Skip to main content
Coronavirus (COVID-19): Latest updates.


New Ashcroft Poll of Vale of Glamorgan

6 March 2015

The Political Blogosphere was aflame on Wednesday evening with news of Lord Ashcroft’s latest set of published polls. Most of the attention, understandably enough, focused on the results from constituency polls in eight seats in Scotland. Very much in line with all the national polls there since October, and also with his earlier batch of Scottish constituency polls, these showed huge swings to the SNP. The party continue to be on-course to make substantial gains in Scotland at the general election.

Amidst all this excitement, there was rather less attention paid to the somewhat less dramatic results of four other polls conducted in marginal seats outside Scotland. One of them was in a Welsh seat, the Vale of Glamorgan. The Vale is a very interesting seat in many respects. It was held by Labour throughout the Tony Blair years, but lost to the Conservatives in 2010: the former regional AM Alun Cairns won the seat by a fairly clear 8.9% margin on a substantial (6.1%) swing from Labour to the Tories. However, Labour held the seat the following year in the Assembly election: Jane Hutt retaining the seat she has had since 1999 with a clear 11.4% margin, on a 5.6% swing from the Conservatives to Labour. The Westminster Vale seat is not one that Labour necessarily need to gain to become the largest party in the House of Commons (as they do with, say, Cardiff North), but it would be hard for them to win an outright parliamentary majority without winning seats like the Vale.

The fieldwork for Lord Ashcroft’s poll here was carried out, as with all his other constituency polls, by telephone; it was done in early February, so the findings are not too dated. What did he find?

As before, Lord Ashcroft reports two sets of headline vote intention figures. The first comes from a standard ‘generic’ voting intention question; the second comes from a follow-up question where respondents are specifically prompted to think about their ‘own parliamentary constituency’, and ‘the candidates who are likely to stand for election to Westminster there’. The table below produces both figures for the main parties, as well as their actual result in the Vale of Glamorgan in the 2010 general election:

Party 2010 Result Generic Question Constit. Question
Conservatives 41.8% 40% 38%
Labour 32.9% 32% 32%
Lib Dems 15.2% 5% 4%
Plaid Cymru 5.5% 8% 12%
UKIP 3.1% 10% 10%
Greens 0.9% 5% 3%
Others 0.5% 1% 1%


For the constituency vote figures, which most people seem to be interpreting as the main ‘headline’ numbers to take from these polls, the Conservatives are therefore six points ahead of Labour. While this is hardly a comfortable margin – one couldn’t imagine Alun Cairns looking at these numbers and deciding to take a week off the campaign trail – it is nonetheless fairly encouraging for him. You’d always rather be six points ahead than behind. What is perhaps particularly interesting is that, in this seat, the Tories appear currently to have held Labour to a very small swing, of less than 1.5%, since the 2010 general election. This is notable because many of Lord Ashcroft’s constituency polls, including some of his latest batch, have been showing swings to Labour since 2010 that are at least as big or even bigger than the average swing suggested by the GB-wide opinion polls – which is around 3.5%-4%. On the evidence of this survey, that has not been the case in the Vale of Glamorgan.

The poll also suggests that the contest in the Vale is very much a two-horse race. That is interesting, because in last year’s European Parliament election, UKIP actually won the Vale of Glamorgan. OK, that was in the Vale of Glamorgan local authority, which does not have quite the same boundaries as the parliamentary constituency. Nonetheless, given UKIP’s victory there last May, and their continued strength in the national polls, I would have been pretty confident in predicting that any poll in the Vale constituency would place UKIP in, at worst, a clear third place. That expectation is confounded, however: the constituency question results place UKIP in a rather poor fourth, behind Plaid Cymru. Meanwhile, the results for the Liberal Democrats are predictable awful, and there is no sign of any Green surge in this particularly green part of Wales.

