Author Archives: Roger Scully

Roger Scully

About Roger Scully

Professor of Political Science in the Wales Governance Centre at Cardiff University. Jazz fan and horse-racing follower. Dog-walker.

National Assembly Voting Intentions in the New Barometer Poll

Most of the attention of all political observers is, of course, currently focussed on the forthcoming general election. Nonetheless, our new Welsh Political Barometer poll has also continued to ask people about how they would vote in an election to the National Assembly. What did it find?

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


Labour: 37% (-1)

Conservatives: 23% (+2)

Plaid Cymru: 20% (+1)

UKIP: 11% (no change)

Liberal Democrats: 6% (-1)

Greens: 3% (-1)

Others: 0% (-1)


On the standard assumption of uniform national swings, 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.

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


Labour: 34% (-2)

Conservatives: 23% (+1)

Plaid Cymru: 20% (+1)

UKIP: 11% (no change)

Liberal Democrats: 6% (+1)

Greens: 5% (no change)

Others: 2% (no change)


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


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

Conservatives: 14 seats (6 constituency seats +8 list seats)

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

UKIP: 5 seats (5 list seats)

Liberal Democrats: 2 seats (2 constituency seats)


The numbers in this particular poll project the Greens just losing out on a regional list seat in Mid and West Wales, in contrast to some other recent polls that have had them narrowly winning a seat there. Instead we have five parties winning seats – but with Labour remaining a long way ahead of all the other parties.

Overall, the new Barometer poll shows for the National Assembly, as it does for Westminster, mainly small changes in the support levels of each of the parties. All of the changes on the previous poll reported here are well within the standard 3% ‘margin of error’, so they could well be the result simply of the normal variation that one would expect to see between the different samples that individual polls are able to gather. Nonetheless, it is always better to see your numbers going up rather than down, and so one would expect the Conservatives and Plaid Cymru to be modestly encouraged by the numbers here. While Labour might be a little disappointed to see their support levels slip marginally, they will also be pleased to remain well ahead of all the competition. The poll adds to the body of evidence suggesting that while UKIP are not attracting quite the same level of support now that they did late last year, they are currently on course to be a significant presence in the next National Assembly. And the poll continues to provide evidence that the Liberal Democrats are not yet recovering their support base; for them to remain a substantial force within the National Assembly after May 2016, something will have to change.

The Latest Welsh Political Barometer Poll

Today’s new Welsh Political Barometer poll gives us the very latest guide to how the people of Wales are intending to vote in May’s general election. Crucially, it is the first poll to be conducted in Wales since the ITV Leaders Debate of 2nd April. How has that debate, and the campaign so far, impacted on each of the parties?

This is what our poll, conducted by YouGov, found in terms of voting intentions for the general election. (Changes on YouGov’s previous Welsh poll, carried out for The Sun newspaper in late-March and the very start of April, are displayed in brackets):


Labour: 40% (no change)

Conservatives: 26% (-1)

UKIP: 13% (no change)

Plaid Cymru: 12% (+3)

Liberal Democrats: 6% (no change)

Greens: 4% (-1)

Others: 0% (-1)


Thus, we see in the main only quite small changes since the last Welsh poll, with most of those changes being well within the margin of error. The only remotely substantial shift is the increase in Plaid Cymru support, which is up by a third since the last YouGov poll; however, that poll had shown Plaid support at an unusually low level, and may thus simply have been an outlier.

If we apply the swings implied by this poll from the May 2010 general election result uniformly across Wales, this produces the following outcome in terms of parliamentary seats:


Labour: 28 seats (keeping the 26 seats they won in 2010, and gaining both Cardiff Central from the Liberal Democrats and Cardiff North from the Conservatives);

Conservatives: 8 seats (losing Cardiff North to Labour, but gaining Brecon & Radnor from the Liberal Democrats);

Plaid Cymru: 3 seats (no change);

Liberal Democrats: 1 seat (losing both Cardiff Central and Brecon & Radnor, and holding on only to Ceredigion).


This new poll adds to the evidence that the decline in electoral support experienced by the Labour party during 2013 and 2014 has stopped, and even been partially reversed. As the election has come closer, Labour support has firmed up from the levels we saw at the end of 2014. Now polling at a level almost four percentage points higher than the vote share that they won in 2010, Labour look in good shape to continue to win the clear majority of Welsh parliamentary seats in this year’s election.

