by Stephen Tall on February 11, 2014
A fascinating piece of polling research from YouGov’s Peter Kellner in today’s Guardian, looking at how votes have churned since the 2010 general election.
My working assumption looking at the headline poll ratings has been that there’s been relatively little movement between Labour and the Conservatives, with most of the movement from the Lib Dems to Labour and from the Tories to Ukip. YouGov’s research shows how simplistic that assumption about votes lost/gained in the last four years is:
Three quick points drawn from this table:
- What immediately jumps out is how many people who voted Lib Dem in 2010 now say they won’t do so – just 1.8 million voters (27%) have stayed ‘loyal’ to the party, with fully 5 million of our 2010 voters currently saying they won’t vote Lib Dem.
- However, of those 5 million, just 1.8 million moved from Lib Dem to Labour. (Yes, I know ‘just’ isn’t the right word really… but, still, I’d have thought that far more than one-third of our lost voters would’ve switched to Labour.) To put in in perspective, half a million 2010 Lib Dem voters have moved to Ukip.
- We have actually gained a significant number of new voters since 2010 – one-quarter of our current support (600k out of 2.4m) is voters who have switched from other parties or who didn’t vote in 2010.
Probably the most important point for Lib Dems to note is this – the importance of winning back the ‘Don’t knows’:
One of the things that dents the Lib Dems’ vote share, but could be rectified, is the proportion they have lost to “don’t knows”: more than one in five, compared with one in eight Conservatives and one in ten Labour voters. One thing all the parties will be trying to do is revive the loyalties of these doubters; proportionately, the Lib Dems have the greatest opportunity. In a number of the seats they are defending, their success winning back these voters will make the difference between victory and defeat.
That is nearly always true of Lib Dems. We have the smallest core vote of the three main parties, which is one of the reasons why the party’s vote tends to dip between elections. This is sometimes put down, a bit simplistically, to us being starved of media attention. While I think that’s a part of the explanation, more important is that Lib Dem voters are more likely to ‘firm up’ as the election approaches, especially in areas the party is targeting its campaigning.
Kellner’s final paragraph is interesting:
For the parties, the lesson is: don’t target your messages too narrowly. Voters will make up their minds, and decide whether to shift or stay put, in a host of ways, some of which you may well regard as bizarre or irrational (but they will consider perfectly sensible). Don’t take any given group for granted – or assume that every member is a lost cause. And in your target seats, take care to listen to as many voters as you can; and listen hard without imposing your own agenda, for many of them will tell you things about their choice of candidate and party that you won’t be expecting.
This is undoubtedly true, though I suspect its main targets are those Conservatives narrowly focusing on winning back Ukippers – just 1.4m of the Tories’ 4.2m lost 2010 voters say they’ll vote Ukip – or those Labourites thinking they can win simply by holding onto 2010 Lib Dems.
As I’ve highlighted before – most recently on Friday here – the Lib Dem strategy for 2015 is much more targeted than in previous elections, focused particularly on the Lib Dem ‘market’ of those 25% of voters currently saying they would vote for us (c.10%) or who would consider voting for us (c.15%) – with the latter c.15% drawn roughly equally from persuadable Conservative/Labour voters and from persuadable don’t knows. That makes sense, especially for what we all know will be a defensive election. What Kellner’s analysis demonstrates is that the 2015 election will be about a whole lot more than simply trying to win back 2015 voters from Labour.
* Methodology note from Peter Kellner: “I have combined the results from all the voting intention polls that YouGov conducted in January. This gives us a total sample of more than 37,000 people. I have then compared the way they voted in 2010 with the way they say they would vote now. (For around two-thirds of the 37,000, their 2010 vote is what they told us at the time; the other one-third have joined our panel since then and we have used their recalled vote. Some of this one-third may have misremembered how they voted, but the impact on this analysis should be small.) … I have made a rough estimate of the number of supporters of each party who are likely to have died since the last election.”
* Stephen Tall is Co-Editor of Liberal Democrat Voice, and editor of the 2013 publication, The Coalition and Beyond: Liberal Reforms for the Decade Ahead. He is also a Research Associate for the liberal think-tank CentreForum and writes at his own site, The Collected Stephen Tall.