Sooner or later, a pollster gets something wrong. It happens to everyone if they are in the game for long enough. There are two responses to that, there is to deny there is any problem and blame it all on a late swing, or you can go away, work out what went wrong and put it right. The good pollsters do the second one – so when all the companies got it wrong in 1992 there was an industry inquiry, and ICM in particular came up with new innovations that addressed the problem and led to many of the methods companies use today. In 2008 when MORI got the London election wrong they went away, looked at what had happened, and made changes to put it right. A pollster that gets things wrong, admits it, and then puts it right isn’t a bad thing.

Anyway, in the US election last year the most venerable of all polling companies, Gallup, managed to get things wrong, showing a small lead for Mitt Romney rather than the eventual victory for Barack Obama. They put their hands up, invited in some academics to help and went away and looked at their methods – the result is here. Gallup tested about twenty different hypotheses of things that could have gone wrong, and found that in the majority things were working okay and there was no issue to address. They ended up with four issues where they think things went wrong and caused the overestimate of Romney’s support.

Most or all of the actual problems Gallup identified aren’t directly relevant to British political polls – different system, different challenges, different methods, different solutions – but it’s still an interesting look at what can go wrong with a poll and how a company should dig through its methods if something has gone wrong.

Likelihood to Vote – In Britain pollsters have a relatively simple way of approaching likelihood to vote: they ask people how likely they are to vote, and then weight and/or filter people’s responses based upon that, either giving people’s answers more weight based on how likely they say they are to vote or excluding people below a certain threshold. The only exception to this is ICM, who also include whether people voted in the 2010 election in their likelihood to vote model. American pollsters tend to use much more complicated methods, they ask people how likely they are to vote, but also whether they voted last time, how interested they are in politics, whether they know where the polling station is and so on – there are seven questions in all, which they use to work out a likelihood to vote score and then include only those most likely to vote. Other American pollsters do much the same, but Gallup’s method put more weight on whether people voted in the past, and their adjustment ended up being more pro-Romney than some other companies. Gallup are going to go away and do more work on turnout, including whether the sort of people who take part in polls are more likely to vote anyway (something that I would certainly expect to be true).

Sampling – Most telephone pollsters in the USA get their phone numbers in a similar way to British pollsters, by using random digit dialling. This ensures that people who are ex-directory are not excluded from samples, but at the cost of getting lots of dead telephone numbers, faxes, modems, business numbers and so on. In 2011 Gallup started doing something different. Like most companies they do a fair amount of their interviews on mobile phones, and noted that the majority of ex-directory people did have mobile phones, so theorised that it was safe to randomly generate their landline sample from telephone directories, while bumping up the mobile phone sample to catch those ex-directory people on their mobiles (mobile people who reported being ex-directory were weighted up to account for the tiny percentage of ex-directory people without mobiles). In theory it should of worked. In practice, it probably didn’t – before weighting the RDD sample was more democratic, younger and more pro-Obama than the listed one, so Gallup are going back to using the more expensive RDD method.

Time zone skews – this is an interesting one. As you might expect, Gallup sample within and weight by the regions in the USA. But within some of those regions there are different time zones, and because Gallup started polling at 5pm local time, it meant that in regions that covered more than one time zone they ended up doing more interviews in the eastern part of the region. Correcting this problem would have increased Obama’s support in Gallup’s final poll by 1%. Of course, in Britain we don’t have different time zones to worry about, but it illustrates a problem that can effect any methodology design – skews within the categories you weight by. A pollster can have, for example, the correct proportion of people in the DE social class or people over the age of 55, but what if within those categories you control for people are skewed towards more affluent DEs, or people only just over the age of 55?

Race – the final problem was a rather specific one on how Gallup asked about race – instead of giving people a list of race categories and asking which applied to the respondent, they asked them one at a time and got people to say yes or no, which produced some rather odd effects like overstating the proportion of Native Americans and mixed race people.

The full Gallup review is here and if you’re interested I’d also recommend reading the verdict of Mark Blumenthal (who spotted some of the problems before Gallup did) here and who has obviously followed it infinitely more closely than me.


65 Responses to “What Gallup got wrong in the USA”

1 2
  1. 1st

  2. Did they allow for voter fraud? For example, I seem to remember some wards achieved 140% turnout with every single elector voting for Obama.