The poll also confirms that Conservative support in Wales remains robust, and at a level that should ensure that the Tories retain the vast majority of the seats that they won in Wales in 2010. The contrast with the fortunes of their coalition partners continues to be stark. Liberal Democrat support remains at below one-third of the level that they won five year ago. The Lib-Dems hope is that their vote will prove much more robust in their existing seats than across Wales as a whole. It will have to do so for the party to hold all those seats.

The poll adds further to the evidence that the surges in support experienced by UKIP and, to a slightly lesser extent the Greens, in the last twelve months have now ebbed somewhat. Both parties are now polling several points below their peak ratings of a few months ago, and neither would appear to have realistic hopes of winning a parliamentary seat in Wales in 2015.

The one party clearly moving forward in this poll is Plaid Cymru. Their support has risen by three points from YouGov’s previous Welsh poll. While that last poll may have given the party an unusually low score, at 12% the current poll equals Plaid’s highest rating in a YouGov poll since the 2010 general election. This improvement in Plaid’s position may reflect Leanne Wood’s showing in the first televised Leaders’ Debate; if so, it is possible that her presence in this week’s second debate may also help her party. Still, Plaid remains in fourth place in Wales, and on uniform swings would be struggling to add to its existing three seats in Wales.

Postscript: My by-now customary postscript, for you cognoscenti of the Blog: Ratio Swing seat projections of the numbers from this poll:


Labour: 28 seats (keeping the 26 seats they won in 2010, and gaining both Cardiff Central from the Liberal Democrats and Cardiff North from the Conservatives);

Conservatives: 8 seats (losing Cardiff North to Labour, but gaining Brecon & Radnor from the Liberal Democrats);

Plaid Cymru: 4 seats (keeping the three seats they currently hold, and gaining Ceredigion);

Liberal Democrats: 0 seats (losing Cardiff Central to Labour, Brecon & Radnor to the Conservatives, and Ceredigion to Plaid Cymru).


The poll for ITV and the Wales Governance Centre had a sample of 1143 Welsh adults and was carried out by YouGov on 13-15 April 2015. Full details will be published on the blog soon.

Some Bed-Time Reading

As we get ever closer to the election, many of you will be wanting to learn more about the coming election, past elections, the electoral battleground and so on.

Obviously, my main recommendation is that you should carry on reading Elections in Wales! And, of course recommend it to all your friends and family. But in addition to this site, there is a wealth of other material available – with much of it easily accessible on-line. Here are a few recommendations of sources that I have found particularly useful.

Seat Guides: There are a number of efforts at guides to the 650 constituencies that will be fought in the election:



UK Polling Report ( is simply in a class of its own for the sensible discussion of opinion polling in the UK. The site is run by Anthony Wells: although Anthony is someone of firm political convictions and works for YouGov UK, the site is rigorously impartial both between the different parties and the various pollsters. A particularly useful post was this recent one ( about the different methods used by pollsters. : A very good News Statesman site; it includes profiles of many seats, plus latest state of play on polls, and more details on many of the detailed questions asked by pollsters. As the web address indicates, a site run by the London School of Economics; includes plenty of excellent short essays and analytical pieces.


Twitter Accounts:

Among Twitter accounts that you may not be aware of, I’d particularly recommend following @UKGenElection and @NCPoliticsUK . @LordAshcroft is also worth keeping an eye on.


Finally, for those of you who have not yet bought a copy, you really should read Sex, Lies and the Ballot Box

YouGov Welsh Poll for The Sun

So there I was a few days ago, enjoying a quiet Bank Holiday Monday, when news starting seeping through about a new YouGov poll in Wales, published by The Sun. To those of you who were wondering why I had not flagged this up previously, I can only say that this was the first I had heard of it! It was not one of YouGov’s regular Barometer polls, and I hadn’t been forewarned about it.

The poll was conducted between 26-31 March – the full tables are now available here. The dates of the fieldwork are important: the poll’s fieldwork overlapped slightly with that of the most recent Barometer poll, and was also all conducted before the seven-way Leaders’ Debate on 2nd April. We therefore cannot use this poll to infer anything about reactions in Wales to that debate. For that, we will have to wait until the next Barometer poll – which isn’t too far away!

