Just one poll published tonight (so far at any rate):
-
YouGov in the Sun … CON 34%(-1), LAB 28%(+1), LIB DEM 30%(+2)
So much for the charge – led by its sister-paper – that the Lib Dem bubble had burst. The poll suggests the party is continuing to hover at the 30% mark, and continuing to push Labour into third place.
There was incidentally one other poll published we omitted to report last night: BPIX (the firm which refuses to comply with polling industry standards) showed CON 34%(+3), LAB 26%(-2), LIB DEM 30%(-2).
It does seems as if the party is heading into the final full week of the election campaign in its healthiest poll position in modern history. Interestingly, it’s usually in the final 10 days of the campaign that the Lib Dems begin to put on a few points in the polls. Now the TV debate and ‘Cleggmania’ may mean that’s already been priced in to the polls. We shall see. Anthony Wells’ UK Polling Report ‘poll of polls’ continues to show:
-
CON 34%, LAB 27%, LIB DEM 29%
Mori’s Bob Worcester has an interesting article in today’s Observer looking at where the new Lib Dem voters have come from based on his firm’s polling in Labour-Tory battleground marginals. Here are the findings:
In these battleground seats, quite a lot more from women (+16%) than men (+9%), more from middle-class voters (+15%) than working-class (+9%), and more from the young, 18-34s (+16%), than middle-aged, 35-54 (+13%), or from older people, 55+ (+9%).
The retention rate is also a factor in the rise of the Lib Dems. From the data before the prime minister called the election and after, the Tory retention rate fell eight points, from 77% of those who recalled voting Conservative in 2005; Labour held its own and a bit more, from 52% before, and 55% this past week. The Liberal Democrats, however, bounced from a lowly 43% up 18 points to 61%.
This volcanic activity seems to be persisting; voters and now bookies signal a hung Parliament. That means Clegg, while not becoming politically king, could certainly become a kingmaker.



4 Comments
Your reporting of the movements differs from the Sun’s Political Editor, who says Cameron “stays firm” on 34, you say its down one, Labour “slump”… but you say they are up one, and you say Liberal Democrats are up two, yes, The Sun acknowledges that on 30 they are up…but only one point. Will be interesting to see who is “up” and who is “slumping” this time tomorrow.
On the subject of polls, the American poll-watchers at fivethirtyeight.com have come up with a very plausible alternative to uniform national swing. The results are not pretty for Labour, and something of an improvement for us by comparison: http://www.fivethirtyeight.com/2010/04/labour-danger-uniform-swing.html
There’s quite a few more sophisticated prediction algorithms been proposed over the years; I don’t think anybody really believes UNS is very accurate. The problem is that they all rely on measuring things which the pollsters don’t measure. We stick with UNS because, in a typical UK election (which shows only a few percent movement in any direction) it’s reasonably accurate, and because it is the only thing which we can generate daily forecasts with.
If the pollsters wanted to be really useful, what we need is large, properly weighted samples of voting intention for each constituency.
In the absence of Pollwatch Day 21, here are today’s polls:
YouGov/Sun CON 33%(-1), LAB 28%(nc), LDEM 29%(-1)
Opinium/Express CON 34%(+2), LAB 25%(-1), LDEM 28%(-1)
ICM/Guardian CON 33%(-2), LAB 28%(+2), LDEM 30%(-1)
ComRes/ITV/Independent CON 32%(-2), LAB 28%(nc), LDEM 31%(+2)
It’s interesting that (if my sums are correct) the average of today’s polls is identical with Anthony Wells’s more sophisticated average of recent polls:
Con: 33 Lab: 27 LD: 30
And his national projection based on a uniform national swing has Labour 60+ and the Tories 70+ short of a majority, with the Lib Dems on just over 100 seats. After Peter Kellner’s analysis yesterday, there’s less reason to think that the Tories will do all that much better than that.