Whose votes have we lost? And why?

It has undeniably been a rubbish couple of months for Labour in the national polls – though not in the London polling where Ken has now taken a lead neither in two polls, nor in real elections where we gained council seats from the Lib Dems in Redcar and St Albans last week.

Initially the decline in Labour’s poll lead was because of the populist appeal of Cameron’s EU “veto”. This then caused us to get wobbly and speculate about our woes, which created a downward spiral. More recently the greater emphasis placed on deficit reduction by the two Eds has seen a further dip but there is no evidence to tell us whether we lost support because people don’t like the more austere (or more realistic, depending on your view) policy, or don’t like Len McCluskey and his reaction to it.

Rather than the causes, I wanted to try to understand whose votes we have lost, as that seems a basic requirement of winning them back.

So I took a look at the dataset of both this Sunday’s YouGov poll which represents a low point for us, and the same company’s poll of December 2nd which represented a high point.

On December 2nd Labour was on 43%, and is now on 36% (-7%). The Tories were on 35% and are now on 41% (+6%). The Lib Dems are flat lining: 9% now and then. Other parties are on 14% (up 1% from 13%).

A partial explanation is some churn from Labour to minor parties and then from them to the Tories. UKIP are down 1% from 6% to 5%,
presumably going to the Tories. The Greens are up 1% from 2% to 3%. The SNP and Plaid Cymru are also up 1% from 3% to 4%. So 5% of the Tory lift doesn’t come from UKIP it looks like it comes from us. And 5% of our drop hasn’t gone to the Greens or SNP or Plaid Cymru it has gone to the Tories. Unless there has been some massive change in relative propensity to vote, it looks like there has been a straight swap from Labour to Tory of 5%. This loss of votes to our right suggests that being too firm on economic prudence is not the problem.

Where has the loss occurred? Let’s look first of all at the way the sample voted in 2010. The Tories are now retaining 91% of their 2010
voters, compared to 82% in December. So their European hardline has hardened up their core vote. We have only managed to attract 2% of Tory General Election voters.

The Lib Dems are also holding onto a bit more of their 2010 vote (though still only 40% of it, compared to 32% in December). Labour’s
share of 2010 Lib Dems has dropped from 43% to 31%. I don’t fully understand the maths of that, as the overall Lib Dem vote is static,
but it suggests to me we need to renew our message to 2010 Lib Dems that their party has betrayed them, otherwise the Lib Dem vote might drift up at our expense – other pollsters suggest this is already slowly happening.

Labour is now retaining 85% of its 2010 voters, down from 90% in December. Virtually no Labour voters have switched to Lib Dem (1%) but 5% have gone Tory (up from 2%), and 8% have gone to other parties (up from 6%). It’s extraordinary that at a time when we are still scoring 7% higher than in the last General Election we have lost 5% of our General Election voters to the Tories. It’s also very dangerous – those people are more likely to live in marginal seats than defectors to the Greens, SNP and Plaid, and a lost vote to the Tories has a double impact in a Lab vs Con marginal seat, boosting their vote as well as decreasing ours. Since December we seem to have lost more of our 2010 voters, who were “core” enough to back us in our hour of need, to the Tories than to the small parties to our left. This effect has been masked in the headline figures by the continued large slice of the Lib Dem 2010 vote that we are attracting.

In terms of gender we are down 4% amongst women but 8% amongst men, so we are suffering a differential hit amongst male voters.

On age, we have taken a massive 18% drop amongst the volatile and non-tribal18-24 age group. There the Tories are up 12%, the Lib Dems down 8% and the minor parties up 14%.

With 25-39 year olds we are down 8%, the Tories are only up 2%, the Lib Dems are up 4% and the minor parties are up 1%.

With 40-59 year olds we are down just 2%, the Tories are up 6%, the Lib Dems are down 1% and the minor parties are down 2%.

Amongst the over 60s, with their very high propensity to vote, we are down 5%, the Tories are up 7%, the Lib Dems are up 1% and the minor parties are down 2% (presumably UKIP pensioners going back to the Tories).

Historically Labour’s vote by age has looked like a ski slope – the younger you are the more Labour. Now it looks like a bell curve, with
weakness at the youngest and oldest ends (though it is important to note there are lots fewer 18-24 year olds than over 60s, and they have an incredibly low turnout rate).

By social class unfortunately YouGov only does a crude spilt into ABC1s and C2DEs whereas other pollsters show you the splits for AB,
C1, C2 and DE.

Amongst middle class ABC1s (well over half the electorate) Labour has dropped 7% from 40 to 33%. The Tories are up 5% to 44% and the Lib Dems are down 1% to 10%.

Amongst working class C2DE voters Labour has dropped Labour 6% to 41%. The Tories are up 8% to 37% and the Lib Dems are up 1% to 7%. In this segment there has been a big drop in UKIP support from 9% to 5%.

By region, YouGov has Labour actually up 2% to 44% in London, down 4% to 28% in the rest of the South (where we were already unpopular), down 4% to 37% in the Midlands and Wales, down 13% to 47% in the North, and down 13% to 24% in Scotland.

These splits all come with the health warning that small sample size once you get to subsets of the main poll could cause dodgy results.
If they are accurate though it looks like we have lost votes everywhere except London (a coattails effect from the intensive fares
campaigning the Ken campaign are doing?), but particularly amongst men, young people and northerners. I’m off to try to figure out how we get them back.

Ideas gratefully received in the comments section.