YouGov March poll

YouGov’s monthly poll for the Telegraph unsurprisingly shows voting intentions largely unchanged from their last poll, conducted just a week ago – CON 39%(nc), LAB 32%(+1), LDEM 17%(+1). The poll was conducted between the 26th and 28th March.

On the “forced choice” question, asking if people would prefer a Labour government under Gordon Brown or a Conservative goernment under David Cameron, the Conservatives lead by 10 points. Taking the responses of current Lib Dem voters as a possible guide to how tactical voting will pan out at the next election, Lib Dem voters would still prefer a Brown government to a Cameron one by 40% to 32%. This is also reflected in a question asking people who their second preference party would be – more Lib Dem voters said Labour (21%) than Conservative (16%), although the party who Lib Dem voters were most likely to name as a second preference was the Greens with 30%. Conservative voters were most likely to name the Liberal Democrats as their second preference (23%), but large sections of Tory supporters named UKIP (18%) or the BNP (12%) as second choices. There was less of a challenge to the Labour party from fringe parties – 33% of Labour voters named the Lib Dems as their second choice, 14% the Greens and 9% the Conservatives.

Asked who would make the best Prime Minister, Tony Blair is once again marginally perfered to David Cameron, perhaps reflecting the recent improvement in the public’s opinion of Blair recorded in YouGov’s brandindex trackers over the past three weeks. Cameron’s lead over Gordon Brown has risen to 5 points, with 30% naming Cameron as the best PM compared to 25% for Brown.

YouGov also carried out a voting intention poll in Scotland. Constituency voting intention, with changes from the last YouGov Scottish poll, are CON 13%(nc), LAB 29%(nc), LDEM 14%(-4), SNP 35%(nc). Regional top-up votes stand at CON 15%(+1), LAB 27%(-3), LDEM 12%(-2), SNP 33%(+1), GRN 6%(+1), SPP 2%(+1), Solidarity 2%. Weber Shandwick’s swingometer translates this into a Scottish Parliament with 46 SNP seats, 39 Labour seats, 18 Lib Dem seats, 19 Conservative seats, 5 Greens and 2 others.

47% of respondents in YouGov’s survey said that if the SNP emerge as the largest group they would like to see them form an executive through a coalition with a smaller party (10% would prefer a minority SNP exective, 18% a continuation of the Labour/Lib Dem executive). There was majority support for a referendum on Scottish independence with 64% of people respondents saying they support one. If a referendum is called, 28% of respondents said they would vote for independence, with 51% favouring the status quo.

57 Responses to “YouGov March poll”

1 2
  1. Anthony, your readers may wish direct links so that they can peruse the detailed datasheets for themselves: – YouGov data for Scotland poll – Telegraph’s Scotland story – Alan Cochrane’s Telegraph column, re Scottish voting intention poll

  2. Certainly it now looks like- bar an SNP blunder–that the SNP will be the largest party in Scotland. This becomes fascinating–I have always argued that accidentally Blair will be the biggest radical in UK history,.

  3. I welcome any polls which show Labour on either 32% or 33% because I’m pretty convinced that’s what they’ll get at the next election.

  4. How can you welcome a poll just because you agree with its findings?

    You may well be right, but since it is 3 years (probably) before a general election, how can you possibily lay claim to predict the future from this poll?

    “events dear boy”, as a great man once said

  5. All the German opinion polls agreed at their last General Election , the snag was they were all wrong .

  6. Does this not highlight as bogus the lazy assumption that disaffected Labour supporters are drifting to the BNP?

    According to this poll, it is Conservative voters who are more likely to support them.

  7. Well, Mark, when a result is pretty close, pollsters often get it wrong. When looking at the difference between two large aggregates, small errors in opposite directions can change the outcome.

  8. So, that’s 3 Scottish polls in 2 days. The averages are: SNP 35.7/33.3, Lab 28.3/28.3, Con 13.0/13.0, LD 13.0/12.3, Greens 3.0/6.7, others 7.3/6.4.

