In the months after the US election last November, I often found myself arguing against people in America who thought that Trump and Brexit were the same phenomenon. On my view, Brexit was in fact an innoculation against Trump and the politics of the radical right. The reason I thought this was a conviction that leaving the EU would kill the kind of right-wing politics in England that Ukip represented, since it was largely driven by a sense of powerlessness. The feeling — and it need be no more than that — that the political process could, after all, be responsive to what people wanted, even on fundamental matters, would immediately remove the emotional force from the radical right’s message, and that duly seems to have happened. Compare Ukip’s performance in the election with Trump’s, or with Marine Le Pen’s, or the radical right’s performance in almost any Western country today. As in the 1930s, Britain may have dodged the bullet of a kind of fascism, and largely because its political structures once again permit rather than constrain radical politics. This is a lesson that needs to be learned more widely: the more one attempts to use constitutional or cultural power (these being largely the same thing) to suppress dangerous and distasteful political movements, the stronger they grow, for the members of the movements now possess a justifiable case against their rulers.

Related to this is what I think is a widespread misunderstanding about the role immigration from the EU played. The general right of EU citizens to come to Britain is a very clear example of powerlessness on the part of the British authorities. There are many other examples, as we shall see, which to a liberal on immigration like myself are more important, but it is perhaps the most visible and concrete case, and we have carefully and responsibly to distinguish between a general hostility to immigrants and the desire to have an immigration policy. It is, I believe, quite strictly parallel to the hostility to illegal immigration in the US; there, it is (at least to my mind) anxiety about its illegality, and the fact that certain kinds of liberals do not seem to care about this, which angers many people, since it, too, is an example of the ordinary political processes ceasing to have any effect. I do not see in the US any serious desire, for example, to repeal the 1965 Immigration Act — which removed racial quotas from the US immigration system — which one might expect if there was a revival there of pre-1965 racism in immigration; one should not forget that the same voters who voted for Trump had, in many cases, four years earlier voted for a black President. In this respect, I think that both Jeremy Corbyn and Boris Johnson, who want a generous immigration policy after Brexit, may have a better instinct about British public opinion than the people who simply accuse the British voters of racism — though one can see why that accusation is a very useful one to make for opponents of Leave.

However, the weeks after the general election have seen signs of a retreat from the clarity offered by the Brexit vote. Remainers, particularly of a familiar world-weary sort, say that this is because the clarity was an illusion, and the full implications of Brexit are only now dawning; but to think this is essentially to make the same mistake which British politicians, and to some degree the whole of British society, have always made about the character of the EU. It is to confuse what one might call policy with constitutional principle. The vote in the referendum was a vote on a constitutional issue, and questions of policy have now to be decided within this new framework — though the framework allows a very wide range of options. Indeed, the striking and unusual fact about the vote is that it was a vote to put in place a less restrictive constitutional framework than has been the case since 1973.

The British have always shied away from considerations of constitutional structures, apart from a familiar type of crank with over-detailed schemes for electoral reform, etc; but the EU has to be thought about in these terms. To adapt Trotsky’s thought on the dialectic, you may not be interested in constitutions, but constitutions are interested in you. This is something that it is easier for Britons to see if they think about the USA. They are used to understanding American politics against the background of the constitution, partly because Americans famously keep emphasising that themselves, with oaths of loyalty to it and so on, and they are familiar with the idea that a critical issue in a Presidential election is the ability to determine political outcomes for a generation via the appointment of justices to the Supreme Court. But close to seven hundred years of a very different kind of political system (I say seven hundred years, since the essential principles of parliamentary legislation and taxation were largely the creation of Edward III in order to fight the Hundred Years’ War with public support) have left the British with very different instincts about their own politics.

So deeply imbued have we been with the idea that parliament, and therefore general elections, can, in principle, change any features of our common life, that the argument about the EU has almost entirely been an argument about what kind of policies we want to pursue at the moment. The most striking feature of the referendum debate itself was that it was to a great extent conducted as if it was a normal British general election, in which matters of policy were to be decided for the next five years or so. The argument about levels of immigration, which came to dominate the debate, at least in some quarters, exemplified this: it was largely concerned with the desirability or otherwise of specific numbers or types of immigrant, as if what was at stake was the British government’s immigration policy over the next few years. I was even told explicitly by a number of anti-Brexit friends that what mattered was preventing a Tory victory in the referendum, and that the issues in debate could be sorted out later. This approach was reassuring, in a way, since it showed that at an instinctive level, the British still thought of politics as something that was open to change at the ballot box; but in this particular setting, the old instincts proved to be an impediment to clear thinking about the issues.