Plaid’s relatively strong performances in the Ashcroft constituency polls continue to be one of the minor mysteries of contemporary Welsh psephology. This is the fifth of the forty Welsh seats to have been polled (see discussion here); none of the five could remotely be described as a Plaid target seat in the context of a Westminster general election. Yet in every case the Ashcroft polls have shown a significant rise in support for Plaid from the level they won in 2010. At 6.5%, the Vale poll actually shows the largest such Plaid rise, although all have been at least 4%. This is puzzling for at least two reasons. The first is that such constituency-specific rises in support are not being matched by rises in Plaid’s support in the national polls. The second is that such evidence as we have from other electoral contests, such as last year’s European election, is that Plaid’s recent electoral performance has been strongest in its traditional areas of strength. These are not the places in Wales where Lord Ashcroft has been polling. And yet he has consistently shown them rising in support. Curious indeed.

We cannot explain this puzzle through a particularly strong Plaid presence on the ground. Among the other questions asked in this series of constituency polls, as in previous ones, was whether respondents had been contacted ‘over the last few weeks’ by any of the political parties. A number of possible types of contact (leaflets, direct mail, email, and telephone or in-person canvassing) were mentioned. The results were as follows (the percentages add up to more than 100 because respondents could, and some did, report having been contacted by more than one party):

Party % reporting contact
Conservatives 33%
Labour 33%
Lib Dems 5%
Plaid Cymru 10%
None of them 44%
Don’t Know 1%

Here we see the two parties in serious contention for the seat well ahead of the others in the reported impact of their ‘ground game’, and dead-level with each other in their overall rates of contact. None of the other parties even come close, although Plaid Cymru do best of the rest. So Plaid’s apparent rise in support since 2010 is certainly not down to them out-working the two main parties on the ground. Nor is Plaid’s relatively strong performance obviously down to the methodology of Lord Ashcroft’s polls. If anything one would have expected the constituency-specific question to hurt Plaid in the five seats where Ashcroft has polled. Ashcroft’s use of recalled-2010 general election vote to weight his data also hurts, rather than helps, Plaid.

We have now had constituency polls in one-eighth of all the seats in Wales. These polls have focussed on seats relevant to the main UK parties. Overall, Lord Ashcroft’s polls suggest that we are in for a very close election here in Wales. I think most of us would have thought that anyway. But it’s always good to have more evidence, even when it confirms what we already believed to be the case.


  1. Welshguy

    Lots of Welsh speakers have moved into the east of the constituency, which is very much in the Cardiff orbit, in recent years. Plaid do relatively well there in council elections, so the rise here may just reflect that.

    But it is interesting to note how, as you say, every seat polled is showing a noticable Plaid increase, despite the national polls suggesting no such thing is taking place.

    3 possible explanations, each of them unlikely in their own way –

    1 – The increase outside Y Fro Gymraeg is being counterbalanced by a fall inside it. This would explain the unexpected maths, but seems unlikely to me – if they’re up 3-4% all over the rest of Wales, they’d have to be down *significantly* in the 4-5 constituencies of their heartlands to balance it out, and, as you suggest, other polling has suggested that they’re doing stronger than ever in their heartlands.

    2 – All these constituency polls are overestimating Plaid’s support somehow. Possible, but if so, why? Presumably they use a similar methodology to the national polls?

    3 – All the national polls are underestimating their support somewhow. But again, why? There are plenty of reasons to expect a rise in their support nationally – all three traditional UK parties are unpopular, events in Scotland have made UK constitutional matters more prominent in the media than ever before, they have a relatively charismatic new leader, and UKIP seem unlikely to effect them as adversely as they do the other parties. Indeed, that the polls show no increase seems genuinely surprising, yet time after time, they don’t, so unless there’s some systematic problem with the methodology then this can’t be the answer.

    It would be nice to have a different company do some Wales polls. But I suspect that won’t happen any time soon.

    • Roger Scully

      Yes, it is puzzling, for the reasons you elaborate upon here. As for more polls, by other companies – well, one can hope, though I’m not aware of anything at present.

  2. R Tyler

    BBC/ICM St David’s Day poll
    Could someone please supply the link to the above.

    • Roger Scully

      I’ll check with ICM as to whether the data has been posted on-line yet; when it has, I’ll link to it here as well.

Leave a reply

Your email address will not be published. Required fields are marked *