There are unusual findings in a MORI poll in the Sun (presumably MORI’s monthly political monitor). The topline figures are CON 30%(-4), LAB 30%(-9), LDEM 25%(+6), Others 15%. The Sun does not report a figure for the BNP, though it may turn up on the MORI website. The 15% score for the “others” could mean that there been an increase for the BNP in this poll, and indeed the Sun’s report does suggest this.

The change in the Labour vote is obviously particularly extreme for a single month. MORI’s polls are far more erratic than any of the other pollsters. Since the general election the average change in a party’s level of support in MORI’s monthly polls is 3.2 points, compared to 2 points for Populus (who have have changes to their methodology), 1.7 for YouGov and 1.5 for ICM. MORI’s volatility is probably because they do not use any form of political weighting on their samples, weighting which serves to dampen down sample error for the other three pollsters.

While the MORI changes may be quite extreme, the movements do back up movements seen in the most recent YouGov and ICM polls. The BNP are up, Labour and Conservatives are down, and the Liberal Democrats are sharply up. I suspect that there are two separate processes going on here – a rise in BNP support after the publicity they received from Margaret Hodge’s comments (presumably at the expense of the Tories), and secondly a sharp fall in Labour support, for whom the Liberal Democrats are the clear beneficiaries.

UPDATE: The full figures are up on MORI’s website here. The BNP are in fact on 3%, so there is a slight rise but not to the extent reported by YouGov – once again this is not unexpected, people will be less willing to admit to supporting the BNP to an interviewer, as opposed to a computer screen. The poll was carried out between the 20th and 22nd of April, so would have come prior to Patricia Hewitt’s comments on the NHS and, obviously, the revelation that foreign prisoners were not being deported on release.


YouGov have carried out an experiment for Sky News, asking two groups of people if they agreed with a list of BNP policies. The first group were not told they were the policies of the BNP, the second group were. Unsurprisingly, it found that public perceptions of the BNP drove down support for policies when they were identified as being from the BNP. On average 6% fewer people supported a policy when it was attributed to the BNP, and 6% more people opposed it.

The survey found strong support for BNP policies on accepting fewer asylum seekers (supported by 77%, or 74% when associated with the BNP) and giving priority to British families in allocating council housing (83% support, or 77% when associated with the BNP). There was overwhelming support for making criminals serve their full sentences (91% and 87% support). The question of whether all immigration should be halted provided the strongest contrast – 59% supported the policy, but when it was presented as a BNP policy support fell to 48%.

There was a more mixed view on Europe – 35% of people supported withdrawal from the European Union, with 36% opposed (the 1 point gap grew to a 10 point gap once the policy was associated with the BNP).

There was however strong opposition to more extreme policies on race – asked if they agreed that non-white people were inherently “less British”, only 16% of people agreed, with 68% opposed. When identified as a BNP support dropped to 11%, opposition grew to 76%. A majority also opposed the government encouraging immigrants to leave Britain (52% opposed, rising to 58% opposition when associated with the BNP).

Finally, YouGov asked people if they would seriously consider voting for a party that supported all these policies. In the unattributed group, 37% said yes, 48% said no. In the attributed group the figure is lower, 20% say they would seriously consider voting for the BNP (a figure comparable to the ICM survey in JRRT report that found 18% of people saying they might vote BNP), 66% said no.

What does this tell us? Firstly some of the BNP’s policies – particularly on crime and prioritising housing – are very popular indeed. Despite that a majority of people would not consider voting for a party that allied these policies with the BNP’s stance on race and repatriation. If you actually mention the BNP by name, with all the negative connetations that come with it, the proportion of people willing to support it falls even further.


-->

ICM’s monthly poll for the Guardian (a week late, presumably to avoid the Easter bank holidays which make it tricky to get a decent sample) shows a sharp fall for Labour and an advance for the Liberal Democrats. The topline figures, with changes since ICM’s last voting intention poll at the beginning of April, are CON 34%(-1), LAB 32%(-3), LDEM 24%(+3).

