We’ve six months to go until the European Elections, but rather surprisingly we already have the first voting intention poll for them. A YouGov poll for the Taxpayers Alliance/Global Vision has European Election voting intention figures – with changes from the 2004 election – of CON 35%(+8), LAB 29%(+6), LDEM 15%(nc), UKIP 7%(-9), GRN 5%(-1), BNP 4%(-1), SNP/PC 4%. The poll was conducted between the 6th and 8th January.

Taking into account the reduction in seats on a uniform swing this would lead to the Conservatives winning 30 seats (up 3), Labour 24 (up 5), the Liberal Democrats 9 (down 2), the SNP and PC one each and UKIP 4 (down 8). The Greens would lose both their seats, while the BNP would fail to secure one. What isn’t clear from the newspaper report is how the question was asked, and which parties were included in the prompt. Back in 2004 YouGov included minor parties like UKIP in the main question prompt and, as a result, ended up over-stating the level of UKIP support – it will be interesting to see exactly how this question was prompted. Either way, how well UKIP will do is a hard one to predict. At the moment they get practically zero publicity in the mainstream media, so a huge drop in support is not a particular surprise. At the last European elections they received a lot of publicity because of Robert Kilroy-Silk, something that won’t be a factor this time. On the other hand, given their level of support last time the broadcasters may be forced to give them a lot more publicity in the run up to the election.

Accorded to the Sunday Telegraph report 10% of those who say they would vote Tory in a general election would vote UKIP in a European election, which implies that straight voting intention questions were also asked. There are no figures from this is the report, so again, we’ll have to wait for the full tables to see if they were.

The rest of the survey showed the usual Euro-sceptic feeling amongst the general public. Support for British entry into the single currency stands at 24%, with 64% opposed. On Britain’s relationship with Europe 16% support withdrawal, 48% a looser relationship and 22% the status quo. As ever, it is worth looking carefully at polls commissioned by pressure groups and we haven’t seen the wording of these questions yet, but there is nothing particularly surprising. The Sunday Telegraph’s report also says “Conservative support could fall still further if Mr Clarke makes a front-bench return, using his position to make high-profile interventions on European matters”. I have no idea if this is referring to something in the polling – the report doesn’t mention anything – or just the Sunday Telegraph speculating.

36 Responses to “First Euro-election poll”

  1. Anthony I agree it is surprising but also encouraging.

    I have linked to your thoughts on my blog and wonder if you could give me a link on your blog lists.



  2. These elections are for the politicians who take part in them.

    They confer no real executive power to those elected.
    They excercise no representative democracy for those who vote.

    What relevance have they for ordinary UK citizens?

  3. Anthony,

    Is there a long-run series for British attitudes to the EU, i.e. back 25 years (or whatever is available)?


  4. In the East Midlands the real issue is to whom will the anti-EU vote go, especially in the abscence of Kilroy Silk. Will Roger Helmer continue to con voters that the Conservatives are anti-EU with the meaningless phrase ‘a looser arrangement’ whilst also supporting Britian’s continued membership of the EU? On the basis of this You Gov poll is looks as though the ANTI-EU parties are in decline. But, we will have to see what happens for real in June. Remeber this is the one election when voters are free to vote to leave the EU without being influenced by their usual partisan opinion relating to Westminster. Interesting times.

  5. If UKIP is down 8 seats, and the Conservatives up 3 and Labour up 5 seats
    Then where have the 2 Liberal seats and the 2 Green seats gone to ?

  6. Gordon – the total number of seats has been reduced.

  7. Anthony – I’d heard their was going to be a reduction in the number of seats but can’t find any details – do you know which European Parliamentary constituencies are losing seats? Does this mean the threshold for minor parties has increased?

    I’m assuming if the Greens will lose theirs then seats in London and the South East are being reduced.

  8. Interesting first pitch in the campaign for the European Elections. Reflects mainly the state of the parties now. It will be interesting to see this mould to the particular circumstances of the campaign in the weeks running up to the vote.

    The fact that UKIP take part in this election at all reflects the moribund state of British understanding of the European Union and how it works.

    No doubt there will be a ramp up in anti-EU coverage in the more eurosceptic press as we approach May, which may affect UKIP polling numbers positively, although given the power of the EU in many areas I think it foolish for voters to waste any influence they do have in the Parliament on a set of politicians who have no intention of engaging (let alone understanding) properly the huge issues that the institution deals with.

  9. Having asked that, I’ve now found the answer on the Wikipedia page for the 2009 Euro elections:


  10. Anthony,

    Are your figures based on 72 or 73 seats for the UK ?
    As Lisbon has not yet been ratified – and this is now unlikely before June (hopefully never) – then the Nice figure of 72 should apply to these elections.

