Is Labour’s poll lead actually 3% not 10%?

by Atul Hatwal

So what is the real Labour lead? Sure, we’ve all seen the polls, and they tell a consistent tale across different pollsters. Looking at YouGov, the latest results from the weekend have Labour 10 points up over the Tories, 42% to 32%.

It’s a commanding lead but for those who remember the 1980s and 1990s, there remain nagging doubts.

At the end of 1980 Labour was registering week after week of double digit leads, peaking at 24% for Gallup in mid-December. But we all know what happened in the 1983 election.

Almost a decade later, it was déjà vu.

In 1990, Labour was once again posting massive poll leads. Between the end of February and end of April, Labour averaged a 22% lead across nearly 20 different polls. Impressive. Except, once again, we all know the result of the 1992 election.

The purpose of this trip down a rather painful stretch of memory lane isn’t to be a Cassandra. The future is not written and any form of poll lead is better than a deficit.

But caution is needed. Taking these leads at face value can breed complacency and for Labour, the experience of the past thirty years is clear: as the actual general election draws near, the poll leads have regularly evaporated.

Since those heady days of Dave and Nick in the rose garden, there has been a fundamental shift in how the public regards the government; and David Cameron in particular. The question is how would this translate in the polling booth? Would voters turn away from the Tories, and more pertinently, would they choose Labour?

The problem with attempting this judgement has been the absence of polling data that can be compared to an actual election, outside of the general election.

While there is a regular cycle of local council elections punctuated with by-elections, the pollsters rarely poll these specific areas, and even on those rare occasions when they do, only after the campaign is underway. So it’s almost impossible to compare like with like.

But regional elections offer a new opportunity. London has been polled by YouGov regularly since 2010 and recently voted.

Looking at the polling following the start of the government’s omnishambles with the budget, but before the election campaign was front and centre in London, gives us a snapshot of poll support comparable to current mid-term national polls.

Comparing this polling against the final results in London then gives us a percentage by which current national polling for Labour and Tory can be adjusted, to reflect a likely actual vote.

The YouGov London poll after the budget had Labour at 50% with the Tories on 33%. Taking the Assembly vote as the best proxy for actual voting (mayoral performance being subject to the personality distortions of Boris and Ken) the final result was a Labour victory on 41% to the Tories 32%.

This means the polling following the budget over-stated Labour’s vote by almost 20% and the Tories by 3%.

Adjusting YouGov’s latest national poll proportionately by these percentages transforms the electoral picture: Labour still leads, but by 3%, on 34% compared to the Tories’ 31%.

A lead is a lead, but it’s close. Almost within the poll’s margin of error. And these are the worst of times for the government: political incompetence, economic recession and a judge-led inquiry roaming around the minutiae of their private dealings with Murdoch.

It could be that this is the pattern for David Cameron’s government, and the current shambolic performance continues all the way to the next election. Could be.

More likely, by 2015, there will be some form of economic upturn, and Leveson will have concluded. In this more benign political environment for the government, even allowing for Number 10’s serial bungling, would Labour’s 3% lead hold?

The lesson for Labour is that the party cannot rest on its current lead. History, and London’s election, shows it is likely to be much closer in reality. To resist a Tory fightback, Labour needs to press the advantage. Now.

Atul Hatwal is editor at Uncut

Tags: , , ,

21 Responses to “Is Labour’s poll lead actually 3% not 10%?”

  1. Sue Marsh says:

    What kind of psephological wizardry is this? It’s like Osborneomics 2+2 is 73 but only when you take of 6 for deflation when inflationary is at 3.6%

    Labour leads currently sit somewhere between 10 and 18%. Even the latest ICM poll, the gold standard for a very long time, so accurate they’ve been spot on for decades had Lab lead at 12% last time, 6% this time (May) so with MOE, perhaps 9% a fair guess.

    What do we know in actuality? That given all the results of local elections – I know, breathe, there is a world beyond London – Labour’s results were around 9% up too. YouGov adjusted their polling some time ago to reduce Labs share, yet still fairly consistently give a 10 – 12 point Lab lead. Probably overstating still. But entirely consistent with local results.