(By the way – for those of you thinking that it seems a little strange for someone to pay for a poll to be conducted before the debate, and then hold back the results until after the debate, and then release them on a Bank Holiday Monday when almost no-one will notice… well, I agree with you. It does all seem a bit odd to me as well. But as I’m not on close terms with the editorial team at The Sun, I can’t reveal any of the thinking behind this.)

Anyway, enough of that; what did the poll actually say? Here are the figures for general election voting intention (with changes from the previous YouGov poll in Wales, the most recent Barometer poll, indicated in brackets):


Labour 40% (no change)

Conservative 27% (+2)

UKIP 13% (-1)

Plaid Cymru 9% (-2)

Liberal Democrats 6% (+1)

Greens 5% (no change)

Others 1% (no change)


Applying the changes since 2010 implied by this poll, using the standard assumption of Uniform National Swing, then the figures from this poll would suggest the following outcome in terms of seats:


Labour: 29 seats (gaining Arfon, Cardiff Central and Cardiff North)

Conservatives: 8 seats (losing Cardiff North but gaining Brecon & Radnor)

Plaid Cymru: 2 seats (losing Arfon)

Liberal Democrats: 1 seat (losing Cardiff Central and Brecon & Radnor)


On the alternative assumption of Ratio Swing, we get the following seat projections:


Labour: 30 seats (gaining Arfon, Cardiff Central, Cardiff North and Carmarthen East & Dinefwr)

Conservatives: 8 seats (losing Cardiff North but gaining Brecon & Radnor)

Plaid Cymru: 2 seats (losing Arfon and Carmarthen East & Dinefwr, but gaining Ceredigion)

Liberal Democrats: 0 seats (losing Cardiff Central, Brecon & Radnor and Ceredigion)


So what can we make of this poll? The fact that its fieldwork is rather out-of-date, and precedes the recent leaders’ debate, makes this poll distinctly less useful than it might have been. Nonetheless, it is now the fourth YouGov poll this year which indicates that the decline in Labour support that persisted more-or-less throughout 2013 and 2014 has come to an end, and has even been slightly reversed. The last three YouGov poll have put Labour on 39%, 40% and 40%, so they have all had Labour some 3-4 points above the vote share that the party won in 2010.

The 27% support reported for the Conservatives is their highest reported vote intention for Westminster in a Welsh poll since autumn 2010. They appear to have begun the campaign at, or possibly even slightly above, the level of support that they won in Wales in 2010. That is a rather impressive, not to say surprising, achievement for the main party in the UK government implementing an austerity agenda over the last five years. The contrast with the fate of their coalition partners continues to be striking. This poll does actually show a very small up-tick on Lib-Dem support levels. But they are still well below one-third of the level of support they won in the 2010 election.

UKIP’s very small drop in support from the previous YouGov poll in Wales may be of little significance. What is probably more noteworthy is that this poll confirms the somewhat more substantial fall in support that the party in Wales have experienced since late last year, when they were polling in the high teens. However, even more disappointed by the results may be Plaid Cymru. It is possible that they were not favoured by the timing of the poll – with the fieldwork starting on the day of the Cameron-Miliband non-debate that would have focussed attention on the two largest UK parties, and concluding before the leaders’ debate that included Leanne Wood; nonetheless, Plaid really would have hoped to be doing better than this, which is their poorest figure for Westminster voting intention since July 2013.

The Sun poll also carried questions on voting intention for the National Assembly – to the best of my knowledge these have not actually been published yet. The figures for the Constituency ballot (with changes from the last Barometer poll) were:


Labour 38% (+1)

Conservative 21% (-2)

Plaid Cymru 19% (no change)

UKIP 11% (-1)

Liberal Democrats 7% (no change)

Greens 4% (+1)

Others 1% (no change)


The figures for the regional list vote were:


Labour 36% (+2)

Conservative 22% (+1)

Plaid Cymru 19% (-1)

UKIP 11% (-1)

Liberal Democrats 5% (no change)

Greens 5% (-1)

Others 2% (no change)


Assuming uniform national swings, the only constituency seat to change hands would be Llanelli, won by Plaid Cymru from Labour! Taking account of these constituency seats, and the allocation of regional list seats, a uniform national swing projection produces the following result:


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

Conservative: 13 seats (6 constituency seats + 7 list seats)

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

UKIP: 5 seats (5 list seats)

Liberal Democrats: 1 seat (1 constituency seat)

Greens: 1 seat (1 list seat)

A Few Things You Might Have Missed

We’ve all been having such a poll-tastic time recently that it is easy to miss out on some things. Last week’s latest Welsh Political Barometer poll had a few results that might have slipped past you. So this post is to help you all catch up.

