1. A Divisive Decision
Referendums are supposed to provide decisive interventions in the affairs of state. They are designed to produce clear ‘yes or no’ answers to large political questions. And as these answers also come with a rare level of popular endorsement, this should facilitate their effective and timely implementation.
That, at least, is the theory. And the tone and rhetoric of the UK Brexit referendum campaign, which reaches its climax on June 23rd, would seem to back this up. Perhaps the only thing that both sides – Remain and Leave – are agreed on, and which accounts for the ever more febrile atmosphere of claim and counter-claim, is that the choice before us is a stark one. In terms of constitutional futures a binary vision of ‘in’ or ‘out’ increasingly dominates the UK debate, a crude black and white landscape with all shades of grey banished.
Yet the referendum theory is in fact deeply flawed. In the first place, regardless of the result, we can look forward to protracted and uncertain post-referendum negotiations before any definitive way ahead emerges. This is so whether the decision is taken to remain, in which case the bare terms of the special settlement reached between UK Government and the European Council in February will have to be fleshed out and made fit for implementation; or, alternatively, the decision is taken to leave under Art. 50 of the Treaty on European Union, in which case of one of several exit routes – the Norwegian-style EEA model, the Swiss-style EFTA-plus bespoke bilateral Treaty package, the Turkish-style Customs Union, or the lowest common denominator WTO fall-back option – will be pursued. In either case, the road will be long, arduous and unsympathetic, its precise destination unclear.
However, the exhaustive (and exhausting) detail of post-referendum negotiation is not my main focus. Rather, I want to examine the prior and more basic flaw in the binary vision. To put matters bluntly, at least according to most shades of political opinion in the UK and beyond, European ‘membership’ for a country of Britain’s size, influence and location should be less a matter of ‘yes or no’ than one of ‘more or less’. The reduction of a complex and graduated choice to a basic dichotomy simply does not reflect the position or serve the interests of the vast majority who will be affected by the outcome. What is more, and worse, the crude logic of either-or plebiscitary politics threatens to become a self-fulfilling prophecy. Having donned the referendum straitjacket, the British people – or rather their political masters – have been increasingly inclined to frame their arguments in mutually reinforcing all-or-nothing terms.
For their part, the Remain camp have over the course of the campaign placed decreasing emphasis on their initially much vaunted special deal from the European Council. Faced with the anti-immigration onslaught of the Leave camp, they have from time to time made defensive reference to their negotiation of an emergency brake in circumstances where public services or the employment market come under pressure from migratory inflow. But for the most part their focus has been on the (somewhat grudgingly acknowledged) general benefits of EU membership, with little mention of other supposedly hard-won concessions, such as the British exemption from ‘ever closer Union’, the new economic governance safeguards for non-Euro counties, or the enhancement of the powers of national Parliaments to oppose European legislation on grounds of subsidiarity. On the Leave side, the reduction of the terms of debate to bare essentials has been even more pronounced. An initial attraction towards half-way models such as the Norwegian or Swiss has given way to a much more emphatic distancing. Sustained by a belief in the continuing economic leverage of the UK as an independent trading power, but also recognising that continued full access to the Single Market would likely come with many of the same old regulatory and financial strings attached, and anticipating that any such formal association from a position outside the EU would in any case have to be negotiated in an inhospitable political environment, the Brexiteers seem to have opted ultimately for an approach which makes a virtue of detachment from the legal and institutional world of the EU.
Yet this hardening of positions and closing off of options pays little heed to the complex architecture of European integration, or to Britain’s own special position within that architecture. Once this is taken into account, I argue, the question posed by the referendum is exposed as the wrong one, or at least one that has been posed in the wrong way – with potentially serious consequences both for Britain’s place in Europe and for the European project more generally
2. The Reality of Differentiated integration
The idea conveyed by the unlovely term ‘differentiated integration’ that some states should be subject to different rules than others within the framework of the EU is hardly novel. To be sure, in the early days of European integration there was much resistance to the concept of differentiated integration as it was seen to detract from the founding ideal of a uniform set of commitments to the pooling of national sovereignty. Since the 1990s, however, in response to the development of a deeper and wider range of EU competences beyond the Single Market core, as well as to successive waves of enlargement from the original 6 to the present 28 members, differentiation has emerged in the majority of areas of EU policy. The most prominent cases are Economic and Monetary Union, with membership of the Eurozone gradually increasing from 11 to 19, and the Schengen Agreement on open borders, signed by 5 out of the then 10 members 30 years ago and today comprising 26 members, only 22 of whom are EU members. Yet this is only the tip of a much larger iceberg, with as many as 50 different cases of differentiation provided for under the EU treaties.
