Monday, December 8, 2014

Piffle pertaining to Populus

Even if it wasn't for my past history at Political Betting, I would have sighed heavily at the latest post from Mike Smithson, which makes a big drama out of the fact that Labour are "only" 3% behind the SNP in a published aggregate of Populus subsamples from the month of November -

SNP 35%
Labour 32%
Conservatives 17%
Liberal Democrats 9%
UKIP 4%
Greens 2%

This of course only tells us what we already knew, because the individual Populus subsamples were all published and were freely available for anyone to tot up (but only if that person could "be arsed", of course).

Smithson does add a couple of half-hearted caveats to the figures, but still comes to the silly conclusion that Labour "will be delighted".  As we've been saying for weeks, Populus stick out like a sore thumb as the one and only pollster that has produced subsamples since the referendum putting the SNP behind Labour - and they haven't just done it once, but several times.  By contrast, every other pollster has put the SNP ahead in every single post-referendum subsample to date - and that includes YouGov, who produce at least five subsamples per week.  If you aggregated subsamples from every firm other than Populus, or even if you aggregated subsamples from every firm including Populus, you'd see a very handsome SNP lead - as this blog's Poll of Polls demonstrates.

There's no mystery about why Populus are different - they downweight the SNP much more severely than anyone else does, based on target figures for party identification from way back in 2010, when of course the SNP had much less core support.

In other words, the above figures are a red herring.  Nothing to see here.

On a vaguely related point of pedantry, Smithson keeps saying that there have only been three full-scale Scottish polls since the week of the referendum.  That's quite simply wrong, but naturally I haven't been able to correct him because of my random lifetime banning from his site.  There have in fact been six full-scale Scottish polls since September 18th - two from Survation, two from Panelbase, one from Ipsos-Mori and one from YouGov.  Even if you discount the first Survation poll because it was conducted immediately after the referendum, that still leaves five.

26 comments:

  1. Great article as usual James. Do you know when the next poll that will ask respondents about their opinion on scottish independence will be done? Obviously I expect this to be a very casual affair post-referendum with pollsters, but it would be interesting to see if there has been any increase since the last opinion poll on the issue done by Survation.

    ReplyDelete
    Replies
    1. I can go along with that format for a poll,good idea.

      Delete
  2. Thanks James. Saw PB, panicked a bit, and came straight on here in the hope you'd lay that one to rest!

    ReplyDelete
  3. Populus don't just downweight the SNP, they also do the same for UKIP, for the same reason. The result is they have typically much higher Con + Lab share, usually in the high 60s across GB, than the other pollsters (who are usually in the lower 60s). The problem with this kind of weighting is that you are effectively discounting the views of people who haven't just shifted their vote for temporary reasons (tactical voting, dissatisfaction with a leader), but have completely changed their identification.

    e.g. today's poll (which has SNP 32, Lab 28 and Con 24 in the Scottish subsample) has SNP identification down-weighted from 64 to 46, Green from 78 to 44 and UKIP from 274 to 82. The only group that is significantly up-weighted is none of the above, from 268 to 492. I think this reflects a problem with online polling, which is that the people who are more likely to complete them are more likely to be politically motivated.

    http://www.populus.co.uk/wp-content/uploads/OmOnline_Vote_08-12-2014_BPC.pdf

    ReplyDelete
  4. "In other words, the above figures are a red herring. Nothing to see here."

    Er... no. Demonstrating that a polling company uses a different methodology from other polling companies isn't the same thing as demonstrating that company is wrong. In this case they quite probably are wrong, but your "this polling company gives results I don't like so they're flawed/biased" attitude to polling analysis is, as usual, completely ridiculous.

    ReplyDelete
    Replies
    1. I don't think James was criticising Populus for their weighting techniques, which may be proved correct, as much as he was criticising Mike Smithson. The post on PB presented the Populus results as evidence that the position in Scotland is close, without adding the context that the sub-samples of every other pollster and the handful of Scotland-only polls have shown a clear SNP lead.

      Delete
  5. Populus poll (aggregation) for Edinburgh:

    35(+23)% SNP
    25(+6)% Con
    24(-13)% Labour
    10(-19)% Lib

    http://www.edinburghnews.scotsman.com/news/general-election-2015-poll-gives-snp-major-boost-1-3628337

    ReplyDelete
    Replies
    1. These are the figures for Glasgow:
      http://www.eveningtimes.co.uk/news/snp-close-in-on-labour-for-general-election-191455n.26048597

      "The poll found Labour had 38 per cent support, while the SNP was on 36 per cent, a cut of 18.2 per cent for Labour and an increase of 18.7 per cent for the SNP."

      That suggests a swing of around 18%, pretty close to the poll of polls swing.

      Delete
    2. I've also realised that the Edinburgh figures also show around an 18% Lab-SNP swing

      Delete
    3. Yes, will be interesting to see tables and whether they have used the same weighting system which results in huge SNP respondent down-weighting.

      Delete
    4. I think those aggregates disabuse the notion that the increase in SNP support / fall in Labour support would be most heavily concentrated in places that voted yes (i.e. greater Glasgow). It appears that the movement is more evenly spread.

      Delete
    5. Of course if the same methodology that was used for the national aggregate was used for the city polling, it implies a bigger Lab-SNP swing in urban areas than nationally.

