ICM Welsh poll

BBC Wales had a new ICM poll of Wales out this morning, conducted in the aftermath of the Scottish referendum. Westminister voting intentions in Wales, with changes from the previous ICM Welsh poll in February, are CON 23%(-1), LAB 38%(-4), LDEM 7%(-2), Plaid 13%(-1), UKIP 14%(+7). This puts UKIP up into third place in Wales, though on a uniform swing wouldn’t give them any seat (on his blog Roger Scully projects these figures would produce 28 Labour MPs (up 2), 8 Conservatives (unchanged), 1 Lib Dem (down 2) and 3 Plaid Cymru (no change)). Asked about Wales’s constitutional future just 3% would support Welsh independence, 49% would support extra powers for the Welsh Assembly, 26% support the status quo, 12% would like the Assembly abolished.

Meanwhile tonight’s YouGov GB poll for the Sun has topline figures of CON 33%, LAB 37%, LD 7%, UKIP 13%

528 Responses to “ICM Welsh poll”

1 2 3 11
  1. Are there any other regions besides Scotland and Wales where the Tories don’t poll either first or second? I have my suspicions about the North West and North East.

  2. Mr N

    For Wales, see above.

  3. Oops, misread the post.

  4. And neither Scotland nor Wales are ‘regions’, thank you very much! They each have regions themselves, as has England. Not sure about NI, but then NI doesn’t seem to occupy much attention amongst those contributing to this site.

  5. I didn’t mean it in the administrative sense, but in the sense of “parts of a geographic whole” – in this case the United Kingdom, or, if you like, the island of Great Britain.

  6. Mr N

    “the island of Great Britain”

    You got something against the Isle of Wight?

  7. Regions have regions, and so on ad infinitum.

  8. Alec (fpt)

    The problem Scotland has is not that everyone dies on average younger than in the UK as a whole. Once Scottish people get to pension age, they are more or less as likely to live to the UK normal age as elsewhere in the UK.

    Do you have anything to back this statement up? The most recent ONS publication on life expectancy ( http://ons.gov.uk/ons/dcp171778_360047.pdf – page 6 for the relevant table) would suggest otherwise, with a gap of around a year and a half (on figures of around 20 years) between Scottish and English life expectancy from age 65.

  9. British Rail had a Scottish Region.

  10. I think we’ve had some of the few(?) Welsh contributors on here suggesting that people in Wales really dislike the Senedd, and think it should be abolished.

    Seems 12% of those in Wales agree with them, while 52% want to see its powers enhanced.

    There is obviously a demand for what Gordon Brown described as “a modern form of [Scottish] home rule within the UK … as close to a federal state as you can be in a country where one nation is 85 per cent of the population.”

    Though nobody has a clear idea of what he meant – or why he didn’t deliver such when in power.

  11. Old Nat,
    I am not Welsh but have lived here for nearly sixteen years.In my experience
    People are not very politically motivated but are broadly supportive of the Senedd.Which rather agrees with the polling.

  12. Ann in Wales

    The great advantage of polling is that it either confirms that the people you talk to are broadly representative – or that you live in a self-created bubble.

    I’m not surprised that you aren’t bubble-wrapped! :-)

  13. The Wales poll chimes with my experience too. There does nonetheless seem to be a certain amount of resentment at the way in which Cardiff appears to have captured a disproportionate share of the devolution dividend – not entirely unlike the way people elsewhere in the UK view London.

  14. @Oldnat

    “Though nobody has a clear idea of what he meant – or why he didn’t deliver such when in power.”

    Or what he meant when he was in power for that matter.


    Has anyone noticed how the big party total VI is dropping? Ten-poll moving average since 2011:


  15. Mr N
    It depends which elections you’re talking about. In the recent EU elections (the easiest to find regional totals for) you can add West Midlands and Yorkshire and Humber to your list.

    But note, Labour finished outside the top two in East Midlands, East of England, South East England and South West England.

  16. Statgeek
    Yes. Figures for last few GE

    Lib/Lab/Con as % of total votes
    1987 96%
    1992 94%
    1997 91%
    2001 91%
    2005 90%
    2010 88%

    Lib/Lab/Con as % of electorate
    1987 72.1%
    1992 73.2%
    1997 64.8%
    2001 53.9%
    2005 55.1%
    2010 57.3%

    Some of this is because of the rise of the SNP and to some extent Plaid, but a lot is UKIP, Greens etc.

    One interesting feature of this is that though the votes for the three main parties (as % of votes cast) has been in slow but steady decline, this has gone into reverse for the last couple of elections as a % of the electorate.

