Sunday, March 2, 2014

YouGov poll shows support for independence rising to its highest level of the campaign so far

Apologies for being a bit later than usual with this poll - I've been having the week from hell.  Among many, many other calamities, I - a) fell awkwardly and twisted my knee, and because of circumstances haven't been able to rest it as much as I should, b) have had my faith in one particular type of public servant rocked to the core, and c) have just five hours ago lost my favourite woolly hat, which I've been inseparable from for months!  Ah well, in the words of the theme song to a 1980s sitcom, life goes on, right or wrong...

So down to business. There was a time, not so long ago, when this kind of late night tweet from Blair McDougall...

"New #indyref poll tomorrow. More first thing."

...would have been a portent of doom. I did go to bed on Friday night wondering if I would awake to a poll showing a drop in the pro-independence vote, but I needn't have worried - instead it was the second YouGov poll in succession to show the Yes vote rising to its highest level of the campaign so far. It seems that 'bad is the new good' for McDougall. If this is the type of poll he likes to crow about these days, he must be getting a tad jittery.

Should Scotland be an independent country?*

Yes 35% (+1)
No 53% (+1)


*YouGov always add a preamble to the actual referendum question - hopefully when the datasets are published it will turn out to be the more neutrally-worded preamble that was first used in September.

Presumably the reason that the No campaign felt they could at least have a stab at spinning this as a good news poll for themselves is the fact that they've increased their own support by 1%, just as Yes have done. The snag is, though, that if the headline No lead remains the same but the number of undecided voters fall, that means by definition that the true lead is actually declining. Here is the position when undecideds are stripped out -

Yes 40% (+1)
No 60% (-1)


So Yes have broken through the psychological 40% barrier for the first time with a pollster that is traditionally extremely unfavourable to them. And that's the key point that must always be borne in mind when assessing YouGov's headline numbers - Yes don't necessarily need to be in the lead by the end of the campaign. It's quite conceivable that Yes could still be four or five (or even more) points behind with YouGov on the eve of polling day, and yet be level pegging or ahead on an average of all the pollsters. As David Halliday pointed out a few hours ago, YouGov's final poll of the 2011 Holyrood campaign had the SNP ahead by just 3% on the list ballot. The actual election result was an SNP lead of 18%.

When the datasets are published, the first thing I'll be looking out for is the gap in voting intentions between higher and lower income respondents - it's been suspiciously low in recent YouGov polls, leading me to wonder if the firm has a serious problem in its sampling process.

Probably the biggest significance of this poll is that it's the final piece in the jigsaw that proves the coordinated announcements on the currency from the three London parties failed to have the desired impact, and if anything probably led to an increase in the Yes vote. It also demonstrates beyond a scintilla of doubt that the Press & Journal regional "poll" that the No campaign and half the mainstream media embarrassed themselves by getting so excited about a few days ago was not worth the paper it was printed on.

* * *

SCOT GOES POP POLL OF POLLS

The pro-independence campaign's progress with YouGov means that they also rise to 34.9% support in this blog's Poll of Polls - the highest Yes figure recorded to date.

MEAN AVERAGE (not excluding Don't Knows) :

Yes 34.9% (+0.2)
No 48.9% (+0.2)

MEAN AVERAGE (excluding Don't Knows) :

Yes 41.6% (n/c)
No 58.4% (n/c)

MEDIAN AVERAGE (excluding Don't Knows) :

Yes 42.0% (n/c)
No 58.0% (n/c)


(The Poll of Polls is based on a rolling average of the most recent poll from each of the pollsters that have been active in the referendum campaign, and that adhere to British Polling Council rules. At present, there are seven - Angus Reid, YouGov, TNS-BMRB, Survation, Panelbase, Ipsos-Mori and ICM. Whenever a new poll is published, it replaces the last poll from the same company in the sample. Changes in the Poll of Polls are generally glacial in nature due to the fact that only a small portion of the sample is updated each time.

For clarity, the Poll of Polls takes no account of polls conducted by bridalwear companies.)


It's only rounding issues that prevent Yes creeping up slightly on the mean average with Don't Knows excluded. The median average is unchanged because YouGov remain firmly on the No-friendly end of the spectrum, with Angus Reid continuing to provide the mid-point.