    For this the WS swingometer gives SNP 48, Lab 42, LD 18, Con 14, Green 7.

    Concerning the Greens, I find it strange that ‘Scottish opinion’ still allows a Green constituency vote. And the range of 4-10 % for the list vote looks large – can that be explained by sampling error? Or are they using different methods?

  9. Clive – when a result is pretty close, people notice that the pollsters get it wrong. The polls in 1997 for example weren’t actually very good (with the honorable exception of ICM), but because they predicted a landslide and there was a landslide no one noticed.

    If there was only 1% between the parties and the polls got it wrong by 3% people would notice. If one party was 10 points ahead and the polls got it wrong by 3% no one would be bothered. Such is life.

  10. Well confirms it’s reputation as the most consitent and accurate of the pollsters – that’s within 0.1% of the WMA and your Std is +/- 1.6% which is lower than the others.
    BTW have you picked up that Shaun Bailey was adopted as a Conservative candidate last night?

  11. In other news, Baxter’s site now has the Tories just one short of a majority:

    So frustratingly close.

  12. NBeale – been a busy week, quite a few candidates I need to add to the guide.

  13. Well Greenpousse if the currrent WS swingometer prediction turns out to be correct on the day then the Greens could find themselves in an interesting position and one wonders what their price will be for supporting whatever administration emerges as a result. A Labour defeat will of course be a watershed in Scottish politics but if the Tories fall to 14 seats many of their activists will finally give up the ghost and a realignment of the moderate centre right will surely take place. Is there any other country in Europe that has so feeble a centre right party?

  14. Re-alignment? Where exactly would the Scottish centre-right go to?

    I stand by my opinion that the Scottish Tories need an independent Scotland. An independent Scotland will not be able to maintain such a dependency culture and so will see a resurgent centre-right.

    I can’t think of any independent country has such a week centre-right Nick, but there are plenty of regions across Europe where there is high dependency on the state like in Scotland where it is common.

  15. NBeale – YouGov are the most consistent, but not necessarily the most accurate. The “benchmark” is the (unknown) state of public opinion, not a weighted moving average, or other combined measure of the opinion polls themselves, which may all be inaccurate and share a common bias.

  16. The yougov result on independence is interesting becuase it differs slightly from previous questions by specifically mentioning the “Scottish Parliament”.

    This shifts the emphasis from Edinburgh Rule v London Rule, and perhaps gives the unions a powerful weapon in any referendum campaign.


  17. Ernie- yougov are the most consistent and have consistently shown to be the most accurate! Communicate shouldn’t even be discussed, and quite frankly neither should Mori. ICM and YouGov are the only reliable polls and this is shown plainly through their accuracy on election night

  18. Ben Murphy. Two points re Conservative second choice. Some of us have a second choice, but nil likelihood of voting that way so your extrapolation is unfounded.

    Secondly, polls can be unreliable. You’ll hardly find a lesser-spotted Tory in any of the places where the BNP has won council seats (Sandwell, Chelmsley Wood, Dagenham). Traditional Labour supporters have revealed their second choice often enough in practice.

  19. Harry, I don’t dispute that, based on a variety of evidence, YouGov may well be the most accurate poll. I have not studied the evidence. My point was that the accuracy of a particular poll can’t be judged by comparison of its results with the mean (however calculated) of all polls.

  20. the clever conservatives should now be talking brown up,the last thing they need is a new,untarnished figure, like milliband(even though he has had a hand in the current mess).its an outside bet with milliband,with the current labour constitution,but worth a punt.

  21. Anew TNS poll has the SNP even further ahead than before with a prediction of 51 seats.

    It will be unvailed in full today before an STV television debate between Jack Mcconnell and Alex salmond.


  22. Anthony,

    When will we see the “scottish voting Intensions” table updated, and is there enough information yet for a graph?

    It’s interesting that people are predicting 4 seats for the Greens, as it’s concieveable that on the current percentage they could fall short in almost every region and get as few a 1 or as high as 7.

    There are two explainations. Firstly that they are just giving them 7% of the 56 lists seats about four, or they have done some kind of regional calculation and decided what four regions the greens will get over 7% in.