    I’m a bit puzzled as what this means. The total votes cast has risen for the last three GE (presumably because of freer postal voting). So do these figures suggest that most of the additional voters have not been supporting the old 3 main parties? Any other explanations welcomed.

  17. Roger Scully’s analysis is interesting in terms of assumptions about the contribution Wales might make to Labour winning a UK election.

    “ICM’s findings reinforce the point, made previously on this blog, that Labour support in Wales has slipped considerably over the past 18-24 months. In the four polls conducted in 2012, Labour’s general election vote share was always at or above 50%. Both the last two have had it below 40%. Indeed, it is notable that while Labour across Britain as a whole is running generally well ahead of the 29.0% vote share it won at the 2010 general election (as of this morning, Labour was on 35% in the UK Polling Report running average), in Wales Labour’s support level is now only 2% points above that gained in 2010.

    Two years ago, Ed Miliband could have confidently looked forward to Wales delivering him several seat gains at the general election; now, Welsh Labour’s seat harvest looks likely to be much smaller. That is probably the most important single message to come out of the recent polls on general election voting intention here in Wales.”

    Labour is doing well in England. Maybe EV4EL should become Ed’s battle cry?

  18. Pete B,

    The changes in the last two elections were small and can presumably be overanalysed. I think you’re roughly right in your surface explanation, although I don’t know the precise figures for postal voting. It conceivable (but speculative) that postal voters tend to be older and more aware of alternative parties, or that they spend longer checking out the ballot paper and might change their minds more readily than someone whose first encounter with UKIP is in a polling booth.

  19. Oldnat,

    There are certainly some interesting rumblings in Wales and Scotland. In both nations, UKIP don’t seem to be taking a majority of their support from the Tories, given that the Tories are roughly where they were in 2010 in Scotland (at least) and they’re down only a little in Wales.

    So where are the New Kippers coming from in Scotland and Wales? It’s especially interesting in that both PC and the SNP are up, in the latter case by a lot. So the New Kippers can only be coming from new/returning voters and Labour/the LDs.


    Has anyone noticed how the big party total VI is dropping? Ten-poll moving average since 2011:

    No but we do now thanks to your wonderful graph.

  21. The projection of seats (while not any sort of surprise), baited me into airing an opinion which I’ve held for quite a while but haven’t seen an appropriate place to raise it so far.

    When people talk about lack of fairness in the voting system they inevitably talk about Labour and the Tories, but it goes further down the tree than that. Despite the fact that the Lib Dems in my constituency are fielding the only candidate that could convince me to vote for them, and despite how torn I am about UKIP – happy to see anyone but the big two gaining traction, nervous about which way that party will go in a way that I’ve never been about any UK-wide party currently in the Commons – it’s still perfectly plausible that UKIP could get 15% and 1 MP (though personally I predict 3), and yet if the Lib Dems were to recover to 10-11% [increasingly unlikely but too soon to say it can’t happen] they could be looking at 30 MPs – still 30 seats short of a proportionate representation in the Commons.

    However little you think of UKIP relative to any of the other UK wide parties, or indeed the Lib Dems relative to whichever one of Labour or the Tories you are more inclined to, surely the scale of over and under representation is one that needs to be looked at, even if only in a very modest way. In my opinion this could be done by topping up ~620 constituency MPs with ~30 who are there to rebalance the parliament ever so slightly in the direction of how the public actually voted

    Arguments against some sort of proportionate list system include:

    Hung parliaments – there’s a significant chance that by 2020 we will have had a decade without an overall majority anyway.

    Parties sticking their cronies in – specifically link it to electoral performance. Keenly fought contests are in general good for democracy, so reward those who have played their part in making it happen, and who in any case have a decent mandate from their constituents.

    People who have been booted out sneaking in through the cat flap – Only allow first-time MPs into parliament via the list system.

    Breaking the constituency link – it wouldn’t break the constituency link but suppliment it, much like the Scottish parliament.

    Letting overt racists into parliament – This argument only holds for as long as the BNP or similar are unable to win a constituency under the current rules. Bring in a requirement that you have to get an MP elected by FPTP to be eligible for the top-up list – if one day a far right party is able to actually win a constituency and get a second MP this way, the problem would lie in the fact that they won the constituency in the first place. Besides, if the BNP of similar were to win a seat, it can reasonably be assumed that their national share would be above 2%,

    Creating a massive surge for one party with minimal FPTP representation – Could equally be applied to UKIP, the Greens and Welsh and Scottish nationalists in some years: cap any party’s list MPs to 50% of their total MPs. If UKIP + Green were to win a combined 4 seats on a combined 18% of the vote, I don’t think giving them a grand total of 8 out of 650 MPs would be distorting things too much. While on the basis of recent weeks I’d frankly prefer that their representation were reduced rather than increased, I don’t think a potential total of 8-10 SNP MPs would be an inflated representation of their standing in 2010.

