Sunday’s News of the World carries a new ICM poll of 192 Labour held marginal constituencies where the Conservatives are in second place. The poll shows a swing of 11.5% in these seats, which according to ICM would equate to 164 Conservative gains and a majority of 78 seats. ICM did a similar poll back in April, and found a 9% swing in these seats – though they are not directly comparable since that poll covered only 145 seats.

Theoretically if there was an 11.5% swing across the whole country it would equate to a 20 point Conservative lead in the polls, though it is more likely that there is just a larger swing in these seats (we know, for example, from the PoliticsHome poll of different groups of marginal seats that there is a lower Labour to Conservative swing in Liberal Democrat seats) so in reality we can’t extrapolate a national vote share from this poll.

Back in April ICM asked people in Lab/Con marginal seats to say whether people rated Cameron or Brown more highly on various things: the only issue where Brown lead was dealing with the credit crunch. They gave the same list of things today, and nothing has changed. People in these marginal seats still trust Gordon Brown more than David Cameron to deal with the credit crunch by a margin of 43% to 35%. Unfortunately for Brown, they still rate David Cameron more highly everywhere else: on dealing with the rising cost of living, setting taxes, fighting terrorism, moderning the NHS, schools, immigration and generally being Prime Minister.

There was, however, strong support for Gordon Brown remaining Prime Minister for now. 58% wanted him to carry on, with only 36% saying he should step down. Amongst Labour supporters in these seats only 15% now want him to go.

ICM also asked about some of the policies that were announced at the party conferences, without identifying them with any particular party. The Conservative pledge to freeze council tax for two years was supported by 71% of people in these seats, the Labour promise to extend subsidized child care to two year olds was supported by 62%, and the Liberal Democrat promise to cut the rate of income tax but cutting spending by £20 billion was supported by 51%.

22 Responses to “ICM poll of marginal seats”

  1. Although polls in marginal seats can be regarded as a reasonable guide, it is woth remembering that we are still in mid-term territory. More cautious voters begin to wonder whether to stick to the devil they know come election time. DC has never been subjected to the rigour of an intense election campaign. Will those on the right of his party remain silent about their determination to win tax cuts? He may be found wanting when under the glare of unremitting questioning. Matthew D’Ancona, editor of the right-leaning Spectator, thinks that, in private, Tories will be worried about the return of Peter Mandelson, knowing his skills in developing effective election tactics. We have seen oppositions fall at the final fence before, much to their chagrin.

  2. I wonder if this is why Geoff Hoon wasn’t sent off to Europe in the latest cabinet reshuffle? Of course, Hoon’s Ashfield constituency is hardly a marginal – it has been a Labour stronghold for many years – but a recent poll suggested it would be lost to the Tories if an election were called now.

    A tricky by-election in the constituency coming just after an equally awkward one in Glenrothes might have brought the Labour leadership issue to the fore again.

  3. Does anyone know if this predicted Conservatives majority of 78 includes gains from the LibDems?

    It looks like they have just added present Conservatives MPs and the predicted 164 gains from Labour together.

  4. Richard, I think it is just about impossible to determine that from such a poll. As it only concerns Labour held seats obviously the Conservatives wont be gaining any of these from the LibDems, so we have to extrapolate. As Anthony suggests there are few of these marginal constituency polls and they are not consistent in terms of which marginals they choose… so we’re into guesswork. Best guess – probably some given the net swing in popularity from LibDem to Conservative based on other national polls.

    Mind you the majority suggested here sounds like a fair guess
    for the 2010 result.

    Oh and looks like tax is going to be an issue next time. Well, isn’t always?

  5. Incidentally AW’s list of the top 200 Conservative targets includes only 155 Labour held constituencies (with 41 LibDem, 3 SNP and 1 Ind).

    Assuming that Conservative gains from Labour followed the actual list of marginality (which would be unlikely to happen in reality) then every single Labour constituency on AW’s list would be gained by the Conservatives plus 9 others.

  6. >John @ 7.16pm

    “He may be found wanting when under the glare of unremitting questioning”.

    Cameron has faced some pretty unremitting question in the past and seemed to hold up pretty well. I’m not sure if he is the only party leader to do so, but last year he allowed himself t be grilled for about 3/4 hour by a pretty tough Newsnight panel of four interviewers, and coped pretty well I thought.


    There is a feeling on that whenever Cameron is featured heavily in the media for several weeks, his polling always improves.

  7. When ICM release a poll like this, are they under no obligation to publish which constituencies the poll covered?

  8. “When ICM release a poll like this, are they under no obligation to publish which constituencies the poll covered?”

    That would be very helpful!

    Keith Jenner pointed out on that the poll seems to be only in Labour marginals where the Conservatives were second. That would be presumabley discount those consituencies on AW’s list where the Conservatives were a close third.

    Taking account all factors the poll suggests that Labour is on course to lose around 200 seats.

  9. Gooey Blob,

    Precisely. It would seem that Gordon Brown did at least check his history book.

    In 1977, the then MP for Ashfield was appointed to an EU post. Labour lost the ensuing by-election to the Tories on one of the biggest Lab-Con swings ever.

    The question remains, if Geoff Hoon is earmarked for an EU post next summer, will they be holding the by-election on the same date as the European Parliament (and County Council) elections ? Perfect day for burying bad news if seat is lost ?

    Paul H-J

  10. James – it’s the sort of thing they should really answer if asked. For their last poll of marginal seats they published a list of the seats covered on their website, so I’d expect them to do that again.

  11. Anthony,James.

    Whrn the full figures are up it will be intesrsting to see the regional variations as they may well give an additional indication as to the likely situation across the country.

