The twice-weekly Populus poll is out and has topline figures of CON 33%, LAB 38%, LDEM 12%, UKIP 9%. Compared to the previous Populus poll on Friday the Conservatives are four points higher, Labour two points lower and UKIP three points lower… though of course, the reality is there almost certainly hasn’t mean any movement. Populus’s online polling so far has on average shown the Conservatives on around 32%, Labour on around 39%, the Lib Dems around 12% and UKIP around 10%. Last Friday’s poll was a bit of an outlier, and this is a reversion to the mean.

177 Responses to “Populus – CON 33, LAB 38, LDEM 12, UKIP 9”

1 2 3 4
  1. ALEC

    Thank you.

  2. That’s really turned my stomach.

  3. Colin
    Agreement on the last post , let’s not make it a habit!

  4. The 15-tonne mass of festering food fat mixed with wet wipes and sanitary products threatened to send raw sewage spurting onto the leafy streets of Kingston upon Thames.

    -Now that would have been a shame!

    Think what it might have done to property prices.

  5. STEVE

    Will try not to!

  6. The polls are sometimes up, sometimes down.

  7. Colin,

    Hope your wife gets better soon.

  8. “The 15-tonne mass of festering food fat”

    Leave Prescott out of this. ^^

  9. @ statgeek

    “The 15-tonne mass of festering food fat”

    Leave Prescott out of this. ^^

    How witty!

  10. Still on the subject of an incumbency boost….of course, a first time incumbency boost would not help the Tories so much to get an overall majority…it would primarily just make it harder for Labour to make gains and become the largest party/win an OM themselves.

    But it’s all theoretical for the moment, I guess.

  11. So polls closing now…Hello Ed Miliband,is anyone there?

  12. I’m sure if someone took the trouble to count them you would find more leaves in the Borough of Wigan than in the Borough of Kingston- even if you included Richmond Park.

    @ Colin

    Hope all goes well for your wife.

  13. @ statgeek
    “The 15-tonne mass of festering food fat”
    Leave Prescott out of this. ^^
    How witty!


    There are plenty of other forums where we can make silly remarks about peoples appearance let’s face it there are obvious candidates in the Tory Party who would have fitted this description.

  14. Well Eric Pickles crossed my mind, but Prescott seemed more apt.

  15. @Ambivalentsupporter,

    Whilst I do not disagree with your comments re- Tory incumbency bonus I think its effect will be reduced by two factors.
    Firstly , in quite a few marginals the former Labour MPs defeated in 2010 will be seeking to regain their seats – eg Broxtowe, Warwickshire N, Stroud and Waveney. In such constituencies I would expect the Labour candidate to hang on to the bonus formerly enjoyed – or at least neutralise any potential bonus enjoyed by their Tory successors.
    Secondly, arising from the failure of the boundary review to be enacted , the existing boundaries will be more favourable to Labour in 2015 than was the case back in 2010 – simply because of demographic change in the intervening years. As a result , a repeat of the 2010 result – Con 37 – Lab 29.7 – LD- 23.6 – would be likely to see a smaller Tory lead in seats over Labour when compared with 2010. A given boundary distribution of seats becomes more favourable to Labour over time.

  16. @Shevii

    Indeed, much of Wigan was officially classed ‘rural’ because of the lack of built spaces.

    Now, in my old manor of Makerfield, that’s because it had more mining spoil heaps than anywhere else in the country and you can’t build on ’em, but it turns out Wiganers are ingenious and hence we have the Three Sisters, a recreational beauty spot quite literally made from industrial waste (and very nice it is too), and a landmark example of brownfield reclamation.

  17. @Graham,

    You may be right. Will be interesting to see if March’s marginal poll was just an outlier and pollsters like Kellner are wrong. It’s possible.

  18. I understand that the main taste problem with the new stem-cell burger was that it was all cultured muscle and no fat.

    Stem-cell à la graisse berger wouldn’t present any problems for the dedicated carnivore would it?

  19. @Graham,

    Either way, like I said before, I think Kellner is vastly overestimating the incumbency effect. It is more likely that it will be worth 2-3% to the Tories IMO….Labour needing a 6-7% lead for an OM seems a bit OTT IMO.

    The last marginal poll seems to indicate this.

  20. I think another valid point is that the incumbency effect will be of more benefit to the Tories where Labour’s vote stays higher (i.e. 34%+) and the GE is close. If the Tories were to be over, say, 3% ahead, more of the first time incumbency effect would be rendered irrelevant anyway.

    That’s why it wouldn’t so much help the Tories to get an OM….just make it harder for Labour to gain seats and be the largest party or win an OM.

    Anything like 2010 polling and a lot of the incumbency effect will be wasted.