    Too complicated – it would have no direct effect on the voter, barring those who are nuanced enough to consider tactical voting who are generally engaged enough to adapt to the new system. It would be of comparable complexity to the way EU seats are allocated, the difference being that this system would determine the election of ~5% of MPs, and those MPs would have substantial support from their contested constituency. Contrast that to the EU, where 90% of those elected are virtually unknown representatives of a political party, even among those who actually voted them in!

    Bad for independents – what isn’t?

  22. OldNat

    You are reading the Welsh poll through the lens of your own prejudices, I fear, and overlooking its most important and unexpected finding which is that in stark contrast to Scotland where 45% voted for independence, just 3% support that option in Wales. This is the result which is correctly headlined in the parent news story on the BBC Wales. 3% is believed to be the lowest support for Welsh independence ever recorded in an opinion poll.

    So here we have two countries both of which have had the same period of devolution, moving in opposite directions on the most critical issue of all. You also need to delve into the detail to see that the 49% ( not inflated to your 52% by adding the lonely 3% independence supporters !) are far keener on issues like policing and welfare being devolved than taxation which would bring major pitfalls for Wales where, unlike Scotland, there is no room for argument against the fact that there is no way Wales can support its public services at UK levels out of taxes raised here. Indeed Prof Holtham pointed out with his customary objectivity on Wales TV tonight that what some in Wales consider to be underfunding does not look that way to many looking from the other side of Offa’s Dyke.

    I can also tell you that in years of canvassing in SE Wales I have never, repeat never, come across a single voter who has ever requested more devolved powers for Cardiff Bay. I have met rather too many who have argued the opposite in terms which would put me into moderation if repeated here.

    The prosaic truth here is that the majority of people who don’t think much about politics just want the Wales Assembly to get on with its major tasks of improving our NHS and schools, avoid blowing our whole capital expenditure on a crazy motorway by/pass vanity project, and to report back for possible extra duties when they have made some progress, instead of constant blathering and bleating about extra subsidy, extra powers and extra Assembly members.

    The polls and data on actual achievements, especially education, show that devolution has so far been more successful in Scotland than Wales. In the short term this is unlikely to change much. The case for dishing out more powers and cash to an Assembly bureaucracy which is struggling to cope with its existing duties is non-existent once you think about it. Take a look at our PISA Reports and a collapsing ambulance service, as examples. You may also see from this week’s news that our AMs can’t even run their own Assembly office operation without losing over £100,000 of our money paid into a fake bank account. What the shameless politicians in the Bay say is needed is no guide to popular sentiment as they all want more power and further tranches of cash to spend without a rigorous Treasury to exercise control. Which is of course why the latest poll shows that all 4 Parties have lost votes to UKIP whose support is now comparable to several less prosperous English regions, with which Wales probably has rather more affinity in the 21st century .

  23. @OldNat

    “Labour is doing well in England.”

    A week is a long time in the politics of Scottish nationalism. Seven days ago the line was that secession was necessary because they weren’t.

  24. @CH, you obviously put a lot of work into that post. FPTP is undemocratic, therefore i suggest PR.

  25. @ Welsh Borderer,

    What the shameless politicians in the Bay say is needed is no guide to popular sentiment as they all want more power


    Popular sentiment:

    3% – Welsh independence
    49% – extra powers for the Welsh Assembly
    26% – status quo
    12% – abolition of the Welsh Assembly

    The shameless politicians seem to be channeling popular sentiment quite effectively. (Whether or not it’s actually a good idea is of course a separate question.)

  26. @ Chris Hornet,

    I don’t think we should be designing an electoral system to keep a specific party in or out, even if that party is the BNP. And some of these rules are silly. Having a rule that only first-time MPs can go on the PR list is not going to free it from party placemen, it’s going to turn it into the exclusive province of a rotating clique of nonentity party apparatchiks because no politician with any national standing could stand.

    PR systems strengthen the party system by imposing big penalties for pissing off whoever draws up your party list. They just do. It’s one of their intrinsic properties, like how safe seats are an intrinsic property of FPTP systems. This is the cost of PR and if England isn’t willing to pay it England can’t have PR.

    Which brings me to Pete’s point- FPTP isn’t undemocratic. It prioritises the democratic principles that every locality should be represented by the politician who commands the highest level of local support, and that parties should try to have a broad enough appeal that they can win the seats to form a single-party government.