    So far I haven’t seen any regional breakdowns, but there was an interesting piecein Scotland on sunday about a libdem document telling local parties how to attack the Greens as they represent a particularthreat to the libdems.

    This included the well tried LibDem tactic of printing a stretched graph in election handouts that portrays the greens as in fourth place in a two horse race and effectively a wasted vote.

    Given that on the YouGov breakdowns the LibDems are the most likely to give their list vote to another party and the greens arguably the most likely benefactor, they are clearly taking the Green threat seriously.

    Not only might it cost them seats but there bargining position post may 3rd, would be greatly weakend if they were to be part of a rainbow rather than two party coalition.

    Having said that the greens have made it clear that they would rather support a new Zealand type deal where the supported the execuitive on the budget and confidence motions but were not in a formal coalition.

    All in all the weekend polls seem to confirm that the mood is one of “time for Change” and that we are seeing all the paties other than Labourand the SNP being squeezed, with the Tories and LibDems well down on figures from a month or so back.


  23. I wonder if the two situations in scotland and uk wide are signs of incumbent governments being punished/mid-term unpopularity or just a labour party problem(?)

  24. Keith, I’d say a bit of both.

    The populus web site has breakdowns by region for Scotland for it’s last poll which are interesting as it gives some indication of regional variation. Ofcoursethe sample sizes are very small so margin oferror is an issue, but it can be comparedwith the national average to give some idea of how things are going.

    Even so I findd it hard to convince myself that what it’ showing for up here in the Highlands, the SNP agead of the LibDems in both FPTP and list cna actually be true, muxh as i would like it to be.

    The regionaal figure for Jack v Alex are also interesting, although I doubt if Nicol Stephens will be happy with his Glasgow support for First Minister, it’s an asterx, which I think means he has the support of no one that was polled…..


  25. I am also puzzled by 4 seats for the Greens. They have one dead certain seat – Lothian. Glasgow and South are also pretty safe, decent results in 2003 and lots of ex-SSP votes to go round. The 4th is probably Highlands. Mid Scotland, North East and the second Lothian seat are marginal. Possible pick-ups are West, Central and a second Glasgow seat (all on the back of SNP constituency gains, which increases proportionality in these regions, and the SSP collapse). More than 10 seats looks near-impossible.

  26. Interesting to note that the final sets of questions in the yougov poll would appear to indicate that core SNP support is 25%, with 43% opposed regardless.

    The surge in headline support is thus based on disaffection with the government rather than desire for independance. If that is a primary driver, we could see some interesting tactical voting, which will make extrapolation from national figures harder.

    LD figures seem to have slipped a bit in the latest batch of polls – are they likely to suffer from association with Labour ? That could spell bad news on the regional list votes, which would make any FPTP seats lost all the more painful..

    And the campaign has only just begun…

  27. Greenpousse,

    We don’t yet know if the latest poll has been weied by likelyhood to vote, certainly previous TNS polls haven’t. It’s possible that what we are seeing is people who said they would vote Labour, but were highly unlikely to vote, are now saying they’d vote SNP but probably still won’t vote.

    In effect the SNP lead is now strong enough to have influenced the opinions of non voters, which might indicate either a sea change in scottish politics or absolutely nothing at all.


    Asa matter of interest on the latest Scottish Westminster Voting Intention results available how many seats are you predicting for the SNP in the next UK general elections and what seats would the SNP gain or lose.


  28. “The regionaal figure for Jack v Alex are also interesting, although I doubt if Nicol Stephens will be happy with his Glasgow support for First Minister, it’s an asterx, which I think means he has the support of no one that was polled…..”

    I thought generally an asterix means something but less than can be measured. Eg not 0 but

  29. … less than 1%

    Post didn’t get posted correctly as I used a less than sign. Site must have thought it was HTML code and blocked it.

    I’ve seen the asterix used like that in statistics before when something is so low its almost negligible, but not quite zero. Rather equivalent to saying “trace” on a foods’ statistics list.