First, the Barometer poll has continued to ask questions about voting intention for the two referendums that are most like to occur in Wales in the next few years: on Britain’s membership of the EU, and on income tax-varying powers for the National Assembly.

For the EU, the results (with changes from our previous poll, conducted earlier in March) were:


Remain in the EU: 44% (+1%)

Leave the EU: 38% (+2%)

Don’t Know / Wouldn’t Vote: 18% (-4%)


This is how the latest poll looks, in relation to the previous ones we have conducted:


Wales, EU Referendum Polls

Poll% Remain% Leave% DK/ NV% ‘remain’ Lead
ITV-Wales/YouGov, February 20134235227
Western Mail/Beaufort, June 2013293735-8
WGC/YouGov, July 2013394021-1
ITV-Wales/WGC/YouGov, December 2013384022-2
ITV-Wales/WGC/YouGov, Feb 201444332311
Walesonline/YouGov, June 20144138223
ITV-Wales/WGC/YouGov, June-July 20144136245
ITV-Wales/WGC/YouGov, September 20144337206
ITV-Wales/WGC/YouGov, December 20144239193
ITV-Wales/WGC/YouGov, January 20154436208
ITV-Wales/WGC/YouGov, early-March 20154336227
ITV-Wales/WGC/YouGov, late-March 20154438186


Thus we can see that although the polls on this question have tended to be very close, those wishing Britain to remain in the EU have led in every one of the last eight polls conducted by YouGov on this topic, and in every poll conducted during 2014 and 2015.

What about income tax? Although the likelihood of this referendum actually taking place continues to recede, we have persisted in asking about it. Out latest poll’s result (with changes from the previous Barometer poll again in brackets) was:


In Favour of income tax-varying powers: 37% (no change)

Against income tax-varying powers: 40% (+4%)

Don’t Know / Wouldn’t Vote: 22% (-5%)


And here, again, are the run of previous polls on the topic:


Wales, Income Tax Referendum Polls

Poll% Yes% No% DK/ NR% ‘No’ Lead
ITV-Wales/YouGov, February 2013393427-5
ITV-Wales/WGC/YouGov, December 20133538263
Western Mail/Beaufort, December 2013323038-2
ITV-Wales/WGC/YouGov, Feb 201431422811
ITV-Wales/WGC/YouGov, May 20143339286
Walesonline/YouGov, June 20143441257
ITV-Wales/WGC/YouGov, June-July 201432422610
ITV-Wales/WGC/YouGov, September 20143839241
ITV-Wales/WGC/YouGov, December 20143738251
ITV-Wales/WGC/YouGov, January 20153739242
ITV-Wales/WGC/YouGov, early-March 2015373627-1
ITV-Wales/WGC/YouGov, late-March 20153740223


Here again the balance between those in favour and those against remains tight, but those opposing tax powers have resumed their narrow lead after being very slightly behind in the previous poll.

One further topic that was explored in the recent poll was the importance of different issues. YouGov asked a pair of questions:

  •  ‘Which of the following do you think are the most important issues facing the country at this time?’; and
  • ‘Which of the following do you think are the most important issues facing you and your family?’

Respondents were invited to choose up to three issues from a lengthy list (hence the percentages below add up to well over 100). In the tables below I’ll produce the results, in order, for both questions from among our sample in Wales. For comparison I’ll also show alongside these the findings from a GB-wide poll that asked the same question that was conducted approximately ten days prior to our Barometer poll.


Most Important Issues Facing the Country:

Immigration and Asylum49%51%
The economy46%45%
Welfare benefits27%27%
The environment10%9%
Family life and childcare7%7%
Increased powers for Wales5%-
None of these1%1%
Don’t Know3%2%


Most Important Issues Facing You and Your Family:

The economy39%43%
Welfare benefits20%15%
Immigration and Asylum18%19%
Family life and childcare13%14%
The environment9%10%
Increased powers for Wales5%-
None of these4%2%
Don’t Know5%4%


Thus, we see few major differences between Wales and the rest of Britain. Health is a little higher in the list of both personal and national priorities in Wales. And housing is a little lower – which is perhaps understandable if you consider that a significant proportion of the GB-wide sample are having to cope with London/South-East England property process. But, overall, the issue agendas are strikingly similar.

