What could happen at the Assembly Elections?

With sampling for our last Welsh Political Barometer poll being completed on the day before the general election, it was understandable that all immediate attention focussed on the implications of the poll for that election. However, as with other Barometer polls, we also included questions on a number of other matters. In this and a following post, I’ll discuss those other results.

I’ll start here with the polling numbers for the National Assembly. Please bear in mind: these numbers come from the same poll that, as I discussed in another recent blog post slightly over-stated Labour electoral support for the general election and under-stated Conservative support. These errors will probably lead to YouGov revising their sampling and weighting procedures in Wales in future, in a manner that makes their polls produce results that are a little less favourable to Labour and a little kinder to the Tories. But that hasn’t been done yet. So I am simply reporting here the figures that YouGov supplied to us last week. They have not been adjusted in any way. You may, therefore, want to mentally revise downwards the Labour vote share figures a small amount, and revise upwards the Conservative ones a bit. Of course, any such changes could potentially have implications for the allocation of seats in the National Assembly.

For the constituency vote, these were the figures (with changes from the most recent previous YouGov poll in Wales indicated in brackets)

Labour: 35% (-2)

Conservatives: 22% (-1)

Plaid Cymru: 21% (+1)

UKIP: 12% (+1)

Liberal Democrats: 6% (no change)

Greens: 2% (-1)

Others: 1% (+1)

On the standard assumption of uniform national swing, this poll would project only two constituency seats to change hands from the last Assembly election in 2011: the Liberal Democrats would narrowly regain Cardiff Central from Labour, while Labour would also lose Llanelli to Plaid Cymru. On Ratio Swing assumptions, we also find that only two constituency seats would change hands – but they are not the same two! Plaid would still capture Llanelli from Labour, but the other seat to change hands would be Carmarthen West and South Pembrokeshire, which would be gained by Plaid from the Conservatives.

The figures for the regional list vote were as follows (with changes from the previous Barometer poll again indicated):

Labour: 32% (-2)

Conservatives: 22% (-1)

Plaid Cymru: 20% (no change)

UKIP: 13% (+2)

Liberal Democrats: 6% (no change)

Greens: 4% (-1)

Others: 3% (+1)

Again assuming uniform swings from 2011 across Wales, and after taking into account the distribution of constituency seats when allocating the list seats, this gives us the following projected overall outcome:

Labour: 28 seats (26 constituency seats + 2 list seats)

Conservatives: 12 seats (6 constituency seats + 6 list seats)

Plaid Cymru: 11 seats (6 constituency seats + 5 list seats)

UKIP: 7 seats (7 list seats)

Liberal Democrats: 2 seats (2 constituency seats)

If we model the poll numbers according to Ratio Swing, then the final seat allocations come out like this:

Labour: 29 seats (27 constituency seats + 2 list seats)

Conservatives: 12 seats (5 constituency seats + 7 list seats)

Plaid Cymru: 11 seats (7 constituency seats + 4 list seats)

UKIP: 7 seats (7 list seats)

Liberal Democrats: 1 seats (1 constituency seat)

Once again, the Barometer poll has produced a projected result (under either modelling method) that gives us five parties with representation in the National Assembly. The good news for Labour is that, under either UNS or RS, they are currently projected to win well over twice the number of seats won by the next-largest party. The much less good news for Labour is that this poll put them, on both votes, within one percentage point of their lowest vote share for Assembly voting intentions in any Welsh poll conducted since May 2010. And that in a poll that was probably, for the reasons outlined above, slightly over-stating Labour support. Perhaps the best news for Labour here is that the poll hardly offers unambiguous good news to anyone else either.