Continuing troubles for Labour

Labour have fallen to their lowest levels of support in Wales, for both Westminster and the National Assembly, since before the 2010 general election. This is the stand-out finding to emerge from the latest Welsh Political Barometer poll, the first opinion poll to be conducted in Wales this year.

As always, our new poll asked respondents about voting intentions for both a general election and an election to the National Assembly. What did we find? First, here are the results for a UK general election (with changes from the last Barometer poll, conducted in mid-September, indicated in brackets):

Labour 33% (-2)
Conservative 28% (-1)
Plaid Cymru 13% (no change)
UKIP 13% (-1)
Liberal Democrats 9% (+2)
Others 2% (-1)

While Labour remain some way in the lead, the two-point fall in their support from our previous poll takes them to their lowest point in any Welsh poll on Westminster voting intentions since a YouGov poll in April 2010 – the last days of the Gordon Brown government. For most of the other parties we see minimal, or no, change. However, Liberal Democrats will doubtless be encouraged to see a two-point rise in their support level. While still within the ‘margin of error’, and thus plausibly reflecting simply normal polling fluctuations, the Lib-Dems’ nine percent actually equals their best score in a Welsh poll since the signing of the coalition agreement in May 2010. Times since then have been tough for the party, and this new poll will put at least the faintest sparkle in the eyes of some of their supporters.

If we apply uniformly across Wales the changes to party support since the May 2015 general election that are implied by this poll, on the current seat boundaries, we get the following projected result (with all seats won by a party at last year’s general election remaining in their hands unless stated otherwise):

Labour: 24 seats (losing Ynys Môn)
Conservative: 11 seats (no change)
Plaid Cymru: 4 seats (gaining Ynys Môn)
Liberal Democrats: 1 seat (no change)

What about the new electoral map for Wales proposed by the Boundary Commission at the end of the summer? Under those boundaries – and using the ‘notional’ 2015 results for those boundaries produced by Anthony Wells of YouGov – this poll projects the following outcome:

Labour: 14 seats
Conservative: 11 seats
Plaid Cymru: 3 seats
Liberal Democrats: 1 seat

These projected results would represent a net loss for Labour of four seats from the 18 that Wells projects them, on the new boundaries, to have won in 2015. The Cardiff North, Flint and Rhuddlan, Gower and Swansea West, and Wrexham Maelor seats, all projected by Wells to have been narrowly won by Labour on the new boundaries, would on the figures in our new poll now be won narrowly by the Conservatives. Labour would therefore remain the largest party in Wales, but fall short of winning a majority of Welsh parliamentary seats – something that the party last did in the debacle of the 1931 ‘National Government’ election.

What about for the National Assembly? I’ll begin as usual with the constituency vote. Here are the findings from our poll (with changes from the last Barometer poll once again indicated in brackets):

Labour 31% (-3)
Conservative 25% (+1)
Plaid Cymru 21% (+1)
UKIP 12% (-1)
Liberal Democrats 8% (+2)
Others 3% (no change) 

Here again we continue to see the Labour party some way in the lead, but with their support ebbing from last time. And once again Labour slip to their lowest point in several years: this is the party’s lowest poll rating on the Assembly constituency vote since YouGov’s first ever Welsh poll, conducted in July 2009 at the very nadir of Gordon Brown’s fortunes.

In last May’s Assembly election, Labour sustained a large fall in its level of constituency vote support from 2011 but without suffering major seat losses – Leanne Wood’s victory in Rhondda being Labour’s only loss. But the situation indicated by this poll would see Labour begin to reach the sort of ‘tipping point’ where they start losing significant numbers of constituency seats. If the changes from last May’s Assembly election indicated by our new poll are applied uniformly across Wales, then we project Labour to lose eightconstituency seats that they won in May 2016:

·         The Conservatives would gain Gower, the Vale of Clwyd, the Vale of Glamorgan, and Wrexham;

·         Plaid Cymru would gain Blaenau Gwent, Cardiff West, and Llanelli;

·         And the Liberal Democrats would gain Cardiff Central.

The findings for the Assembly regional vote show a broadly similar picture to those for the constituency ballot:

Labour 28% (-1)
Conservative 22% (no change)
Plaid Cymru 20% (-1)
UKIP 14% (+1)
Liberal Democrats 7% (+1)
Others 7% (-3)

Applying once more the assumption of uniform national swing, and also taking into account the projected constituency results just mentioned, our poll provides the following projected outcome for the regional list seats:

North Wales: 2 UKIP, 1 Labour, 1 Plaid
Mid & West Wales: 2 Labour, 2 UKIP
South Wales West: 2 Plaid, 1 Conservative, 1 UKIP
South Wales Central: 2 Conservative, 1 Plaid, 1 UKIP
South Wales East: 2 UKIP, 1 Conservative, 1 Plaid

This, in turn, gives us the following overall projected outcome:

Labour 22 seats (19 constituency, 3 regional)
Conservative 14 seats (10 constituency, 4 regional)
Plaid Cymru 14 seats (69constituency, 5 regional)
UKIP 8 seats (8 regional)
Liberal Democrats 2 seats (2 constituency)

Any such outcome would be Labour’s worst-ever National Assembly election result by a considerable margin.

Overall, our new Barometer poll shows the Labour party’s continuing troubles across the UK having a notable impact even in Wales, their longest-standing bastion. The saving grace for Welsh Labour continues to be the lack of a single strong opponent: while Labour’s performance in our new poll is weak, none of their opponents are exactly achieving glittering ratings either. But as our seat projections for the National Assembly show, even without its opponents doing that well, if Labour’s support continues to ebb then the party may pay a very heavy electoral price.