  21. To return to the subject of the Populus poll, if you look at the unweighted figures they show a rather different story from those in the headline (in brackets) [1]:

    Con 27% (33%)

    Lab 37% (38%)

    Lib Dem 9% (12%)

    UKIP 18% (9%)

    Others 8% (8%)

    Which is interesting much more like the sort of VI we see from newer on-line pollsters such as Survation and Opinium.

    Now some of the weighting effects will be uncontroversial for things such as age. But the biggest effect seems to be in response to Populus’s Q4:

    Regardless of which party, if any, you are likely to end up voting for at the next General Election due in May 2015 or are leaning towards at the moment, which political party would you say you have usually most closely identified yourself with?

    In this poll 11% identified with UKIP (and the figure is similar in other recent Populus polls), but this was weighted down to only 1% of the sample[2]. The reason for this is explained in Populus’s methodology:

    [Rather than using past vote]Populus has therefore sought a different way of ensuring that the political make-up of each of its online samples is more politically representative and leads to results that are more consistent with samples drawn randomly by phone. We have done this by asking respondents which political party – if any – they generally identify with and comparing these answers to those on party identification given to the British Social Attitudes Survey in the year of the last General Election.

    The trouble with this approach is that it assumes the current Party-id profile of the population is unchanged from 2010. In Q4 they are asking what people alliegence is at the moment not what it was 3 years ago. Given the changes since (the rise of UKIP, the drop in Lib Dem support) this seems very unlikely. It assumes that only 1% of the voters should say that they would identify with UKIP and that any over that is due to something being wrong with the sample that has to be corrected.

    Now I’ve said myself here that there probably is a problem with most on-line polls being too “UKIP-friendly” for various reasons. BUt I suspect that Populus are over-compensating for this by using this method (it may also supress Labour’s VI by a point or so). So these figures should be treated fairly cautiously

    [1] There’s no unweighted figures for Parties other than the named four, so I’ve left that constant.

    [2] If you’re wondering why UKIP’s VI isn’t reduced to about a tenth rather than half of its unweighted figure, it’s because people who gave a VI of UKIP but a different Party-id (or none) are still counted in full and not down-weighted (similarly the vote of someone who gives a UKIP-id but ‘votes’ elsewhere will be reduced by 90%, but at the moment there probably aren’t many of these). Also other factors, such as UKIP voters having a higher likelihood to vote that other Parties will boost the UKIP VI a bit.


    Thank you.

  23. @Roger

    Well spotted on behalf of all of us who don’t look at each and every set of data tables.

    I agree that the problem lies in the wording of the question that Populus pose to identify party ID compared to how they use it:
    “Regardless of which party, if any, you are likely to end up voting for at the next General Election due in
    May 2015 or are leaning towards at the moment, which political party would you say you have usually most closely identified yourself with?”

    The use of “have” does nudge an alert respondent to bring in some retrospection, but I’d be surprised if many people referenced their response back to the party they were identifying with in 2010, without a very clear and unambiguous steer towards that date. You could easily interpret the question as asking about the party you have most identified with over say the past year or six months as opposed to the one you are leaning towards currently. So it seems to have been a leap of faith for Populus to calibrate the poll by using the 2010 BSA Survey weights, and the nonsense that the responses of 9 out of 10 UKIP identifiers have had to be ignored via reweighting seems to bear out how wrong that leap of faith was.

    In the opposite direction it’s also worth noting that the LDs have been significantly reweighted upwards, albeit to a lesser extent, which is also exactly what you would expect if many people were again basing their party ID on a timeframe after 2010.

  24. @Roger
    I’m thinking that this issue around identification would apply to all three parties, but is likely to most impact ex-Tory identifiers who now see themselves as UKIP – so perhaps it might be circumspect to consider shaving a point off the Lib Dems and two or three off the Tories?

    I’m not remotely a good enough statistician to know if this is at all valid analysis…

  25. Who’s most to blame for the cuts? The new thread monitor.

    I will cut and paste the last post to the new thread.

  26. alec

    “@Turk – “I see the retail figures are on the up, perhaps people are begining to feel the effects of an up turn in the economy.”

    You need to be very careful in making such bold assertions”

    Why ???????????? What happens if one is wrong?


    Re polls I will repeat what I wrote towards the end of another thread, comparing 2015 to 2010

    Lab WILL poll higher
    Ukip WILL poll higher
    LD WILL poll lower

    {Dunno if anyone would debate that?}

    Cons MUST poll MUCH higher for an OM

    But where wil they get the extra votes from? [, given that they will certainly lose some to UKIP, they need to first cover that loss before effecting ANY increase.]

    This is the best time, finacially, to predict and bet on the outcome of the 2015 Tory leadership contest…….. IMO

  27. May I also add, that there is nothing in Islam that says a Muslim must eat meat – they can be a veggie or a vegan if they wish, or anything else as long as they don’t eat non halal meat or haram (prohibited) food.

1 2 3 4