On The Spectrum

The news that Oxford Dictionaries had selected "youthquake" as its word of the year for 2017 was met with widespread incredulity, and not just because the smart money was on milkshake duck. It's the lexicographical equivalent of dad-dancing, combining a toe-curling lack of self-awareness with the impression that the publishers of the OED still rely on newspapers for their impression of popular discourse. The centrist attempt to explain the upsurge in support for Labour at the last general election exclusively in terms of the youth vote (aka the "millennial awakening") was echoed this week in the patronising claim that the Alabama Senatorial contest was won by black women. Talking of the Alabama vote, Corey Robin correctly noted that political victory usually depends as much on encouraging abstention by the other parties' supporters as it does on maximising the turnout of your own: "In an electoral democracy, the way to break your opponents – especially opponents like these – is to demoralize them, to make them feel they are a small and isolated minority, that their cause is a loser".

This observation runs counter to the prevailing orthodoxy that politics is increasingly dominated by "values" or "culture", which implies strongly-held beliefs no more likely to be shaken by evidence than by moral suasion. Much of the commentary around Brexit is still fixated on the idea that there is a fundamental division in society, represented in terms of personality types such "somewheres" and "anywheres", that the political superstructure must eventually align to. The idea of an intrinsic temperament has an ancient history in Western thought, going back at least to the four humours of Greek medical theory, and it lives on in modern marketing and self-help psychology from where it bleeds into politics. The idea that voters make judgements on values that trump material self-interest has been around since the advent of democracy (the patriotism card, the race card etc), but it doesn't really gain traction in political science until the postwar period when we start to see the traditional left-right linear spectrum of political mapping augmented by biaxial (or Cartesian coordinate) models that combine the economic with a values dimension.

One reason for this development was anticommunism, which sought to establish a common ground between Nazism and Communism centred on the denial of "personal freedom". Given the lack of social autonomy and the extent of cultural conservatism in the 1950s, which was highlighted by the "rebellion" of the 60s, this required a particular interpretation of personal freedom to achieve the desired ends. As a result, early theorists in the field, such as Leonard Ferguson and Hans Eysenck, tended to emphasise values that were politically salient, such as attitudes to abortion and religion, rather than necessarily rooted in psychological predispositions. An unintended consequence of the biaxial model was the boost it gave to both anarchists and right-libertarians (e.g. the Nolan chart) who shared an interest in extending the freedom axis beyond the simple absence of a police state. Over time, both the economic and personal axes were increasingly articulated in terms of an American conception of "liberty", reflecting the neoliberal shift in the premises of conventional political science and specifically the Chicago School emphasis on legalism and markets.

A cruder consequence of this was the emergence over the last twenty years of the centrist Horseshoe theory, which sought to reinsert the "both as bad as each other" idea into a traditional linear spectrum in which the extreme left gradually approaches the extreme right. This started out as a trope of those, like Nick Cohen and Jonathan Chait, who have built a career mapping the supposed moral equivalence of everyone who diverges from liberal orthodoxy, but it was soon adopted as a tactic by right-wing authoritarians, notably in Eastern Europe, who sought to position themselves as equidistant between the Nazis and any variety of leftism. More broadly across the West, this has fuelled the inflation of mild social democracy into an existential threat. In the UK, it has resulted not only in Tories accusing Jeremy Corbyn of being a Marxist, as you might expect, but in liberals insisting that Momentum is Militant reincarnate. An inevitable twist to this general trend was the Israeli right's insistence that the equivalence of the left and the Nazis meant that any leftist criticism of Israel must necessarily be antisemitic.

The problem with these models (and there are a lot of them) is that they seek to map the field of politics from personal opinion (or the psychoanalysis of historical figures) rather than through the interpretation of social action or institutions. They are subjective rather than objective, which means they reflect not only the prejudices of the researchers but the received wisdom of popular debate, and thus media bias. Factor analysis is used to identify the discriminating values for plotting purposes, but this often leads to begging the question (a high-profile  example of this is Jonathan Haidt's The Righteous Mind). The ideological bias isn't difficult to spot. In most biaxial schemes the economic dimension reflects the individual's attitude to the degree of state control over the economy, rather than attitudes towards the relationship of fiscal and monetary policy or workplace democracy. Likewise, the liberty dimension reflects the degree of social coercion experienced by the individual rather than the extent to which the individual profits from collective action. In this reading, working class solidarity is considered to be authoritarian and government borrowing is assumed to be left-wing.

A recent example of the problems this gives rise to was an FT article that tried to explain the lack of "Bregrets" by casting the EU referendum result as a "cultural" divide between authoritarians and libertarians. The conclusion was that "Brexit has become a form of identity politics". Leaving aside the oddity of that "become", which implies that Brexit was once something other than a matter of identity, the impression given by the chart below left is that remainers are broadly libertarian and leavers broadly authoritarian. But this only makes sense if you imagine a definition of authoritarian that excludes Tony Blair and George Osborne, or assume that a right-libertarian like Daniel Hannan is utterly unrepresentative of leaver opinion. In fairness, they are all eccentric in their own ways, but I doubt you'll get Blair to admit he is anywhere other than solidly in the middle of the remain camp, while Hannan would be justified in claiming that there is a significant number of leavers who are convinced that unilateral free trade and a bonfire of regulatory red-tape is all that stands between Britain and a restored greatness.

The juxtaposition of the chart on the left with the one on the right is misleading on two counts. First, it implies a stronger relationship between party affiliation and voting in the referendum than is shown in the data. Second, and related to this, it suggests that leave won because a pivotal chunk of Labour voters deserted. This is partly a trick of visualisation: the two zones aren't proportionate to the vote share, but you'd instinctively think they are because there is no variation in shading to reflect the density of the data-points (i.e. the individual coordinates). What the two charts don't show is the relationship of the referendum vote with other party support. Bear in mind that 33% of the vote in the 2015 general election went to parties other than Labour and the Conservatives. The implied relationship between party support and the referendum also doesn't account for habitual non-voters who were tempted out to the polling booth in 2016 for the first time in many years. The turnout between the 2015 general election and the EU referendum increased by almost 3 million, from 30.7 million to 33.6 million. Given than the winning margin in the referendum was a little under 1.3 million, and recognising that there will have been some people who voted in 2015 but not in 2016, it should be clear that leave won because it mobilised habitual non-voters, not Labour supporters.

As was evident in high leave-voting areas like Sunderland, this was a national constituency not so much of self-identifying authoritarians but of social conservatives who had become politically disengaged after the millennium. The increase in the general election turnout of 1.5m between 2015 (30.7m) and 2017 (32.2m) owed something to the "youthquake", but this was dwarfed by the number of older voters who turned out for the first time in years in 2016 and then went to the polls again this year. Remember, the Tory vote increased from 11.3m in 2015 to 13.7m in 2017, despite their poor performance among the young. I draw two implications from this. First, the 2016 referendum is probably the high-water mark for leave, not so much because their older voters will die off, but because many will retreat to non-voting as the Brexit process grinds on to a disappointing end (though I think a second referendum remains unlikely). Second, and echoing Corey Robin, Labour's chances in 2022 (or earlier) will depend on Tory voters becoming demoralised, which is likely both because of the compromises entailed by Brexit and the inevitable neglect of other policy issues. None of these dynamics are evident if you view politics through a spectrum, but they are what will ultimately matter.