    Any idea how the reduction will be allocated ?

    Depending on where the reductions are made this could have a bigger impact on UKIP and LD than on Con and Lab as they fail to meet the hurdle for first or seat in some smaller regions.

    Is your prediction of only 1 for SNP accurate ? That would surely only hold true if Scotland were reduced to 5 seats and both Con and LD qualified for one each.
    Based on the regional break in this poll, six seats would go 2:2:1:1, but if there were only five seats it’s too close to call betwen either:
    – one of Con or LD failing to meet the hurdle and the five Scottish seats being shared 2:2:1; or even
    – both Con and LD failing to meet the hurdle so seats are shared 3:2 between Lab and SNP.
    Either way, I make that 2 MEPs for the SNP, unless,
    – All four parties qualify for at least 1 seat, and SNP losing to Lab for the fifth seat.

    Elsewhere in the UK, I suspect that UKIP will be a bigger loser than the LDs, and may even face a wipeout.

  11. On voting intention, analysis of the tables would suggest that the headline EURO figures of 35/29/15 would translate to about 40/32/17 for a UK general election.

  12. Paul is correct that the SNP must win two seats in Scotland even if their support remained the same as in 2004 – in fact the poll implies a large increase in support. Indeed on the basis that Scotland and Wales cast one in eight GB votes in 2004,a similar proportion in 2009 would imply that a third of voters in the two regions combined must vote for the nationalist parties.
    This leads me to question what weighting was given to those areas which are having local elections on the same day as turnout is likely to be significantly higher in those. As these areas are on the whole southern shire counties, turnout could be twice as high in Surrey or Somerset than in South Yorkshire or Strathclyde. If this factor was not taken into account this would overstate the likely Labour share and understate both Conservatives and UKIP. The large national share achieved by SNP/PC leads me to believe that undue weight may have been given to those regions.

  13. The Poll shows the SNP on 35% of the vote, although as it’s a UK poll the regional samples may not be accurately weighted.

    However as the SNP is averaging about 33% in polls it doesn’t seem that out of line. We got two of the seven seats last time with less than 205 of the vote, so i can’t see us getting one seat out of six with more than a third.

    Having said that I don’t see us getting three as even on seven seats the rule of thumb is add one and divide in to 100 to get the percentage needed for a seat which would be 12.5%.

    On that basis we would need to poll 38% to get three seats. Not impossible just unlikely.



  14. East Midlands loses one seat from 6 (2 Con, 1 Lab, 1 Lib-Dem, 1 UKIP 1 Kilroy Silk) to 5. Thus for me the question as to where the anti-EU and Kilroy Silk vote will go. People do vote differently for Europe over Westminster. The English Democrats are very active in the region.

  15. The SNP figures are plausible enough, although as you say the regional sub-samples can be dubious and it does seem unlikely that the Conservatives and LDs will both be as low as 11-12% there.
    My main question was whether it takes account of likely differential turnout between and within regions. In 2004 there was all postal voting in the three northern regions and the East MIdlands as a consequence of which turnout was around 5% higher than in the other regions in England. Also there were local elections on the same day in Wales and the Metroplitan boroughs together with a minority of shire districts and Mayoral elections in London which would have had the effect of boosting turnout in these places. In 2009 there will not be all postal voting and none of the Met authorities have elections, nor does London or wales but the shire counties of England do.
    This means that whereas in 2004 the turnout in Rotherham was scacely lower than it was in Rother (East Sussex), in 2009 it is likely to be very much lower (in 1999 for example it was half as much).
    I note also that 14% of the sample say they will not vote and a further 19% are undecided which means that two thirds of them say they will vote in a particular way. The reality is that probably only around a third of the total will actually vote at all – that is half of those who are saying who they will vote for will not come out to vote on the day. Of course these abstainers could be drawn proportionately from all parties, but one suspects that Labour supporters (and this is born out by previous expereicnce) are likely to be disproportionately represented.

  16. I doubt the Greens will lose there South East seat as there is no reduction in the number available there and they came 8/9 anyway in 2004. London loses one seat and the Greens may be vulnerable but again they had the penultimate seat so it is more likely Labour will lose their third seat.

  17. “A YouGov poll for the Taxpayers Alliance/Global Vision has European Election voting intention figures – with changes from the 2004 election – of CON 35%(+8), LAB 29%(+6), LDEM 15%(nc), UKIP 7%(-9), GRN 5%(-1), BNP 4%(-1), SNP/PC 4%. The poll was conducted between the 6th and 8th January.”

    You’re missing something. These numbers are 5 months before the election,. So comparing them directly to the election numbers is a tad off. Clearly, a party like UKIP, which comes into its own at euro election times, is going to find a change in numbers as people start to concentrate on hte euros.

