On 23 February, Donald Trump stood before a rally of cheering supporters to celebrate a thumping victory in the Nevada Republican caucus – his third consecutive win, in defiance of the naysayers who had predicted that his bubble was about to burst. “If you listen to the pundits, we weren’t expected to win too much – and now we’re winning, winning, winning the country,” he bragged. “We won with young. We won with old. We won with highly educated. We won with poorly educated. I love the poorly educated.”

The possibility that education has become a fundamental divide in democracy – with the educated on one side and the less educated on another – is an alarming prospect. It points to a deep alienation that cuts both ways. The less educated fear they are being governed by intellectual snobs who know nothing of their lives and experiences. The educated fear their fate may be decided by know-nothings who are ignorant of how the world really works. Bringing the two sides together is going to be very hard. The current election season appears to be doing the opposite.

Trump continues to poll far ahead of Clinton among voters who did not go to college, while Clinton still leads by a considerable margin among college graduates. This is a significant change from 2012, when the picture was far more mixed. Four years ago, the college-educated vote was almost evenly split, with graduates favouring Obama over Romney by a narrow margin, 50 to 48. Recent polling puts Trump’s lead over Clinton among white men without a college degreeat a sobering 76 to 19.

Of course, there are other factors at play here. Race is one; gender is the other. The overwhelming majority of Trump’s supporters are white, regardless of their education levels. However, white men with a college degree split much more evenly between the candidates, whereas white women without a college degree still strongly favour Trump. Less educated voters who support Trump are not necessarily poor: many earn more than $50,000 (£39,000) a year. Trump scores particularly well among small business owners who did not go to college. These polling numbers – which are only indicative, since no one has actually voted yet – can be unpicked a hundred different ways. But it is hard to avoid the conclusion that how people vote is increasingly being shaped by how long they spent at school.

What is happening in the United States has also been happening in the UK. The Brexit campaign had its own Trumpian moment, courtesy of Michael Gove, who told Faisal Islam in an interview on Sky News on 3 June that “the British people have had enough of experts”. Gove was also widely mocked – if not experts, who was he proposing to get to repair his car, fix his teeth, teach his kids?

But what he said struck a deep chord, because it contained a large element of truth. The experts Gove was deriding had been telling the British public that the risks of Brexit far outweighed any potential benefits. Gove insisted that the voters should decide this for themselves, on the basis of their own experiences, rather than listening to elite voices that had a vested interest in the outcome. Those voices came trailing educational qualifications, which had put them in their positions of authority – at the IMF, the Bank of England, the Treasury. Gove was asking voters lacking anything like the same educational qualifications to feel empowered to reject what they were being told. And in the referendum on 23 June, that is what they did.

Voters with postgraduate qualifications split 75 to 25 in favour of remain. Meanwhile, among those who left school without any qualifications, the vote was almost exactly reversed: 73 to 27 for leave. A report from the Joseph Rowntree Foundation last month confirmed that “educational opportunity was the strongest driver” of the Brexit vote. Again, there were plenty of other factors at work – including a significant generational divide. Older voters were far more likely to vote leave, which partly helps to explain the education gap, since the rapid expansion of higher education in recent decades means older voters are also much less likely to have attended university. But the Rowntree report concludes that educational experience was the biggest single determinant of how people voted. Class still matters. Age still matters. But education appears to matter more.

A post-Brexit electoral map of Britain starkly illustrates this new divide. Scotland voted remain for its own particular reasons. But in England and Wales, many university towns emerged from the referendum as isolated outposts of pro-EU sentiment in a sea of Brexit. Newcastle, York, Nottingham, Norwich, Cambridge, Brighton, Warwick, Exeter, Bristol, Reading, Oxford and Cardiff all voted remain. I live in Cambridge, which voted remain by a margin of 74 to 26. There was consternation here following the result. It was accompanied by a barely suppressed feeling that ignorance had won the day. I lost count of the number of times I was told that one of the top trending searches on Google in the immediate aftermath of the vote was: “What is the EU?” The implication was that we had been taken out of Europe by people who did not even know what it was they were being asked to decide about.