* * *

UPDATE : I've just posted this as a comment below, but on reflection I thought it was worth adding to the main post -

On a related topic, I've been having an exchange with a unionist Twitter troll who has a bee in his bonnet about the possibility that Yes may only have appeared to break the 40% barrier with YouGov due to their vote being rounded up. Well, we'll never know, because YouGov only publish percentages in their datasets, not raw numbers. However, ICM do provide raw numbers, and having had a look at them I'm delighted to report that their last poll was a touch better for Yes than we originally thought -

Yes 37.4%
No 48.7%

So the true lead was only 11.3%, which is hardly a million miles away from the 7% lead ICM reported in their sensational January poll. With Don't Knows excluded, it's -

Yes 43.4%
No 56.6%

So rounding certainly worked against the Yes side in the way that poll was reported. Hopefully those numbers bode well for future ICM polls, though.

6 comments:

  1. Thanks, James, helpful analysis as always. What was your view of the Scotpulse poll?

    ReplyDelete
  2. Thanx James,

    I think the Youguv polls are weighted using the last general election results, therefore giving No a big much bigger percentage than is credible.

    Youguv is owned by a Tory...need I say more?

    In spite of this we see another 2% swing to Yes and Yes breaking through the 40% barrier for the first time, with that pollster.

    Blair McD will just be happy that it wasn't like the more credible polls and showed a 6% boost for Yes.

    Not sure if you saw it on another post James (your wooly hat accident might have ment you missed it) but an SNP poll recorded that the people of Scotland said that Ed Balls joining the currency bullying attempt by Osborne has meant 30% say hey are less likely to vote Labour now.

    18% of labour voters have also said they are now less likely to vote Labour for the same reason.

    Take aim at your foot...fire!!!

    ReplyDelete
  3. "I'll be looking out for the gap in voting intentions between higher and lower income respondents - it's been suspiciously low in recent YouGov polls, leading me to wonder if the firm has a serious problem in its sampling process."

    Not only YouGov but all of them. It doesn't seem credible to me that almost all of the straw polls at meetings/debates etc show YES with substantial leads but is still struggling to break 40% in these accredited polls.

    You have already told me the manner in which these pollsters choose their sampling universe, ie, by asking for volunteers via the I'net. And this I am sure means they end up with mainly anoraks, retirees and better off people in their samples. Ordinary voters have neither the time nor inclination to indulge in what they see as time wasting exercises.

    I know some of the errors which would be produced by such sampling can be adjusted by weighting but surely only up to a limited point?

    These polls right or wrong, are very important because the media and the BBC can and do take the view that currently the NO campaign has a large majority and thus their support implicit or not for the NO campaign is justified.

    ReplyDelete
  4. "Not only YouGov but all of them."

    Well, YouGov are different in this respect. Their last poll actually had No in a slightly stronger position among lower income voters, which is totally at odds with every other poll I've seen recently.

    I read an interview with an American pollster the other day who made the point that even telephone polling shouldn't work in theory (because there isn't a high enough response rate to be truly scientific), but it does in practice. So online polling should be judged on the same basis - if they have correctly predicted election results in the past, they should be taken seriously in the future. The problem, of course, is that YouGov don't have a great recent track record in Scotland (oddly enough they were the most accurate pollster for the Holyrood election of 2003, but that was a very long time ago).

    Michael : I presume the Scotpulse poll wasn't scientific, but it might be of some limited use in identifying trends.

    On a related topic, I've been having an exchange with a unionist Twitter troll who has a bee in his bonnet about the possibility that Yes may only have appeared to break the 40% barrier with YouGov due to their vote being rounded up. Well, we'll never know, because YouGov only publish percentages in their datasets, not raw numbers. However, ICM do provide raw numbers, and having had a look at them I'm delighted to report that their last poll was a touch better for Yes than we originally thought -

    Yes 37.4%
    No 48.7%

    So the true lead was only 11.3%, which is hardly a million miles away from the 7% lead ICM reported in their sensational January poll. With Don't Knows excluded, it's -

    Yes 43.4%
    No 56.6%

    So rounding certainly worked against the Yes side in the way that poll was reported. Hopefully those numbers bode well for future ICM polls, though.

    ReplyDelete
  5. I know you occasionally like to be kept up to date with PB's incompetent Moderator James, and this is truly priceless stuff.

    After having another hissy fit over me describing SeanT as a sex tourist, which as you know SeanT has done innumerable times on PB and at great length himself, he's now temporarily banned me for saying, and I quote.

    "I'd point out the obvious untruth but I'm respecting the mods wishes and instructions right now".

    So now respecting the moderators wishes is a banning offense! :-D

    It's so wonderfully incompetent and stupid only TSE could possibly do it. All the while ignoring some far-right nutcases ranting on about aerosol-sniffing and changing user names. Which TSE earlier said was forbidden but let's not worry about consistency from a hypocritical bet welcher.

    ReplyDelete