    A better marker is what is support for UKIP now as against what it was 5 months before the last election. On that basis support is double last time around.

    As you know, yes, I work for UKIP

  18. Have you got any European election voting intention polls published around Jan 2004 Tim?

  19. No, I’m going purely on what I was told by the people who conducted this survey (they work about 20 metres away from me).

  20. Well I’m not about to draw comparisons from unpublished internal polls! :) The earliest published stuff I can find from 2004 was in April.

    Generally speaking polls are snapshots of now. It’s possible to speculate at what might happen in the coming months – like I have in the orginal post, but it’s just speculation. Comparing to “points in the cycle” implies some sort of inevitability.

  21. Anthony,

    Are you still standing by this predicting only one seat for the SNP, because I can’t see how it does with a projected vote 50% higher than the one that got us two.


  22. Peter – no, Scotland isn’t done properly yet. I just banged together a quick spreadsheet to do the projection using a uniform swing. The drawback of that was the swing towards the snp/pc was spread evenly across great britain, when obviously it would actually all be in Scotland and Wales. The projection therefore very probably underestimates how well the snp would do

  23. I agree with Peter cairns on this. if accurate (and as Peter says the figures are broadly in line with Scottish polls elsewhere), both the SNP and Labour should get two each. I’d predict the Conservatives will definitely get one seat (although the Scottish sample of the euro poll has them just behind the lib dems). The really interesting question is who gets seat number six – The SNP should be bullish about their chances of picking it up if they are polling at +35%, Lib dems are at c11% and the tories are in the mid-teens. Labour could also rally a bit a push the SNP all the way for top spot, thus creating a four-way battle for seat number 6. For tartan anoraks, this is going to be exciting!

  24. If Brown had been hoping to bury bad County Council results under “improved” European Parliament results by having both on the same day, he may have badly miscalculated.

    Of the 12 Euro regions, five have no County Council elections. In the other 7 regions (all in England) turnout is likely to be higher in those areas which do have County Councils than those that don’t. A cursory glance at the 2005 County Council election results – which took place on the same day as a General Election – confirms that these areas are – unsurprisingly – disproportionately Conservative. Thus, even if there had been no change in national party support since 2004, this would lead to Labour losing MEPs to the Tories. Given that, even though the Tories polled 4% more than Labour in the 2004 Euros, they were still well behind Labour in the national opinion polls at that time, it is hard to see how Labour could recover ground when they are now substantially behind in the national polls. Adding differential turnout to the mix could prove an electoral disaster for Labour in the European Parliament results.

    While Labour may hope to draw some comfort from the fact that the reduction in seats from 78 to 72 would, on 2004 results, hurt both the Tories and LD more than Labour, the reality is that the high UKIP vote in 2004 is likely to collapse, returning predominantly to the Conservatives. (Even if Ken Clarke may be viewed as a “gift” to UKIP, this will not save them from falling below the threshold to win seats in many regions.)

    The media will have fun for days after 4th June.

  25. Tam / Peter,

    FWIW, I predict that the Scottish result will be SNP/Lab/Con/LD in that order, but I am not (yet) going to stick my neck out on teh figures.

    This could lead to either SNP 2; Lab 2; Con 1; LD 1 or SNP 3; Lab 2; Con 1 depending on whether LDs fall below 11% and how far above 30% SNP go.

    Don’t think Tories could pick up the sixth seat. To do so they would need to have more than twice LD vote.

  26. Paul,

    I can’t follow your logic. Differential turnout makes no difference as the seats per region are fixed.


  27. In fact, if one assumes that in a region which has locals at the same time the vote increases for all parties in that region, then differential turnout will change the UK-wide precentages (in favour of the party that is strongest in the regions with higher turnout) but not the number of seats for each party.

    In other words, a UK-wide poll that correctly predicts the voting intentions of the voters across the UK and does not correct for differential turnout would get the percentage wrong but the seats right.


  28. IF one is pondering how turnout is going to change in the european elections, its worth remembering that for several regions last time round there were all-postal pilots, which presumably aren’t being repeated this time around.

  29. Christian,

    Your point would eb valid if there were differential turnout as between entire regions.

    However, the point is that in 7 of the English regions there will be County Council elections in part-only of a larger region. Since the County Council elections will be in areas where Tories (and to a lesser extent, LDs) are stronger, while the urban areas where Labour is stronger will not have local elections, differential turnout could distort the results in those regions. This could be particulalrly true in the West Midlands, North West and Yorkshire.

  30. Utterly meaningless considering how important “regional” variations in voting are under this system. SNP will get 2 seats in Scotland (might sneak a 3rd) Labour should hope to get 2, Tories 1, Liberals???? probably one as well.