The New Welsh Political Barometer Poll

As we enter the last full week of campaigning for the National Assembly election, Labour retains a clear lead in the latest Welsh Political Barometer poll. But its support has slipped to the lowest level seen for six years. Meanwhile the poll contains some encouragement for both Plaid Cymru and even the possibility of good news for the Welsh Liberal Democrats.

Those are the headline findings from our new poll, conducted by YouGov between 19-22 April. The poll once again asked a representative sample of people in Wales about their voting intentions on both the constituency and regional ballots for the National Assembly election. Here are the figures for the constituency vote (with changes on the previous Welsh Political Barometer poll, conducted at the beginning of the month, in brackets):


Labour: 33% (-2)

Plaid Cymru: 21% (no change)

Conservatives: 19% (no change)

UKIP: 15% (-2)

Liberal Democrats: 8% (+2)

Others: 3% (no change)


So here we see Labour remaining well ahead of the field, despite its support slipping two points since the previous poll. Plaid Cymru will be encouraged that they retain a clear second place ahead of the Conservatives; while the Tories can hardly be pleased by these results, the decline in support for UKIP at least means that, unlike in our previous poll, they do not appear under threat of falling into fourth place. And, for the first time in a long while, the poll also contains some encouragement for the Welsh Liberal Democrats: their two point rise here comes on top of a further slight increase in the last poll.

Applying uniformly across Wales the changes since the 2011 National Assembly election indicated by this poll, the figures project three constituency seats to change hands. The Liberal Democrats are projected to capture Cardiff Central from Labour; meanwhile Plaid Cymru are projected to gain Llanelli from Labour and also Carmarthen West & South Pembrokeshire from the Conservatives.

For the regional list vote, the figures in our new poll are as follows (with changes from the last poll again in brackets):


Labour: 29% (-2)

Plaid Cymru: 22% (+2)

Conservatives: 19 (-1)

UKIP: 15% (-1)

Liberal Democrats: 8% (+3)

Greens: 4% (no change)

Others: 4% (+1)


Again assuming uniform swings since the 2011 election across Wales, and after taking into account the distribution of constituency seats, this gives us the following projected distribution of the regional seats:


North Wales: 2 UKIP, 1 Conservative, 1 Plaid Cymru

Mid & West Wales: 2 Labour, 2 UKIP

South Wales West: 2 Plaid Cymru, 1 Conservative, 1 UKIP

South Wales Central: 2 Conservative, 1 Plaid Cymru, 1 UKIP

South Wales East: 2 Plaid Cymru, 1 Conservative, 1 UKIP


Combining both sets of figures produces the following overall outcome for the National Assembly:


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

Plaid Cymru: 13 seats (7 constituency seats + 6 list seats)

Conservatives: 10 seats (5 constituency seats + 5 list seats)

UKIP: 7 seats (7 list seats)

Liberal Democrats: 2 seats (2 constituency seats)


Labour thus remain some way ahead of the field. But they must be somewhat concerned that support has edged downwards further: 33% on the constituency vote is Labour’s lowest level in any Welsh poll since before the 2010 general election. A similar YouGov poll conducted in late April 2011 put Labour on 45% for the constituency vote and 41% for the list vote. So Labour are now twelve percentage points lower, on both ballots, than they were at this stage in the electoral cycle before the last Assembly election. Were Labour’s election performance this year to fall some way below their poll rating – which it has normally done in Wales in recent years – then Labour could be under threat of achieving their worst ever vote share in a National Assembly election.

And yet the divided opposition leaves Labour well in the lead. Plaid Cymru must be encouraged that this poll now has them in a clear second place on both ballots, and projected to finish three seats ahead of the Conservatives. Plaid would need to make further progress, or achieve very strong local swings in particular constituencies, to make substantially greater seat gains. Nonetheless, given that they have run behind the Conservatives on both Assembly ballots for most of the last five years, recent trends are at least moderately encouraging for them. The findings are much more concerning for the Conservatives; the optimistic talk at the party’s March Welsh conference of making significant seat gains now appears significantly less likely to be realised.

