Labour polling: Tooth Decay, not Lego

A couple of years ago, 29 plus 6 was a popular sum among those interested in Labour’s likely election performance.

It represented what some regarded as a ‘floor’ in Labour support, made up of the 29% of the electorate who voted Labour in 2010, with the 6 representing the roughly a quarter of 2010 Liberal Democrat voters who shifted to Labour after the Coalition was formed.

These seemed, according to most research, to be even more certain to vote Labour than former Labour voters.

Think of this electoral coalition as being, effectively, me and Sunny Hundal. The argument went that this coalition, while it would benefit from a little augmentation from Tory converts, the young, or previously disillusioned non-voters, represented a relatively solid base, from which Labour could reliably build.

Today, with Labour hovering in the low thirties with most pollsters, we don’t hear that argument much any more.

So what went wrong?

The first thing to point out is that the argument was not a stupid one. It remains right. 2010 Labour voters overwhelmingly still vote Labour. A significant proportion of 2010 LibDems will vote Labour at the next election.

It’s just that the ‘overwhelmingly’ and ‘significant’ are a bit smaller than they were.

The problem is that Labour people thought of these groups as robust lego blocks, when in fact they were more like a row of teeth, subject to gradual, almost imperceptible decay.

To demonstrate this, I’ve been tracking Labour’s polling internals with YouGov.  Let’s compare the first half of December 2012, close to Labour’s peak support, with the same period last year and the current polling.1

What do we see?

Most obviously, a decline in Labour share.  Over the last two years Labour share of the vote has fallen 9.1 points. Interestingly, Labour’s lead has only fallen 9.7 points, indicating almost no Tory revival.


What’s happened to 2010 Labour voters?

Basically, Labour are holding on to a few less of them.

Slightly more former Labour voters now say they won’t vote at all, but this may be margin of error stuff.


Of those that will vote, Labour’s retention rate has fallen from 91.8% to 78.6%.


Who has benefitted?

The Tory share of Labour voters has increased a little, but the big jump has been in support for others.

14.9% of 2010 Labour voters now say they’ll vote for someone other than the big three. About half of these say they’ll vote UKIP, the rest backing the SNP or Greens.

What about the ‘Red Liberals’?

There has also been a small, but tangible, decline in Labour’s support among 2010 LibDems. In 2012, Labour averaged 37.7% of 2010 LibDems who said they’d vote. Now that number is 31.9%. That’s still a big wedge, by anyone’s standard, but it is a decline.


Perhaps surprisingly, There is no difference in Labour’s decline by gender or age (all about 9 points down).

However, over the last two years Labour’s rate of decline has been much higher among C2DE voters than among ABC1 voters.

Resist the temptation to conclude that this represents a loss of the Labour ‘core vote’, however.

YouGov’s final poll in 2010 had Labour on 31% among C2DEs. That number is around 36% today. This suggests that the C2DE voters who joined and left Labour in the interim were perhaps not ‘core’ supporters!

Indeed, since the General Election Labour’s ABC1 and C2DE support has increased by approximately the same proportion. Labour’s C2DE boost in 2012 now appears a little ‘Easy Come, Easy Go.”


By region, there’s good news and bad news. Labour has barely lost any support in London since 2012. PPCs in the metropolis can afford a smile. London is now nearly Labour’s most ‘core’ region (Food for thought for the more anti-metropolitan elite Labour thinkers, there).

However, Scotland is a different story, now being the only area of the UK where Labour is currently performing worse than in the 2010 election. This is an unusual time in Scottish politics, and there is little certainty that current polling will predict next May accurately. That said, given Labour’s outstanding performance in Scotland in 2010 (42% of the vote and no seat losses) it is likely Labour will need to focus resources on defending seats in Scotland as well as attempting to gain them elsewhere.


To describe how Labour’s vote is performing, throw out any mental imagery of blocks of voters, whether ‘2010 Lib Dems’ or ‘Working Class Core voters’. That’s not how politics is working now.

What is happening instead is a gradual decay of the Labour vote in several directions. The row of teeth that makes up Labour’s coalition is each a little less lustrously enameled. The teeth haven’t fallen out, but there are gaps and subtle cavities. Sunny and I are still there, but we are more lonely than we were.

As a result, the 29 plus 6 formula that made Labour strategists smile has become a 27 plus 5 sum that should make them frown.

