The pre-election wave of the Welsh Election Study included a very large number of questions. The results of many of these will be subject to detailed academic analysis, and not be published for some time. But in addition to the figures on National Assembly voting intention, we are publishing now the latest evidence on two matters: voting intentions for Westminster and for the forthcoming EU membership referendum.
First, Westminster. The numbers from our latest survey are (with changes on the February Welsh Political Barometer poll again indicated in brackets):
Labour: 36% (-1)
Conservatives: 25% (-2)
UKIP: 16% (-2)
Plaid Cymru: 14% (+1)
Liberal Democrats: 6% (+2)
Others: 3% (+1)
So here, as for the National Assembly, we see Labour remaining some way ahead of the pack; we also observe Plaid Cymru moving forwards and UKIP losing some support, although to a rather lesser extent in both cases than we do for the devolved election. In addition, for Westminster the Conservatives’ support seems to have slipped backwards a little bit since February; the extent of the change is well within the ‘margin of error’, so it could be down simply to sampling variation, but it may also be the case that Tory disunity over the EU Referendum is starting to harm their support for Westminster election even if it has not yet started to have such an effect at the devolved level. We also, interestingly, see just about the first even vaguely good news for the Liberal Democrats in a Welsh poll for some time. This is hardly a major fight-back yet, but it may, conceivably, be just the first stirrings of life.
(A side-note: checking over my file of past Welsh opinion polls – I know how to have fun – it would appear that Plaid Cymru’s 14% in this poll is their highest Westminster score in a poll with YouGov since March 2010. That’s hardly something to crack open the champagne over, but must still be encouraging for Plaid).
Following the standard Uniform National Swing assumption, this poll projects three seats in Wales to change hands from the result in the general election: for Labour to regain Gower and the Vale of Clwyd from the Conservatives, but to lose Ynys Môn to Plaid Cymru. Under the alternative assumption of Ratio Swing, the same three seats are also projected to change hands.
For the EU Referendum we have some rather more dramatic numbers. The figures (with changes from February again indicated in brackets) are as follows:
Remain: 41% (+4)
Leave: 36% (-9)
Don’t Know / Wouldn’t Vote: 24% (+5)
So, after a record eight point lead for Leave in the last Barometer poll, our new survey places Remain back in the lead – albeit by a far from comfortable margin. Why do we see such dramatic change since the previous Welsh poll? This is the same survey agency, YouGov, and the same survey question was used both times – one based very closely on the actual question to be used in the referendum itself. So what could be happening? It may that the change is partially explained by sampling variation – the February Welsh Political Barometer poll featured very strong scores both for Leave and for UKIP; this time around both do significantly worse, and that may reflect simply differences in the samples that even YouGov’s weighting procedures have not been able to iron out. Or it may be the case that this is, at least in part, a genuine change, and that the warnings against a Leave vote, and its particular potential impact on Wales, have started to hit home. We cannot be sure, but this is certainly something that the Barometer polls, and further waves of the Welsh Election Study, will be keeping a close eye on.