Few felt comfortable, at least in public, following this thought through to its logical conclusion. Yet it was hard to escape the sense that a long buried suspicion of democracy was fighting its way back to the surface. If politics has turned into a contest between ignorance and knowledge, then places like Cambridge may be starting to feel dismayed by the realisation they are now on the losing side. But that would be to fall into an old trap.

Elite anxiety about the consequences of political ignorance is nothing new. In the long history of intellectuals worrying about democracy and its failings, two basic fears keep nagging away. The first is that democracy will mean rule by the poor, who will use their power to steal from the rich. The second is that democracy will mean rule by the ignorant, who will use their power to do the dumbest things. Both these worries go back at least as far as Plato. The ancient Greeks understood full well that democracy meant letting the have-nots get their claws into the haves. For Aristotle, that’s what the word meant: it was rule by the poor (the demos) over the wealthy. But if class conflict came with the territory, the deeper fear was what the masses might do out of sheer foolishness.

For Plato, democracy suffered from the basic defect of putting decision-making in the hands of people who were not competent to decide. Politics was a skill – and most people were simply clueless. Worse, that made them prey for hucksters and demagogues who would promise the earth and get away with it. Democracy was fertile ground for fantasists with a taste for power. If you tell the people that up is down, and the people believe you, then who is going to let them know that they are wrong?

These fears have never really gone away, and they reassert themselves at times of political crisis. In the 1920s, the American political commentator Walter Lippmann updated Plato for the 20th century by arguing that modern citizens simply lacked the mental capacity to process the information needed for intelligent decision-making. Lippmann had worked in American propaganda during the first world war and had seen at first hand just how easy it was to manipulate public opinion.

“It is no longer possible,” he wrote, “to believe in the original dogma of democracy: that the knowledge needed for the management of human affairs comes up spontaneously from the human heart.” People will vote on the basis of anything that grabs their attention in a passing moment, filtered by whatever deep prejudices they harbour beneath the surface. Evidence means little to the average voter; reasoned argument means even less. Lippmann concluded that democracy could only be rescued by establishing a cadre of specially trained experts, whose job was to steer politicians away from the dubious instincts of the people and back towards what the evidence required. Otherwise, the manipulation of public opinion would become the be all and end all of democracy, which is all the encouragement demagogues ever need.

Lippmann’s fears chimed with the growing worry of some prominent mid-20th century economists that too much democratic decision-making would lead to financial ruin. A series of future Nobel prize winners, from Friedrich Hayek to Milton Friedman, channelled the ancient critics of democracy by arguing that putting the voters in charge meant short-term impulses would prevail over long-term prudence. Ultimately, they thought, democracy produced inflation, which was just another way of the poor getting their own back on the rich. Hayek and Friedman both put their ultimate faith in the power of markets. But first, they felt experts had to be empowered to rein in the self-destructive impulses of the voters. Hayek at one point advocated restricting the franchise to those aged 45, in order to cut out the young, who don’t know what’s good for them, and the old, who have a vested interest in the past. Hayek did not believe that anyone, however expert, could know the future. But he wanted to put self-knowledge back at the heart of democracy: the worldly wisdom of understanding what is possible, and what is just wishful thinking.

Lippmann, Hayek and Friedman insisted they were trying to save democracy, not destroy it. But what they thought democracy needed saving from was itself. The voters had to be rescued from their preference for fairytales and the people peddling them. In the late 20th century, partly under the influence of such thinkers, democratic politics did carve out spaces where experts could be better insulated from the impulsive decision-making of the masses, as independent central banks and other unelected bodies were entrusted with more and more decision-making capacity, away from the glare of public opinion. Now, in the 21st century, these are the experts who find that the voters no longer want to heed their advice.

In the year of Trump and Brexit, it is tempting to think that democracy is reverting to type and that popular ignorance is once again being set against expert knowledge. Trump is what you get when demagoguery is allowed to run unchecked, egged on by a craven media that simply enjoys the show. Brexit is what you get if you ask people a question that they lack the basic information to answer. This is the view that has been doing the rounds in the circles in which the highly educated move. But to think this is a big mistake. The educational divide that is opening up in our politics is not really between knowledge and ignorance. It is a clash between one worldview and another.