For some years before the referendum I had been trying to get clear in my own mind how to theorise the US constitution, and I came to think that Britain’s relationship with the EU made sense, rather surprisingly, in the same terms. I tried to explain the approach in lectures delivered at Cambridge in 2012, which subsequently appeared as my The Sleeping Sovereign in 2016; I didn’t discuss the EU directly in the lectures, but it was already at the back of my mind, and my Dissent piece drew on the thoughts I had had four years earlier. Briefly, what I argued was that we should take seriously the distinction that some major seventeenth- and eighteenth-century political theorists drew between sovereignty and government. For hundreds of years it had been assumed that democracy of the ancient kind was impossible in a modern state, since the population could not meet to deliberate in a nation the size of France or England. All that might be possible was a system of representation (hailed as the great modern — i.e. medieval — invention by eighteenth-century historians), but that was not democracy in the ancient or the natural sense of the term, in which the people legislate: Aristotle, for example, had described election as an aristocratic principle, since it picked out a limited set of legislators.

What eighteenth-century theorists realised — above all Rousseau, but many of the American founders as well — was that popular legislation on fundamental matters was not impeded by the size or character of a modern state. Government, to use their term, had to be conducted by small groups or even a single person, able to deliberate and devote all their time to the issues; but sovereignty could be expressed in the occasional creation or amendment of fundamental laws that would form a constitution. The referendum naturally followed as a means of occasional popular legislation on constitutional matters, the very first in the world being in Massachusetts in 1778 when the new constitution of the independent state was put to the vote of all the citizens. Other American states followed suit, and even the Federal Constitution, though not put to a referendum, was designed to be ratified in a series of popular assemblies. Revolutionary France then embarked on the most extensive experiment with constitutional referendums, and though they fell into abeyance after the Revolution, interest in constitutional referendums revived in the late nineteenth century and again after the Second World War, until they became the norm in almost all European countries and in all but one of the states of the USA.

Once the distinction between acts of sovereignty and acts of government was in place, it would be possible to assign the role of constitutional legislator to non-democratic institutions as well — indeed, the two earliest theorists of the distinction, Hobbes and Bodin, assigned it to monarchs; but the distinction was always more relevant to democracy than to any other system of fundamental legislation, for the obvious reason that a monarch or an aristocratic board was not impeded in exercising acts of government as well as acts of sovereignty, and both Bodin and Hobbes seem, surprisingly, to have understood this.

Historically, Britain had kept out of this story, retaining its medieval representative institution and treating what elsewhere would be constitutional laws, passed in a special way, as merely ordinary statutes. (Incidentally, it is often said that Britain has an unwritten constitution and the US a written one. This is not really true: there are many constitutional conventions in the US, as there would have to be, and there are written laws which are patently constitutional in the UK, such as the Act of Settlement, the Act of Union, and the European Communities Act. The difference is not whether the rules are written or unwritten, but who does the writing, and whether they are a different and more democratic body than the one that writes the ordinary legislation of the country). But accession to the EU changed this.

The right way to theorise the EU, I argue, is as in effect a coordinated set of constitutional structures for each of the member countries. The EU is not a ‘superstate’, nor can it easily become one, juridically: it has always been maintained by the highest legal authorities in each country that, at least at the moment, the countries are sovereign entities, able in the last resort to decide their own futures. This is not empty rhetoric since, among other things, it is the justification for the continued representation of each EU country separately at the UN — something they are extremely unlikely ever to renounce. For this reason, much of the use of the term ‘sovereignty’ in the referendum debate was indeed as unhelpful as its critics complained. Moreover, the fact that the EU is not a state is the source of many of the problems it itself faces, as well as the problems conventional states face dealing with it; the thousands of deaths in the Mediterranean are testimony to the dangers of its current anomalous character and the fact that it is stuck in a half-way house, neither able to be a state with its own borders, nor an alliance of states that control their own. It is also why negotiating with it is not like negotiating with a normal state, but more in some ways (though one should not push this analogy too far) like negotiating with a Supreme Court — the picture Varoufakis paints in his gripping memoirs of his dealings with the EU institutions illustrates their strange character, and the mistake we make if we treat the EU either like a unitary state or an ordinary international grouping.