      Delete
  6. The other thing Smithson has been banging on about in recent weeks is the need for a Scottish constituencies poll from Ashcroft. Does Ashcroft not use Populus? If so, would such a poll not show the low-end marker of the SNP vote in these constituencies?

    ReplyDelete
    Replies
    1. No (or at least not necessarily) - Populus seem to do the fieldwork for Ashcroft, but he has his own weighting scheme.

      Delete
    2. I know for a fact that there have been constituency polls for Dundee West, Glasgow East and one of the Fife seats (poss Glenrothes) From my informant, the SNP lead in Glasgow and Dundee is "significant", the gap slightly less in the Fife seat but SNP are apparently still well ahead. Apologies for lack of further information, hopefully this will become apparent in due course, however it is interesting that. IPSOS Mori, YouGov, Panelbase and Survation are all concurring as far as Scotland is concerned. I believe the movement is absolutely genuine and very real, I'm a seasoned campaigner and feel the wind of change sweeping through the streets. As for the Edinburgh figures, I have predicted for a good couple of years that it will eventually be the Torys who come second to the SNP. Another point to be aware of, a new phenomenon of the "die hard" Unionists voting tactically has risen its head. The SNP surge and Labour collapse have been the points of interest, but be aware, the Tory base appears to have strengthened slightly since the referendum.

      ..nice site, keep it up Mr Kelly.
      regards, ( from an ex Labour CLP office holder, was Yes, now SNP/SSP!!!!!!!!)

      Delete
    3. Should have said I have long predicted eventually the Torys would second to the SNP in Edinburgh and Labour would face a big squeeze in the Capital. Too early to be affirmative, but trends are moving that way.....

      Delete
    4. "Another point to be aware of, a new phenomenon of the "die hard" Unionists voting tactically has risen its head."

      I don't think this would be a bad thing for the SNP at all. One of the SNP's arguments is that there is no real difference between Labour and Tories. What greater evidence could you have of this than Labour people in Perth saying "vote Tory" and Tories in Glasgow saying "vote Labour".

      It would have the effect of hardening the blocks of yes and no into party voting patterns. This would clearly be to the benefit of the SNP because far more people voted yes than have ever voted for the SNP.

      Delete
    5. From my own (Unionist) side, I've certainly been hearing many suggestions of tactical anti-SNP voting myself. I campaigned for Better Together. I'm not affiliated to any party (Nor do I have any plans to be) but I'm planning on doing a little campaigning for some of the non-SNP candidates who are the best hope for keeping the SNP out.

      I never imagined that I'd become a political campaigner, but then the referendum happened. I actually enjoy it.

      Delete
    6. I really do hope there's tactical unionist vote campaigning. Should boost the SNP big style. Imagine a wee labour activist on doorsteps telling an old devo max inclined labour voter to vote Tory. Excellent.

      Anyway, no sign of it in the polls. Or at least if there has been some effect, it's an increasing SNP lead.

      Delete
    7. I like to think we'd be slightly more subtle than that, Skier.

      Delete
    8. How do you subtly advise Labour supporters in an SNP/Tory battleground to vote Tory?

      Delete
    9. Keaton,
      Stoat is far more subtle. I am sure they will frighten little old ladies that they will lose their pension if they vote SNP. Or perhaps really ill Scots will not get transplants any more.

      I am sure he will be very "subtle" - but this time we will be ready for it.

      Delete
  7. Thanks a lot James. I was fearing some pretty dire spin around the figures. But they should be pretty genuine if Ashcroft's weighting is sound.

    Anonymous. Looking forward to the figures being disclosed. I can't for the life of me understand why the 2010 Labour majority was so high in Glenrothes. It seems to me like the SNP should be strong there, as they seem to be growing across all the new towns.

    ReplyDelete
  8. @Anon 6:23

    Well ahead in Glasgow East? That'll be Mags Curran's home turf and one of the safest Labour seats in Scotland, as is Glenrothes.

    If this is true it would mean utter meltdown for Labour in Scotland!

    ReplyDelete
  9. I'm going on hearsay, albeit reliable sources. Labour's nether regions are literally involuntarily manufacturing maltesers in Scotland, Gordon knows this fine well. As a power hoarding control freak ( yes, I've dealt with him in my Labour days) the last thing he would be doing is stepping down if he thought there was a genuine opportunity in the foreseeable future for him and Labour in Scotland. Murphy is being set up for the fall of his political career by Milliband, who's attitude is "if I go down, I'll take someone with me I don't like". Darling, Tory to the core, just like his family, political career of convenience with NewLabour, standing down, destined for the lords. Possibly if he thought there would be something for him he'd hang on another term, clearly he is reading the ground well, he "saved" the Union for the short term, time to take the money, walk the Axminster carpets and park his aristocratic posterior on the red leather benches.

    As for tactical Tory/Labour/Liberal votes, you raised an interesting point. I too happen to believe the SNP will benifit greatly from this scenario. Note:- BBC promote Rennie and Ruth Davidson with regularity, frequently to either make a point degrading the SNP and/or attacking Scotland and indeed Scottish public sector institutions like the NHS with the backing of a semi litterate, barely coherent drone from NewLabour or indeed vise versas, Labour with Tory/Liberal backing.

    People, working class especially, have realised Labour have moved beyond recognition, and are wary of the continued promotion by the right wing bourgeoisie in the media.

    ReplyDelete
  10. You might want to look up the changes they are making to MP's pensions to understand the traffic heading for the exit door.

    ReplyDelete