There are only two voting intention polls in the Sunday papers – the regular weekly Opinium and YouGov polls for the Observer and Sunday Times respectively.

The Opinium/Observer poll has topline figures of CON 35%, LAB 33%, LDEM 6%, UKIP 15%, GRN 7%. This is the first time that Opinium have shown a Conservative lead since back in 2012, just before the Omnishambles budget. As ever though, it’s just one poll – taking a broad average of the polls suggests that the actual position of public opinion is a very small Labour lead, so it’s inevitable that normal sample variation will spit out some Tory leads from time to time. Doesn’t mean much unless they start getting more frequent. Tables are here

The YouGov/Sunday Times poll has toplines of CON 33%, LAB 34%, LDEM 8%, UKIP 13%, GRN 6%. The one point Labour lead is very much in line with YouGov’s average. The 13 percent figure for UKIP is equal to the party’s lowest from YouGov this year, but a lowest we’ve already seen a couple of times, so again, not necessarily anything new. Tables will be on the YouGov website tomorrow.

428 Responses to “Latest Opinium and YouGov polls”

1 3 4 5 6 7 9
  1. This is what happened in 1979.
    ‘On 1 March 1979 a Referendum on the Scotland Act saw a majority vote for devolution, but a threshold imposed by anti-devolution MPs requiring 40% of the electorate to be in favour was not reached due to low turnout. When the Government refused to implement the Act (per its special threshold legislative clause) the Scottish National Party MPs put down a motion of no confidence. After consulting with the Liberal Party to confirm that they were still supporting motions of no confidence, Leader of the Opposition, Margaret Thatcher put down an Opposition motion which took precedence. The debate was scheduled for Wednesday, 28 March 1979’

  2. Thanks for the clarification Graham. It’s a long time ago…

  3. Good afternoon everyone from a very wet West central Scotland.

    Great result for the Hoops..Hail Hail and YNWA.

    On the vote of no confidence…………… If a party is running a minority government and won’t budge an inch on any proposals one of the other parties would like to see introduced and at the same time expects that party to toe the line (sort of taking their support for granted but giving nothing in return) then it’s perfectly reasonable for that party to put the boot in and get shot of them.

    I know the polls for rUK are neck and neck but I will put my neck on the line and say David Cameron will be PM after May’s election and that could be in the form of another coalition, minority or although unlikely, a majority.

    The principle opposition party is making no headway what so ever and at best their hopes for power hinges on a Labour + SNP + PC + Green + Respect + Sinn Féin + SDLP coalition with c&s from the Lib/Dems.

  4. @ Old Nat,

    the judgment of the Speaker seems to be the only definitive ruling

    If it comes down to a question of whether Bercow wants to bring down Cameron the SNP can rest easy- that confidence motion will be tabled for 9:00 on the first day of the new parliament.

  5. I have been reading Piketty again this morning which is a rare treat after nine years of having no time to read at all;

    Returning to the UK polls, compared to the 2010 GE result, the 11 polls between February 13th and 20th reveal:

    Labour +3.8%
    Conservative -4%
    UKIP +11.7%
    LD -15%
    Green +5.9%
    SNP/PC +3.1%

    Unless by some miracle LD can hang onto more seats than SNP obtain Cameron is finished as of May 8th.

    Do the math. Cameron has to ensure that the Conservatives basically keep or come close to the 303 they now have, which assumes UKIP will obtain no more seats and/or any losses to Labour are made up by gains from LD.

    Then LD has to obtain a minimum of 23 seats if the Conservatives can hang onto 303, which means no losses to Labour or UKIP and the Conservatives winning every LD seat LD loses in E&W..

    Further if SNP take 9 seats from LD in Scotland that means the coalition loses 9 seats in their column, as happens if either Labour or Green take Bristol West.

    In fact LD cannot lose more than 24 seats in E & W if SNP take 9 in Scotand, and the Conservatives cannot lose any seats in E & W at all.

    If Conservative and LD fall below a combined 326 then all bets are off, unless Sinn Fein abstain then the required majority fall to 323. Assuming Sinn Fein keep 5 seats.

    Then if DUP agrees to support Consevative that means if LD have 23 and Sinn Fein abstain, the Conservatives can lose no more than 11 seats.