This is perhaps why Labour is finding a response to decline so difficult to set out. It is not a straightforward task. To secure one group more firmly (say to retain C2DE labour voters) endangers the least convinced of another group (perhaps ABC1 LibDem converts). To attempt to do better in the Midlands might put London noses out of joint.

In other words, Labour strategists have a thankless task, as the data does not point to a single, easy solution.

The advice is simply appeal to more people, of all sorts. (Oh, and do better in Scotland).

That is, perhaps not a problem an election planner can solve. It is a strategy and direction issue, not a tactics and execution one.

The Labour campaign team do have one crucial advantage to console themselves with though.

Their Tory rivals have an even less attractive set of numbers to digest.


  1. Health warning: The current period, being conference and by-election related may well be distorted. Still, as we edge closer to the election, there will be few politically fallow periods to compare []

10 Responses to “Labour polling: Tooth Decay, not Lego”

  1. Newmania

    Wow imagine having to worry that much about what other people think !
    I once saw a map of the UK adjusted to render journey times as actual distance. It was a very odd shape , much fatter than tall and with weird bulges and distortions.
    Thats rather what you need to imagine when you look at needing , let us say , 5% more voters.
    The journey required for that 5% may be further than you have gone for the preceding 30% so the quoting of numbers alone may have the appearance of being much closer to a majority than in fact you are.
    My sense is of a Labour Party in its comfort zone which likes the discussion of numbers so as to avoid a discussion of how far they have to go.

    I continue to think that Labour have a real problem with Ed Milliband by the way and it may be worse when a real decision has to be made.

  2. Martin McG

    “…with Labour hovering in the low thirties with most pollsters…”

    There have been 22 polls since the start of October (according to UK Polling Report) – 4 polls have shown Labour under 33%.

    The poll of polls shows Labour on 34% – or within spitting distance of 29+6.

  3. Natasha

    I’d be interested to know how exactly you would recommend that Labour “do better” in Scotland, especially as for the first time since 1974 Scottish votes might actually make a difference to the overall result.

  4. Jon

    This is a real problem – it’s also a problem for the Tories, incidentally (I think what Matthew Parris rather inelegantly touched on in his column).

    But what are those ‘see-saw issues’ (for want of a better phrase) that put off metropolitans while exciting non-metropolitans, and vice-versa?

    Probably most likely to come to mind are issues like welfare immigration; Europe as well, though it is less salient.

    That doesn’t mean there aren’t areas where Labour could find unity if it wanted, though. A few ideas that command near-universal approval:

    It is something to be aware of, but it is by no means insurmountable; and probably less so for Labour than for the Tories trying to balance non-metropolitan conservatives and pro-business liberals. Labour is far from paralysed here.

  5. Josephine Bacon

    The exciting thing about the next election is that, with the collapse in the LibDem vote, the rise of UKIP and the alleged unpopularity of Ed Milliband in his own party (which may be something dreamed up by the media, and anyway the Labour front bench consists of lots of other people) the next election is totally unpredictable. Will UKIP win Rochester and Strood? Will their recent and possibly future success just be a flash in the pan or will they do well in the General Election? It makes the next General Election really exciting and I think that will bring voters of all persuasions out to vote, I suspect the percentage of voters will be higher than it has been for years.

  6. Sunny Hundal

    Hundal and Sen – the last men standing :)

    TBH I think some of this was inevitable, and the polling comparison from 2012 is flawed IMO because the economy was in a double-dip then, and the Tories were getting hammered over the Omnishambles Budget.

    I think the small bleeding is entirely expected given the economy has improved since (but it hasn’t been worse because its a superficial recovery).

    Many would say the best way to get a hearing among people who have left is to offer something big that they could get excited about.

    • hopisen

      Last of the mohicans, us!

      A couple of years ago those of us who said that Labour’s polling lead was soft, that we’d lose support when the economy recovered and the coalition wasn’t the only party having to set out how they’d handle problem (rather than just objecting to them) were told we didn’t understand the way the new progressive consensus was reshaping politics. Nice to know it was inevitable the whole time- if so, perhaps we should have tried to broaden our initial coalition a bit!

      On your last point, yes many would say that. What they neglect is that often their ‘something big’ is someone else’s ‘something scary and expensive’


Leave a Reply