Three-D Politics 2021

Posted 22 March 2021

Politics has seen a lot of volatility in recent years. These have been driven by Brexit and its aftermath, identity politics, and wild political opinion swings during the Covid crisis. But have underlying political attitudes moved a lot too, or have people been stable in what they believe.

A few years ago Electoral Calculus pioneered a unique way of measuring political attitudes using a three-dimensional scale (details). This gave each person's attitudes an individual score (from left to right) on three separate axes: economic, global and social. The combination of all three scores gives a person's position in this three-dimensional political space. People with similar views could be grouped together into like-minded "tribes" who thought in similar ways.

We identified seven tribes: strong left, traditionalists, progressives, centrists, somewheres, kind young capitalists, and strong right. Each tribe had identifiable political views, and its members often shared some demographic characteristics.

This analysis was based on British Election Study polling from 2016-7, which was before Brexit (and before the Brexit deal), before at least one general election, and before Covid. How have attitudes changed since then?

Luckily, the British Election Study has continued to run polls which ask similar questions. Their most recent poll was a large-scale survey of over 31,000 people which took place in June 2020. Of those respondents, 27,500 answered all the relevant questions, which is an impressively large sample size.

We had a look at these responses to see what has happened to the nation's political attitudes and to check up on our seven tribes. Here's what's happened.


Let's recall the three political dimensions we use:

Each axis is measured on a scale from -100 (extreme left position) to +100 (extreme right position). This sign convention is chosen so that the left-wing position is on the left-hand side and the right-wing position is on the right-hand side.

Due to considerable overlap between the questions asked in the 2016-17 polls and the 2020 poll, we can measure directly the change in opinion on each axis. The Global axis had the largest change, with a swing of 16 towards the left-hand side. This means that, on average, people became a bit more globalist in their attitudes and a bit less nationalist. Scores on the Economic and Social axes were very stable, which indicates no big overall change in those opinions.

But there was another important trend seen. As well as looking at the average score, we can also look at how closely or widely people's views are spread about that average. In statistical terms, this is called the standard deviation of responses and shows whether people's view are tightly clustered around the average (low standard deviation) or more widely spread (high standard deviation). In the old sample, the Economic axis had the largest standard deviation, making it the biggest driver of political attitudes. But the recent poll has reversed that, and the Global axis has the greatest standard deviation with the Economic axis in third place.

This indicates that the Global axis is becoming more important politically. This is further confirmed when the poll asks people to place themselves on a left-right scale. Previously this answer was most correlated with Economic attitudes, but now it is most closely related to Global ones. This is an important shift which shows the public now puts questions of international/national identity ahead of economics as the main driver of political attitudes.

An interesting sidelight is that some economic questions (on deficit reduction and generosity/necessity of state benefits) are now more correlated with the Social score than the Economic score. A striking figure is that there is now zero correlation between wanting to reduce the government's fiscal deficit and views on other economic questions. Perhaps this is further evidence of the crumbling power of economic ideology to determine political attitudes.

Seven tribes

The original seven tribes are still very much visible, and have the same broad attitudes. The first chart shows each of the seven tribes positioned according to each one's Economic and Global score. Whether the tribe is typically socially liberal or conservative is shown by its colour, and its size represents its prevalence in the population.

Seven Tribes by Economic and Global position 2020

This chart shows that people don't have consistent left or right attitudes across all three dimensions. The "Traditionalists" are left-wing economically and socially, but are more neutral on the global questions. The "Somewheres" are right-wing nationally and socially, but left-of-centre economically. The "Progressives" are left-wing globally and socially, but neutral on economic questions. And the "Kind Young Capitalists" are right-wing economically and globally, but more socially liberal. Only the "Strong Left" and "Strong Right" tribes have total ideological consistency. Most people don't.

Tribes have gained or lost some members over the years. The "Strong Left" and "Progressive" tribes have each gained members, while the "Centrists" and "Kind Young Capitalists" have lost members. Overall the tribes are more equally sized than they were before.

There has also been some movement in the tribes' positions. The second chart shows how they have changed since the previous poll. For simplicity, the global and social scores have been averaged to form the second (vertical) dimension, since these are quite correlated. Each line represents the old (tail) and new (head) position of each tribe.

Seven Tribes change in position 2017-20

The main trend is that almost all tribes have moved closer to the centre economically. This is consistent with the reduction in the standard deviation of economic scores, and the decrease in political salience of economic attitudes. On the combined Global/Social axis, there has been more stability, albeit with some drift to the right by right-wing tribes.