  30. * in tables means that they got somewhere between 0 and 0.5%, so a figure that would be rounded down to zero, but wasn’t actually zero.

    (Now you’ve said it, I think using trace would be nicer, e.g. Veritas recieved but a trace of support)

  31. Okay I’ll conceed the point,

    There is probably somebody in Glasgow that wants Nicol Stephens for first minister. I wonder where his mum lives?


  32. ICM’s new poll for scotland is out this Morning (Tue 3rd April).

    It’s largely unchanged from the last one with,

    SNP 32/31 (-2/+1), Labour 27/27 (-2/-1), LibDem 19/17 (+3/+5), Tory 13/12(-3/-3).

    It’ still gives the SSP and Solidarity five points on FPTP on the WS predictor as it can’t split them, and for some reason the Scotsman quotes the greens at 7% but WS at 5%. If the 7% is correct the greens are up 3% from the last Scotsman ICM poll.

    Disappointingly for what is supposed to be a quality paper the Scotsamn story focuus es on how disappointed the SNP must be by the fact that there momentum has stalled.

    Given that the real change seems to be a Libdem jump and a Tory slump, you’d think they would mention that. They are also predicting only 2 green MSP’s on a 7% list vote.

    On WS with 7% list I get at least 6 greens and up to 8 using the same figures. I have a suspicion that the fact that the ICM poll still has the SSP and Soildarity as part of FPTP is having an effect

    What is also becoming an issue for all the polls is the wide variation in the LibDem and Tory votes. I have seen over the last week the Libdems get anything from 16 to 24 seats…..


  33. I can understand the problems of predicting the Green seats that Peter Cairns has pointed out, but I can’t understand why anyone should believe a poll result showing such a dramatic fall in their fortunes without some obvious explanation.

    The Greens, as did SSP, improved their position in the second election as the electorate became more aware of the effect of the second vote, and this should continue. The recent troubles of SSP/Solidarity might also help the Greens.

    The Conservatives are at the tail end of a generation long decline which is still continuing, though now very slowly. I can’t see them losing more than one MSP.

    I have no idea how many independents may be successful, but I’d be surprised if Margo Macdonald was the only one. The reason is the same as the reason the Greens will do better than predicted: the electorate are now more experienced, and the change for council elections will draw attention to the possibilities.

    If the Conservatives/Greens/Independents get in total more seats than predicted, which party loses seats? SSP/Solidarity will lose votes in total and may both fall beneath the threshold for seats except in Glasgow. Even so, the changes in both constituency and regional seats may be much fewer in number than predicted.

  34. I suspect the Greens on May 3 will take a seat in most regions and that this will restrict the apparent the Liberal Democrat advance.

  35. Large sections of Tory supporters named UKIP (18%) or the BNP (12%) as second choices.

    Are separate figures for Scotland availble?

    UKIP & BNP are below the horizon in Scotland simply because they are English nationalist parties.

    In England, many of their natural homophobic anti-PC racist supporters – the C2’s that brought you Margaret Thatcher – confuse voting for an MP with POP idol or Big Brother. They don’t want to vote for losers so, if they vote ar all, they vote Conservative as a second choice anti-foreigner party.

    These people do not exist in Scotland in significant numbers, and the Scottish Conservative party did not recieve a boost from this segment of the population during the Thatcher years. Indeed, many economic Conservatives have shown an interested in the potential for an independent Scotland and have joined the SNP.

  36. Peter , You may have the advantage of reading the actual newspaper but the online article is mixing up 2 sets of figures from the ICM poll and a private SNP/Yougov poll . It seems to be the latter that mentions Greens at 7 % .
    There is certainly a variation in LibDem and Con support with ICM/Yougov and TN Sys 3 and Progressive . On the 2003 performance we should perhaps be ignoring the latter 2 companies .

  37. Mark,

    I think the TNS can be looked on as dubious as as far as I am aware it still doesn’t filter by likelyhood to vote.

    Similiarly predictions on ICM for FPTP have the problem that they are still quoting figures for the SSP.