What is more, there are a number of different general models of differentiated integration vying for ascendancy within today’s complex and ever-shifting supranational institutional machinery. The most conservative such model, albeit increasingly outstripped by practice, is known as ‘multi-speed Europe’, where a core group of countries moves forward at a faster pace than the others, but with the expectation that those others will catch up in due course. A Europe of ‘variable geometry’ or ‘concentric circles’ offers a more radical – but in the view of many increasingly more realistic – vision in which the core or inner circle opts for deeper levels of integration than the outer circle(s) of membership on a permanent or indefinite basis. A third model, commonly referred to as Europe ‘à la carte’, is most flexible in its accommodation of difference. It would allow states to opt for more or less integration depending on the issue area, provided they adhere to a general set of common EU objectives.
To a greater or lesser degree across these various models, differentiation promises a degree of flexibility that offers both micro-political and macro-political advantages. At the micro-level, it tracks particular national preferences for more or less integration in various policy areas. At the macro-political level, it allows from some kind of accommodation between ‘supranationalists’, who seek deeper economic and political integration and do not want that aim to be diluted by successive waves of Enlargement, and ‘intergovernmentalists’, who want to guard against a political imperative which requires all to integrate at the rate of the most Europhile states.
The UK has always benefited from differentiation from the intergovernmentalist side of this divide. It is a member neither of Schengen nor of the Eurozone, possesses other significant opt-outs in immigration and criminal justice co-operation under the Area of Freedom, Security and Justice, and has from time to time enjoyed other exemptions – for example, from the Social Chapter between 1992-7, and, today, from the full effects of the Charter of Fundamental Rights of the European Union. What is more, the decision of the EU Heads of States or Government, made as recently as 2014, to endorse the idea of ‘different paths of integration for different countries’ as being perfectly compatible with the maintenance of a common commitment to an ‘ever closer Union’ was intended, inter alia, to assure the UK that it did not need to take a unilateralist stance to accommodate its particular relationship with the EU. 
As we now know, that particular gambit failed. Even some of the Remain supporters in the referendum do not consider the existing multilateral framework of differentiated integration sufficient to meet Britain’s needs and concerns; hence David Cameron’s insistence in negotiating the February agreement on a future exemption from ‘ever closer Union’ as part of a new customized membership model. Yet the picture remains a complex one, and the distinctions between the many positions on the table are not clear-cut. It is certainly the case that at an earlier stage of negotiations before the in/out referendum was proposed, the Conservative Government did seek to advance its arguments for Treaty reform in multilateral terms, favouring a two-tier, ‘variable geometry’ conception of the EU with a Eurozone core and the UK included in a less integrationist outer circle. And, importantly, even now there remains a significant degree of continuity between, on the one side, the various multilateral models of differentiated integration that are available and, on the other side, not only the unilateral conditionality pursued by the UK government in support of continued membership but also some of the options available and under consideration on the Brexit side of the argument.
Cameron’s new deal for remaining, although unilateral in initiative, is actually multilateral in content with the exception of the exemption from ‘ever closer Union’. And, even from the Leave perspective, some of the options for continued external association that were initially considered, though subsequently discarded, assume the retention of quite a high level of common regulation with the remaining members, as is already the case with a number of other non-member neighbours of the EU. Tellingly, indeed, in the wider literature on differentiated integration, some commentators talk about the difference between ‘internal differentiation’, which embraces members only, and ‘external differentiation’, which extends to non-members, as being one of degree rather than kind (the key Schengen agreement which excludes some EU members and includes some non-members being a prominent case in point). It follows that some of the general models of differentiation considered above may, within limits, be extended to cover the situation of non-members as well as members.
In sum, where the legal architecture is in any case complex and non-uniform, the distinction between the different combinations of common and distinctive regulation available on either side of the membership divide, and so even between full members and non-members, becomes less precise, and certainly more fluid and more subject to fluctuation, than is often appreciated, or least conceded, from the binary yes/no perspectives of the referendum debate.
3. British Choice, European Consequences
The point of my remarks is not to deny the legitimacy of the British referendum on Europe. After all, it was a choice made by a democratically elected Government to resolve a major constitutional decision in a democratic manner. And given the fault-line of Euro-scepticism running through British politics for as long as it has been an EU member, there is a case – albeit perhaps built more on hope than rational expectation – that Britain needed its second referendum moment – 40 years on – to put its ambivalence about the European project behind it once and for all.
