This weekend we got three new polls, Opinium polled Wednesday to Thursday for the Observer, YouGov/Times (Friday to Saturday) and YouGov/Sun (Saturday to Sunday).
Opinium/Observer: CON 36 (+3) LAB 34 (+1) LIB 7 (=) UKIP 11 (-3) GRN 6 (-1) Fieldwork 8th-9th #GE2015
— NumbrCrunchrPolitics (@NCPoliticsUK) April 11, 2015
YouGov/Times CON 34 (+1) LAB 34 (-1) LIB 7 (-1) UKIP 13 (=) GRN 6 (+1) Fieldwork 10th-11th #GE2015
— NumbrCrunchrPolitics (@NCPoliticsUK) April 11, 2015
YouGov/Sun CON 33 (-1) LAB 36 (+2) LIB 7 (=) UKIP 13 (=) GRN 5 (-1) Fieldwork 11th-12th Daily roundup to follow #GE2015
— NumbrCrunchrPolitics (@NCPoliticsUK) April 12, 2015
Of the pollsters to have surveyed entirely after the non-dom row started, Survation showed a two-point shift to Labour, Opinium a two-point shift of the Conservatives, Populus no change in the lead and YouGov's multiple polls have shifted in both directions, most recently showing a three-point Labour lead. So cutting through the statistical noise, things seem to have shifted to Labour a bit since the campaign started, but not to the extent that might have seemed plausible on Thursday. Perhaps more important is that it has shifted sentiment and the (much-narrated) narrative Labour’s way.
Some polling averages (and associated forecasts) have shifted violently in response to this week’s polling, but others have been more stable. The reason for this happening now is purely mathematical – some averages take the most recent poll from each pollster, others average them up over a longer period. As we’re now in campaign mode, with polling becoming much more frequent, the former can shift quite sharply when a number of polls are released at the same time.
There was also a report in the Mail on Sunday of another private poll (this time in Rochester & Strood) suggesting that UKIP’s Mark Reckless is “narrowly behind” his Conservative challenger Kelly Tolhurst. Because the numbers haven’t (yet) made it into the public domain, the pollster (even if, as is likely, it’s a BPC member) is not required to publish the details.
It’s not too surprising that UKIP trail in this seat, given what some of the models have been predicting. The problem is that defections and by-elections are troublesome for forecasters to account for, and this seat has had both.
. @NCPoliticsUK best guess at numbers: Con 39; UKIP 26; Lab 26; LDem 3 @Election4castUK http://t.co/RiTou6v56w
— Chris Hanretty (@chrishanretty) April 12, 2015
Tomorrow we should get Populus, Ashcroft National, ICM/Guardian and YouGov/Sun, in roughly that order. It’s also the day of Labour’s Manifesto launch.