    For the coalition to stay toegther Conservative and LD combined have to obtain 315 seats, and not lose more than 44 seats to SNP, Labour, UKIP or Green.

    Remember every seat UKIP obtains is a seat that Conservative/LD do not have, and cannot be added to the Conservative/LD column, because LD have said that, just like SNP, PLaid Cymru will not work with Conservative, they will not work with UKIP.

    If you think Conservative and LD can lose fewer than 44 seats between them then Cameron will remain PM. But any less than 315 combined and he is no longer PM.

    The fact that SNP takes seats from Labour in Scotland really does not matter, in terms of who forms government.

    But the loss of LD seats in Scotland to SNP or Labour, impacts the coaltion vote requirement in E and W. The same applies to Wales where any gain by Labour or Plaid Cymru from Conservative and LD is a plus.

  6. Spearmint


  7. AC
    ‘No headway whatsoever’ is overstating it – Labour are 4% up in GB as a whole compared with 2010 – and 5-6% up in England.

  8. @OldNat

    Absolutely agree with you :)


    “Labour are 4% up in GB as a whole compared with 2010 – and 5-6% up in England”

    Is that it? In Five years they managed +4%!! What about the polls that gave them leads well into double figures? Where are they now.

    Don’t answer that..I see them up on the right hand side in the archives . :-)

  10. Since this is a coalition government is not the fact that LD are down 65% on their 2010 result a significant factor here?

    In 2010 the coalition government supposedly had 59.1% of the electorate supporting it.

    In February 2015 that is at best 40%. So this government has lost 32% of it electoral support, in five years.

    The fact that 60%, the reverse of 2010, do not support the two governng parties should be seen as significant, should it not?.

    The problem is that those 60% cannot agree on a clear alternative to the coalition.

    The dilemna for Cameron and Clegg, if the latter retains his seat, will be what to do if between them they do not have 326 seats, and I would have thought less than 50% electoral support.

    If Clegg cannot support bringing in Farage, he has to leave the coalition and it breaks up, and Cameron resigns.

  11. Graham

    Yes, and we all waited with baited breath trying to discern which way Frank Maguire was going to vote. After travelling all the way from Northern Ireland for one of his rare visits to the Commons, he abstained and the Government lost by one. It all came down to his vote. The only time in his life he held so much power!

  12. Andy Shadrack

    Unless we have the skills of Unicorn, Election Forecast etc we’re kind of stuck with these fairly useless %s of party support that the pollsters are set up to supply, and which the media commission (although these two factors should probably be the other way around).

    No one has ever suggested that a UK UNS exists. It is clear that a GB UNS no longer exists.

    The question is whether there is a useful English UNS anymore.

    As long as we have FPTP, voting shifts in safe English constituencies tell us nothing about May’s parliamentary arithmetic.

    There have been calls for proper English regional polls, but I’m not sure that answers the problem.

    The extent to which Lab or Con outdoes the other in terms of MPs is entirely down to the English marginals.

    If they demonstrate a single pattern, then the UMS (Unified Marginal Swing) within them would serve a useful purpose.

    If they vary, due to demographics or political balance, then they could presumably be grouped to provide samples which could identify relevant UGMSs (Unified Group Marginal Swing). If the likes of Election Forecast have built such into their modelling, it would be good to know the details.

    ““Ed hasn’t said whether he’s a Protestant atheist or a Catholic atheist yet.”
    He’s a Jewish atheist, surely. Still, I suppose we don’t know if he’s a Protestant Jewish atheist or a Catholic Jewish atheist so fair comment.

    You can be sure that Papa Miliband will have read the boys The Protestant Ethic and the Rise of Capitalism as one of their bedtime stories, so Ed will almost certainly have plumped for not believing in a God that requires him to demonstrate his salvation as against not believing in one that offers salvation in return for a hundred Hail Mary’s. I believe, however, that he will keep not believing in Jehovah in reserve for those occasions when political circumstances require him to tip his opponents into the pit of everlasting hellfire.

  14. John Pilgrim

    Clearly many MPs in formerly safe seats subscribed to the 18th century Scottish Presbyterian doctrine of predestination.