Meanwhile, for the first time in a very long while the fourth party in the Assembly, the Liberal Democrats, have some moderately good polling news. This poll suggests that the Assembly election campaign has enabled them to increase their support levels. Their position is still highly precarious, but the 8% they score on the list vote here is equal to their vote share in the 2011 election, and suggests that they may just have a chance of holding onto some of their current regional list seats. Meanwhile UKIP have seen their support slip slightly in this poll. But unless they suffer a further substantial erosion in support, UKIP are still likely to win regionalist list seats on May 5th across most, if not all, of Wales’ five electoral regions.

The poll for ITV and Cardiff University’s Wales Governance Centre had a sample of 1001 Welsh adults and was carried out by YouGov from 19-22 April 2016.



As has become customary, for the benefit of the cognoscenti who read this blog, I have also computed Ratio Swing projections from our poll.

For the National Assembly, Ratio Swing, as with UNS, projects three constituency seats to change hands. But they are not the same three seats. Ratio Swing still projects Llanelli to be gained by Plaid Cymru from Labour, and Carmarthen West & South Pembrokeshire to be gained by Plaid from the Conservatives. But it also projects Plaid to gain Aberconwy from the Tories. For the first time in some while, however, a Ratio Swing projection of this poll does not project the Conservatives to gain Brecon & Radnor from the Liberal Democrats.

For the regional list seats, Ratio Swing produces the following projections:


North Wales: 2 UKIP, 2 Conservative

Mid & West Wales: 2 Labour, 2 UKIP

South Wales West: 2 Plaid Cymru, 1 Conservative, 1 UKIP

South Wales Central: 2 Conservative, 1 Plaid Cymru, 1 UKIP

South Wales East: 2 UKIP, 1 Conservative, 1 Plaid Cymru


Overall, a Ratio Swing projection of this poll therefore generates the following outcome for the Assembly election:


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

Plaid Cymru: 12 seats (8 constituency seats + 4 list seats)

Conservatives: 10 seats (4 constituency seats + 6 list seats)

UKIP: 8 seats (8 list seats)

Liberal Democrats: 1 seat (1 constituency seat)


  • David Willis

    I’m afraid I’m not one of the cognoscenti. What is the difference between Ratio swing and uns? And which has usually been closest to the actual result in previous elections?

  • Graham Burnby-Crouch

    Taking the mean of the last 3 polls and taking the midpoint of the two methods uniform and ratio would see just Llanelli, change hands I think.

  • Glenn Page

    Hi Roger,

    Is the data accessible anywhere?

    Very interested to know the breakdown of list seats, particularly in the three southern regions; how many votes decided the fourth seat and which party was runner-up…

    • Roger Scully

      Data now available in the Opinion Polls section here. (Just scroll to the bottom of the page).

      Surely you could use my instructions on how D’Hondt works from last week to work out the numbers yourself?!

      • Glenn Page

        Great. Cheers.

        Well, I am going to try but I am no ‘pollster’. 🙂

  • Akash

    Shouldn’t the Liberal Democrats get at least one regional seat because it is equivalent to their regional percentage in 2011?

    • Roger Scully

      No – there is no necessary link between vote share and list seats.
      Basically, the LDs are suffering because the Labour list vote (which was all wasted anyway outside Mid and West) is down, and UKIP is up. Because UKIP don’t win any constituencies their list vote converts very efficiently into list seats.

  • YoYo

    How close is the Lib dems to a list seat in North wales – they need just over half the UKIP vote I belive?

  • Ian Williams

    I fail to see the relevance of this poll and what we are seeing and being told on the streets. UKIP only getting 1 seat in South Wales East is a nonsense – we fully expect 2.

    • Chris Johnes

      @ Ian Williams; let’s see what the relevance is on 6 May; these kind of sites are full of political activists denying the veracity of polls they don’t like; the pollsters are far far more liable to be proven right than the activists – but only time will tell

    • Christian Schmidt

      When I did canvassing mans tallied canvassed returns, I sometimes shared the overall picture with the agents of the other parties on election night (after polls we closed, before votes were counted). The total votes between us was always above the actual total vote…

  • Ian Williams

    Being told by an eminent pollster that polls are ‘opinion formers’ more than ‘opinion gatherers’ shows collusion within the political sphere. He did not apologise for misleading!!

Leave a Reply

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