The church hall in Ludlow smelt of instant coffee and damp coats. Around six tables sat forty residents who had been chosen, more or less at random, to spend two Saturdays discussing a question that had divided the town for years: whether to close the high street to traffic. On the first morning, the facilitator, a softly spoken woman named Priya Nandakumar, asked each table to write down what they thought the people who disagreed with them believed. Then she asked them to check.
The results were, by her account, typical. 'People almost always overestimate how extreme the other side is,' she told me during the break. 'The drivers thought the cyclists wanted to ban cars from the whole county. The cyclists thought the drivers didn't care about children's safety. When they actually talked, most of them wanted roughly the same things in a different order.'
Researchers who study polarisation distinguish between two kinds. Ideological polarisation is a widening gap in what people actually believe about policies. Affective polarisation is something else: a growing dislike, even contempt, for people on the other side, regardless of how far apart the positions really are. In several countries studied, the second kind appears to have grown faster than the first. We may not disagree much more than our parents did, but we seem to like each other less for it. A related finding – sometimes called the 'perception gap' – is that people's picture of their opponents is often a caricature, drawn from the loudest and least representative voices.
Why this should be so is fiercely contested. The most popular explanation blames social media, whose algorithms are said to trap users in 'echo chambers' that only reflect their own views back to them. The evidence is more equivocal than the phrase suggests. Several studies have found that most people encounter a fair amount of opposing opinion online; the trouble is the form in which they encounter it – usually as the most inflammatory example, shared by someone on their own side as proof of how unreasonable the others are. Exposure, in other words, is not the same as understanding. Some experiments have even found that simply showing people more posts from the other side can harden, rather than soften, their views.
Other scholars look further back. They point to the decline of institutions where people of different views once mixed as a matter of course – trade unions, churches, sports clubs, the local pub – and to the tendency of people to cluster, by choice or by cost, in neighbourhoods and professions where most others think as they do. On this account the internet is less a cause than an accelerant, pouring fuel on divisions that were already smouldering.
Not everyone thinks the picture is so bleak. Daniel Esterhuizen, a political theorist who spoke at the Ludlow meeting, warned against nostalgia for a supposed age of consensus. 'Some periods looked calm because certain people weren't allowed in the conversation,' he said. 'Disagreement is not a sickness. A society where everyone agreed would be either very small or very frightened.' What matters, he argued, is not how much we disagree, but whether we can still lose an argument without feeling that we have lost our place in the community.
By the second Saturday, something had shifted in the hall – not a conversion, but a thaw. A retired lorry driver named Colin and a young mother named Aisha, who had sat at opposite ends of their table on the first day, ended up drafting a compromise together: closure on weekend afternoons, a trial for six months, and a review led by residents. Neither was entirely happy. 'I still think she's wrong about the buses,' Colin said, grinning. 'But I don't think she's daft any more.'
It would be naive to imagine that a few dozen people in a church hall can mend the divisions of a nation. Ms Nandakumar does not claim they can. Her ambition is smaller and, perhaps, more realistic: to remind people that the opponent in their head is usually a straw man, and that the real one, met face to face over bad coffee, tends to be more reasonable, more uncertain and more human. Polarisation, she suggests, thrives on distance. Its most reliable antidote may be nothing more sophisticated than proximity.