    Believing themselves to be the elect was the downfall of more than Holy Wullie!

  15. Norbold
    There is a further interesting tale re the March 1979 vote. One of the SNP MPs – Hamish Watt – having voted with his colleagues in the Opposition lobby decided that he would cancel out his vote – ie effectively abstain – by also voting in the Government lobby.As he approached the lobby the Division Doors slammed shut in front of him so preventing his second vote. Had he been five seconds quicker his vote would have counted and Callaghan’s Government would have survived.What a difference to history those five seconds made!

  16. In 2010 the coalition government supposedly had 59.1% of the electorate supporting it.
    In February 2015 that is at best 40%. So this government has lost 32% of it electoral support, in five years.

    another bit of clarity…there’s precious little chance that the tories and the liberals will have 326 seats together, even with the vagaries of FPTP…

    Old Nat isn’t right. a 4% swing from C to L in England is a 4% swing. If that happens, the tories will lose lots of seats to labour…

    the argument isn’t about whether seats a lost given a certain swing%, the argument is about what the actual swing% will be….some say there will be swing back, or that the polls are inaccurate and hide some deeper factors, while others say the polls won’t change that much. believe me, if the swing is anything like what the polls currently show, i.e a 4% swing from the tories to labour in England & Wales, the tories will be out on the rear.

  17. @ON

    Currently I doubt that there can be an English Marginal UNS that makes sense either. Those UKIP-ready coastal marginals aren’t much like other marginals elsewhere – there are midlands marginals that are immigrant rich and overall unfriendly to Farage. The SW goes its own way. Northern marginals are more traditionally split LabCon. Without doing a lot more marginal polling in individual seats we’re guessing.

    This is where Nate Silver had it easy. There are a few marginal states and all were polled individually and extensively. I doubt Ashcroft has enough dosh to provide for the equivalent here.

  18. Andy: The record low level of support for the two main parties is a consequence basically of the rise of UKIP and, to a lesser extent,, SNP. Both phenomena are temporary IMHO and likely to be reversed in the long run. I still think we might see a 5-6 % drop in UKIP support before the GE, which would put the Tories in the position to win a plurality in Parliament, though not necessarily a majority. Despite the scepticism of most commentators here, I don’t see a Labour government taking over in May.

  19. Interesting graphic from the Economist from earlier this week tracking population movement vs party holding the constituency

    They conclude that Labour held constituencies have had the most population growth, so the old in built Labour lead is diminishing.

    “We found that the old pattern of faster population growth in Conservative areas has inverted. Labour constituencies are now growing more rapidly than Tory ones.”

  20. Postage Included

    That is as I suspected. Even in the English marginals there is too much variation for even that level of Unified swing to make much sense.

    As to the cost – Number Cruncher’s fascinating input on his analysis of the microdata from the YouGov Scottish crossbreaks was pretty accurate.

    I wonder if that is a low cost database that could be interrogated to discern useful patterns.

  21. Thanks Graham, I didn’t know that story.

  22. Richard

    The Economist has an interesting view of “Britain”. The northern bit is already grey with the ashes of Smaug’s breath, so has no people. :-)

    Other than that, they say it is “mainly London that is driving that”.

    Sounds rather like it is simply the effect of gentrification of poorer London boroughs due to silly house prices.

  23. MBruno
    “The record low level of support for the two main parties is a consequence basically of the rise of UKIP and, to a lesser extent,, SNP. Both phenomena are temporary IMHO and likely to be reversed in the long run.”

    In the elections since 1992, the percentage of votes for Lab and Con as a % of the total vote is as follows:

    1992 76%
    1997 74%
    2001 72%
    2005 68%
    2010 65%

    So it looks as though there is a long term trend away from the two main parties, and there will soon (i.e. 2 or 3 elections) be a tipping point where they no longer get a disproportionate number of seats.

  24. @Oldnat

    I did notice that. That must be why they called the chart “Red Faced”.

  25. @ Richard

    I was interested by your 1.10pm analysis of different pollsters’ speeds of response to game-changing events. I note that you apparently don’t use any statistical criteria to detect step-change so any of the drops you identify could well have been ‘lucky breaks’. Of course, if you insisted on statistical stability then YouGov – with its much higher polling frequency – would have a huge advantage over the other pollsters and must surely win the gold medal in this category.

