Schroedinger’s Backstop

A symmetric approach

The EU was founded on replacing the principle of direct reciprocity with compliance to common rules: instead of slapping tariffs or sanctions on each other member states would bind themselves to, say, common competition law rules. This is what happens within the EU. And yet, when it comes to third countries, the EU has had no compunction in requiring unilateral adoption of its standards by the rest of the world as a condition for access to its market, without others having a say in such standards. Fair enough, this is what you do when you are a regulatory hegemon. The problem with Brexit is that the UK is making this fundamental asymmetry visible and contestable. Remainers may be right that the best way to avoid becoming a  “vassal state” is to stay around the table. But short of a reversal, the EU must come to terms with the fact that the UK cannot acquiesce to business as usual. 

While this is a challenge for upcoming negotiations on the future relationship, Brexit 2.0, the problem is foreshadowed in the withdrawal protocol on the backstop.

This is in effect the issue that Joseph Weiler, Daniel Sarmiento and Jonathan Faull seek to address in their Verfassungsblog proposal. They suggest that the best way to avoid a no-deal Brexit, even at the 11^th hour, would be to adopt “a regime of dual autonomy” (the Ulster Unionists put forth similar proposals over the weekend). To summarise, this implies that each side is expected to enforce the other side’s standards for exported goods. This would mean that the EU and UK would each maintain their own customs and regulatory regimes but use all the mechanisms under their control to protect each others’ standards. Compliance would be ensured through mechanisms such as non-frontier spot checks and making non compliance a criminal effect. 

This approach amounts in effect to changing the backstop in two ways: i) the UK would not stay in the single market and customs union (avoiding asymmetry); ii) the integrity of the single market would still be guaranteed but only by bringing to bear the force of the UK’s rule of law’ on exports to the EU and the EU would do the same (establishing symmetry).

It may be the case that this is “inadequate and not anywhere near the landing zone” as a senior EU source told the BBC News. No ZOPA (zone of possible agreement) at this stage of the negotiations would mean that we can only head towards no deal Brexit or no Brexit at all. 

But we can also entertain the thought that reciprocity or symmetry is indeed a necessary if not a sufficient condition for the backstop compass to lead us to a landing zone. 

Six ingredients need to be added to the mix: a dynamic backstop, a compatibility assumption, a new reading EU of history, turning the precedent concern around, empowering local institutions, distinguishing positions and interests.

First: a dynamic backstop 

The parties could agree more clearly that the backstop is a dynamic mechanism.  For one, the EU can do more to reassure the UK governemnt that the backstop is only relevant to the extent that so called “alternative arrangements” relying on technological means to avoid a border in Ireland are yet incapable of avoiding a border in Ireland. 

Prime Minister Boris Johnson has abundantly  mentioned in his selling pitch around Europe the Prosperity report on alternative arrangements, drafted by a Commission chaired by MPs Greg Hand and Nicky Morgan. The report proposes a range of solutions – without transposing other border arrangements to NI. These include avoiding one-size-fits-all arrangements and futuristic high-tech solutions; use of Enhanced Economic Zones; trusted trader programmes; sanitary and phyto-sanitary checks carried out away from the border; building upon existing common all-island regimes; use of the Norway/Switzerland Transit system at the border. 

The EU is right that the alternative arrangements do not do the same job as the backstop. But everyone also agrees that a paradigm shift is in the making in border management, for better or worse. These are not fantasy unicorns, only unicorns to be born. The parties only disagree about timing and the extent to which these are sufficient to guarantee the integrity of the single market. The drafters of the AAC report recognise that their proposals cannot be deployed fast. They say within 2 or 3 years while others say it will take longer. Let an impartial referee adjudicate when the time comes even if the two parties retain ultimate control. New technology can be introduced and progressively trusted, while the regulatory approach to alternative arrangements is deployed.

Second: a deal based on core convergence and broad compatibility. 

Turning then to the regulatory side, since the backstop is a hypothetical proxy for a future relationship deal,  it is not unreasonable to design it in its long shadow as article 50 spells out, albeit ambiguously.

Clearly, this amended backstop requires mutual trust – but as Weiler rightly argues that is already the case if the UK were to stay in the single market as the backstop envisages.  But since it is difficult to assuage the further concern that it would be possible to diverge from each other’s standards under the proposal, the amended backstop needs to distinguish between core convergence and broad compatibility.

Few disagree that core convergence and common standards need to be retained where the island can only be one single ecosystem, especially by continuing to respect EU law on food standards and protect human health. There would be no need to check such products at the point of entry (eg the NI border) under the backstop as agreed. This would not change.

More broadly, however, it is not unreasonable to assume that a former member state like the UK will remain EU-compatible, not only on day one but thereafter, probably more so than many less developed member states – this is true for the existing version of the backstop or for whatever could replace it.  