Youthful left-wing politics

Another striking feature of contemporary politics are the strong left-wing attitudes expressed particularly by younger adults. Often combined with identity politics, these 'woke' ideas have attracted considerable social and political debate.

In political terms, a common theme of these attitudes is to form a coalition of identity groups which are assumed to be left-leaning, such as ethnic minorities, lesbian and gay voters, and religious minorities. These groups can, in theory, be combined with traditional Labour-leaning groups such as the working-class and the less well-off to form a winning coalition.

Let's look at this, to assess the consistency of the political allegiances and attitudes of these groups.

We'll begin with the left-hand side of the political spectrum. The chart below shows the Labour support at the 2019 general election (for those voting) by membership of various demographic groups.

The groups selected as potentially pro-Labour are: the traditional working class (those with an approximate socio-econonomic grade of D or E), those in the least well-off quartile of the population, ethnic minorities, Muslims, people aged 18-24, current university students, those who have completed university, those identifying as LGBT, and Londoners.

Labour support at GE2019 by selected sub-populations

The national Labour vote at the election (33%) is shown by the horizontal pink line. Most groups selected are indeed Labour-leaning. The main exception is those people in the DE socio-economic categories, who are no more pro-Labour than the general population. This striking change from previous decades has been much commented on, but it is still remarkable. Similarly, those in the least well-off quartile are not overwhelmingly supportive of the Labour party, despite its historic association with this group. The university-educated are mildly more likely to vote Labour, which is also a change from previous generations.

The other groups are clearly more supportive of the Labour party, from Londoners (46%) all the way up to Muslims (80%).

On the right-hand side of the spectrum, we have a similar chart where we have selected groups that might be expected to be Conservative-leaning. These groups are: the solid middle classes (socio-economic groups A and B), the richest quartile of the population, those with only a school education, the retired, those aged over 65, those identifying as Christian, and those living in the South West of England.

Conservative support at GE2019 by selected sub-populations

The national Conservative vote at the 2019 general election (44%) is shown as a light blue horizontal line. Most of the groups are Conservative-leaning, except the solidly middle classes and the richest quartile, which are bit less Conservative than the national average. Again, this underlines the change in the political order over recent years as these groups would previously have been more Conservative than average.

The school-educated, Christians and South-West residents are mildly but not strongly more Conservative. Only the retired and those aged over 65 (groups with an obvious common overlap) have an overwhelming preference for the Conservatives.

The Electoral Calculus political attitude scores let us look beyond the headlines of party affiliation to see into the political attitudes of these groups. Let's look at the Social axis, because it is the closest proxy to identity politics.

The final chart shows the various groups from both sides, positioned on the social axis according to the average score of each group's members. The national average has been calibrated to be zero, so groups to the left of zero are more socially liberal than average, and groups to the right of zero are more socially conservative. The average Social scores of Labour and Conservative voters respectively are also shown for reference. The size of each group's circle is proportional to its size in the population.

Social Score by selected sub-populations

The most socially liberal group by far is that of university students with a score of -28 (on a scale from -100 to +100). That is far ahead of the second most liberal group, which is those aged 18-24 years of age, even though the latter shares many members with the former. Those identifying as LGBT are also strongly socially liberal. All these groups are well to the left of the centre-of-gravity of Labour voters (at -16). Interestingly, many other Labour-inclined groups are much more centrist on social questions. The university educated, Londoners, Muslims, members of ethnic minorities and the least well-off all have social scores between -10 (only mildly social liberal) and 0 (neutral). The working class groups are even found on the mildly conservative side of the social axis.

Conversely, many Conservative-leaning groups are less socially conservative than the average Conservative voter. Only the over 65s and the retired are as socially conservative as the Conservative voters' centre-of-gravity. The better-off and the ABs, who are not even politically Conservative any more, are no longer socially conservative either.

At the moment, the extreme social liberalism of the young is concentrated among university students and is not widely shared by the rest of the population. Most other demographic groups have members with a range of social views, whose averages are closer to the national average. The main exception is those aged over 65 who are clearly socially conservative.


Our three-dimensional political framework has held up fairly well through the changing political weather. The country as a whole is slightly more internationalist than it was, and cares less about economic issues. The seven tribes are still broadly similar and continue to provide a useful way to segment the national population.

University students are showing a strong social liberalism, but one that is not fully shared by other groups, even those who are Labour-inclined. Labour's challenge will be the hold its varied coalition together and stop parts of its "right flank" being peeled away by the Conservatives. The Conservatives' worry may be that these socially liberal attitudes could cross over from the student minority into the public mainstream.

You can also take a short survey to find out which tribe you belong to.

Return to the home page.