  26. Richard

    btw Thanks for answering my query about lost deposits on the previous thread. As you will have guessed – I totally missed it. :-(

  27. @Unicorn

    Sadly I am not gifted like you and others here. I did stats101 and was glad to be rid of it! So I just look at simple number sequences and am happy to be educated when necessary.

    I was just testing a theory that I had that party id weighting means that real changes will be not be detected in a timely manner, because pollsters keep weighting back to a past point. For that example, my theory was wrong… I do want to repeat the same thing though for the emergence of new parties – UKIP and Greens – to see which pollsters seemed to detect that first, and which ones were lagging. But hard to put an exact date on when those parties really did start emerging or got a big boost.

    I think we are going to see some large swings later in the campaign – possibly after the debates, and it will be useful to know which pollsters detect real change first, vs which ones assume that change is MOE and weight it away.

  28. Even if between now and May 7th UKIP support drops by 4% to 6.7% and all goes to the Conservatives that still only gives the Conservatives the same vote as last time.

    Any shift from Green or SNP will likely only go to Labour, as both parties and Plaid Cymru are staunchly anti-austerity.

    If Labour support does not move between now and May 7th, they will still win some Conservative and LD marginals.

    If you look at Election Forecast you can see the number of seats SNP and Labour stand to gain that are above an 80% chance with current standings.

    In contrast LD have no such gains in the offing and the number of Conservative gains all come from LD at 80%.

    So from my perspective at least, the odds are not in Cameron’s favour.

  29. MBRUNO
    Andy: The record low level of support for the two main parties is a consequence basically of the rise of UKIP and, to a lesser extent,, SNP. Both phenomena are temporary IMHO and likely to be reversed in the long run. I still think we might see a 5-6 % drop in UKIP support before the GE, which would put the Tories in the position to win a plurality in Parliament, though not necessarily a majority. Despite the scepticism of most commentators here, I don’t see a Labour government taking over in May

    Almost agree with you except for the drop in the SNP VI. Polls show they have the highest retention of all the parties with those polled now and planing to stick with them at polling day.

    Also it was at this stage in 2011 when the SNP were ahead of Labour they went onto win a landslide at the election so where I agree with you on UKIP I disagree with regards to the SNP.

    I’m not saying the SNP VI will not slip but I can’t see it slipping that much. As far as Scotland is concerned the election is over.

  30. BTW – Tonight in the Cricket World Cup (22.00 GMT), England play Scotland at Christchurch.

  31. On

    “which ones assume that change is MOE and weight it away.”

    That should probably read – which pollsters assume they have an unbalanced sample and weight the change away.

    A prime example of that is Scotland – you can see how many pollsters still down weight all those over eager SNP voters in national polls, when we know from Scotland only polls that there has been a real change there.

    I think I’ll also look at who picked up the rise of the SNP in national polls as well…

  32. Richard

    I think I’ll also look at who picked up the rise of the SNP in national polls as well.

    Not possible because there were very few of them before the referendum:

    (I seem to have spent most of last year complaining about this). And all of them came from Survation. Loads of people asked about the referendum and some about Holyrood, but no one polled for Westminster because everyone ‘knew’ that Labour was bound to win most of the seats.

    You could use crossbreaks from GB polls I suppose, but then it gets confused by the variation caused by small sample sizes. You also wouldn’t know how much of the differences were caused the pollster not getting the crossbreaks right rather than them not reflecting changes.

  33. @Allan Christie

    I am not disputing what you are saying about Scotland so that means that LD will likely lose 9 seats there.

    In the Northeast Labour can only pick up 4 seats, two from Conservative and 2 from LD. So the question becomes whether UKIP will make it more likely that the Conservatives lose their 2 seats.

    In the Northwest Conservatives have 22 to lose and LD 6, and depending on the UKIP split Labour may do relatively better.

    In the East Midlands LD have no seats and only 2 in the West, so this truly remains a battleground between labour and Conservative.

