On finding political axes using maths

Drifting in the sea of political beliefs, left and right don’t seem to have any meaning any more.

It’s possible to make up new axes. Alas, these aren’t grounded in real views of the population, instead they’re distorted by the politics of their creators.

Luckily, there is a way of finding actual axes by experiment, using opinion polls and maths.

Back in 2005, my friends Chris and Tom persuaded a youthful YouGov to add a complete set of political position questions to their demographically weighted UK poll.

The results are summarised in these slides they made at the time.

(See methodology details with discussion in a blog post by Chris)

So to summarise, there are only two axes of political belief they observed by using maths to find which beliefs are correlated together.

The most powerful axis combines opposition to the EU, support of capital punishment, dislike of state funded international aid and wanting to tax working people less. I’d like ideas for what we should call each end of it – trust / distrust, soft / hard, open / ordered, internationalist / nationalist, remain / leave?

The second axis, which is quite a bit weaker, combines agreement with the Iraq war, love of free trade, support of genetic engineering and not wanting to tax rich people more. This is neoliberal / not neoliberal – it’s tempting to call it left / right or socialist / capitalist but I think that loses important nuance.

Those first two axes only account for 27% of the variation in beliefs. There are lots more political beliefs, but those are independent of each other. They don’t form useful clusters for the purpose, say, of forming political parties.

Referendum on EU membership

As a remain voter, the political establishment has largely backed my soft, trusting, open, internationalist viewpoint for the last 20 years. Pro immigration, pro international aid, against capital punishment.

My view is that this left half the population disenfranchised. They tried voting UKIP, but because of our electoral system, that got no power. In the end, the only way they could express their views was by threatening to split the Tory vote, forcing a referendum, and taking that one opportunity to exert one binary digit of power.

This is backed up by Eric Kauffmann’s research, summarised by the BBC as “The link between Brexit and the death penalty“.

In short, voting leave is highly correlated with being nationalist on the most significant axis of political belief.

Situation the Labour party is in

The Labour party has leadership and unity troubles at the moment. Would this were the worst of its problems.

The real problem is that the Labour party have lost lots of seats to the SNP, will lose more due to the reduction in number of MPs to 600 (analysis), and are in danger from UKIP in the North of England (analysis).

The referendum result, combined with knowledge that the main axis of political grouping is internationalist / nationalist (not some meaningless left / right), show a significant realignment of UK politics. Any Labour strategy to form a Government needs to take account of this – blindly supporting, say, a second EU referendum without would be foolhardy.

My personal view, especially as a green party member, is that if the Labour party continues to have internal divisions it might be time for them to unify on a platform of implementing proportional representation. The tactic would then be to join a progressive alliance to win a majority in favour of PR, then safely split the party based on other divisions later.

Hacking democracy when not represented

The first past the post electoral system hangs over lots of drama in UK politics.

It makes it impossible to form new parties, meaning any political view opposed by Labour and the Conservative party gets no political representation.

I think of the desperate strategies people have to take in response to that as “hacks” on the political system. They aren’t signs of a functional democracy, which takes on board significantly held views.

Here are three such strategies that have at least somewhat worked:

  1. Form a regional party, and so get lots of seats like the SNP (they have the least votes per seat).
  2. Wait nearly 30 years of pain being under-represented in Parliament like the Lib Dems, until there is a hung Parliament, then join a coalition with a voting reform referendum as the prize.
  3. Form a party like UKIP representing a cluster of unrepresented beliefs on an important political axis, get a couple of MPs by defection (and none from later winning 12.6% of the vote), use that to threaten to split (the vote of) another party, thereby forcing a referendum on leaving the EU.

The Green party’s call for a progressive alliance to implement PR is a new variant of 2. Given the situation of UKIP, it is a bit more plausible now than when the Lib Dems last tried their variant of it.


Ideally, I’d like the research Chris and Tom did to be run again, so we can double check they’re still the axes in 2016.

Meanwhile, I’d like commonly used names for positions on these two axes. Just to improve the quality of political conversation, as it is pretty incomprehensible without such names.

What names would you use?

5 thoughts on “On finding political axes using maths

  1. On the rebranding side, the national front in France seem well aware of the national/international axis, and have been making many statements along that theme (possibly/probably to distance themselves from the traditional ‘extreme right’ political location which renders them unvotable for many people they are trying to woo)

    1. That’s interesting! I haven’t looked closely at how UKIP campaign in northern England, where they’ve had some success at swinging Labour voters to them. I assume they need to do something similar…

Leave a Reply

Your email address will not be published. Required fields are marked *