    PR prioritises other democratic principles, namely that everyone should be represented by a politician of their preferred party even if that party has few supporters in their area, that small parties should have a chance to win seats, and that parties should work together to govern on a cross-party basis.

    All of these are democratic principles. You can make up your own mind about which to prioritise, but that make that doesn’t make the other ones or the other electoral system illegitimate.

  27. @Allen C

    “UKIP are to appeal to Scots with new logo.”

    Does it look like a busted flush?


  28. A bucket of tartan paint I would have thought.

  29. I should add- I say all this as a supporter of PR. I think it’s a better system.

    But it’s important to recognise that all the possible electoral systems have costs and benefits, and FPTP accomplishes things are objectively good as well as making life difficult for small parties and rendering votes in safe seats meaningless.

  30. * things that

  31. There was actually another Welsh poll (YouGov for ITV Wales and Cardiff U, f/w 8-11 Sep) released about a week ago, which we all missed in the tartan storm:


    Westminster VI was pretty similar to ICM:

    Labour 38% (-3)

    Conservative 23% (-2)

    Plaid Cymru 11% (no change)

    UKIP 17% (+3)

    Liberal Democrats 6% (+1)

    Greens 5% (+2)

    BNP 1% (no change)

    Changes from June/July. In seats terms this gives the same minimal nett movement as ICM (Lab +2, L/D -2).

    For the National Assembly, Constituency:

    Lab 36% (-1)

    Con 21% (n/c)

    PC 19% (-1)

    Lib Dem 6% (+1)

    UKIP 12% (-1)

    Green 4% (+1)

    Other 2% (+2)


    Lab 31% (-3)

    Con 21% (n/c)

    UKIP 17% (+1)

    PC 16% (-2)

    Green 7% (+3)

    L/D 5% (n/c)

    Other 3% (n/c)

    Changes again from last poll (26 Jun- 1 Jul).

    Roger Scully reckons this would give an Assembly breakdown of:

    Labour: 29 (-1); 27 constituency AMs, 2 list AMs

    Conservative: 11 (-3); 6 constituency AMs, 5 list AMs

    Plaid Cymru: 10 (-1); 6 constituency AMs, 4 list AMs

    UKIP 9 (+9); all 9 would be list AMs

    Liberal Democrats: 1 (-4); 1 constituency AM


    What a splendid advert your post is for devolution of powers -and money-to local politicians,

    Actually your point about an imbalance between tax potential & public spending in Wales is very interesting.

    I see that EM has disowned any responsibility for NHS Wales.

  33. @Mr Nameless

    “A bucket of tartan paint I would have thought.”

    Could be. They’ve certainly being pouring a bucket of something else all over the Tories south of the border!


  34. It’s worth pointing out that FPTP can be just as dominated by central Party hierarchies as a list system, depending on the rules by which a Party selects or imposes its candidates. Effectively it’s a Party list system as well with a list of one and, unless you have some sort of Primary election, the electorate have no choice over the contents of that list either. It’s also almost as hard for independents to break into.

  35. Having a rule that only first-time MPs can go on the PR list is not going to free it from party placemen, it’s going to turn it into the exclusive province of a rotating clique of nonentity party apparatchiks because no politician with any national standing could stand.

    I’m not talking about standing separately, or even the parties choosing the list. I’m talking about having a formula to allocate the seats to those who ran the constituency MP close in the GE.

    And if a politician with national standing loses a local constituency race despite the massive advantage of their high profile that they have over other candidates, their democratic mandate to continue is weaker than a previous unknown in the same situation, not stronger.

  36. @ Roger Mexico,

    While that’s true in theory, in practice under the current system at least for the two main parties the selection process is somewhat decentralised and constituency associations (usually) have the power to pick candidates against the wishes of the national leadership. There are inevitably strings the central party organisation can pull, and we all know of cases where people have been parachuted in, but even so, the potential exists for local control. The deselection of Anne McIntosh is a recent example.

    I don’t see how that would ever be possible under a full PR system.

    @ Chris Hornet,

    I’m talking about having a formula to allocate the seats to those who ran the constituency MP close in the GE.

    That seems very arbitrary and unfair to the politicians, since most of that is determined by the marginality of the seat. Let’s say you run a brilliant campaign in a safe seat and shave 4,000 votes off the incumbent’s majority, and some other schmuck runs a horrendous campaign in a marginal your party was poised to win, underperforms the national swing by 6% and loses by 10 votes. He goes to Parliament and you don’t?

    Plus, this means marginal seats have two shots of getting their PPC into Parliament and safe seats have only one. Isn’t a big part of the point of PR to give people in safe seats a voice, instead of having the parties pander exclusively to the swing voters in the marginals? Giving marginals two MPs is going to do exactly the opposite!

