Tonight’s YouGov poll for the Sun has topline figures of CON 37%(nc), LAB 34%(+2), LDEM 17%(nc). Leaving aside that 2 pointer for the Sunday Times, which in hindsight was probably something of a rogue, it’s the lowest Conservative lead YouGov have shown since October 2007 and the election-that-never-was.

The movement is well within the margin of error. What is actually more significant is that the Conservative lead has been below 6 points in every one of YouGov’s polls this week. For the first two weeks of their daily polling, it seemed quite evident that the lead was around about 6 points and the daily polls were just random sample variation around it. This week the lead seems to have narrowed.

UPDATE: There is also a new Angus Reid poll for Political Betting out tonight, which shows a completely different picture. There the topline figures are CON 39%(+1), LAB 26%(nc), LDEM 18%(-1). Others are presumably somewhere around 17.

Clearly there is a methodological difference here – while the levels of Conservative support are within the same ball park, Labour’s figure down in the mid 20s stands out in contrast to most other companies except for newcomers Opinium. Part of this will be down to Angus Reid’s politicial weighting, but that’s probably not enough to explain all of it. The other difference, as I mentioned earlier in the week, is on support for the minor parties. The newer online companies (Angus Reid, Opinium and Harris) are showing much higher levels of support, up in the region of 17 or so, than the more established companies like ICM, Populus, YouGov and MORI, who all have them down near 11 or 12.

401 Comments »
Filed under: AngusReid, Voting Intention, YouGov, YouGov Daily Polls

Polls in northern Ireland are very rare creatures, not least because they have a rather poor record. There is a strong tendency for them to under-report the proportion of people voting for parties at the more hardline ends of the political spectrum, and over-report those in the centre.

I am not aware of any recent published polls, however, in the run up to the vote on the devolution of policing powers this week two polls were released, one commissioned by the Northern Ireland Office, the other by the Office of the First Minister and Deputy First Minister (OFMDFM). No figures for party support were officially released, but Mark Devenport the BBC’s Northern Ireland political editor has them “unofficially”.

These come with HUGE, TOWERING caveats. We don’t know if they are true, we don’t know if what the companies involved did to get them, and the record of polling in Northern Ireland really isn’t great to begin with. For what it’s worth though:

Red Circle/OFMDFM
TUV 2%
DUP 30%
UCUNF 19%
All 11%
SDLP 19%
SF 16%
Others 4%

Northern Ireland Office
TUV 1%
DUP 26%
UCUNF 16%
All 8%
SDLP 21%
SF 17%
Others 9%

I would not put too much trust in them. Sinn Fein look very low indeed, the Alliance very high. Jim Allister’s Traditional Unionist Voice is almost completely absent despite managing 13% in the European elections. However, unless someone commissions a proper poll in Northern Ireland, it is all we have.

UPDATE: I haven’t managed to track down the proper poll Ruairi mentions in the comments, but I have found another proper voting intention poll commissioned by the Belfast Telegraph. A face-to-face quota poll by Inform Communications, weighted by age, gender, community background and class (and it does at least have Sinn Fein ahead of the SDLP). NB – it’s Assembly election voting intention, not Westminster.

Inform/Belfast Telegraph (5th-9th Feb)
TUV 6%
DUP 18%
UCUNF 15%
Alliance 7%
SDLP 13%
Sinn Fein 21%
Others 6%

43 Comments »
Filed under: Northern Ireland

There is a Populus poll of marginal seats in tomorrow’s Times, Peter Riddell’s commentary is here. The headline figures in the poll is CON 38%, LAB 38% and this has resulted in a flurry of excitement from twitterers, but what it means depends upon which seats were polled.

Populus’s poll sampled Labour held Conservative targets from 50-149, so excluding the 50 Labour-v-Conservative seats with the smallest majorities. By Peter Riddell’s calculations, these seats had shares of the vote of CON 31.4%, LAB 45.3% in 2005. On that basis this poll represents a swing of about 6.7%, just short of the sort of figures the Conservatives would need for an overall majority.

Of course, we haven’t had a national Populus poll for a month so we can’t say for certain if this poll suggests a larger lead in the marginals than at a national level. For the record though, the swing this poll suggests in the marginals is the equivalent of a 10 point lead nationally, a larger lead than most polls from other companies have been showing in recent weeks.

366 Comments »
Filed under: Populus