Two types of outlier

This morning TNS released a new poll showing figures of CON 28%, LAB 35%, LDEM 6%, UKIP 18%, GRN 7% (tabs). The seven point Labour lead is striking, and out of line from the broader trend. Usual caveats apply, but for once I haven’t seen many people over-react to it. Most have sensibly enough seen it as the just the other side of the coin to the ICM poll earlier this week showing a solid Tory lead – two outliers in opposite directions. However, it is worth looking at the different reasons why these two polls went against the trend.

On average ICM produce figures that are slightly more Con/less Lab than average, but only by a tiny bit, generally they show Labour -v- Tory leads that are much the same as those from other companies. The reason that their poll on Monday showed a four point Tory lead when other companies are showing the race almost neck-and-neck is almost certainly just random sample error. Samples vary from day-to-day, month-to-month and sometimes you get one that’s a bit Laboury or a bit Conservativety in ways that weighting does not correct. I expect ICM’s next poll will be in line with those from other companies.

TNS is a different sort of outlier. The seven point Labour lead may be different from that shown by other pollsters, but its actually in line with TNS’s previous polls. In their previous four polls they showed Labour leads of 6 points, 0 points, 7 points and 7 points. Over on my chart of house effects TNS’s polls on average show a Labour lead three points larger than other companies (and if anything the gap is growing!) This isn’t random sample error, this is a consistent methodological difference between TNS and other companies, and unless they switch methods I expect their future polls will continue to show bigger Labour leads than other polls.

Two outliers, but one is probably just a random blip from a pollster who normally shows the same as other companies, one looks like a typical poll from a pollster who regularly produces polls that show a bigger Labour lead than the pack.

311 Responses to “Two types of outlier”

1 5 6 7
  1. @ JohnB160 and @ ChrisLane

    There are good reasons for taking with a pinch of salt any projections you get either from Electoral Calculus or from the advance swingometer (=regionalised UNS model).

    Both fail to take systematic account of the constituency polling data that is making such an important contribution in this election cycle and it really doesnt seem to make a great deal of sense for supporters of any party to build up hopes (or have them dashed) on the basis of models that are ignoring crucial evidence.

  2. Northumbrianscot


    I’ll maybe do a bit of database construction that would allow some teasing out of the assumptions behind whatever model Scotland Votes eventually produces.

    Thanks for the tip; I am guided by the Gospel of the UKPR SWINGOMETER.

  4. @ Richard W,

    The really astonishing number to me was the Lib Dems’ £3,038,500. How is it that the Lib Dems on ~8% of the vote, with no rich hedge fund donors or union backing, collapsing membership and no prospect of dictating government policy in the near future, can rake in nearly half the funding Labour or the Tories have on 33%? And twice what Ukip on ~15% are getting as well?

    Where is this money coming from? Oakeshott can’t bankroll the whole party, surely, and it’s definitely not their MEP levy.

  5. JayBlanc

    It’s not about funding. It’s about people. Funding gets glossy adverts, it does not produce motivated political activists. And the ground game needs those people for canvassing, locally tailored leafleting, turning up for events and what’s more to give a local-legitimacy to candidates.

    All that is to some extent true, but good funding means that these traditional, labour-intensive activities can be supplemented by more expensive ones such as direct mail and to pay for staff to design and plan such strategies. Paid leafleters can be used to free activists up for more skilled work such as canvassing. Even the traditional things, such as leaflets, can be done more frequently.

    Of course there’s no substitute for local knowledge, a good and energetic candidate selected early, numbers of helpers and prolonged campaigning. And outside ‘experts’ can often do more harm than good, especially if coming in last minutes. But money can help fund all these things long term and be used to make the best of your assets during the campaign.

  6. More on Green/LibDem crossover.

    There is a fair amount of UKPR discussion of ‘crossover’ but no agreement on how to identify that its happened. Green/LibDem crossover may be imminent, but I don’t think that anyone believes that it would make the slightest bit of difference to the number of seats won by either party in May. In the hope that this means that it is not too politically charged, I am going to dwell on a couple of technicalities.

    Back in November, I fitted linear equations to the VI trends for each of the parties (using figures from all polls since the beginning of 2014). For these two parties the relevant equations were:

    LibDems: Predicted VI = 7.37 – 0.0079 x D
    Greens: Predicted VI = 5.41 + 0.0098 x D

    Where D is the number of days since Nov 21 (when I did the calculations)

    Solving these equations indicates that they suggest that crossover should occur on March 12th, although this date should obviously have a wide margin of error associated with it.

    Is this likely to happen, and if so when? Charts posted by @ Statgeek already show the Greens having VIs above those reported for the LibDems. But these averages are based on YouGov polls alone, and in the spirit of Anthony’s Polling Averages I think it is fairer to use polls from all the different houses. Each month up to and including January, the LD VIs have been reliably higher than those for the Greens. Wikipedia shows the two VI plots converging very recently, while Electionforecast still has the LDs comfortably higher than the Greens. To complicate things further, although the LDs are still numerically ahead there is no statistical difference between the VIs over the last 30 days.

    The best we can say is that the plots are now very close. Does this mean that LD/Green crossover is now imminent (in line with my trend predictions)? Perhaps not. In recent weeks the LibDems have been performing better than their 2014 trends and to the extent that this continues they may be able to keep their level of support slightly above that for the Greens. So, we might be heading for what I called a “kiss and recoil” event rather than a crossover. The only way of distinguishing between the two would be to find a future period over which the Greens’ VI is reliably higher than that for the LibDems, and it may be several weeks before that happens.

    Turning to more contentious matters, we may not know about Labour/Conservative crossover (and therefore swingback) until the actual results start to come through early on May 8th.

  7. @Spearmint

    “Where is this money coming from? Oakeshott can’t bankroll the whole party, surely, and it’s definitely not their MEP levy.”

    Especially now that he’s giving as much (£300,000) to Labour candidates as to the Lib Dem MPs he hasn’t taken exception to.

  8. @ ChrisLane

    I am guided by the Gospel of the UKPR SWINGOMETER.

    I’m sure you will choose your own oracle, but – given the constituency polls just published – are you aware that the swingometer would place Ukip about 20 points behind the Tories in both Boston and Skegness and Castle Point?

    You can obviously ignore the fact that these projections are wildly different from the neck-and-neck situation seen in the polls themselves or you might just stop and consider whether there are better sources to consult.

  9. @ Jayblanc and Spearmint

    Yes, I was surprised at LD donations being double the UKIP ones. I guess it tells us something about the relative lack of wealth of UKIP’s supporters and the absence of both business (suspicion of an EU referendum) and union support for UKIP.

  10. Rafael Behr of the Raga Nudi gets good reviews on these pages, and I was struck by his column yesterday which for all intents and purposes concedes there will be a Tory victory in May and goes on to talk about the ‘inevitable’ EU referendum nightmare.
    His take seems to be that:
    a) the Tories are very chipper and
    b) the Labs are very miserable and
    c) the Tories look at the labs and get even more chipper.

    This rather chimes with something someone (I think it was RolandGateaunose) was saying on here but doesn’t chime at all with the vibes I, as an ‘umble and not very active activist, pick up. It also doesn’t seem to chime with the polls, though the punters seem to agree with Rafael.


  11. test

1 5 6 7