as to whether the regional development justification was a good one (viewing the matter from an EU-wide perspective) and whether the subsidy was proportionate, limited to what was necessary, and justified overall in the EU public interest. That analysis could be challenged by an interested party before the EU courts, but only on narrow, judicial review, grounds (manifest error of fact, irrationality, error of law), and few such challenges succeeded given the wide margin of discretion given to the Commission in this area.
The structure of subsidy control under the Bill will be very different. The starting point is that it will be for the granting authority itself, under clause 12, to assess any proposed regional development subsidy against the subsidy control principles in Schedule 1 and to be satisfied that the subsidy complies. Where the subsidy is one that would have been covered by the “regional aid” provisions of GBER, that requirement to assess imposes a new obligation on the granting authority: it is no longer enough to tick the GBER boxes. On the other hand, where the subsidy is one that would not have fallen under GBER, there is no longer any need to wait for the subsidy to be notified and cleared by the Commission: subject to the provisions in Part 4 requiring large grants (“subsidies of particular interest”, the scope of which is yet to be defined but may well be confined to subsidies of over £5 million) to be referred to the CMA for it to advise – advice that the granting authority is not bound to take – the granting authority can proceed to make the assessment itself and, if the proposal passes that assessment, to proceed. The only prohibitions of likely relevance to most regional development subsidies will be on rescue and restructuring aid without a credible path back to viability (clauses 19-20) and a narrow provision in clause 18 that prohibits subsidies that are conditional on relocation of activity from one part of the UK to another (a provision that does not, as the Minister confirmed in the public bill committee debates, page 128, apply to cases where the subsidy is not conditional as a matter of contractual obligation on cessation of activity elsewhere but merely has the effect of causing that cessation as an economic consequence of starting activity in the subsidised location). It remains to be seen whether the clause 18 prohibition on express relocation conditions is included in the final Bill.
What all that means is that (in contrast to the regional aid provisions in EU law) there is nothing in the Bill that expressly favours regional development subsidies to promote development in Hartlepool rather than in Hertford, making the former acceptable in cases where the same subsidy in the latter would be prohibited, or at least subject to more stringent controls. Rather, the arena in which the boundary between acceptable and unacceptable regional development subsidies will be drawn is the assessment by granting authorities of proposed regional development subsidies against the subsidy control principles.. That gives rise to two critical questions: what the subsidy control principles have to say about regional development, and how that boundary is to be defined and policed. That second question is particularly important in the context of regional development, because (a) territorial granting authorities (devolved governments, the UK government when it acts for England, and local authorities) will have every reason to find that the subsidy control principles are consistent with a subsidy that promotes investment in their area rather than elsewhere and (b) granting authorities elsewhere will have every reason to be concerned by subsidies that displace activity from their area or cause it not to be located there in the first place.
First, then, what do the subsidy control principles have to say about regional development? Principle A deals with the objective of a subsidy: they should have a policy objective of either remedying an identified market failure or of “addressing an equity rationale (such as social difficulties or distributional concerns)”. Though the former phrase might apply in some cases to regional development aid, it is the latter phrase that seems particularly apt. But that phrase is not free from difficulty: as the former Lord Chief Justice Lord Thomas of Cwmgiedd asked during the Bill’s second reading in the House of Lords, “what does that actually mean?” In particular, what does “distributional concerns” mean in the context of regional development? Would it cover a concern by a local authority for a relatively prosperous county (say, Surrey) about an area that was (on various measures) disadvantaged compared to the rest of Surrey but rather more advantaged than most areas of the United Kingdom? How free is a local authority able to be in choosing measures that tend to point to the area at issue being disadvantaged as opposed to other measures that do not?
Principle B requires subsidies to be proportionate and limited to what is necessary to achieve it. That principle is likely to be particularly important in the context of “subsidy races” (contests between different territorial authorities for particular investment): the effect of principle B should be to stop an authority winning such a race by offering more than what is needed to meet the policy objective of addressing the regional development rationale.
Principles F and G both require territorial authorities to consider the effects of the proposed subsidy across the United Kingdom: principle F requires that the subsidy minimise negative effects on competition or investment within the United Kingdom and principle G requires that the benefits of the subsidy exceed its negative effects on (among other matters) competition and investment in the United Kingdom. In the case – very likely in the case of regional development subsidies – where the counterfactual is investment in another part of the United Kingdom, those principles require a territorial granting authority to carry out the task of evaluating and giving weight to the harm caused by the proposed subsidy outside its territory: a task which is somewhat awkward given that any territorial authority in a democracy is properly accountable to, and serves the needs of, its own voters.
The subsidy control principles are therefore capable of permitting regional development subsidies, but their application is likely to be deeply contestable. As pointed out above, given the potential for real conflict between different territorial authorities, that makes the question of who defines and polices the boundaries of proper application of the principles rather critical.