  3. See my post on last thread. I could have added (should have) that it is not unknown to hear vox pop voters giving opinions on economy at variance with facts.

  4. @Stutter: before joining the tinfoil hat brigade, please recognise that in many places in America, you’re able to register on polling day but turnout figures are calculated on pre-election rolls.

    Also recognise that some precincts are tiny and located in areas with 100% minority voters (and yes, I know how oxymoronic that sounds). Such areas were never going to vote for the whitest, richest, least charismatic Republican candidate in my memory so a blanket vote for Obama isn’t remarkable.

  5. @AW

    Interesting stuff. Is there any value in asking people if they voted in 2005 or 2001? Folk who turned out then (the lowest turnout elections) might be die-hard ‘always vote’ folk.

    One other tiny point…”Of course, in Britain we don’t have different time zones to worry about, but it illustrates a problem that can effect any methodology design”

    I think it’s ‘affect’. I’ve started using MS Word to spell check my own stuff, but unfortunately it doesn’t correct for wrong, but valid words.

  6. I haven’t got time to read the full report I’m afraid, but it was a very interesting summary. Point 3 was particularly interesting to me. The concept of skews within a category could well apply in Britain, even though the specific time-zone problem wouldn’t.

    For instance, a lot of young people have mobiles but no landline. If a particular social group tend to turn their phones off when they get home from work that could skew the results.

  7. @Stutter

    You seem to be watching (and remembering) to much Fox News. That just didn’t happen.

    http://snopes.com/politics/ballot/2012fraud.asp

  8. Sorry “too much” sp

  9. Anthony,

    I would presume that although Yougov ask what age category people are in you know there actual dates of birth etc. so you can to an extent avoid category skewing problem.

    However it does raise the issue that in theory given the database YouGov has you could probably create an app that let you input three or four criteria (age, gender, class, location) and get a very specific breakdown for that group.

    Equally you could put in other criteria (party, issue, etc) and get the profile of those most likely to be attracted by it. I have no doubt that you could even take an individuals basic details and fairly accurately give percentages for which parties they are likely to vote for.

    Nearly 300 polls a year at 1,500 a time over ten years is almost 5 million pieces of data to crunch and tabulate.

    How’s about giving us a few hints to what you have learned beyond what the regular polls show us.

    Peter.

  10. “In theory it should of worked.”

    Oh, Anthony… (smiley thingy)

  11. Peter, shh! If we had that tool there’d be no point to this website!

  12. Two points:

    1. “Gallup’s method put more weight on whether people voted in the past, and their adjustment ended up being more pro-Romney than some other companies.” Interesting in a UK context. The company that stands out in fine tuning more and more its likely voter model eventually ends up overdoing it and coming a cropper. ICM likewise put more weight on whether people voted in the past. Their model is more complicated than any other here, but it’s not necessarily the most accurate just because of that.

    2. “the final problem was a rather specific one on how Gallup asked about race”.
    Not an issue that’s relevant to a UK context, because we don’t ask about race here, at least not in national polling. But that’s the point. In terms of BME voters, race in the UK is an extraordinarily strong indicator for differentiating VI, as it is in the US. Which is why I consider YouGov and other pollsters to be missing a trick by excluding ignoring ethnicity as a potential weighting factor. That is, the fact that Gallup can come a cropper but getting ethnicity wrong in the US confirms the need to try and get it right there. If so, why not make more effort to get it right here as well?

  13. “by” not “but” in “getting ethnicity wrong”

  14. Are there also lessons to be learned for the UK from Italy’s polls?

    http://www.reuters.com/article/2013/02/26/us-italy-vote-pollsters-idUSBRE91P0UF20130226

    “Another prominent pollster, Renato Mannheimer, said the huge shake-up in politics caused by Grillo’s rise, combined with a sharp increase from 20 to 35 percent in those making last-minute decisions on how to vote, had thrown computer models based on previous elections and the methods used to analyze them out of kilter”

    Seems similar to what we are seeing with the rise of UKIP recently. Different polling companies are coming out with massively different projected vote shares. This graph shows it quite nicely

    http://upload.wikimedia.org/wikipedia/en/9/96/UK_opinion_polling_2010-2015.png

    Those dots seem to be getting further and further away from the trend line, seems we can only guess within a range of 4-5% either way what the level of support for each party is.

