Values and voting: Part II

In a previous piece I looked at the measurement and relationships between value measures and how these related to EU Referendum voting. This shorter piece aims to show these patterns more clearly, highlighting how they combine across different value positions.

Firstly, to reiterate the way values are measured. The data are taken from the British Election Study Internet Panel, Wave 10 (December 2016). The two ‘core value’ scales are measured using items that have been well-tested and used for this purpose in the UK for almost 30 years.

Measuring populism among the electorate is much more recent. The survey uses the following measures.

In each case people are asked to express how much the agree or disagree with the statement on a five-point scale. The items are added together to form a scale (running from one to five) with low values representing the ‘left’, ‘liberal’ and least populist positions respectively. To represent the value space, the ‘left-right’ and ‘liberal-authoritarian’ scales are divided into three groups then combined to create a value space with 9 positions (Left/Liberal, Left/ Centre, Left/Authoritarian, Centre/Liberal, Centre/Centre, Centre/Authoritarian, Right/Liberal, Right/Centre and Right/Authoritarian). The distribution of the electorate in these groups is shown in Figure 1 below. Note that exactly where the boundaries of these groups should be drawn is not fixed and so whilst the broad shape of this distribution is indicative of the shape of the electorate it should not be used as a definitive measure of the size of the groups.

Using these measures, around 55% of the electorate are on the ‘left’ of the ‘left-right’ scale. But these are split with the largest group on the ‘left’ being those who are also authoritarian. Overall, the largest of the nine value groups are those who are in the centre on both measures, with around 1 in 4 of the electorate in this position.

To look at how populist sentiments vary across this value space, the populist scale is split into two groups. Those who score highly (a score of 4+) on the populism scale and those who do not. Using this measure we can see how ‘populist’ different parts of the value space are.

Populist sentiments are more common on the left than the right of the economic dimension, and authoritarians are more likely to be populist than liberals but the patterns are complex. Among the ‘left-authoritarian’ group almost 7 in 10 are also populist; among the ‘left-liberals’ this is fewer than 2 in 10 and among ‘right-authoritarians’ fewer than 3 in 10. This is an important caveat for thinking about populist politics; it appears to find its greatest support among those who are authoritarian and on the left of politics; in other words an interplay of economic and social values.

We can use this same framework to see how these values interact to produce political behaviour. Firstly, in the case of the EU Referendum vote. The chart below shows the percentage of leave voters within each of the 9 value space positions but broken down into populists and not populists.

In each case, there is a greater proportion of leave voters among the populist group than the not populist group. In some sense the EU referendum vote was motivated by populist sentiment. But this is not the whole story. In all cases the more authoritarian groups have a greater proportion of leave voters than the liberal groups. Such that regardless of left-right position, at least 6 out of 10 authoritarian voters voted Leave. This is further modified by populist sentiment. Among authoritarian voters who were also populist 8 out of 10 voted leave. Again this is not (quite) the whole story as there is also a (much smaller) influence of left-right position within each group those on the right are a little more likely to be leave voters than those on the left.

We can use the same framework to look at how these interactions between value positions influenced voting behaviour in 2017. Unlike the leave vote, the largest differences here are on the ‘left-right’ scale. Populist sentiments have little effect here (the two bars are roughly equal across all groups). But there again is a more complex pattern with the liberal authoritarian dimension than might be anticipated. The Labour party win a clear majority of the votes of those on the ‘liberal’ and ‘centre’ left. But their vote share is lower among the ‘authoritarian’ left than it is among the centre liberal position. In fact, the Conservatives gained greater share of the left-authoritarians (regardless of populist sentiment).

By thinking about value positions spatially rather than as a single dimension, we are able to better describe and understand positions within it. In particular we can see how the interactions of these values are complex and not uniform across value space. Populist positions further complicate the picture. They are central to understanding the EU Referendum vote but play virtually no role in the 2017 General Election. And whilst ‘left-right’ position has only a limited impact on Referendum voting, it remains the strongest influence on General Election vote choice between the two major parties.

It would be exceptionally premature to declare the death of the ‘left-right’ divide. But again it is no longer the whole story; particularly for those voters who we might think of as ‘cross-pressured’, aligned with Labour on economics and the Conservatives on social issues. Successful appeals to this group of voters may well be critical to future electoral outcomes; even though this group is often the least ‘tuned in’ to the messaging from Westminster they are numerous and can be appealed to by each party on different grounds. Perhaps finding the messaging for this group which is not that naturally of the party is key (May’s ‘just about managing’ seeking to connect with the authoritarians ‘left-ness’, while Corbyn’s ‘workers’ Brexit’ perhaps connecting to authoritarianism on the left). However, this appears a trickier balancing act for the Labour party seeking not to alienate its ‘liberal’ support base (with both the Greens and Liberal Democrats waiting in the wings) than it is for the Conservatives for whom there is no equivalent ‘right-wing’ electoral challenge.