What is the EU? Hearing educated remainers mock those who asked that question the day after the vote was an uncomfortable experience – and not just because the story about Google searches was largely apocryphal. After all, the question is not as straightforward as it seems: it is simple enough to say what the letters refer to, but far more difficult to know what they really mean for our politics and our future. Education is not the same as knowledge. Nor is knowledge the same as knowing which way to vote. The split between the university towns and other parts of the country did not arise because one set of people understood what was truly at stake and the others were just taking a wild guess. Both sides were guessing.

Even now, no one truly knows what is going to happen. The better-educated cleaved to one set of predictions because these chimed with what they already believed in. Polling carried out before the referendum, which asked people what they thought was likely to happen in the aftermath of a Brexit vote, found that university graduates thought that it would produce an immediate financial crash, whereas those with fewer qualifications thought it much more likely that things would carry on as before. Prior political preferences shape what we think the evidence shows, not the other way round.

None of this would have surprised Lippmann. When he argued that ordinary voters were incapable of judging complex policy questions on their merits, he did not exclude the educated from that judgment. He meant everybody. As he wrote: “I have not happened to meet anybody, from a president of the United States to a professor of political science, who came anywhere near to embodying the ideal of the sovereign and omnicompetent citizen.” He thought that we needed experts who were trained in eliminating their own biases – and whatever a regular university education does, it does not do that.

“The historical record leaves little doubt that the educated, including the highly educated, have gone astray in their moral and political thinking as often as anyone else,” write the political scientists Christopher Achen and Larry Bartels in their new book Democracy for Realists, echoing Lippmann. What the educated are better at is sounding like they know what they are talking about, because they have been trained in how to make an argument. “Well-informed people are likely to have more elaborate and internally consistent worldviews than inattentive people, but that just reflects the fact that their rationalisations are better rehearsed.” Education gives you the ability to tailor your arguments to suit your personal preferences, which is why it is a big asset on the job market. But it does little to help tailor your personal preferences to suit the best arguments.

In the current political climate, a lack of education is sometimes blamed for the spread of conspiracy theories, which run like wildfire through many populist movements. Trump’s campaign, with its nods and winks to “what’s really going on”, and its pre-emptive warnings about the vote being stolen, has been feeding and feeding off the appetite for conspiratorial thinking. Yet to blame this on the absence of education is just another version of the old fear of the credulity of the untutored masses: they will believe anything.

Possessing a university degree does not alter the probability of someone believing that global warming is a hoax – that tracks prior political commitments, not superior knowledge. Nor are conspiracy theories confined to the right. Jeremy Corbyn’s support among Labour party members contains a high proportion of university graduates. Yet in a YouGov poll, 55% of Corbyn supporters agreed with the statement that intelligence services such as MI5 have been working to undermine Corbyn since he became leader. That figure compares with only 19% of the wider public who give credence to the same claim. People will believe what their political loyalties incline them to believe, regardless of how much education they have received.

This doesn’t mean that our political beliefs are simply a reflection of our narrow self-interest. It is true that we all have a tendency to favour the worldview that enhances our future prospects. The preference of university graduates for remaining in the EU echoes the benefits that EU membership gives them: the free movement of labour and easy access to European networks is better for those with the qualifications to take advantage of a knowledge economy. But it is not the case that the education divide is just another version of the class divide, with the winners from globalisation lined up against the losers.

As with Trump supporters in the US, Brexit was not simply the cause of the disadvantaged and left behind. Many Brexiteers come from the affluent middle classes, particularly in the south of England, outside of the university towns. Meanwhile, universities are producing their fair share of losers these days, as students leave burdened with debt and confronted by a bleak job market. Yet these students – rich or poor, in work or out, elite or not – overwhelmingly favoured EU membership. Education does not simply divide us on the grounds of what is in our interests. It sorts us according to where we feel we belong.

Why does education do this? Political scientists have been aware of the growing education divide for decades. A report on the 1983 general election – whose authors included a young John Curtice, the current eminence grise of the UK polling business – noted that education levels were emerging as a significant indicator of voting patterns. The report put this down to the distinctive social attitudes that were picked up at university. “Educational experience,” it noted, “is an important source of values distinct from class experience.” These values often related to issues that were not straightforwardly economic.