  15. Off-topic:

    New Ashcroft poll in The Times:

    Lab 37
    Con 27
    UKIP 15
    LD 9

  16. I don’t think that’s off-topic at all Andy! Thanks for reporting it. We now await YouGov.

  17. Latest YouGov / The Sun results 6th June – CON 32%, LAB 39%, LD 10%, UKIP 13%; APP -34

  18. Harry Enten has his own take on this:

    h
    ttp://www.guardian.co.uk/commentisfree/2013/jun/06/gallup-2012-election-polling-errors

    When it comes to Rasmussen, he thinks they are now overcompensating for their dire 2012 performance:

    h
    ttp://www.guardian.co.uk/commentisfree/2013/jan/31/weighting-game-rasmussen-predicts-democratic

    His articles are collected on the marginoferror website

  19. http://lordashcroftpolls.com/wp-content/uploads/2013/06/State-of-the-parties-data-tables-Telephone.pdf
    Tables here –
    Final sample size of 488 once all the adjustments are made.

    Sample of 642 before adjustments, VI of –
    Con 24 (-3), Lab 40 (+3), Lib 9 (-), UKIP 15 (-)

    Sample of 458 before reallocation of DK but after turn-out –
    Con 25 (-2), Lab 38 (+1), Lib 8 (-1), UKIP 16 (+1)

    So turn-out weighting boosts Con and UKIP (hurting Labour), then reallocation of DKs boosts Con and Lib (hurting Labour and UKIP).
    Which we already knew would be the case.

  20. Ashcroft failing to find the Labour voters, who we keep being told, are now “increasingly turning to UKIP”.

    He finds the following 2010 voters now saying UKIP:

    Con 26%, Lab 2%, LD 13%.
    In terms of VI that would be something like 18:1:6

  21. @”the Salafist insurrection in Syria”

    When one reads that sort of assertion , the heart sinks, and a big neon sign lights up saying-“Stay Away from these people”

  22. The latest Yougov is probably slightly overdoing the Tory VI and I suspect the Labour lead with Yougov is probably around 8-9%, but I think we can now safely conclude that the Labour lead has fallen in the last couple of weeks – as has the Labour VI.

  23. Colin
    Do you mean stay away from Daodao?

  24. I think recent polls have shown a slight softening of the UKIP support, a whole week without Farage smugging on the TV is this a record?

    Possibly some of the Tory Defectors to UKIP have concluded that the Conservatives are actually sufficiently Swivel Eyed and Loony after all.

  25. BB

    “Ashcroft failing to find the Labour voters, who we keep being told, are now “increasingly turning to UKIP”.
    He finds the following 2010 voters now saying UKIP:
    Con 26%, Lab 2%, LD 13%.”

    Well, well, well. That’s anotger brick in the wall of the case I’ve been making for the last couple of weeks. The recent Lab supporters who have moved to UKIP are the “bugger the lot of ’em” ones who left Lab for the LDs between 01-10, came back to Lab in 10-12 and have now joined the latest repository of ill-defined protest.

  26. The Ashcroft poll is fairly close to the latest YouGov in terms of the VIs for Labour, the Lib Dems and UKIP but, as with other pollsters, there’s quite a big variance on the Tory VI. 32% plays 27% is a sizeable difference and while it’s possible that the more recent fieldwork of YouGov might have detected a late Tory surge, YouGov are starting to stand out now as the only pollster showing the Tories regularly in the 30s. Whenever we get the odd Survaton, Opinium, TNS, Comres, ICM, Ipsos/Mori and Angus Reid, the Tory VI tends to range from the mid to late 20s with, admittedly, the odd strange one like the Ipsos/Mori on the 13th May. That’s starting to look like an outlier now in the context of the many polls that have been conducted since then.

    THE UKPR rolling average, up to and including 03/06/13, of Con 28 Lab 37 LD 10 UKIP ? feels about right to me and, in fairness to Lord Aschroft, more or less mirrors his latest poll. In that context, today’s YouGov looks a tad dodgy.

  27. @CrossBat,

    Yep, I think Yougov has a tendency to overdo both Tory and Labour VI slightly.

  28. The latest Yougov poll also seems slightly odd in that the Tories are also leading with the youngest voters. Today’s cross-sectional breaks seem a tad dodgy to me.

    That aside, I think we can safely conclude that the Labour lead with Yougov is probably around 8-9%.

    I can’t really see us seeing much in the way of significant polling shifts over the next few months.