In this spirit, it makes sense that both the backstop and the broader future deal recognise the extraordinary level of supervisory collaboration and trust that already exists between regulators on both sides. It is a mistake to believe that a few months of acrimonious negotiations can erase almost four decades of working together in the EU.

I refer in detail to this mindset and its concrete implications as the compatibility model acknowledging on both sides that access to the EU’s Single Market is neither about ‘managing convergence’ as with enlargement, nor about ‘managing divergence’ as the Brexiters dream and the EU fears, but about ‘managing differences’. Compatibility is not conformity. The compatibility paradigm suggests that it is wrong to decide a priori that potential future differences in regulatory approaches will necessarily overshoot the bounds of legitimate differences. The European legal imagination has for many decades populated the space between regulatory independence and subservience with numerous notions labelled approximation, inter-operability, regulatory coherence proportionality, balancing and functional equivalence, which apply differently to different sectors. Why should it be afraid of this legal philosophy?

Third: the parties need to engage with the EU history of managed mutual recognition 

Which leads us to the third ingredient for a successful deal. An EU official commented that the Verfassunsblog proposal was a repeat of Theresa May’s doomed Chequers plan, rejected by Brussels.

This concern must be engaged with. For one, by acknowledging that the philosophy of the amended backstop of course is not new. When prime minister Theresa May first spoke of mutual recognition in her Mansion Speech House in London on 2 March 2018 spelling out British aspirations for the future relationship, some of us (including my colleague Stephen Weatherhill on this Verfassungsblog blog, and myself here and in my book) sought to strike a balance between those who praised it uncritically and those who only heard unicorn-speech. 

I believe that the Verfassungsblog proposal falls into this middle category, putting forth a minimalist version of a managed mutual recognition regime with maximal enforcement guarantees, which respects the integrity of the single market.

True. When the EU rightly seeks to protect the “integrity of the single market” we are referred to the OD definition of integrity as “the state of being whole and undivided.”  But integrity is also about “adherence to consistent principles over time.”  In this spirit, the EU ought to recognise the margins of freedom offered by the history of the single market which provides subtle guidelines rather than fixed rules for interpreting the contours of such integrity. 

There are two lessons in particular. 

Lesson one is that the single market has been built over time in a piecemeal and pragmatic way around the principle of mutual recognition, underpinned by rigorous adjudication of disputes where they arise. Countries of destination are asked to trust the home state to provide the right stamp for goods and services crossing borders.

But because trust is never blind between states, there is no such thing in the EU as pure mutual recognition. Instead, the EU Single Market has become a complex and layered system of managed mutual recognition, a recognition which can be partial, conditional and reversible, and involve more or less alignment. Even when more harmonisation is introduced, we still at a minimum need mutual recognition of how the rules are enforced.

This is an ingenious dynamic process, involving trade-offs that may change over time. Ironically, it was devised to a great extent by Brits in Brussels precisely to avoid one-size-fits-all standards and supervision, allowing for a high degree of national regulatory autonomy. 

By invoking mutual recognition out of context, Brexiters failed to acknowledge that they have set out to reinvent the EU wheel but without the gears and spokes that make it work, otherwise known as institutions. They failed to recognise what it means to access an entire regime of mutual recognition rather than a single country that would simply recognise you. 

This is in part why Brussels has been so suspicious of the British offer in the past two years. Yet, the EU might consider the idea that to take away recognition ought to be harder than to grant it initially. It might thus decide imaginatively to explore a new dynamic version of regulatory managed mutual recognition for Britain. If it does, the minimalist version suggested by the Verfassungsblog proposal would merely be a hint in this direction (since it is not meant to be invoked anyway).

Lesson two is that the EU has been seeking to export the Single Market model to the rest of the world for decades. It is fair to say that this ambition has worked better as an asymmetrical exercise, for example in Norway, than a symmetrical exercise, as in the United States. And that the so-called mutual recognition deals the EU did manage to strike with outside partners are but pale imitations of the original, involving what the minimalist Weiler et al. proposal puts forth, e.g. the recognition that exporters certify to the importing country standards (6 such sectorial mutual recognition agreements were signed with the USA in the 1990s). If every external trade deal negotiated between the EU and third countries in the last two decades has included a chapter on regulatory cooperation, these have remained long-term horizons. 

Here is the thing. With Brexit, the EU can now experiment with a country that could live on both these horizons. As a former member state the UK will not simply be a new exemplar of EU Story Two. It will have been an integral part of EU Story One. If there is no precedent for such a state of affairs, we can invent a new paradigm consistent with the EU’s own history and principles. 

Negotiators may object that we are not currently negotiating the future relationship, so these considerations are irrelevant. Yet, the rebuttals to the amended backstop are in fact also commentaries on the future relationship. If the political declaration can be amended to reflect a bold and ambitious vision of that relationship, one whereby Brexit is a staircase between EU Stories One and Two, this opens a space for amending the backstop in the same spirit. 

The backstop is but a first stone in this edifice.