    In Eastern the Conservatives took 52 to LDs 4 and Labour’s 2. So the question again becomes how big of a swing away from Conservative to UKIP does it take for Labour to become competitive and is this a region where some tight three way races with UKIP will occur?

    In the Souh East Conservatives took 74 to LDs 4 and Labours 4. So again the question becomes how big a swing from Conservative to UKIP does it take for Labour to become competitive and does this also become a region of tight three way races with UKIP.

    Interestingly I see London as quite tight between Labour and Conservative, with the big question given that the LDs have lost 50% of their supoport in the capital, if not more, to whon do their 7 seats go.

    While the Green have obtained more support in every election since 2010 than LD I am unsure if that translates into any seats, and am unclear about UKIP too. The immigration leaflet put out by Labour suggests they think they are in trouble in London.

    I also found this article on social housing in Barnet very instructive:

    And this interview with a former Labour member now running for the Green Party:

    And I can think of similar interviews with former LD supporters who have gone Green or to Labour and UKIP supporters from Conservative.

    Which is why I am having such a hard time with swingback.

    That leaves the southwest where Conservative had 36, LD 15 and Labour 4. In 2010 LD took 34.5% of the vote, but in recent weeks all I have seen for LD in crossbreaks is highs of 11% to 14%.

    I am also surprised that UKIP are not a factor in this region, but the Green Party has sometimes appeared as competitive with LD as UKIP.

    I suspect a lot of LD support originated with Labour and has now decided to go to Green rather than back to Labour.

    In the 2014 European election this was the only region of significant growth for Green and the one region where Green and Labour were within 2 points of each other.

  34. Mr Bruno
    I still think we might see a 5-6 % drop in UKIP support before the GE, which would put the Tories in the position to win a plurality in Parliament, though not necessarily a majority.

    the ukip dip from 15% to 10%, if it happens, won’t necessarily mean the tories will jump 5%… as commentators here pointed out ages ago…ukip’s rise in 2012-13 from 4 to 10% or so was almost entirely at the expense of the tories, from 10-11% to 15% they got vi from other parties.

    The unwind of ukip, should it happen, and i am not convinced it will, won’t purely benefit the tories…indeed there are signs that labour are benefitting too, as i suspect the lib dems might.

  35. Adam Boulton says “I hope months of mess suits you:thats what you’re voting for.

    Heres the link but you will have to buy the paper to read adams amazingly perceptive article.

  36. @ Richard

    But hard to put an exact date on when those parties really did start emerging or got a big boost.

    Yes – it’s a tricky task to detect an uptick in VI (or the reverse). @Catmanjeff uses his CUSUM method for this but this depends on a package that costs a bit. To answer a similar (but not always identical) question I have been calculating departures from long-standing trends. From the observations you make at the end of your last comment, I suspect that the most interesting shift look for over the coming weeks would be a rise in the Conservative VI. As it happens, either of the methods can probably deal with this. The Tory long-term trend is close to horizontal. (There is a tiny, though statistically reliable drop, but this only gives away less than 0.01% a month, and so you can safely treat their trend line as being clamped firmly to 32%). Given this, it should be possible to pick up a week’s worth or fortnight’s worth of upturn by using a t-test to compare the 20-40 recent polls with the 500-600 that preceded them. Different approaches are needed for all the other parties where the previous year’s worth of polls cover a great deal of genuine variation.

  37. Boulton…..perceptive???? Never. The man runs with the hares and the hounds.
    A lightweight.

  38. A 4% swing from incumbent to opposition is actually quite good: the average since the war is about 3%, and even considering that since Mrs Thatcher came to power, swings have generally been larger, it’s still about the average for recent times.

    It’s only slightly less than the average swing from “kick out” elections, where the party of government is replaced. As the Conservatives didn’t really win in 2010, thus a bit less swing is needed to do that.

    The election predictions have been rather boringly stuck at low 280’s apiece for the two largest parties for some weeks now. It’s not looking good for DC unless that supposed swingback actually appears. He’s going to need quite a lot of it.

  39. @ James Peel

    The unwind of ukip, should it happen […] won’t purely benefit the tories…indeed there are signs that labour are benefitting too, as i suspect the lib dems might.

    Quite so. That was why I asked @CMJ for his updated churn analyses. Recent plots have provided good support for your observations here.