  29. Meant to add that Ashcroft’s numbers also support the amateur quantitative assessment that I’d done.

    Based on looking at the Lost LD numbers from AW’s polls, and the trends of Lab & LD VI over the last couple of years, I’d concluded that somewhere between 2-4% of the 2010 Electorate had switched from LD to Lab then from Lab to UKIP in 10-12, then 12-13.

    Ashcroft is saying that 13% of those who voted LD in 2010 (23%) are now supporting UKIP. That is 3% of the 2010 Electorate. Since these voters have not suddenly switched from LD to UKIP (because the LD VI has held steady – maybe even risen a notch – during the recent UKIP surge) it’s a fair guess to say that these have taken the journey from LD to UKIP via their 10-12 dalliance with Lab (which is where the majority of those who left LD post-10 went in the period 10-12).

  30. You gov now moving slightly in favour of the Comservatives. I think this might well continue given the confusing speech by EM yesterday which seemed to move Labour to a more Conservative position around a three year benefit cap.

  31. Good work Lefty – 3% of the electorate is a large chunk of swing voters and where they end up may well determine who ends up as the largest party.

  32. Morning everyone.
    As far as the Yougov polls go it certainly has a slight ‘trend’ appearing.
    Con 30-32
    Lab 37-39

    Labour Lead of around 7-9%

    Only a tad as they say but a tad is a tad and Labours regular double digit lead is largely a thing of the past.

  33. You gov now moving slightly in favour of the Comservatives. I think this might well continue given the confusing speech by EM yesterday which seemed to move Labour to a more Conservative position around a three year benefit cap.

    -It might have been less confusing if the BBC and Sky didn’t interrupt it for 30 Minutes to cover the rest of the news. We then had their political pundits interpretation of what they thought He meant.

    Haven’t seen them do that when Cameron makes a key note speech.

  34. f I have got it right watching the Speech from EM between interventions regarding crucial news such as a Teacher is still allowed to work and the weather is nice I think the gist of what Ed was saying was the central point in the welfare issue is shifting the mentality of all people, so that they once again take as much responsibility as they can for supporting themselves. That will still leave many people who are in genuine need and we should continue to support them.
    We should also do all within the power of a government to discourage redistribution for the poor to the rich in the form of high rents and subsidising low paid jobs.
    Frankly it is difficult to see much wrong with this attitude and it should appeal to LD’s and some moderate Tories as well.

  35. Jim Jam

    It is, true. Trouble is, I don’t know how you go about chasing them. If they are prepared to switch from LD to UKIP over 3 years, then where is the policy anchor that will secure them?

    The obvious one for Lab is immigration. But the trap here is that this risks alienating a bigger constituency of Lab supporters.

    As I’ve said before, my advice would be NOT to chase these voters for 15, but for Lab to hold its nerve and keep its current constituency together. BUT, at the same time, consider what it is that has alienated these ex-core Lab supporters, and develop the long-term policies to re-secure them, which, in the long run means a fairer distribution of the proceeds of future growth than we have seen for the past 25-30 years.

  36. I don’t think it’s Ed’s speech….it’s mainly down to the rise of UKIP in the last few months and maybe the fact that we are just moving beyond mid-term.

  37. Interesting piece Anthony, thank you. It would seem to point to online polling (over 24 hours), as your organisation does, removing some of those problems.

    I just wondered about ethnicity, religion and culture factors with online polling. I don’t suppose there are groups of people who regard using the internet as sinful are there? I wonder (that was a half-joke suggestion) if there are other problems in online polling?

    Today’s polls are consistent with previous ones and so i don’t agree with Rich (0959); we would need a few more like this one to draw his conclusion.

    I confess that the extract of EM’s speech chosen by the news vendors did nothing for improving his physical image, in my case anyway. If this is repeated among many voters, I would assume Labour’s focus groups will have revealed this factor. I hope for EM’s sake that I am in the tiny minority. My bias (geeky stuff apart) is the same one I have with policemen, namely, he looks so young. :-)

  38. Slight side track, anecdotedly the case for a quick EU in out referendum is building in the Labour Party.

    Eds position that the uncertainty is damaging and that until the Euro Zones impact on the EU is clear a referendum is too early is credible imo.