    It’s 2012 and Lab are not popular. As u point out yourself. E-Mili’s ratings are dire. Those saying they wld vote Lab just now say they will do so through gritted teeth. This is no ’91.

    Perhaps re-write this entire article in one sentence – “Labour, don’t be complacent”

    No need to construct a false, convoluted narrative based on a few metropolitan voters.

  2. Andy says:

    YouGov asked separate voting intention questions in their London polls for assembly, mayoral and general elections. Which set of figures did you use when manipulating the facts to fit your preconceived conclusion?

    Never mind, either, that polling methods have changed (generally for the more accurate) since the 1980s, and eve-of-election polls at the 2010 election were generally within a few points of the actual result.

    It’s true that there’s no cause, yet, for complacency, and that governments tend to enjoy a swing towards them as an actual election approaches (though the “economic upturn” you predict seems unlikely at the moment), but to claim that the pseudo-mathematics of Atul Hatwal is a better gauge of public opinion than established, credible pollsters is stretching credulity somewhat.

    Focus on Ed Miliband’s personal ratings instead – a much happier hunting ground, considering your agenda.

  3. Felix says:

    And of course we all remember how Labour’s strong lead in the polls in the mid 1990s to the mid noughties translated yet again into catastrophic defeat in 1997, 2001, 2005.

    Talk about being selective with history just to suit your argument.

  4. Dan Judelson says:


    “The YouGov London poll after the budget had Labour at 50% with the Tories on 33%. Taking the Assembly vote as the best proxy for actual voting (mayoral performance being subject to the personality distortions of Boris and Ken) the final result was a Labour victory on 41% to the Tories 32%.

    This means the polling following the budget over-stated Labour’s vote by almost 20% and the Tories by 3%.”


  5. Matthew says:

    Mmm, the YouGov national poll before the Mayoral election split 42-38 in London region for Labour. This would imply polls understate Labour lead by about 250%,.

  6. Richard Livingstone says:

    This is an exercise in voodoo psephology, as UK Polling Report explain (, who I think we can all agree know a bit more about polls.

    What is true is that a 10% lead now does not automatically mean a similar lead in three years time, and your final point about pressing home the advantage is therefore right. This could be a period of mid-term blues for the Coalition – that doesn’t mean the polls are wrong, it just means that people change their mind.

  7. Anon E Mouse says:

    Sue Marsh

    Did you actually read this article before you launched into your (usual) diatribe?

    Why is it the left hates any opinion that isn’t what they believe anymore. Has the New Labour control freakery not gone away?

    How I long for the old days of free debates in Labour instead of the response Sue Marsh shows here…

  8. Tom says:

    London is not a good measure of the country as a whole. It has a much greater plurality of views than the rest of the country. Also, the election was held under AV+ which is distorting. Given that AV+ has FPTP AND PR type elements people are freer to vote for whichever candidate they like, especially on the top-up sheet. Tactical voting, which is extremely common in a GE, is less pronounced. As a result I find it hard to agree with your numbers although I think the ‘don’t be complacent argument is salient’

  9. john P reid says:

    good stuff Atul, But it will fall on death Ears OF Ed miliband’s fan club of Medhi Hasan And Sunny hundal. The real question is If the Tories win the next election with A majority of 20 can Ed be ousted, or will he do A kinnock and hang on to lose in 2019

  10. Brumanuensis says:

    With respect, Atul, I have to disagree, for similar reasons to Richard, Sue and Andy. I agree with your broad point about not being complacent, but the polls that you quoted after the budget were presumably for a GE, whilst the London Assembly likely works on different dynamics. In addition, we have to bear in mind the effect of turnout, which was down on 2008 at around 38% (if I recall correctl ), which might have skewed the figures somewhat.

  11. Paul Scott says:

    Interesting. Also, bear in mind that much UKIP support (currently around 8%) will in practice go back to the Tories in a General Election, since Cameron will make some anti-EU sound bites, and UKIP sympathisers won’t want to let in Labour MPs by splitting the right-wing vote. So in a real General Election, the Tories and Labour would probably be about neck & neck, with the LDs holding the balance of power again, possibly?

  12. Adam Gray says:

    Dan, The difference between percentages and percentage points explains your confusion.

    32 is 97% of 33 – the one point difference = 3%. 41 is 82% of 50% – hence the 9 point difference is “almost 20%”.