The key feature of the EU is that the sovereign authority in each state has enacted a certain rather curious kind of constitutional order for each of them, in which a set of principles and institutions are entrenched in a position beyond the reach of conventional, ‘governmental’, legislation. These principles and institutions are supranational in character, of course, and that is why the states took this course of action, but seem, from within each state, that the supranational character is not, in a way, their key feature: the key feature is rather that they are entrenched within the legal system of each country (this is what makes them different from the other supranational arrangements with which they are often compared, such as NATO or — even — the UN, at least in great part). The curious feature of these constitutional orders, however, is that they cannot be amended by the same process by which they were imposed: a UK Act of Parliament by itself straightforwardly entrenched the EU institutions in UK law, but no UK Act of Parliament by itself can amend them. Only a process of intergovernmental negotiation, issuing in changes that no one country can impose upon itself, can alter the essential character of the EU’s constitutional structure. The only thing an individual state can do is to repudiate the whole structure — as we are finding out.

Most states on the continent already had constitutional structures of some sort before the EU was formed, and their politicians were used to operating inside them, just as American politicians are. But the idea that a constitution could not be amended was new to them also — though, and this may be significant, not to German politicians. The German constitution is a legal oddity: the West German constitution, the Grundgesetz, was technically authorised by three of the four powers in the Allied military government, and included the provision that in the event of reunification a new constitution would have to be ratified by the German people. After the dissolution of the military government in 1991, the provinces of East Germany simply acceded to the Western state and its Grundgesetz, so the German constitution has never actually been ratified properly; moreover, a tradition has developed within German constitutional jurisprudence of supposing that certain fundamental moral principles are enshrined in constitutional law without the need for positive enactment. It is easy to see how a domestic structure of this kind renders the structures of the EU far less problematic for Germany than they are for the UK — or indeed for France, with its long history of popular constitutional legislation.

Britain, by virtue of its desire to join what was then the Common Market, thus found itself forced unwittingly into the default shape of a modern state, with a constitution that lay beyond the power of the government to change. And, as an almost instinctive recognition of this, the Wilson administration, as we all know, decided to use for the first time the default institution of constitutional legislation in a modern state, the referendum, in order to legitimate it. Though constitutional referendums had occasionally been proposed in the UK, notably to deal with Irish — and indeed Scottish — home rule, this was the first time that such a thing had seemed clearly necessary in Britain — about two hundred years after it seemed equally clearly necessary to the English settlers in Massachusetts. Since that time, as we also all know, the constitutional referendum has become a familiar feature of British political life. Characteristically, this has happened without a formal or legal acknowledgment of their fundamental role, and technically they are merely consultative; but the idea that they could be disregarded seems to most people about as fanciful as the idea that the Queen could actually use the power, still technically in her hands, to veto a parliamentary statute. Even in the aftermath of the Brexit vote, few people have advocated simply ignoring the result; the popular anti-Brexit response has been instead to call for a second vote, and that seems to me to be testimony to the obviousness of the change that has come over British politics. The EU and the referendum as an institution in the UK are wrapped in one another’s arms.

I might add at this point that the dangers and disadvantages of these kinds of structures tend to be far less obvious to people who are politically engaged or have some kind of public role. I mix in America with people who are regularly dealing with the Supreme Court, are leading figures in the political parties, or are writing for the press and trying to influence the political agenda. For them, it is easier to think that they will have some effect on politics through these processes than through the old-fashioned process of elections, and it is natural for them to think that their personal experience is something like an objective fact. One of the critical comments on my Brexit piece concluded with something like “perhaps we have had too much democracy”; it struck me reading it that “we” would not have to worry about less democracy if “we” were people like you and me, but for most people the vote is the one way they possess of altering their political circumstances. As a result, I think the general population has always been able to think more clearly about the EU than the political elites, since they have much more to lose.

To repeat: you may not be interested in constitutions, but constitutions are interested in you. They are not neutral, benign forces, however much the lawyers charged with maintaining them pretend that this is so; again, you only have to think about the history of American constitutional jurisprudence (much more familiar to us than continental constitutional jurisprudence, for obvious reasons) to see this. Think about the way the Commerce Clause has been used to extend federal power; think about the Dred Scott judgment and its endorsement of slavery; think about the Korematsu case on the internment of Japanese Americans; think about Citizens’ United. Put against them, of course, Brown v the Board of Education or Roe v Wade; but we will be choosing according to our political preferences. Certain kinds of political programmes are simply impossible in certain kinds of constitutional orders.