Yet we should also resist the notion that such a referendum was the only legitimate way of testing UK scepticism. The normal structures of representative democracy suffice most of the time in most states as a way of resolving even the most significant constitutional issues, including continuing membership of a fast-changing EU. Nor can we ignore the connection between British Euroscepticism, a certain brand of nativist populism (which both the UK Independence Party and certain sections of the Conservative Part purport to represent), and the insistent mobilisation of nationalist political forces around the idea of a plebiscite. And, in turn, we cannot deny the unfortunate polarising effects of the choice of a one-question referendum as the relevant debating frame, and in particular, its tone- deafness to the flexibility and nuance of the long and rich history of differentiated integration, from which Britain itself has been a major beneficiary.
If, on the one hand, Britain votes to leave, then it seems that the supporters of exit have boxed themselves into a corner that commits them to a clean break. In any case, a number of prominent European politicians, including the British Prime Minister himself and the German Finance Minister, Wolfgang Schäuble, have warned that continuing membership of the Single Market is incompatible with a decision to leave the EU. And for those inclined to dismiss this as a mere tactical threat, apt to be withdrawn in the to and fro of post-Brexit negotiations, two factors should be borne in mind. First, continued UK membership of the Single Market, whether through the EEA or otherwise, would require the unanimous consent of all remaining member states – an extremely tall order in a post-Brexit scenario. Secondly, it is arguable that while this kind of membership-lite – this species of ‘external differentiation’ – may be appropriate for aspiring members and other associating neighbours, it is not so for ex-members who have volunteered to leave. Differentiated integration, defenders of the Union will contend, speaks to a flexibility inherent in membership, and at one remove, on the road to membership and in various other forms of more loosely convergent relationship. And it is precisely because this flexibility is already liberally extended to members, and to the UK in particular, that a similar latitude need not and should be allowed where a choice has been made no longer to exercise that insider flexibility but instead seek to renegotiate a more advantageous relationship from outside the tent. In other words, the crude exit option invited by the referendum choice threatens to upsets the graduated logic of differentiated integration in a manner that, from a pro-Union perspective, may be seen as gratuitously excluding many otherwise acceptable points on the spectrum of association.
If, on the other hand, Britain votes to remain, we have seen that the deal negotiated between the British government and the European Council offers an unprecedented variant of differentiated integration. Unilateral in initiative, bilateral in negotiation and conclusion, yet, with the important exception of the exemption from ‘ever closer Union’, multilateral in content and outcome, the distinctiveness of this new form of customized membership lies less in what it does and more in what it may portend. On the one hand, the UK is still signed up to the vast majority of the vast EU legal acquis it was already committed to prior to the new deal; and – to repeat – most of the new commitments entered into in the deal apply generally rather than to the UK alone. On the other hand, the exemption from ‘ever closer Union’ appears to anticipate a different attitude to future forms of integration, posing as a standing reservation in areas where the UK does not share the ambitions of other Member States.
How sustainable is such a settlement? One can envisage different lines of development. One possible trajectory – again a consequence of the rigidity of the referendum’s stark choice – involves a hardening of this framework of semi-detached, customized membership for the UK. While still officially ‘in’ rather than ‘out’, a pattern might emerge whereby the standing reservation of the UK was regularly invoked, and where it became less involved with the European project across a range of policy areas. If this were the case, the result might begin to resemble the Swiss, a la carte version of the exit model discussed above. Formally speaking, the UK would be both more constrained by its existing obligations and in a better position to exercise a bargaining voice over new initiatives if, unlike Switzerland, it remained ‘in’ rather than ‘out’. In practice, however, a repeated pattern of opt-outs might lead to mutual distancing – a drift towards a position where expectations of the UK’s productive participation in the on-going project of integration were scaled downwards on either side.
Alternatively, the UK’s continuing membership, and the multilateral nature of most of the concrete new measures contained in the deal, might allow for its gradual re-integration into a fuller membership role. A ‘Yes’ vote might mark the end, or at least the temporary disappearance, of a viable exit option in British politics. This might in turn signal the re-emergence of a more pro-European approach, and the consignment of the ‘ever closer Union’ exception to the status of purely symbolic legislation. The stratifying effect of the referendum on British public opinion, however, makes this an unlikely way forward.