Enforcement of the principles
The first candidate for policeman is the Secretary of State, as having the power, under clause 79, to issue guidance to which all granting authorities must pay regard when applying the principles. Draft guidance has already been issued. That guidance confirms that an equity rationale may include “Levelling up a deprived or disadvantaged area”. It does not, though, do much to address the question of how deprivation or disadvantage are to be established, noting only that the analysis “should include measures or statistical indicators set against appropriate comparators (such as regional or national averages)” (thus rather dodging the questions of whether the poorest corner of Surrey is to be regarded as a “deprived or disadvantaged area” and how free territorial authorities are to choose particular comparators that suit the case for regional development aid in their area). The draft guidance does not yet address the questions of how to apply principles F and G: but it does have a section that would urge granting authorities to be “cautious about subsidy races occurring as these may lead to a displacement of investment away from locations where the public benefits are the greatest and it may incentivise firms to use their leverage to secure larger subsidies than would have been possible had public authorities not been bidding against each other to secure the investment”, going on to refer to principle B (proportionality and necessity) and suggesting a “more extensive” analysis of the counterfactual, including a comparison of disadvantage between the area sought to be assisted and the area where the investment would otherwise have occurred.
Although the guidance has yet to be finalised, it does not look as if it will resolve the difficulties identified above: and, in any event, it is not binding (though granting authorities must have regard to it and the CMA and Competition Appeal Tribunal (“CAT”) probably will have regard to it).
The second candidate for policeman of the boundary is the CMA. The CMA has the advantage of being a UK-wide body, though the government has so far rejected attempts to give the devolved governments any right to appoint, or to have a formal say in appointing, the members of the subsidy advice unit who will be charged with this work (those appointments being for the Secretary of State, who is also, of course, a Minister with territorial responsibilities for England). It would appear to be open to the CMA to take a prescriptive approach to the way in which granting authorities should assess how the equity rationale applies to proposals for investment in particular areas. However, the CMA’s function is only advisory: a granting authority can, if it wishes, ignore any such advice. Moreover, the CMA is only required to be involved in the case of “subsidies of particular interest”: in the case of other subsidies, the CMA either has no role at all or (in the case of “subsidies of interest”) has a role only if called in by the granting authority or by the Secretary of State. Quite how much that matters will depend on how the Secretary of State defines the categories of “subsidies of [particular] interest”: a policy statement suggests that in the usual case the subsidy will need to be over £5-10 million to be a “subsidy of particular interest”.
The third candidate for policeman is the CAT. Judicial review in the CAT can be started by the Secretary of State or by any “interested person”: though that term does not obviously include a territorial authority concerned by the impact of another authority’s subsidy on its territory (it could indicate a narrower financial interest), the Minister confirmed in the Commons that it was intended to do so (see pages 308-309). However, it is unclear to what extent the CAT will give – or to what extent it would be appropriate for it to give, bearing in mind the economic and policy questions involved – any firm steer on the difficulties identified above: or to what extent it would be prepared to say (or would be appropriate for it to say) that failure to follow the CMA’s position on those issues (if it takes one) would of itself be irrational or an error of law.
The technical route for the CMA to take a prescriptive line would be to read the principles as having a hard-edged legal meaning, despite their open texture. However, such an outcome would create a regime that – in determining the key question of what subsidies are justified and which are not – was more prescriptive and court-oriented than the EU system (which leaves the Commission with a huge margin of discretion). That would seem a somewhat paradoxical consequence of Brexit.
On the other hand, for the CAT to take relatively hands-off approach would allow territorial authorities to interpret and apply the principles in a way that permits regional development subsidies even where there is a strong case that those subsidies are (from a UK perspective) not directed at areas of most need and create more harm than good in terms of levelling-up.
What does all that mean from a policy perspective?
One possible result is that the uncertainties, need to analyse measures within a particular framework, and risk of challenge will deter authorities from granting regional development subsidies in cases where they are clearly the right policy response. The obvious route to avoid that result will be for the government to created streamlined subsidy schemes under clause 10 that provide a safe harbour for large categories of regional development subsidies: and such schemes could sensibly define areas where less strict conditions applied for a subsidy to fall under the scheme, recreating some of the advantages of the State aid system: moreover, the risk of distorting the terms of subsidies so as to fit under the scheme (a real problem in the EU system) is less pronounced because it would always be open to a granting authority to take the, fairly secure, view that a subsidy that fell just outside the terms of a scheme would nonetheless be safe given that it was only just outside the safe harbour.
The other possible result of the problem that the boundaries between acceptable and unacceptable regional development subsidies are neither clear or firmly policed – particularly when combined with the absence of any provisions in the Bill to police cases where a granting authority wrongly decides that what it is doing is not a subsidy at all (for example by taking an implausible view of what a private investor would do and relying on clause 3(2)) – is a real danger that regional development subsidies that are likely to harm the UK economy as a whole and which have a weak distributional justification will be let through. That risk is partly addressed by the role of the CMA – but it remains to be seen how willing the CMA is to put down boundaries in this area or to what extent its boundaries are respected by authorities or enforced by the CAT.
Either way, the overall effect of the Bill on “levelling up” subsidies is both complex and murky.
GEORGE PERETZ QC
(With thanks to Alex Rose for comments on an earlier draft of this post: any errors are, of course, mine.)