The centrist fallacy

One thing I heard a few times at Liberal Democrat Conference was an assertion along the lines of ‘most people are centrists, therefore they’ll want to join us and vote for us if we just give them a chance’. This is normally backed up by some ‘recent polling’ that has shown people are generally centrist, and an earnest belief along the lines that they just need to be told the right words and then they’ll realise they’re just like us.

This is a belief I like to call the centrist fallacy because it rests on a number of assumptions about people, political behaviour and political parties that don’t apply in practice, even if some of them appear to have a basis in fact. Let me take you through some data in an effort to show you why.

First, ‘most people are centrist’ does actually have some basis in fact. Let’s take a look at the most recent British Election Study data which is generally recent, has a very large sample size and (most importantly for my purposes) is free to use and easy to obtain. It asks respondents the question “In politics people sometimes talk of left and right. Where would you place yourself on the following scale?” with a scale that runs from 0 (left) to 10 (right). And when we combine all their responses, we get the following graph:

Source: British Election Study

And yes, there’s a big peak in the ‘centrist’ position — 21.7% of people who give an answer put themselves at the dead centre position 5, with another 9.1% and 9.7% in positions 4 and 6, which means we can probably fairly say that around 40% of the electorate see themselves as some form of centrist. (For clarity, I’ve only included those who gave a definite answer. Just over 5,000 people — 16% of the total survey — gave a ‘don’t know’ answer)

The problem, however, with relying on this as some proof on an unmet demand for centrism is that merely asking people to place themselves on a scale isn’t doing any questioning of their actual political views. If you’re reading this post then it’s highly likely that you’re an anomaly, or at least above average for the UK, in how much you think about politics. Most people spend very little time thinking about politics or how they’d classify their own views on a left-right scale, so when they’re asked to place themselves on a scale like that, they’re quite often going to plump for something near the centre either because it’s a good way of answering without committing to anything, or because most people like to think of themselves as being generally moderate.

Luckily for us, though, the BES asks people a lot more questions than that. (I’m writing this post quickly so I haven’t had time to properly weight and organise the data so some of the numbers — especially on voter intention — below might be a bit confusing as they conflate 2015 and 2017 election data, but I think the broad points they illustrate are valid)

First off, obviously, there’s how people vote. The raw (unweighted) data in the BES gives us the following:
Conservatives 32.6%
Labour 27.1%
Liberal Democrats 7.2%
UKIP 3.4%
SNP 3.2%
Green 2.8%
Plaid Cymru 0.5%
Wouldn’t vote 4.9%
Don’t know 17.3%
Other 1%

Now, let’s look at the same voting question for those we can claim as some kind of centrist (scoring themselves at 4–6 on the left-right scale):
Conservatives 32.6%
Labour 19.9%
Liberal Democrats 10.7%
UKIP 3.4%
SNP 5%
Green 2.2%
Plaid Cymru 0.7%
Wouldn’t vote 3.6%
Don’t know 20.4%
Other 1.4%

So, self-positioned centrists are slightly more likely to vote Liberal Democrat, SNP or Plaid Cymru, but not excessively so, and they’re also more likely to answer ‘don’t know’ when asked how they’ll vote. (And if anything, the figure in there that most makes me want to delve deeper to examine it is the drop in Labour support rather than anything else)

The BES does allow us to look in a bit more depth at what people actually believe, rather than just relying on their self-placement for that. There are a number of questions on people’s opinions of things which include a set of ten questions that work quite well for finding positions on a two dimensional political map. There are five traditional ‘left-right’ questions asking respondents how much they agree or disagree with:

  • Government should redistribute income from the better off to those who are less well off
  • Big business takes advantage of ordinary people
  • Ordinary working people do not get their fair share of the nation’s wealth
  • There is one law for the rich and one for the poor
  • Management will always try to get the better of employees if it gets the chance

And five questions on a libertarian-authoritarian scale:

  • Young people today don’t have enough respect for traditional British values
  • For some crimes, the death penalty is the most appropriate sentence
  • Schools should teach children to obey authority
  • Censorship of films and magazines is necessary to uphold moral standards
  • People who break the law should be given stiffer sentences

The answers to all ten are on a five-point Likert scale (Strongly Disagree/Disagree/Neither Agree nor Disagree/Agree/Strongly Agree) and by combining the two sets of scales we can give every respondent a score from 0–10 on both a left-right and libertarian-authoritarian scale. So, someone answering ‘neither agree nor disagree’ to all five questions would end up with a score of 5, a 0 would indicate a full left or libertarian position, and a 10 a full right or authoritarian one.