Graduates, even in the 1980s, tended to be much more concerned about the environment than other sectors of the population. They were also strikingly more internationalist in outlook. In 1983, 58% of graduates expressed a positive attitude to what was then known as the European Economic Community, whereas only 35% of those who left school at 16 felt the same. Not much has changed since. But at that time, the Labour party was still overtly hostile to the European project, and withdrawal from the EEC was part of its famously radical 1983 election manifesto. Many graduates reported having leftish attitudes on a whole host of other questions – from defence spending to women’s rights. But on this issue their cosmopolitanism took precedence over their party loyalties. The result was that in the 1983 election more than twice as many university graduates voted for the centrist pro-European SDP-Liberal Alliance as voted for Labour.

One reason this was not a bigger issue back then – and the Alliance did not manage to supplant Labour as the main party of opposition – was that there were not that many university graduates. Although participation had been steadily rising, in the mid-1980s fewer than one in five people had benefited from higher education. That was enough to make a difference, but not enough to tip the balance. By 2012, university participation among 18-30 year olds was close to 50%, and it has only dipped recently because of rising tuition fees. That is enough to start splitting the population into two camps.

But it is not simply a question of demographics. The gap has been amplified by certain forms of social mobility, which have reinforced the education divide by enabling the better-educated to start congregating together: socially, geographically, romantically. In a previous generation, graduates often married non-graduates, because their choices tended to be driven by where they happened to live or work. As the cliché has it: bosses used to marry their secretaries. Not any more, and not just because there are fewer secretaries. If you went to university, ask yourself: how many of your friends didn’t go to university? And among your friends, how many of those who did are married to people who didn’t? Greater freedom of movement produces greater freedom of choice. But that does not produce more social diversity, it produces more social stratification.

Social media now enhances these patterns. Friendship groups of like-minded individuals reinforce each other’s worldviews. Facebook’s news feed is designed to deliver information that users are more inclined to “like”. Much of the shock that followed the Brexit result in educated circles came from the fact that few people had been exposed to arguments that did not match their preferences. Education does not provide any protection against these social media effects. It reinforces them.

The growing political divide between the educated and the less educated can be seen across Europe. It is most pronounced in Scandinavian countries, where university attendance is high and levels of education are an increasing driver of voting habits. It is less visible in southern and eastern Europe – in places such as Portugal and Poland – where participation in higher education is lower, and other social factors, including family and religion, still exert a strong grip.

But Britain and the US are different again, because of their political systems. In European countries with proportional representation, smaller parties can offer a home to particular segments of the population, including those at both ends of the education scale. The support for many European parties of the populist right is heavily drawn from less educated voters. Meanwhile, green and liberal parties, especially those that favour immigration, rely strongly on university graduates for their support. That group includes the Liberal Democrats, which is still the UK-based party that attracts the highest proportion of graduates among its voters. But as the Lib Dems have discovered to their cost, with first-past-the-post that is not enough. Under a PR system, targeting the graduate vote can generate significant representation in parliament. Here, it can get you wiped out – especially if you make a promise to students not to raise their tuition fees and then fail to keep it.

The education divide is never going to supplant traditional left-right politics. There is not going to be a “Graduate party” taking on a party of “School Leavers”. Instead the divide cuts across left and right, which is why it is proving so disruptive to our politics right now. Big-tent political parties are struggling to hold their fractured coalitions of voters together. Polling still reveals some shared attitudes between university graduates and those without educational qualifications on economic issues, such as support for trade unions and mistrust of the free market. But the education divide derives from an alternative set of values, which is often characterised as the opposition between libertarians and authoritarians.

Authoritarians are looking for order and control, libertarians want greater freedom and tolerance. Along with education levels, the strongest indicator of likely support for Brexit was shown by attitudes to capital punishment: the more you were in favour, the more you wanted to leave the EU. These attitudes tend to track educational experience. Labour’s support is now split between left-leaning libertarians (broadly pro-union and anti-banks, but also pro-immigration, and often highly educated) and left-leaning authoritarians (also pro-union and anti-banks, but far tougher on immigration and very concerned about crime and community). Each grouping might gather upwards of 20% of the electorate under its wing. Together that would be enough for a parliamentary majority. But they do not fit together any more, and increasingly they neither like nor trust each other.