There are also few substantial differences between the national and personal agendas evident here. Immigration and Asylum, and also Europe, recede somewhat in importance when we ask people to focus on what they see as important to them and their families. But much of the agenda is common. These lists emphasise the over-riding importance of the economy and the NHS – and why we will be hearing a great deal more about them over the next few weeks.

Attitudes to Different Levels of Government

Much of the attention given to the results of last month’s BBC/ICM poll was devoted to their findings on the main UK party leaders, and to the answers given to their regular question on constitutional preferences. However, the poll covered a number of other interesting areas.

One, which I’d like to highlight here, concerned attitudes to different levels of government. The poll used a slightly different question format from those I’ve seen in previous studies, so it is worth explaining in full.

Respondents were asked the following three questions:

“Which one of the following political institutions do you have the most respect for?”

“Which one of the following political institutions do you trust most to do the right thing for people?”

“Which one of the following political institutions do you think is most likely to improve things for you and your family?”

Respondents were then allowed to choose between several options for each question:

  • Your local council
  • The Welsh Assembly
  • The Westminster Parliament
  • The European Union
  • None of them
  • Don’t Know


This set of questions thus probed some interestingly distinct aspects of attitudes to the main levels of government in the UK.

So what did people say? The following table summarises the poll’s results for the three questions:

Most Respect ForTrust to Do Right ThingMost Likely to Improve
Local Council24%26%20%
Welsh Assembly31%35%38%
European Union7%6%6%
None of them9%10%6%
Don’t Know2%2%4%


What can we make of these results? The thing that most immediately stands out from the figures, to me at least, is that the National Assembly scores highest for all three questions. Although its lead over the other levels of government on the ‘Most Respect’ question was rather narrow, on the ‘Trust’ and ‘Most Likely to Improve Things’ questions the Assembly is rather further ahead of both local councils and Westminster. We might want also to note that the Assembly scores these leads despite the fact that local councils actually did pretty well themselves in another question in the same poll, one which specifically asked about local service delivery. So the responses here don’t seem to be just reflecting a sort of ‘best of a bad lot’ set of attitudes.

Instead, I’d be inclined to view these results as being consistent with some others that have suggested that, while most people seem to retain at least some degree of suspicion about all politicians and governmental institutions – an attitude that I’d generally regard as fairly healthy – there has developed a broad, basic goodwill to the Assembly amongst many people in Wales. We must not over-state this: few, if any, are wildly in love with the National Assembly and its members. But as well as being generally supportive of the principle of devolution, the majority of people seem to have come to regard the Assembly as – whatever other faults it may have – an institution that will at least be concerned with and focussed on the interests and problems of the people of Wales.

For the other levels of government, it is interesting that the Westminster Parliament scores least well on the question about trust. The latter half of the fieldwork for the poll was actually administered as the recent Rifkind/Straw ‘scandal’ story was breaking; this may have influenced responses. It may also be the case that the memory of previous expenses problems lingers long for some people. It is also notable, if entirely unsurprising, that the European Union scores least well on all three questions. In fact I was slightly surprised that it scored as high as 6-7% on these three items! More optimistic news for Europhiles came elsewhere in the poll, however: when asked whether or not the UK would be better off remaining in, or outside, the European Union, there was a strong balance (63% to 33%, with 4% Don’t Knows) in favour of remaining in the EU.

Overall, I think this was an interesting and innovative set of questions for the BBC/ICM poll to run. I think they should be applauded for doing so. And I hope they repeat the items at least occasionally in the future, so that we can see whether attitudes on these matters change over time.

The Leaders’ Debate

Many of you will have watched the seven-way party leaders’ debate last night. Personally I thought the format worked slightly better than I had expected – or feared – for which much credit must go to Julie Etchingham, the presenter. It didn’t quite turn into the seven-way shouting match that some had feared, although it verged on it once or twice. I also think that all the leaders, notwithstanding their different styles, performed creditably. But then there are reasons why people become party leaders (i.e. they tend to be quite good at this politics stuff).

Still, no-one really cares what I think. What did the public make of it all? There was the, now customary, rather silly race between the pollsters to deliver their verdicts on who ‘won’ in the instant reaction polls. (For what little it is worth, I believe my friends at YouGov actually got their figures out first…). This race rather obscures the fact that some more interesting data emerged a little later, in the details of the polls carried out by the four companies doing post-debate polls last night.