This is how the population as a whole lines up on the authoritarian-libertarian axis (remember 0 is most libertarian, 10 most authoritarian)

And on the left-right scale (0 is left, 10 is right):

(On both graphs, the blue bars are the actual numbers of responses at that point, the black line is an indication of a normal distribution of data)

In both cases, we can see that the population as a whole isn’t arranged around what we might think is a moderate position. Voters’ views tend towards the ‘authoritarian’ and ‘left’ sides rather than clustering around the middle. (For more on this and how left-authoritarians are perhaps the largest and most unrepresented group in British politics, see the work of Paula Surridge on values and voting).

So, what we can already see is that despite people’s self-description of themselves politically clustering around the middle, their actual political views don’t. So, how do our centrists compare with the rest of the population on these values? Like so:

If anything, self-described centrists are slightly more authoritarian than the population as a whole (an average position of 6.67 compared to 6.32) and very marginally more left-wing (3.09 compared to 3.12) and for those of you who really want to delve into the data, then I’ve posted the crosstabs as Excel spreadsheets: one for the data about centrists and one for the general data. These show how the two scales interact with each other and show that there are more people in the left-authoritarian quadrant of the diagram than any other, and the right-libertarian one is rather empty.

Another example that’s relevant to this post is to look at those people who have liberal views and see how they identify themselves. Specifically in terms of the Liberal Democrat appeal to voters, which emphasises the party’s liberalism — it’s there in the name, after all — are people with liberal values likely to see themselves as centrists. Taking ‘liberal’ as anyone with a score of 4 or less on the liberal-authoritarian BES scale and looking at how they identify themselves on the left-right scale, we find this:

On average, they place themselves at 3.13, with almost half (45.4%) going for 2 or 3 and nearly two-thirds (64.7%) volunteering a position of less than 3. This also applies to their actual views on left-right issues:

Here, they have an average position of 2.43, which is further left that the 3.13 we found for the general population. Their voting patterns are also markedly different from the general population: 47.1% for Labour, 15.9% for the Liberal Democrats, 8% for the Conservatives, 7.2% for the SNP, 6.6% for the Greens, 1.2% for Plaid Cymru and 0.3% for UKIP. They’re also less likely to not vote (2.2%) or say they don’t know (10.3%).

So, this post has been a bit quick and dirty and I need to go into a lot more depth with some of these figures in the future but I think they’ve illustrated a few points that are useful in assessing any claims of the ‘lots of people are centrist, therefore lots of people will vote for us if we say we’re centrist’:

  • ‘Centrist’ means a lot of different things to people, and just because someone puts themselves in the middle, it doesn’t mean they agree with you when you claim to be there too.
  • People who think of themselves as being in the centre have views that tend away from the moderate and more towards the economic left and the authoritarian right.
  • These views are the effective centre of views in Britain, but they’re not really at the centre of political debate and in conjunction they tend to be the most unrepresented.
  • People with liberal views tend to both see themselves as being on the left and have more left-wing economic views. They do not tend to see themselves as centrists and are somewhat outside of the mainstream of opinion.

All those are somewhat tentative, and I need to work on the data much more before being rock-solid on them, but I think they’re in line with other current research on BES data. However, if you do spot any major errors in anything I’ve said here, then do please let me know!

I’d like to do a bit more sometime to see how some of this lines up with views on the parties and their leaders, as well as some of the other views and opinions sections that the BES asks about, but Paula Surridge’s work also covers a lot of that and if you’ve read this far, you’ll probably find it useful and interesting.