    This is one of the tactics Lib Dems use on their leaflets when they can’t actually show a dramatic REAL change in actual percentage points.

  13. John P Reid says:

    Brumanuensis, the 2008 second prederences Mayoral election had Boris on 1.5million Ken 1.03 million, this time it was Boris 1.05 million second preferences Ken 992,000 and both of these were considerably up on2004, In fact the fall came more in outer london this time than the 2004 as literally 10’s of thousands of people who didn’t vote at all in 2004 Mayoral election came out In bexley Bromely to oust Ken by voting boris,

    I know in Havering that Boris vote fell by 4000, and Kens fell by 1000 ,yet Ken still got as many votes in 2012 as he did in 2004, the real reason for the fall was Boirs wasn’t very good and his voters (who ahted ken) stayed at home) but not enough of them stayed at Home to oust Boris.

  14. ThePurpleBooker says:

    I think it is fair to say we should not be complacent but also not unrealistic. I think any sane person can take comfort in the fact that the Tories will not win a majority and that Labour has very good chances of winning next time round. I’d challenge the idea that UKIP voters will rush to the Tories. If Labour is the first party to promise a referendum, then they most certainly will have more reason to rush to Labour (just for a referendum) than to get to the Tories.

  15. Mike Hartley says:

    Two things:

    One, in the London Elections, Livingstone’s support – not the Tories’ – was underestimated, not overestimated.

    Two, read a real polling expert like Anthony Wells, who has answered this here:

    Three, stop endlessly wheeling out arguments (or non-arguments in this case) to bash Ed Miliband. Just say “I don’t like him”.

  16. John P Reid says:

    mike hartley
    it wasn’t Livingstones support was underestimated, it was the Idea that peopel who dilike boris were going to come out and vote for him anyway to stop ken,a nd half of them didn’t that’s why Kens vote went up, the reluctant boris voters stayed at home, not Ken was More popular than had been previously thought.

  17. Mavrodi says:

    Never mind, either, that polling methods have changed (generally for the more accurate) since the 1980s, and eve-of-election polls at the 2010 election were generally within a few points of the actual result.

  18. uglyfatbloke says:

    Regional variations and FPTP can distort national figures very easily; pollsters do ‘weight’ their results in the light of experience, but that will not necessarily provide a reliable forecast. As an example, in 1974 the SNP got about 1/3 of the vote, but only about a dozen seats. The disparity between Holyrood voting and Westmisnter voting may well be a thing of the past now, in which case the gnats could benefit hugely from FPTP – maybe a total of 30 gains – without that being apparent from polling data. The same sort of thing on a shallower, but much wider basis might bring unexpectedly good results for UKIP – they may not make gains, but have a dispropotionate effect on the tory vote.
    Equally, the glib-dumbs are suffering in the polls just now; if they recover markedly that will have an impact on seats without necessarily having much impact on the headline Labour vote. Alternatively, if the glib-dumbs collapse even further (and they may well) it’s hard to estimate who will benefit in vote share or in seats. Probably it would be Labour overall, but the gnats in Scotland…but that’s no more than a guess.

  19. Sue Marsh says:

    You might notice that Anon E Mouse doesn’t like me much.

    It’s my insistence on peppering my comments and articles with lots of those pesky fact thingies (as above). All that evidence and research gives him a headache.

  20. Tom Miller says:

    QTWAIN, surely?

  21. Anon E Mouse says:

    Sue Marsh

    I don’t dislike you at all just the fact that you distort the truth and do the Labour Party no favours with your tribalist approach.

    I also find your childish comparisons to cuts in benefits, (some supported by the vast majority of Labour voters – 70% at the last poll) to the holocaust deeply offensive considering the industrial scale on which human beings were murdered by fascists.

    I think that any merit in your case regarding cuts is ruined by your dishonest approach to the subject but as a Lib Dem supporter I seriously hope you continue along your current path since your distortions of the truth do give cause for amusement and serve to remind Labour voters just how hypocritical and awful the last government actually was.

    You remind me of Gordon Brown that time he claimed to have “Saved The World”.

    Are you sure you don’t work for the coalition government Sue Marsh?

Leave a Reply