brexit-twitter-image

Britain’s choice to vote Leave, we are told, is a protest by those left behind by modernisation and globalisation. London versus the regions, poor versus rich. Nothing could be further from the truth. Brexit voters, like Trump supporters, are motivated by identity, not economics. Age, education, national identity and ethnicity are more important than income or occupation. But to get to the nub of the Leave-Remain divide, we need to go even deeper, to the level of attitudes and personality.

Strikingly, the visible differences between groups are less important than invisible differences between individuals. These don’t pit one group against another, they slice through groups, communities and even families. Our brains deal better with groups than personality differences because we latch onto the visible stuff. It’s easy to imagine a young student or London professional voting Remain; or a working-class man with a northern accent backing Leave.

fig 1

Open and Closed personalities are harder to conjure up: yet these invisible differences are the ones that count most – around two or three times as much as the group differences.

fig 2

A second problem is that many analyses are spatial. Don’t get me wrong: I love maps and differences between place are important, especially for first-past-the-post elections. However, some characteristics vary a lot over space and others don’t.

Figure 1 ranks Local Authorities and England and Wales by the average social grade of their White British residents. The lower the average class position of White British residents, the higher the vote for Brexit. In fact this working class index explains 58% of the variation in the Leave vote across districts. But, according to the 2015 British Election Study Internet Panel of over 24,000 respondents, class only explains 1-2% of the variation in Brexit voting intention among individuals. There may have been a slight shift over the past year, but this won’t have altered the results much.

Figure 1.

fig 3

Source: https://www.nomisweb.co.uk/ (census data); http://www.electoralcommission.org.uk/

Why the discrepancy? Social characteristics such as class, ethnicity or region are related to where people live, so they vary from place to place. Psychology and personality do vary over space a little because they are affected a bit by social characteristics like age. But they’re mainly shaped by birth order, genetics, life experiences and other influences which vary within, not across, districts.

Aggregate analysis distorts individual relationships even when there aren’t problems caused by the ecological fallacy. The sex ratio, for instance, is more or less the same from one place to another, so even if gender really mattered for the vote, maps hide this truth. On the other hand, maps can also exaggerate. If 1% of Cornwall votes for the Cornish nationalists then Cornwall would light up on a map of Cornish nationalist voting with a 100% correlation. But knowing that all Cornish nationalist voters live in Cornwall doesn’t tell us much about why people vote for Cornish nationalists.

As with region in the case of Cornish nationalism, class matters for the vote over space because it affects, or reflects, where we live. This tells us a lot about a little. Notice the range of district average class scores in figure 1 runs only from 1.8 to 2.4 whereas individuals’ class scores range from 1 (AB) to 4 (DE). The average difference from the mean between individuals is ten times as great as that between districts.

Let’s therefore look at individuals: what the survey data tell us about why people voted Brexit. Imagine you have a thousand British voters and must determine which way they voted. Figure 2 shows that if you guess, knowing nothing about them, you’ll get 50 percent right on average. Armed with information on region or their economic situation – income and social grade – your hit rate improves to about 54 percent, not much better than chance. In other words, the big stories about haves versus have-nots, or London versus the regions, are less important.

Age or education, which are tied more strongly to identity, get you over 60 percent. Ethnicity is important but tricky: minorities are much less likely to have voted Leave, but this tells us nothing about the White British majority so doesn’t improve our overall predictive power much.

Invisible attitudes are more powerful than group categories. If we know whether someone supports UKIP, Labour or some other party, we increase our score to over 70 percent. The same is true for a person’s immigration attitudes. Knowing whether someone thinks European unification has gone too far takes us close to 80 percent accuracy. But then, this is pretty much the same as asking about Brexit, minus a bit of risk appetite.

Figure 2.

Picture1

Source: British Election Study 2015 Internet Panel, waves 1-3

For me, what really stands out about figure 2 is the importance of support for the death penalty. Nobody has been out campaigning on this issue, yet it strongly correlates with Brexit voting intention. This speaks to a deeper personality dimension which social psychologists like Bob Altemeyer – unfortunately in my view – dub Right-Wing Authoritarianism (RWA). A less judgmental way of thinking about RWA is order versus openness. The order-openness divide is emerging as the key political cleavage, overshadowing the left-right economic dimension. This was noticed as early as the mid-1970s by Daniel Bell, but has become more pronounced as the aging West’s ethnic transformation has accelerated. Figure 3 shows that 71 percent of those most in favour of the death penalty indicated in 2015 that they would vote to leave the EU. This falls to 20 percent among those most opposed to capital punishment. A similar picture results for other RWA questions such as the importance of disciplining children. RWA is only tangentially related to demographics. Education, class, income, gender and age play a role, but explain less than 10 percent of the variation in support for the death penalty.

Figure 3.

fig 5

Source: British Election Study 2015 Internet Panel, waves 1-3.

Karen Stenner, author of the Authoritarian Dynamic, argues that people are divided between those who dislike difference – signifying a disordered identity and environment – and those who embrace it. The former abhor both ethnic and moral diversity. Many see the world as a dangerous place and wish to protect themselves from it.

Pat Dade at Cultural Dynamics has produced a heat map of the kinds of values that correspond to strong Euroskepticism, and to each other. This is shown in figure 4. Disciplining children and whipping sex criminals (circled), keeping the nation safe, protecting social order and skepticism (‘few products live up to the claims of their advertisers…products don’t last as long as they used to’) correlate with Brexit sentiment. These attitude dimensions cluster within the third of the map known as the ‘Settlers’, for whom belonging, certainty, roots and safety are paramount. This segment is also disproportionately opposed to immigration in virtually every country Dade has sampled. By contrast, people oriented toward success and display (‘Prospectors’), or who prioritise expressive individualism and cultural equality (‘Pioneers’) voted Remain.

Figure 4.

fig 6

All told, the Brexit story is mainly about values, not economic inequality.

Eric Kaufmann is Professor of Politics at Birkbeck College, University of London. He may be found on twitter @epkaufm.