Monday, February 24, 2014

Poll of Polls : the long-term trend towards Yes

As the evidence mounts that journalists in the mainstream media (including the BBC, the Scotsman and the Guardian) have been letting their profession down all day by earnestly reporting the results of what is little more than a voodoo poll, I thought now would be an opportune moment to step back into polling reality and have a look at the long-term trend towards Yes that has taken place over the last few months. One or two people have asked me to provide some sort of graph for the Poll of Polls, which I would really like to do, but I've no experience of that kind of thing and I can't make head nor tail of the online chart generators that I've looked at. If anyone can point me in the right direction that will be great, but in the meantime I'll just give you the numbers. I'm hugely grateful to Ivan McKee of Business for Scotland, who retrospectively calculated the figures all the way back to September of last year and sent me his table.

The No campaign's lead in the Poll of Polls headline figures :

Sep 2013 - 20.2%
Sep 2013 - 20.0%
Sep 2013 - 18.4%
Oct 2013 - 17.9%
Oct 2013 - 17.5%
Oct 2013 - 17.4%
Nov 2013 - 17.5%
Dec 2013 - 17.1%
Dec 2013 - 16.3%
Dec 2013 - 16.2%
Dec 2013 - 15.8%
Jan 2014 - 14.2%
Jan 2014 - 14.8%
Feb 2014 - 14.8%
Feb 2014 - 14.7%
Feb 2014 - 15.1%
Feb 2014 - 13.6%
Feb 2014 - 14.0%


As you can see there was a very steady trend towards Yes until the start of this year, since when the position has oscillated somewhat, but we've nevertheless ended up right now with the second-lowest No lead to date - and that's from a sample that has a substantial post-Osborne chunk. Almost a third of the deficit (6.2% out of 20.2%) has been made up since September.

12 comments:

  1. Encouraging James.

    The P&J voodoo poll you mentioned previously was apparently done by Ideas In Partnership (http://tinyurl.com/ph5urkg), a company that are 'experts in entrepreneurship and have used these skills throughout the UK'. Hilariously one of their other 'skills' is running bridalwear events.

    Unsurprisingly they've become PB's favourite pollster.

    Keep up the good work!

    ReplyDelete
  2. The problem with polls, if I may say so, is the lag time between research and publication.

    For instance, it seems to me that Cameron has shot himself in the foot yet again today.

    Assume that that is true.

    How much of a lead time will we have to endure before the polls either prove me right or I'm shot down in flames?

    What we need to know are whether what we think are significant interventions, (and no, David Bowie isn't one,) and the time that takes to percolate into voter intentions. I suppose I want more certainty in the statistics and that is just not possible, else daily polls.

    ReplyDelete
  3. The SNP have a 'poll of polls' that shows a clear trend towards Yes as well James.

    I've been trying to find a graph that I saw on-line a few days ago, that shows the ups and down of both the Yes and No polling figures.
    It has a black line through both of these figures charting an average percentage.

    It shows a clear downward course for No, as well as an equally clear upward course for Yes.

    Keep it up and both lines will inevitably converge.

    If I do find the chart James, I will post it on your 'Scot goes Pop'

    ReplyDelete
  4. Thanks, Patrick.

    TUD : Thanks for the information. Crikey. At this rate we'll have cake shops joining the British Polling Council soon.

    Might still be a marginal improvement on YouGov, though.

    ReplyDelete
  5. James, are those figures for the elad excluding don't knows?

    ReplyDelete
  6. No, it includes Don't Knows. The current figures are No 48.7%, Yes 34.7% (lead 14.0%), Undecided/Won't Vote 16.6%.

    Excluding Don't Knows, it's No 58.4%, Yes 41.6% (lead 16.8%).

    ReplyDelete
  7. I've just remembered P&J carried out a poll last year on Independence for the Northern Isles. The breakdown of ages polled set alarm bells ringing.

    http://wingsoverscotland.com/a-new-flag-for-scotland/

    ReplyDelete
  8. Calum : Yes Scotland referred to yesterday's P&J survey as a "straw poll", which would generally indicate that the normal demographic weightings were not applied. If so, the results are literally worthless, and the media should be ashamed of themselves for misleading their readers and viewers.

    ReplyDelete
  9. I've put a simple plot of the figures in Excel - https://dl.dropboxusercontent.com/u/18361224/NoLead.xlsx - both with just the data points and with a linear trendline (though it's questionable as to how much use that is). Assuming a linear trend, we do need things to accelerate a little to have a Yes lead in time for the vote, but I'd hope that there would come a tipping point once it starts to look more possible.

    If you're after something that can be posted online and updated live, Google Docs will let you publish in individual chart and the way it can be created is essentially the same as Excel. Once you've got a spreadsheet on google, add a chart and there's a button on it with an option to publish, you then get some html to copy+paste.

    ReplyDelete
  10. Nice to see the curlers getting involved in running down Scotland.

    Winning the World Championships as a Scottish team was rubbish compared to competing for the glorious team gb according to e muirhead. Absolute total and utterly amazing demonstration of the Scottish cringe. And yet people are defending her racist Scots are untermensch views.

    ReplyDelete
  11. I've only seen a couple of quotes from Eve Muirhead today, but if those are anything to go by she seems to have been fairly cagey. I'm not sure she meant her comments to be quite as political as the media have spun them - which brings me back to the feeling of ambivalence that I expressed at the start of the Olympics. In the end I was 100% behind the curlers in every game - but just imagine how much worse this media circus of "let's try to get them to say something that sounds unionist" would be now if they had won the gold.

    One thing we do know, of course, is that lower-income people are disproportionately likely to be Yes voters. Curling (at least in Scotland) is a sport played by the relatively affluent, so my guess is that if you did a poll of curlers there would be a decent No lead - particularly if that poll was conducted by the Aberdeen bridalwear company.

    There was a bizarre article in the Independent the other day claiming that Eve Muirhead must be a Yes voter, on the basis of no other evidence than that she plays the Scottish sports of curling and golf, is a talented bagpipe player, and once did a naked-except-for-tartan photoshoot. I'm not sure what that assumption says about the "you can be proudly Scottish and still want to remain within the United Kingdom" narrative.

    ReplyDelete
  12. Sandy : that's great, thanks very much for your help.

    Yes, I think the hope must be that the move to Yes will accelerate from late May onwards, because from that point on the broadcasters are required to provide strictly balanced coverage (whether they will actually do that is highly doubtful, of course, but the situation should at least improve).

    ReplyDelete