    He can switch though by saying as DCs 4 year time frame is creating 4 years of uncertainty we should have a referendum sooner rather than later. Call for one next year or promise one within 6 months of the 2015 GE if the Coalition won’t agree but how would DC react if Ed backed a referendum in Autumn 2014 so the matter is settles before the GE to allow the parties to draw up their manifestos knowing where we stand.??

    The Scottish Independance poll may scupper my Baldrickesque Plan or at least the timetable.

    One date to avoid would be the GE date itself in my opinion, although I am sure UKIP would like.

  39. @leftylampton

    I tend to think a section of the electorate is absolutely certain to vote, not out of any deep interest in the politicical process, more a kind of commitment extended list ticking. Every six months pay the water rates… every two years clean the gutters… four/five years vote in GE etc.

    These people are billiard balls and more are more likely to be knocked about by a media narrative/personality concerns.

    As you say (according to Ashcroft), maybe 3% thought it was ‘time for a change’ in 2010, didn’t like Brown but didn’t trust the Tories either, they then decided ‘Clegg is a washout’ and went to Labour for a while, now they think ‘that Farage is a sort’.

    My instinct is that there will be a significant accommodation between Tories and UKIP after the EU elections/before the GE. That 3% would be put in doubt by such a tactic, also the 0.5% (Lab 2010s saying UKIP to Ashcroft), but what of the 9% of VI who voted Tory in 2010 but now opt for UKIP… might not some of them be billiard balls too?

  40. @Steve, I broadly agreed with what he said, as frankly it’s much more moderate than the worry of a lurch to a harder left position. My point is that there will be a chunk of labour voters on the more socialist left who won’t like it, so it might chip away at some core support. In fact I saw somebody on sky news yesterday who called for a new party of the left. I can’t remember the name, maybe somebody can enlighten me!?

  41. Steve

    EM’s belief is, I think, that we’re at the once a generation-and-a-half moment when the terms of the socio-economic debate shift. We had it in 45, we had it in 79. I think he sniffs a growing swell of feeling that the approach of the last 30-odd years, where the proceeds of growth went disproportionately to the wealthy and the growth in wages and living standards at the middle-to-lower end of the scale were stunted, is not an acceptable state of affairs for the next 30 years. It is indefensible to allow 1929 levels of income inequality to continue into the foreseeable future.

    New Lab tried to deal with this by redistributing through tax credits. And at the end, they were left accused of allowing the welfare budget to explode.

    It’s a clumsy term, but EM’s “predistribution” kind of captures the idea of where we ought to be going. And his instinct is supported by a flourishing academic field on the centre-left, looking at the effect on growth that letting the wealthiest off the leash over the last 30 years has had.
    (E.g. http://www.epi.org/publication/raising-income-taxes/
    Have a look at the bullet points a few lines from the top – suggesting that higher top tax rates have no significant effects on growth, but do have big effects on reducing inequality – and that the optimal top tax rate on the Laffer curve may be 60-70%. Discuss…)

    I’ve said times many that I see us in a re-run of the 70s, with an un-telegenic weirdo with a grating voice as LoO. But a weirdo who has correctly sniffed the wind and who sees the great socio-economic shift coming, even if he/she is, for political reasons, not able to fully express the policy shift that will come over the next decade. And who may well get to drive through that shift over the next decade because of the incompetence of his/her opponents.

  42. Lefty Lampton
    If your analysis is correct, would you not regard this group as non-voters in 2015? If they do bother, are they not more realistically to be regarded as Independent voters, having no effect on overall results? A UKIP vote from this group will most of the time be a wasted one under FPTP. Only the Con UKIP switchers will not be ‘wasting’ their vote by voting UKIP, as it will lead to the possible loss of the Con seat, where this applies. The law of unintended consequences possibly applying here.

    I suppose a few LD seats could be affected too, but generally, I don’t think sitting LD mps needed those ‘here today, gone tomorrow’ votes. (Perhaps David Heath as example?).

  43. @Rich,

    I don’t think Ed’s speech will either enthuse or put off Labour voters. Let’s be honest here – the vast majority of people know what austerity (or whatever you want to call it) will continue regardless of who is in office.

    The question is not so much how much is going to be cut but what is going to be cut, and which groups are going to face (disproportionately) the biggest burden.

    It’s these kinds of questions, along with how the economy is performing come 2015, which will have a greater affect on who wins in 2015 IMO.