The same forces are threatening to play havoc with electoral politics in the US. Trump’s army of the less educated is on the march inside the Republican party, but that leaves plenty of traditional Republicans – including many college-educated ones – wondering where to go. Some Republicans will be voting for Clinton. Some traditional Democrats will be sorely tempted by Trump’s deeply authoritarian message. It is true that American democracy has proved highly resilient in the past and the party system has adapted to many demographic and cultural shifts. However, it is also possible that the gap between the educated and the less educated is going to become more entrenched over time, because it is not just a question of economic interests. It represents a gulf in mutual understanding.

The conflict between rich and poor still matters a great deal in our politics. But the age-old fear that democracy would end up letting the poor steal from the rich has always looked overblown. Constitutions – including the one carefully crafted by the American founders to guard against just that outcome – have done their bit to safeguard the interests of the propertied classes. So, too, have the increasingly complex institutional arrangements that give power to experts in areas – defence, finance, environment, health – where technical knowledge is assumed to be required. Representative democracy has proved effective at staving off class war. It has done it by finding electoral outlets for popular dissatisfaction – as with the election of Franklin D Roosevelt in 1932, of Attlee in 1945, or of Obama in 2008 – that never quite let the people take charge. When inequality gets out of hand, the system does what it can to correct for it. But it is designed to avoid extreme solutions driven by popular anger. That still holds. The haves do not look in danger of being expropriated by the have-nots any time soon.

The contest between the educated and the less educated is different. Many of the safeguards that have been put in place to bypass popular politics – above all, the authority that now resides in central banks – have had the effect of empowering a new class of experts, for whom education is a prerequisite of entry into the elite. These are not just the bankers, but the lawyers, the doctors, the civil servants, the technicians, the pundits, the academics. Not all of the educated are winners in this world, but almost all of the winners are educated. It gives the impression that knowledge has become a proxy for influence.

When Gove suggested that the experts should not be trusted because they have a vested interest in what they are saying, that was his point: once knowledge becomes a prerequisite of power, then it no longer speaks for itself. It appears to speak for the worldview of the people who possess it. At that point it ceases to be knowledge and simply becomes another mark of privilege.

The education divide has the potential to break apart the careful ties that hold representative democracy together. Regardless of our different interests, we elect representatives to take decisions on our behalf on the understanding that we share certain basic values, including a respect for knowledge, wherever it comes from. Once knowledge is assumed to be just another one of the perks of power, then the basis to trust others to take decisions for us becomes eroded. Asserting the facts and asserting your privilege grow increasingly difficult to distinguish.

Educated v less educated may be even more toxic than rich v poor, because it comes laden with assumptions of moral superiority. These days the rich find it quite hard to get away with the presumption that their wealth is proof of their virtue. When they seek protection from the system, it is pretty clear what they are up to: they are looking after their interests. But when the educated look out for themselves they can dress it up as something ostensibly better than that: expertise.

To those on the receiving end, that stinks. It stinks of hypocrisy, and it also stinks of self-interest. The fact that the educated are not always the beneficiaries of the social attitudes that they hold – Corbyn’s supporters, like Bernie Sanders’s, would rightly insist that many of the positions they adopt are designed for the benefit the socially excluded – does not help. It just makes them sound even more self-righteous.

In the run-up to the EU referendum Jeremy Cliffe of the Economist – the house journal of the educated expert – suggested that although Brexit might win this time, it was possible the UK would vote to rejoin the EU in a second referendum 20 years from now, by which point current levels of university education would have spread to later age groups. Twenty years is a very long time in politics. Before then, the gap between the educated and the less educated has the potential to widen, as each side digs in. The viciousness of public debate, as bad as it is now, could still get worse. Because two-party politics does not map on to this division, mainstream politicians will have to find creative ways to try to harness it. These carry the risks of making it worse.

The EU referendum was seen by educated optimists – including some of the people around David Cameron – as just another way for democracy to let off steam: a means of giving vent to anger without letting it run out of control. That is what the optimists have been saying about Trump too. But the steam is still rising.

• Follow the Long Read on Twitter at @gdnlongread, or sign up to the long read weekly email here.