A third option would see the UK’s exceptionalist initiative neither as confirmatory of its outlier role, nor as a short-lived gestural politics, but as a prompt for further multilateral treaty change. As the general nature of the pro tem solutions adopted under the British deal indicate, the issues raised in the Brexit context – the limits of sovereignty pooling, the reassertion of the powers of national Parliaments, the future of EMU, the response to mass migration and the scope of Social Europe – are hardly questions of interest to the UK alone. Broader concern with these matters, with the financial stability of the Eurozone and the willingness and capacity of the EU to absorb migratory flows from the South and East most likely to provide the catalyst, may lead to a more rounded resettlement initiative – perhaps involving the re-assertion of the Eurozone-centred variable geometry version of differentiated integration first mooted in reform discussions between the UK and the EU.
What this third option makes vivid is that the future of European Union today is in many ways as uncertain as is the future of the UK within the European Union. It also suggests that most of the plausible architectural alternatives for our continent involve a significant degree of differentiated integration – a trend that is only reinforced by the continuing pressure from neighbourhood states to join the club and the consequent need to continue building intermediate structures of attachment. The choice for Britain ought to have been shaped as much by the availability and negotiability of these alternative models of differentiation as by the bald options of ‘yes or no’, but, as we have seen, these categorical solutions have threatened to consign the debate over differentiated integration to the margins. Yet the choice for Europe is bound to be significantly influenced by the precedent set in the British case. On the one hand, as the first state to formally contemplate exit, Britain is a test case for what might be negotiable on the other side of exit, but also, and more urgently and more threateningly, for the EU’s capacity to withstand the shock of a first exit without significant domino effects on other members. On the other hand, if Britain votes to remain, it will be a trend-setter in the evolving possibilities of differentiated membership.
Yet with so much at stake for Europe in general as well as the UK in particular, a pattern of mutual disregard has supplied one of the more unfortunate symmetries of the Brexit debate. The campaigners on either side of the UK debate have been strikingly insular, hardly considering the wider EU interests at stake except insofar as they bear upon specifically British interests. Likewise, for most of the campaign, the rest of the EU has met indifference with indifference and has paid little attention to the introverted British debate. Indeed, many of the strongest European voices have come from ultra-nationalist Eurosceptic forces lending their support to Leave. As decision day draws near, as the Brexiteers grow stronger and as the vote looks too close to call, however, Europe’s attention has at last become more engaged. A recent survey tells us that 70% of EU citizens sampled across nine member states believe that Britain leaving would be a bad thing, apt to diminish Europe’s place in the world and to spread popular disillusionment, while only 16% believe it would be a good thing. Reflecting this sentiment, senior politicians in the EU and national institutions have become more vocal in backing a British Remain. And through initiatives such as Der Spiegel’s instantly famous ‘Please Don’t Go’ headline of last week and the Irish4Europe campaign group targeting of the 600,000 Irish-born voters eligible to vote in the UK referendum, some attempts have also been made to mobilise opinion and communicate support from afar.
Europe, then, is finally waking up to the fact that the British referendum is no local squabble, but a dispute with the potential to reshape the European Union itself, more likely for worse, but – to end on a positive note – perhaps still for better. For ironically enough, the very preparedness of the UK to think outside the box of uniform supranational integration, even if framed by the wrong question, may prove a pivotal moment. Provided relations are not fatally poisoned by an irreversible ‘No’ vote, or by a lingering Euro-cynicism on the UK side and/or an equivalent UK-cynicism on the European side, Britain’s domestic conflict may yet prove to be influential and instructive in the constructive reshaping of European integration over the decades to come.
 European Council Conclusions, 19 February 2016, EUCO 1/16
 The general ‘enhanced co-operation’ provisions in Arts 226-34 of the Treaty on the Functioning of the European Union (TFEU) are the best example of this approach in the present Treaty framework. However, they are concerned exclusively with the option of additional integration beyond the EU norm rather than opting–out from the norm or co-operating below the level of the norm.
 European Council Conclusions, 26-27 June 2014, EUCO 79/14.
 See e.g. R. Liddle, ‘The Risk of Brexit: Britain and Europe in 2015 (London: Rowman &Littlefield, 2015) ch.1.
 See e.g., D. Leuffen, B. Rittberger and F. Schimmelfennig, Differentiated Integration: explaining variation in the European Union (London: Macmillan, 2013).
 At the moment, following the accession of Croatia as the EU’s 28th member in 2013, there are five officially recognised EU candidate countries – Albania, Macedonia, Montenegro, Serbia and Turkey, and two countries – Bosnia and Herzegovina – that have been promised candidate status in due course. This slowly moving conveyor belt is closely connected to the EU’s Stabilisation and Association Process (SAP) towards the Western Balkans.
 See e.g., J-W. Müller, ‘Europe’s Sullen Child’ (2016) 38)11) London Review of Books 3-5.