  40. Keith – The Conservatives did win last time, you sound like Alistair Campbell.

    I assume, to be consistent, when Miliband piolls less than 37% and gets less than 305 MP’s you will say he hasn`t won either.
    The Conservatives outpolled Labour in 2005, I would say that means they won.They have enacted a lot of their policies, again an outcome of winning.

  41. 07052015

    Adam Boulton says “I hope months of mess suits you”

    In reality, as long as a majority of voting MPs agree to continue current taxation for a further 4 months, the UK Government will continue as before till a government that commands the confidence of the House emerges – most matters are handled by executive decision, not Parliamentary vote.

  42. I found the 1974 – 79 parliament most interesting historically, particularly some of the tactics carried out by both govt and opposition whips, it recently spawned a play at the national theatre… another story on the 28th March 79 confidence vote was the Labour MP John Broughton was seriously ill but offered to be taken in by ambulance to the commons, one of the opposition whips – can’t recall his name but he was briefly a speaker in the 1980s – offered to abstain so broughton’s life was not endangered…

  43. ps the opposition whip was Bernard Wetherall….. he was interviewed on bbc parliamnet about it…

  44. Good Evening everyone.
    ALLAN C: Hail, ynwa etc.
    I agree with you; it looks as the centre left will have more than 325 seats on May 8th. Lab plus SNP plus SDLP, plus Green. Not sure whether the Lib Dems will in the end go to the Blue Corner.

  45. Opinium have been pretty consistent in finding that among current supporters they will “definitely” vote as follows;

    SNP 94%
    Conservative or UKIP 83%
    Labour 78%
    Green 72%
    LD 63%
    Plaid Cymru 52%

    If you add in “probably” will vote:

    SNP or Plaid Cymru 100%
    Conservative or UKIP 97%
    Labour 95%
    Green 93%
    LD 89%

    I am not surprised by Conservative and UKIP lining up, and note the two groups of nationalist party supporters are the most motivated to vote.

    This is the first time I have seen Green party supporters more motivated to vote than another “major” party.

    Someone also needs to explain to me if the unweighted count for Green is 146 and that for LD is 73 and the wighted count 146 for Green and 75 for LD how they end up at 7% and 6% in the poll respectively.

    Is a Green polling respondent in an Opinium poll only worth roughly half that of an LD one?

  46. Andy Shadrack

    Opinijm use a weird weighting system on Party ID, which upweights Tories and LDs, and downweights SNP & PC.

    There is no obvious rationale.

  47. @Andy

    I believe that Green party down weighting has something to do with a small party bubble downweighting

    “I’ll leave you with a couple of stats. ICM re-interviewed UKIP and Green intenders after the 2010 election to understand what they ended up doing: only 60% of UKIP intenders voted for them, only 42% of Green intenders voted for their own pre-election choice. For these surges to occur, smaller parties need to convert at much higher levels, or what we have, is just the latest polling bubble.”

  48. You can also read this

    “So, taking all those into account, there is no clear reason for the Lib Dem overestimate in the pre-election polls. My guess is that a little bit was down to late swing, with other bits down to disproportionate response from Lib Dem supporters during the enthusiasm of Cleggmania, and pollsters having samples that are rather too well educated and interested in politics. As I’ve said above though, actually proving this is an entirely different matter!”

    I think the pollsters have learnt some lessons from then.

    Who is more likely to respond to a poll?
    A person who votes Labour/ Tory at every election but doesn’t really like Cameron or Miliband, but will still vote for them?
    Or someone really upset about Labour/ Tory and who has decided to vote for the ‘new’ party, and wants to make that known.

    I’m sure that is a valid thing that happens, but measuring how much is froth vs how much is real change in voting intention must be very difficult.

    You can see by the different levels of Green/ UKIP support being shown across the board that there is no agreement across the industry on the best method.

  49. @Richard

    But a poll is a snapshot Not a prediction. So weighting because you think someone will change their mind is WRONG. Weighting should only be used to get a representative sample

  50. Patrick Wintour of the G has tweeted that a big political story will break later tonight.

    Not sure if it is UK or polling related.

1 3 4 5 6 7 9