    If there are big regional variations it may well update the parties list of target seats


  12. Writing from memory, I recollect that yesterday’s “Times” explicitly stated that Brown had not wanted to cause a by-election through the appointment of a European Commissioner. See several posts I have contributed for the Ashfield seat. In addition, it was suggested that Brown had (in consultation with the European Commission) specifically wanted to appoint a woman as a Commissioner, and had agreed in advance with Europe that the candidate would, like Mandelson, become Trade Commissioner.

    Whether the Prime Minister should exercise political correctness for such a senior position might be a matter for political debate. The implication is that Brown may (or may not, depending on the quality of the male candidates available) have put equality before choice of the person with the best ability and experience to further the interests of the United Kingdom and the European Community. Given the current backlash in places like the “Daily Mail”, regardless of what one thinks personally such an action could have a cost for Brown with ordinary voters. Or he might gain gain votes from people who want more women in high office.

    The European Commission is re-appointed next year AFTER the European Elections. If Brown wanted to appoint Hoon as a Commisioner at that point, it would therefore seem that an Ashfield by-election would have to be subsequent to the Euro-elections. But unless the Commissioner Brown has just appointed makes a complete mess of things, it would hardly seem right for her to be replaced after a matter of months.

    Incidentally, Tony, can we have a seat guide to European seats so that we can debate psephological issues relating to them in the run up to next year’s elections?

  13. A different comment from the one I have just posted.

    Anthony’s point that the Labour to Conservative swing is below average in Conservative/Liberal Democrat marginals is important, and interests me. When the PopulusHome poll of marginal seats came out, I posted on a number of Labour soemwhat marginals (places like Reading West and Cardiff West) that the poll suggested above average swings from Labour to Conservative in Labour marginals. It did cross my mind that this must imply smaller swings from Labour to Conservative elsewhere. Now Tony suggests where “elsewhere” is.

    There are obviously issues relating to tactical voting in marginals that help explain this ICM poll. I also wonder, without looking into the detailed data in detail, whether Labour lost middle class voters before 2005, for instance over Iraq. They may already be at their floor in, for instance, many heavily middle class South Eastern seats outside London (which include a number of Conservative/LibDem marginals such as Eastbourne) But Labour may since 2005 have started to lose skilled working voters who are more interested in economic considerations. This may be particularly bad news in relation to the current financial crisis, as such voters are likely to be in the private sector, perhaps self-employed, and dependent on savings rather than protected public sectr pensions for their future prospects.

  14. I think there’s a general feeling Labour middle managers in the public sector have abused their position. The SNP have very effectively campaigned on this.

  15. John C

    The irony is that DC is the longest-serving leader of the 3 major parties in the country – Gordon Brown, I suppose, might have some experience rubbed off as the long-time 2nd-in-command, but that’s different to being in the full lights of the election that Tony Blair was expert with.

    I can’t really say what they’ll be like, come 2010, but Peter Mandelson had better keep his paws out of places where he’s so innocent, he has to resign for the 3rd time, and the resurgent old Left will have no love for him.

  16. The ICM figures are now up and the first thing i noted was that of the 192 marginals polled only the following are in Scotland;

    Dumfries & Galloway, Edinburgh South, Ochil & South Perthshire, Stirling, Renfrewshire East, Edinburgh North & Leith, Edinburgh SW, Aberdeen South, Ayr Carrick & Cumnock.

    Thats only nine seats some only 5% of the seats as opposed to the 8% you might expect. This is no doubt because they are fewer marginals in Scotland, particularly Labour/Tory ones. That probably means it won’t tell us that much about the key Labour/SNP battle.


  17. Peter Cairns is right. It would be interesting to have more information as to how the 192 constituencies were chosen. Not least to check the statistical validity of this poll.

    It also appears that this was a normal sized, i.e. about 1,000 people, sized sample. So only about 5 people per seat included. Therefore it will not be possible, unlike the very useful PopulusHome survey of marginals, to dig out details for individual seats. For instance, this poll, unlike the PopulusHome one, included Ashfield, where there is a suggestion that Labour in effect shied away from a by-election by not appointing Geoof Hoon as a EuroCommissioner. It would have been interesting to establish which party currently has most support in that seat.

    In relation to Scotland, the inclusion of such a small sample of seats is indeed questionable. But I have previously pointed out on this site that the Scottish and English electorates cannot properly be regarded as a statistically homogeneous population from which to sample. One obvious reason is the presence of an addtional major party in Scotland. In my view, ICM and other pollsters should explicitly survey England, and then do a separate survey for Scotland.

  18. I agree with Peter above, in that any marginal involving Labour and/or the Conservatives should be involved.

    Ignoring the SNP/Labour seats (which the SNP is likely to gain in) and the Tory/LibDem ones (where the Tories would gain under current polling conditions) must introduce a bias in later, country-wide analysis that ICM which would underestimate SNP and Tory gains. This might explain the poor correlation between ICM’s 11.5% swing and simply applying an 11.5% swing to the model on Electoral Calculus (majority of 158).

  19. Frederic/Richard.

    Or in short, Anthony’s poll was better than this one.


  20. Looks like the 12 Tory gains in Scotland are still on the cards that i predicted.

  21. Dealing with the credit crunch even slightly effectively might just help Gordon Brown in the polls when nothing else will. Yes, I know a lot of it is simply posturing and warm words because little can be done to change such a global problem, yet every other pressing national problem is ignored in the media whilst this is goes on. David Cameron’s got to be careful to offer the right amount of support, if the government chooses the wrong option, he won’t gain from it.

  22. The 12 Tory gains include gains from SNP which, in the current climate, is inconcievable.