  44. Lefty – what we don’t know if they are ant-Eu, anti-immigration or really just NOTA voters (which is what I tend think without any evidence most are), probably a mix but in what proportions? Would be very difficult and probably expensive to find out.

    Some have asserted that these could in up Tory based on being anti-Eu but if that was their main concern why would they vote LD in 2010 although perspectives change of course.

    I would not want a LP strategy and platform to move in a certain way just to try to grab these voters but if we can speak to their concerns effectively all well and good. If they are NOTA then there is little Labour can do except be polling well enough to be the home for these voters who don’t want to waste a vote on candidates who can’t win a seat.

    We will gladly take protest votes.

  45. Peter,
    I think Lefty has answered your question to me fairly challenging whether EM is just NuLabour mark 2.

    It is difficult to pinpoint but I recall 94-97 and it was clear we decided that trimming was the deal in order to win a GE but the sense and mood is very different this time.

    I hope this will become clearer by 2015, in fact by next September to inform voters in your referendum.

  46. Jim Jam
    I just agreed (in advance) with your NOTA comment but mine was thrown into mod. I indeed wonder if these should be regarded as non-voters (essentially)?

  47. PS:

    If anyone doesn’t accept that income inequality hugely widened from 1979-2010, have a look at Figs 3.7-3.9 in this report (by the IFS – hardly a bunch of bleeding heart pinkos)

    http://www.ifs.org.uk/comms/comm124.pdf

    Or their conclusion

    “Income inequality in the UK fell sharply in 2010–11, with the widely-used Gini coefficient falling from 0.36 to 0.34. This is the largest one-year fall since at least 1962, returning the Gini coefficient to below its level in 1997–98. Although this reverses the increase in income inequality that occurred under the previous Labour government, it still leaves income inequality at a level much higher than it was before the substantial increases that occurred during the 1980s.

    “Inequality fell in part because the largest falls in income took place at the very top of the income distribution, with income at the 99th percentile falling 15%. The introduction of the 50% marginal income tax rate on incomes exceeding £150,000 per annum is one of the major drivers of these substantial falls in income at the very top, and therefore also explains some of the decline in the many summary measures of inequality that are sensitive to changes in the extremes of the income distribution. Ongoing changes to the taxation of very-high-income individuals will continue to influence when such individuals choose to realise their incomes until at least 2013–14 and may affect reported net incomes for several years after that, meaning that it will be difficult to identify the ‘underlying’ trends in top incomes in the coming years.”

    In other words, we had a huge increase in income inequality in the 80s. The rate of increase tailed off under Major, Blair & Brown, but did not reverse. There was a big blip in 2010/11, but that was mainly due to the very richest managing their income in response to the 50% top rate. Overall, we still have a historically very high level of income inequality.

    As the report says:
    “Although inequality seems likely to increase in the next few years, it is unclear whether the current government has a view on the level of income inequality that would be desirable or acceptable and, if it does, what that level is.”

    This underpins EM’s constant whisperings about the system having failed the oppressed middle.

  48. JimJam (10:36)

    I agree on every point.

  49. @Howard

    My reply to leftylampton also went into mod… I’ll try a bit of it:

    I tend to think a section of the electorate is absolutely certain to vote, not out of any deep interest in the politicical process, more a kind of commitment extended list ticking. Every six months pay the water rates… every two years clean the gutters… four/five years vote in GE etc.

    These people are billiard balls and more are more likely to be knocked about by a media narrative/personality concerns.

  50. @leftylampton

    I tend to think a section of the electorate is absolutely certain to vote, not out of any deep interest in the politicical process, more a kind of commitment extended list ticking. Every six months pay the water rates… every two years clean the gutters… four/five years vote in GE etc.

    These people are more likely to be knocked about by a media narrative/personality concerns.

    As you say (according to Ashcroft), maybe 3% thought it was ‘time for a change’ in 2010, didn’t like Brown but didn’t trust the Tories either, they then decided ‘Clegg is a washout’ and went to Labour for a while, now they think ‘that Farage is a sort’.

    My instinct is that there will be a significant accommodation between Tories and UKIP after the EU elections/before the GE. That 3% would be put in doubt by such a tactic, also the 0.5% (Lab 2010s saying UKIP to Ashcroft), but what of the 9% of VI who voted Tory in 2010 but now opt for UKIP… might not some of them be likely to change their VI in unpredictable ways?

1 2