  37. From this poll, can we assume a devo max for Wales has majority support? (49 + 3)?

  38. Mr N

    Thanks for the reply about Survation poll, I had missed RAF’s posting.

    So much for all that speech dissection by the hacks.

  39. Speermint,

    I think you are absolutely right about party list systems – I need to be careful but the LDs selection process may have had a casting couch element to it – allegedly

    As you say, all systems have strengths and weaknesses.

    For me STV in multi-member constituencies with open lists is the least bad system.

    Keeps the constituency link gives, minor parties in an area a chance of some representation, gives voters a choice of different candidates from the same party and if used wisely can promote inclusivity.

    Also means an OM is possible with 40+% as needing 50% can lead to an FDP type situation of a small centrist party being in permanent Government which is hardly democratic.

  40. Chrishornet

    you seem to be groping your way towards the “single transferable vote” system. A system so simple that even the Irish can understand it

  41. @ Jim Jam,

    I could come around to that. Bigger, multi-member constituencies also have the advantage of creating larger constituency associations, which would help to prevent, er, Falkirk.

  42. FPTP would be fairer if voters of the various parties were more evenly distributed, we’d get more marginals and fewer safe seats where individual votes mean less.

    But that’s not the case in this country at the moment or foreseeable future: it’s very polarised between North and South, Midlands generally being only area of real interest if you want a majority.

    PR would make politicians campaign in (and perhaps, care about) the various parts of the country more equally? Maybe.

  43. Re. voting systems, and the balance between the “direct accountability” of FPTP and the “representativeness” of PR, I actually quite like the modified D’Hondt top-up thingy that we’ve got in Scotland. In a 5-ish party environment it seems to produce results that are not too far from what everyone wants, or at least, I have not heard anyone rail against its iniquities. In particular the second vote allows for a greater expression of preference, so one can make a tactical first vote and a principled second.

  44. @KeithP

    PR would make politicians campaign in (and perhaps, care about) the various parts of the country more equally? Maybe.

    Broadly, perhaps — but it introduces another bias, they might just concentrate on large centres of population and ignore the rest.

    (Of course, I’m not saying this would be any worse than the current focus on marginals!)

  45. “FPTP would be fairer if voters of the various parties were more evenly distributed, we’d get more marginals and fewer safe seats where individual votes mean less.”

    Hardly. Rather, the only reason that FPTP works at all (to the extent that it does work) is that voters are unevenly distributed. If every constituency voted 51% for Party X, 49% for Party Y, then Party X would win every single seat – hardly a form of democratic representation. Even FPTP advocates would probably acknowledge that the main opposition party needs significant representation in Parliament.

    Now, of course, by advocating a more even distribution, you aren’t saying it should be exactly the same everywhere. Nevertheless, I think FPTP can lead to outcomes every bit as absurd with relatively even distributions as it can when there are very uneven distributions. A relatively but not exactly even distribution, for instance, could fairly easily enable the second-placed party (in share of vote) to be the first-placed (in seats), just as an uneven distribution can.

  46. Spearmint, Roger M et. al.,

    Generally “parachuting in” candidates only happens in by-elections. The vast majority of candidates are selected locally. In addition, the single-member constituency allows the electorate to build a relationship with their MP and some have a large personal vote. Others get chucked out in negative circumstances despite being supported by their party – e.g. Neil Hamilton – and there are plenty of other examples.

    I would be very upset if this level of scrutiny were abandoned in favour of some kind of party list.

  47. Sticking with Wales does anyone know who won the Llanfairpwllgwyngyllgogerychwyrndrobwyll-llantysiliogogogoch by-election?

    Was it Dafydd Carduggan-Wigley Caentwn or Bob Hill?

  48. “I would be very upset if this level of scrutiny were abandoned in favour of some kind of party list.”

    Has anyone advocated that, though? I thought most people who wanted a party list wanted a top-up list, as in Scotland, Wales or the Jenkins Commission proposals – not the replacement of constituency elections.

    As someone else said, whether the party list candidacies are decided democratically depends on the rules applied by each party – perhaps they should be pressured to be democratic, or you could incentivise them to have primaries. Or you could enable voters to preference-vote individual list candidates as in Sweden, or to cross out candidates from the list which I think some countries also allow or formerly allowed.

    What is outrageous is the way that EU elections are decided by party lists, and closed ones at that. They should be either open lists or STV (which are both within EU rules).

  49. “The vast majority of candidates are selected locally…”

    But the candidates invariably have to be from a pre-approved central list. Even UKIP requires parliamentary and EU candidates to go through an assessment process.

1 2 3 11