Labour vote share to decline?

Roger Scully assesses the latest voting intentions for the National Assembly elections.

BBC Cymru-Wales have this week published findings from a new poll, conducted by ICM soon after the European elections. The poll covered a number of areas, including attitudes to the policy achievements of devolution, and awareness of devolution, which I looked at earlier this week. Now let’s focus on the poll’s findings on voting intentions for the National Assembly.

ICM asked about voting intentions for both the constituency and list vote; this is the first poll to ask about these to be conducted by a company other than YouGov since the 2011 Assembly election. Before discussing these results, though, a few cautionary words about comparing them with those from YouGov. The BBC/ICM poll used the exact question wording on voting intention used by YouGov since December 2013. We don’t, therefore have concerns about different question wordings causing different results. However, we should recall that the February BBC/ICM poll, which asked about general election voting intentions in Wales, produced a lower figure for Labour support than any YouGov poll that has been conducted the 2010 general election. It also produced an unusually high Plaid Cymru figure.

We may want to bear that in mind when looking at the figures below. This is not an implied slur on either ICM or YouGov; both are, quite deservedly, internationally-respected survey companies. (The Political Betting website habitually refers to ICM as the ‘Gold Standard’ of pollsters, while among YouGov’s many achievements was estimating the recent European election results in Britain closer than any other polling company). My point is simply that the different methods used to conduct the surveys (the ICM polls have been conducted by phone, while YouGov’s use the internet) and weight the data seem to be generating small but distinct differences in the results produced: ICM have Labour support a little lower, and that for Plaid Cymru a little higher, than do YouGov. It is therefore sensible to take this into account when analysing ICM’s new findings alongside those from recent YouGov polls. We are not quite comparing like with like.

The basic figures on voting intention are as follows:

Constituency Vote

List Vote










Plaid Cymru









Translated into a national result, and assuming uniform national swings (with all the usual health warning applied to that assumption), this would produce the following result in an Assembly election (with changes from the seat totals won in May 2011 in brackets):





26 (-2)


28 (-2)


4 (-2)

7 (-1)

11 (-3)


2 (+1)

0 (-4)

2 (-3)

Plaid Cymru

8 (+3)


14 (+3)



5 (+5)


The constituency seats to change hands would be: Aberconwy and Carmarthen West & South Pembrokeshire (gained by Plaid Cymru from the Conservatives), Llanelli (gained by Plaid Cymru from Labour), and Cardiff Central (gained by the Liberal Democrats from Labour).

Even with the qualifications mentioned above about comparing this poll with those conducted by YouGov, ICM’s findings are broadly consistent with the trend over recent YouGov polls for Labour’s support on the constituency vote to be slipping below the levels enjoyed through 2011-13. ICM’s findings are also consistent with the modest upwards trend in Plaid Cymru constituency vote support found in recent YouGov polls; nonetheless, this is Plaid Cymru’s highest Assembly constituency vote score since October 2009.

Looking at the other parties, UKIP continue to perform quite strongly. That said, this was not an outstandingly good poll for them by recent standards; frankly, I was actually expecting a post-European election ‘bounce’ to lift them a little higher than this, especially on the list vote. The Liberal Democrats’ ratings continue to be awful; those for the Conservatives to be steady and resilient at a decent, if not outstanding, level of support.

One slightly surprising feature of the results is that Labour’s vote share is higher on the list ballot than in the constituency one. Labour have always won a higher vote share on the constituency vote than the list in the Assembly elections, while the last poll to put their list vote share as the higher one was an NOP poll conducted in April 2007. This may simply be one of those occasional, one-off findings that polls can produce, although we won’t know this for sure until we have more such ICM polls in Wales.

If we up-date our figures on the average constituency vote scores for each party for the last three years, these ICM findings – as indicated above – strengthen the trends that were already observable for Labour support to be declining, and that for Plaid Cymru and UKIP to be rising.



2014 (so far)









Plaid Cymru











Whether these trends will continue, of course, is another matter. For now I’d just like to thank the BBC and ICM for adding some further information to our understanding of where the parties in Wales currently stand.

Professor Roger Scully is Professor of Political Science at the Wales Governance Centre and Director of Research, Politics of Cardiff University. This piece was originally published on Roger's blog, Elections in Wales (

4 thoughts on “Labour vote share to decline?

  1. The assumption that Cardiff Central would shift from Labour to Lib Dem would need some clarification. I don’t see how Cardiff Central will buck the trend and produce a Lib Dem revival when they face decimation elsewhere.

  2. There are a number of interesting questions that arise from this poll. The first is the decline in support for the Labour Party since the last election. In 2011, the Labour Party had 42.3% of the constituency vote, so the above result shows a drop of 6 points in their support which is significant. As Roger points out, the significance is measured in loss of seats. The question for me would be why the drop in support.

    There are a number of possible explanations. One would be that the Welsh Government have had a series of negative headlines regarding the performance of the Welsh Health Service. The attacks from the Conservative Government may have had some effect but I think it is more likely that there is an image of a Government in Cardiff Bay that is not on top of the problems facing the WHS and that this has been self-inflicted as they have resulted from cuts imposed by the Government itself. Another would be the number of traditional Labour voters, particularly in the Valleys, who have gone over to UKIP as an anti-politics vote.

    The rise in Plaid’s support has been surprising. Up from 19% in 2011 by 5 points, it is an advance that has crept up on the inside rail without any apparent reason why. Perhaps a constituency analysis would offer some help here. They have a new leader who performs well on television but lacks the intellectual depth of an Adam Price. The success of the vote in Anglesey, for example, cannot be ascribed to anything Ms Wood had to contribute.

    If UKIP do succeed in gaining 5 AMs then they will finally have arrived in the Welsh political context. It is hard to see this as anything other than being caught up in the slipstream of the European elections. It will be interesting to see what their spedific policies will be with regards to Wales.

  3. I had the impression that the Welsh electoral cycle followed Westminster so when the Tories were in power in London Labour’s vote in Wales went up. Here we have the opposite effect with Labour’s vote declining. Could it be that some voters at least are reacting against the government in Cardiff rather than the UK government? That would be a healthy development.
    Incidentally, the electoral system is not proportional and works in Labour’s favour. With below 40 per cent of the vote, they are expected to get 47 per cent of the seats. That fact makes a non-Labour coalition very difficult and if the protest vote diverts to UKIP, who are irrelevant in the context of the Welsh Assembly, anything but a Labour government becomes impossible. Vote UKIP, get Carwyn – for ever.

  4. Billy – the seat projections are calculated using uniform national swing: ie you adjust the results from 2011 for each party in each seat by the % change since 2011 indicated by the poll. As Cardiff Central was won by a margin of only 0.2%, and as the poll has Labour falling by over 6% since 2011 while the LDs are down by a little less than 6%, uniform national swing would therefore project the LDs to take Cardiff Central.

    This does raise the obvious question of how credible uniform swing assumptions are – could we really expect the LDs to take a seat when their national vote is falling by more than half? I’ll be considering this in a future blog post. But my basic views is that while of course we wouldn’t expect the next election to produce wholly uniform swings, uniform national swing is probably the least bad methods of projecting poll findings in the aggregate, even if one should have some caution about taking too literally all the individual seat projections.

Comments are closed.

Also within Politics and Policy