Labour’s level of support is one of the lowest recorded in any poll this Parliament (the other being the MORI poll which put the Conservatives 9 points ahead, and which in hindsight was almost certainly a rogue), the Liberal Democrats’ one of their highest. Using their present weighting regime ICM tend to give higher ratings to the Liberal Democrats than any of the other pollsters, when other companies had them falling to the mid-teens after Charles Kennedy’s resignation ICM never reported Lib Dem support below 19%, but even by ICM’s standards this represents a significant advance for them.

There has been little coverage of the Lib Dems in recent weeks, suggesting that their advance is due more to them being the public’s “not Labour” party of choice, gaining support as Labour stumbles. The fall in Labour’s score could be a result of the arrest of Des Smith finally giving some salience to the issue of party funding, or perhaps the first effects of concern over NHS funding. Another poll by MORI for the Sun shows that Labour have only a 3 point advantage over the Tories as the party that “would be most effective in getting good value for the public money it spends”. Given that Labour normally has a solid advantage in any poll asking about public services this poll does raise some questions over whether the question of NHS funding has started to bite.

Of course, as with the vast majority of polls, the changes are within the margin of error and, unless other polls reflect the same trends, it could turn out to be entirely meaningless.

The BNP are up to 2% in the poll – ICM do not normally report the level of BNP support in their monthly polls, but if their previously levels of support were similar to those found by YouGov this is likely to be a slight rise – though obviously not on the scale reported by YouGov last week. This isn’t necessarily a contradiction, the spiral of silence effect that has historically led to Conservative voters being unwilling to admit their voting intention is likely to be far, far stronger for a party with an image like the BNP’s. YouGov polls are self-completed by respondents, which makes respondents more willing to admit to unfashionable or socially unacceptable answers. While ICM adjust for the spiral of silence, their method works on the assumption that some people too embarrassed to admit how they’ll vote will vote in the same way they did last time – it can’t account for people too embarrassed to admit to a new voting intention.


A BPIX survey in the Mail on Sunday has the first signs of how Labour’s “Dave the Chameleon” PPB went down with the public (for those who missed it the ad is viewable here). A straight out attack ad aimed at David Cameron, there has been some debate over whether the advert will backfire, either through being overly negative, through underlining that the Conservatives are changing, or just because the little Dave the Chameleon character is too cute.

The survey suggests that only 1 in 20 people said the advert had made them less likely to vote Tory, while 1 in 7 said it made them more likely to vote Tory. The chameleon itself was just too nice – over a third of people said it was likeable – and the changes it made against David Cameron were too applicable to Tony Blair himself. While 49% of people thought that David Cameron would indeed “do and say anything to win votes”, 68% said it was true of Tony Blair.

Finally BPIX asked respondents which animal they thought politicians most resembles. While this is a valid technique in focus groups as a way of exploring why people think politicians are like particular animals, I’m wary of how useful it is in a quantative poll like this, especially when people are given a list of animals to chose from. Still, it does suggest that the chameleon tag hasn’t stuck on Cameron – only 11% said he most reminded them of a chameleon, coming equal with labrador.


Treat with caution

The Sunday Mirror this morning claims to have a poll showing that 45.5% of people in Barking will vote BNP in the local elections – I would treat it with a buckload of salt. It appears that the poll was conducted by the Sunday Mirror itself, rather than by a reputable polling company. This raises of the questions of whether the poll was properly sampled (the article says it was face to face – according to proper quota samples, or just random people off the streer?), whether it was weighted at all, how professional the interviewers were, how balanced the questions were, how was turnout factored into what will probably be a very low turnout election, etc, etc. Add in the fact that the BNP are only contesting 7 of Dagenham & Barking’s 17 wards anyway and you have even more of a problem.

Local newspapers often do the same sort of straw polls prior to elections and are almost always hideously wrong. The chances are this is the same thing (though that said, it could potentially have been carried out by a reputable pollster who the Sunday Mirror has failed to credit – in which case the polls figures should at least be accurate, though you still have the question of dealing with the wards with no BNP candidate).