You can find the detailed poll findings here: from YouGovICMSurvation, and ComRes. I’d encourage you to look through them for yourself. There are a number of questions in these surveys which strike me as much more useful and informative than those that simply asked for a single ‘winner’.

One of the best questions was in YouGov’s poll, where instead of asking for people to nominate one winner they asked for people to rate all the leaders out of 10. Unfortunately they have not been able to provide a ‘regional’ breakdown on these figures; however, these are the GB-wide averages:


Sturgeon: 6.7

Cameron: 5.9

Miliband: 5.9

Clegg: 5.5

Farage: 5.5

Wood: 5.1

Bennett: 4.7


The other polls do provide some regional breakdowns; although the sub-samples are fairly small, and not weighted for representativeness within those regions (and so should therefore be interpreted with considerable caution), the figures still give some useful indications. For instance, all the Scottish sub-samples show Nicola Sturgeon rating very strongly among Scottish viewers of the debate. It is difficult to imagine that the debate will have done anything other than reinforce her party’s currently strong position in Scotland.

ICM do have a question on the leaders’ performances which is, I think, almost as good as that of YouGov. Respondents were asked to state whether they thought each of the leaders had performed well or badly. Subtracting the percentage of ‘Badly’ responses from the ‘Well’ ones for each leader, we get the following results:


Sturgeon: +38

Miliband: +28

Cameron: +21

Clegg: +16

Farage: +14

Wood: +12

Bennett: +1


Among the Welsh sub-sample (a fairly small one, at only 78 respondents), these were the following net well-badly ratings:


Sturgeon: +66

Wood: +50

Bennett: +45

Miliband: +15

Cameron: +12

Farage: -4

Clegg: -7

Welsh Assembly Figures from the Latest Welsh Political Barometer Poll

As well as exploring voting intentions for the forthcoming general election, our latest Welsh Political Barometer has continued to ask about how people would intend to vote in an election to the National Assembly.

So what did we find? Where do the parties stand right now in devolved voting intention?

For the constituency vote, the results of our new poll were (with changes from our previous poll, earlier in March, in brackets):

  • Labour 37% (no change)
  • Conservative 22% (no change)
  • Plaid Cymru 19% (-1%)
  • UKIP 12% (+1%)
  • Liberal Democrats 7% (+1%)
  • Greens 3% (-1%)
  • Others 1% (no change)

On these figures, and assuming uniform national swings across Wales, two constituency seats would change hands from the results in the last Assembly election in May 2011: Labour would lose Llanelli to Plaid Cymru. They would also (which may surprise some people, but is simply the projection produced by uniform national swing) lose Cardiff Central to the Liberal Democrats.

For the regional list vote, we saw the following results (with the changes from our early-March poll again indicated in brackets):

  • Labour 34% (+1%)
  • Conservative 21% (-1%)
  • Plaid Cymru 20% (-1%)
  • UKIP 12% (no change)
  • Greens 6% (+1%)
  • Liberal Democrats 5% (no change)
  • Others 2% (no change)


Thus, we can see that none of the parties have moved more than a single percentage point on either vote since the last poll! Taking into account both the standard ‘margin of error’ for opinion polls (roughly 3% either way for a poll with a sample size of just over 1,000 respondents), and the fact that YouGov quote figures rounded to the nearest full integer, I think we can call this essentially a ‘no change’ poll. I don’t think, for instance, that we should read very much into the tiny changes in reported list vote support that have now relegated the Liberal Democrats into sixth place for that ballot.

Taking into account both the constituency and list results, and assuming uniform national swings across Wales, this produces the following projected seat outcome for a National Assembly election (with aggregate changes from 2011 indicated in brackets):

  • Labour: 28 (-2): 26 constituency AMs, 2 list AMs
  • Conservative: 13 (-1); 6 constituency AMs, 7 list AMs
  • Plaid Cymru: 11 (no change); 6 constituency AMs, 5 list AMs
  • UKIP: 5 (+5): all list AMs
  • Liberal Democrats: 2 (-3); 2 constituency AMs
  • Greens: 1 (+1): 1 list AM

This is now the fourth time in a row that our Barometer poll has projected an outcome which would mean six different parties being represented in the National Assembly. But I must also issue my customary note of caution before people start talking about what sort of National Assembly we are ‘on course’ for after May 2016. Not only is there a great deal of political water to flow under various bridges before then (with the outcome of this year’s general election likely to do plenty to shape the electoral context for next year). We should also be aware that the results projected above depend on some very tight outcomes for the final list seats in several regions. Very small changes in support could change the balance between the parties rather significantly.