My favourite example of this, and something of great relevance to the general theme of this chapter, is the creation of the National Health Service in Britain. It required a very unusual constitutional order, since its most distinctive feature — and the thing which still sharply differentiates it from the single payer systems found in most developed countries (and even, in many respects, in the USA) — was the fact that it involved a mass expropriation of private property, in the form of the so-called ‘voluntary’ hospitals, some of which, like Barts, had been independent institutions for over eight hundred years. This was the issue that was most fiercely debated within the Attlee cabinet, and the result of Nye Bevan’s victory was one of the most far-reaching examples of nationalisation from those years, and the only one that has survived more or less intact. It is often asked by opponents of the NHS, “If it’s so good, why don’t other countries copy it?”. But in this respect it would be extremely difficult for other countries to copy it, since in most modern states expropriation of private property without compensation would be legally impossible without a far-reaching constitutional amendment, which might be very hard to pass. In Britain in 1946, all that was needed was a single sentence in an Act of Parliament: “there shall, on the appointed day, be transferred to and vest in the Minister by virtue of this Act all interests in or attaching to premises forming part of a voluntary hospital or used for the purposes of a voluntary hospital […]”. What this illustrates is that the achievements of the Attlee government, still the touchstone for left-wing measures in this country, required the kind of ancient omnicompetence that parliament still possessed in the 1940s.

The fact that every schoolchild is taught (or are they still?) that the British labour movement was intensely parliamentary and non-revolutionary, was not the consequence of some magic trait of the British which enabled them to avoid the turmoils of the revolutionary years on the continent, and indeed in America (for what else was the Civil War but a vast revolutionary moment?). Marx and Engels observed from their vantage point in Victorian England that the bourgeoisie had taken different routes in France and England to hamper the industrial working class from gaining power. In France, they had conceded universal male suffrage — first in 1792, and again and permanently in 1848 — but they had bound the legislature with a constitution which secured things such as private property (broadly defined) against legislative encroachment. In England, there were no such constraints on the legislature, and therefore the proletariat had to be denied the vote — which in this country, we should remember, was finally extended to the entire adult male population only in 1918, the same year that the first women received it (or rather recovered it — women lost the vote in 1832). So Marx and Engels concluded that the only thing necessary to bring about socialism in England was the extension of the parliamentary franchise, whereas in France it required revolutionary and extra-constitutional action. Exactly the same logic activated the early leaders of the Labour Party in Britain: they had every confidence that the parliamentary route to power was the right one, since they would then have available to them any measures to promote socialism which they thought fit, and which they could persuade a simple majority of their country (by definition, largely the working class) to support.

In the case of the EU, the overall character of the constitutional order pushes consistently in what we might call a neo-liberal direction. This is the point that Wolfgang Streeck has repeatedly insisted on, and has documented in convincing detail; he thinks that it is largely because of the influence on the institutions of German capital, and that is clearly true to an extent, but I would also argue that the institutions to some degree have a life of their own. Put in place a constitutional order that specifies certain economic freedoms — for the EU, the now notorious four freedoms, the free movement of goods, capital, services, and labour, to which we should also add the lesser-known but very important freedom of establishment; let a group of modern jurists loose on them; and the result will almost inevitably be a series of rules that are tilted towards the market. Constitutional orders are a combination of rules and the people interpreting them (as Hobbes, in particular, understood very well), and the people inevitably develop a certain kind of internal culture, which is usually proudly immune to outside political pressures. The American founders realised this, and were very interested in ways in which the judicial process could be made responsive to the citizens, including, in some states the election of judges, and the elaborate process of nomination and confirmation for federal judges. The worst of all worlds is to have a strong constitutional order and an independent judiciary — something I sometimes fear Britain is drifting towards even outside the EU.

But the ancient institution of parliament is pretty tough, and left to its own devices, it is unlikely once again to permit the kind of constraints on what it can do which the EU represented. The safety valve that democratic institutions constitute will be back in place, and we have no reason to suppose that — for example — even a liberal immigration policy will be met by the kind of political protest that Ukip represented, once it is clear that the policy has been decided through a transparent set of procedures, and, crucially, that it can be altered. Despite what many people particularly in the universities and the media believe, the best course for people in the UK who want liberal policies in a variety of areas is to support and not to oppose Brexit, for reasons that would have been perfectly obvious to any British politician on the left until the late twentieth century: the only secure basis for any liberal or left-wing programme is that it has been enacted through a democratic process, and there is no point in hoping that magically it can be enforced on an unwilling population through constitutional subterfuge.

Richard Tuck is the Frank G. Thomson Professor of Government at Harvard University.