Fourth: ambition not concession, or turning the precedent story on its head

Can the EU contemplate a UK both on the outer ring of its own managed mutual recognition system – Story One – and as the new frontier of its external mutual recognition ambition – Story Two? 

Indeed the political declaration already makes clear that if it wants access to the single market, it will need to respect the spirit of internal EU law, which is not about accommodating ad hoc and idiosyncratic exceptions but about refining a principled approach over time. And because things change over time, including levels of trust, the deal needs to involve a good dose of contingent contracting, predicated on accommodating different expectations. 

Evolutionary clauses are the bread and butter of international law and can provide the basis for a dynamic Brexit as well as a dynamic backstop. 

Managed mutual recognition relies heavily on conflict management, remedies in case of disagreement and political decisions on whether to stick with incompatibilities that may arise or take the hit in terms of market access. Because the UK will be only partially involved in the EU’s eco-system of recognition management, these technologies of conflict will need to be robust and operate in the shadow of the Courts on each side. The EU will have the power to manage incompatible regulations when the problem arises. To be sure, it is wrong to think that mutual recognition from outside can look and feel the same as mutual recognition for members of the club (there will be more asymmetry, reciprocity and reversibility). But it is not clear why it should be denied altogether or reduced to bare-bone unilaterally granted and revoked equivalence.

Would others trade partners consider this as a precedent to follow suit? Not if they are not as EU-compatible as this uniquely compatible former member state. Global non-discrimination clauses do not apply if you are different. It is not true that anything offered to the UK must be offered to others. The unique status of the UK credibly justifies a special partnership.

The narrative here maters. After all, the EU has made it its great mission in life to save free trade, battling against its detractors, all the while overcoming the devastating perverse effects of globalisation denounced by Europhobes. If Brexit is to be the measure of the EU’s free and fair trade credibility, the most unprecedented of events could come to serve as a precedent which may one day inspire the EU’s relations with a broader circle of countries, starting with its neighbours, which over the next decades can progressively be brought within the ambit of the EU’s managed mutual recognition system. This could in turn inspire new post-colonial EU thinking on reciprocity when dealing with third countries. And the compatibility model could eventually inspire global partnerships for economic governance.

In this spirit, the compatibility paradigm can turn on its head the narrative that Brexit is a bad precedent. Brexit could come to stand in the court of history not as a disastrous internal precedent but as an ambitious external precedent, a reinvention of the EU’s managed mutual recognition approach to fit its role in the world as a reliable hub in a turbulent world. Does the EU want to convince the world of its commitment to free trade, cooperation and openness? If so, today the backstop, tomorrow its future relationship with the UK, will be the test.

Fifth: subsidiarity calls for embedding the proposed agreement in local institutions

But of course, who says autonomy calls for trust: I trust you to enforce my regulations.  Who should be thus entrusted? The relevant regulatory authority of course. But I believe that the importance of the local layer in this equation should be further stressed. 

After all, back in 1998, the good Friday agreement (GFA) set up institutions for shared democratic responsibility and mutual adjustment including on the implementation of the single market. Strand One, on the character of devolution to Northern Ireland institutions;  Strand Two, on the framework for cooperation and joint implementation between the North and South of Ireland and crucially mechanisms to channel local views to European institutions; and Strand Three, on East-West coordination, between the Republic and the UK. 

Arguably, these institutions all need new impetus, but if this can happen, they can be the first port of call for the mutual enforcement responsibilities laid out in the proposal. They can be  entrusted with re-building trust around border issues after Brexit, thus more explicitly linking the backstop or its alternative to the operational integrity of the Good Friday Agreement. If there is to be a legal addendum to supplement the withdrawal agreement it could be re-embedded firmly in the GFA, and above all Strand 2.  

In sum, the GFA is not the problem that the backstop sought to solve but can support the solution to the problems the backstop has raised. If the need to implement the single market informed the peace process, the GFA institutions can now inform Brexit’s Withdrawal agreement. The north-south institutions would be strengthened and charged with ensuring the continued compatibility between, on one hand, the no-border commitment, and, on the other hand, single market requirements; while charged with supervising the respective enforcement of the other side’s rules.

Granting the GFA institutions a key role in border management would be the best institutional anchor for the  reciprocal approach suggested by the proposal. It does not obfuscate the EU role in managing one of its external border, but at the same time, all three Strands also underline the role of the UK in the process.

Sixth: how to get there

Even were the parties to agree on some version of an amended backstop, how should they save face given their respective commitment that withdrawal agreement is either dead or un-negotiable? Clearly, they will need to turn it into Schroedinger’s cat, both dead and alive at the same time. The amended proposal goes a long way in achieving this miracle by suggesting that the approach be laid out in the political declaration, with a clause in the (therefore amended) backstop referring back to the wording and granting it legal value as the withdrawal agreement. As others have pointed out, the backstop is currently written as a protocol, separate to the withdrawal agreement, so technically, the EU keeps its withdrawal agreement while the UK has got rid of the backstop as it stands.