    A problem created by list only parties is that the polls seem to be designed to be backwards compataible, which is fine for looking at trends and works pretty well for FPTP Westminster but makes predicting results in Scotland more difficult.

    The other fly in the ointment is that if it is as most people now suspect a straight Jack v Alex election then the tactical voting permutations are massive.


  38. The ICM poll has Greens on 5%, SSP on 5% and others on 3% (according to the SNP website, others as Peter said above). The Scotsman seat distribution is Lab 39, SNP 44, Con 15, LD 24, Green 2, which leaves 5 for others (presumably of which 3 others – as last month – and 2 SSP).

    The SNP’s You Gov poll is Lab 26/27 (list vote first), SNP 33/36, Con 14/13, Lib 14/16, Green 7, SSP 3, Sol 1, other 3/8 (according to the Scotsman)

    ICM strangely has Lab on the same %age for list and constituency vote – I find that very unconvincing, constituency vote should be higher than list vote (same goes for SNP, Lib and Tories of course). ICM and YouGov also still have 8-9% for others in the constituency vote, which seems way too much (2003: 3.5%+6.2% SSP). I am also not sure about ignoring System 3 – I think in 2003 they were the only one who saw the SSP and the Greens coming.

  39. Look at the System 3 record from 2003 – it was pretty appalling

  40. Greenspousse,

    If ICM and YouGov are showing 8% for others then it is Green, SSP, Solidarity, UKIP, BNP, ScotlandsVoice, The pensioners Party and the likes of Margo McDonald and Denis Canavan.

    The fact that some of these parties and indeed Individuals, might not be standing FPTP isn’t something they can filter out. It’s not knowing where these votes will actually go FPTP especially that makes predicting seats difficult.

    Given that YouGov give others as Green 6%, SSP and Solidarity 2% each and 3% the rest, it could be that the SNP could pick up as much as 4% of that.

    If there is a UKIP or BNP vote of 2% then concievably the tories could pick up 1%. The LibDems night also pick up 1 or 2%, with the samller parties still getting 1% to 3%. It’s also possible that some of the Solidarity SSP % could return to Labour.

    So adding 6% of the 9% for “Others” to the Yougov constituency headline figures might give.

    SNP 35%(+1% to 3%), Labour 29%(+0% to 2%)LibDem 14%(+1% to 2%),
    Tories 13% (+0 to 1%). Which I would tentatively suggest means the SNp lead over labourand LibDem over tory is probably 15 higher.


  41. “The fact that some of these parties and indeed Individuals, might not be standing FPTP isn’t something they can filter out. It’s not knowing where these votes will actually go FPTP especially that makes predicting seats difficult.”

    Why would it matter where they go FPTP? They won’t win anyway on a fraction of 8%.

  42. It matters because it could influences who wins in the constituencies, and the distribution of constituency seats has consequences for the number of list seats for any given percentage.

    There is a good example on, if SSP wouldn’t have stood in the constituencies, and the votes had gone to the SNP, then even on an otherwise identical list vote, Lib-Lab would not have had a majority.

    Like in 2003: The two Labour defeats in the Lothians to the Tories and the Libs effectively led to one extra Green and one extra Socialist.

  43. Philip,

    I think Peter’s point is that since the minor parties are not (in the main) standing in the FPTP section, those votes will de facto go to one of the four main parties. How they split could be decisive in some seats – but then, we also don’t know how FPTP preferences will vary by region either. For example, it would be ridiculous to presume that either LD or Cons will have even distribution of their votes – whether it be 13% or 17%. Plus, there will undoubtedly be significant tactical voting. This all makes the FPTP section very diifcult to predict, except in those seats where one party is clearly closest to ousting Labour – eg Dumfries, Dundee West, Eastwood, Govan or Western Isles. Strangely I can’t find any seats where LD were within 10% of Lab in 2003, so despite any increase in their poll ratings, they are more likely to lose FPTP seats than gain them..

    I think the real issue for both LD and Con is how much of their support they can retain on the regional list in those areas where their potential support has voted tactically.

    To a certain extent, regional differences will also affect Greens and SSP. While Greens may struggle to make 7% across Scotland, they will probably cross the bar in several regions. In theory they could pick up anything from 0-8 seats, but I expect 2-5 is more likely.

    Final point, it is possible that the regional composition of polling samples could explain some of the volatility in Tory and LD ratings in recent polls.

  44. Paul H,

    thats about it, where the 7% or so that people seem to be attributing to the Greens, SSP and Soilidarity for FPTP, must go somewhere.

    That is a different issue from tactical voting, where people vote for a candidate other than there first choice, in that it is a case where their first choice candidate doesn’t exist.

    What would be interesting is for a pollster to ask a “second choice” question for the Local seat to see how the green SSP and Solidarity votes actually break down.

    On your point about different regions have a look at this poll from the Edinburgh Evening News.

    It clearly shows the posssibility of the Greens picking up two list seats in the Lothians. It also indicates that the SNP might be the largest group on Edinburgh City council, although of course way short of control.


  45. I am not sure how many polls ave asked about Scottish local government voting intentions, but it could be that a combination of PR and a “time for change” mmod might be taken by many as an opportunity to try something other that labour rule at council level, for the first time in over a generation.


    Just as a matter of interest did anyone in YouGov like the idea of holding a 20,000+ megapoll the week before the Scottish elections.


  46. The EEN survey was not a properly conducted poll and was clearly a survey rigged by the SNP . You should get some clue that something is wrong by the fact that it claims turnout at the elections will be 90% !!!!

  47. These figures from the latest SNP commissioned YouGov poll are on the SNP website, dated Monday 2nd April.

    YouGov poll for the SNP, fieldwork 27-30 March, sample 1,064

    The constituency vote (excl don’t know, won’t vote)
    SNP 36%, Lab 27%, Lib 16%, Con 13%, Oth 8%.

    Regional list vote (excl don’t know, won’t vote)
    SNP 33%, Lab 26%, Con 14%, Lib 14%,Gre 7%, SSP 3%, Sol 1%, Oth 3%,

    Based on WS this would give,
    SNP 48, Lab 38, LibDem 19, Tory 15, SSP1, Green 8, Ind 0, Others 0.

    Again the *% others in the consyituency vote includes some green and SSP votes that will go elsewhere so that isn’t that will effect the result, and the list independants will probably see at least one elected as might one of the minor parties.

    I think the similar figures are on Peter Kellners column.


  48. Mark,

    I hadn’t looked at the details only the story, and you are right the 90% is way out of line.

    I don’t know how it was conducted, so it could well be it was an on line poll without any filtering effectively with a self selecting panel.

    However, that would explain the high SNP vote if it ment that the SNP was far more organised than everyone else, and it also shows a higher vote for the greens and the SSP.

    I think a more probably outcome might be that it reflects people who are both politically active and have internet access.

    I so a comment on a similiar effect with regards to blogging and on line oppinion in both Venezuela and Iran, where the media in the US and Europe were taken aback by election results.

    One reason given was that they paid more attention to what they were seeing on the net than on the ground, and that had ment that they underestimated the impact that the unwired poor would have on the result.

    So it may still have a limited validity in showing us how active and motivated the parties are in Edinburgh and indeed how likely they are to get their vote out.

    If anyone knows how the 2,000 were selected I’d love to know.


  49. On the face of it the results of the Edinburgh ‘Evening News” poll together with most of the others appear to signal disaster for the three Unionist parties.But before SNP stalwarts break open the champagne a word of warning. A lot of voters across the political spectrum want to punish Labour and as the biggest opposition party in Holyrood the SNP is the most obvious vehicle by which to drive their message home. I suspect however that a significant number of these unsettled voters will,in the last 10 days of the campaign think twice about voting SNP. Differential turnout is also a factor to consider. Is it not the case that Lib Dem, Tory and yes Green supporters are usually much more likely to turn out than those of the others? I predict that each of these three parties will do better than the polls are suggesting which means that others will do less well. It is still all to play for!

1 2