Postscript: And for the real hard-core enthusiasts out there, here are Ratio Swing projections from the same poll for the National Assembly:

  • Labour: 28 seats (26 constituency seats; 2 list seats)
  • Conservatives: 12 seats (6 constituency seats, 6 list seats)
  • Plaid Cymru: 10 seats (6 constituency seats, 4 list seats)
  • UKIP: 7 seats (all list seats)
  • Liberal Democrats: 2 seat (2 constituency seats)
  • Greens: 1 seat (1 list seat)

New Welsh Political Barometer Findings!

The latest Welsh Political Barometer findings on how people are intending to vote in May’s general election show the Labour party to be still well on course to win a clear majority of Welsh seats.

When asked by YouGov how they would vote in a general election, our respondents gave the following responses (with changes on our last poll, earlier this month, displayed in brackets):

Labour: 40% (+1)

Conservatives: 25% (no change)

UKIP: 14% (no change)

Plaid Cymru: 11% (+1)

Greens: 5% (-1)

Liberal Democrats: 5% (no change)

Others: 1% (no change)


If we apply the swings implied by this poll from the May 2010 general election result uniformly across Wales, this produces the following outcome in terms of parliamentary seats:

Labour: 28 seats (keeping the 26 they won in 2010, and gaining both Cardiff Central from the Liberal Democrats and Cardiff North from the Conservatives);

Conservatives: 8 seats (losing Cardiff North to Labour, but gaining Brecon & Radnor from the Liberal Democrats);

Plaid Cymru: 3 seats (no change);

Liberal Democrats: 1 seat (losing both Cardiff Central and Brecon & Radnor, and holding on only to Ceredigion).


Clearly, very little has changed since our last poll: this one represents, perhaps, the ‘calm before the storm’ of the full-scale campaign that gets under way this week. Nonetheless, I think even in a poll like this which shows very little change, there are still several features worthy of note.

Having seen their vote share decline steadily in the Welsh opinion polls throughout 2013 and 2014, this is the third Barometer poll of 2015 to show that Labour have halted that decline, and even reversed it to a slight extent. This places the party in a strong position to make at least some gains in the general election. The Conservatives’ poll rating also remains robust at a level very close to that which they won in the 2010 election, while Plaid Cymru will be encouraged to be edging up very marginally in support, again to more-or-less the level they won in the last election. However, Plaid remain in fourth place – behind UKIP, whose decline in our previous two Barometer polls appears to have levelled out for now.

While the Liberal Democrats’ poll rating also remains steady, they will surely be less encouraged by stability than many of their opponents. The Lib-Dems continue to poll at only one-quarter of their 2010 vote share, and have made no ground at all since the previous Barometer poll. About the best thing that can be said for their performance here is that at least they are no longer in sixth place – the slight fall in support for the Greens places both of those parties in a joint, but rather distant, fifth.

As we have done in all the Barometer polls this year, and as we will continue to do right through to election day, our new poll also asked about how likely people were to actually vote in the election. Respondents to the poll were asked to rate their likelihood of voting on a 0 to 10 scale, where 0 means ‘Definitely will not vote’ and 10 means ‘Definitely will vote’. Some 67 percent of respondents placed themselves at 10 out of 10 on this scale.

Particularly interesting are the differences between the parties in how certain each of their supporters are to vote. And here there is further bad news for the Welsh Liberal Democrats. Fully 18 percent of all those who indicated that they would vote Lib-Dem in the general election actually placed themselves between 0 and 3 on the scale of how likely they were to vote; no other party had more than 4 percent of its supporters claiming to be so unlikely to take part in the election. So not only do the Liberal Democrats have relatively few supporters remaining in Wales; even those that remain seem much less motivated than the supporters of the other parties. YouGov do not weight respondents by likelihood to vote when reporting polling numbers; had they done so, the Liberal Democrats’ position in Wales would have looked even worse than it already does.


Up-Date (01/04/15; and this is not an April Fools’ joke…): Apologies for being rather slow on the analysis this week, but I was away at a conference on Monday and Tuesday. Anyway, as has become customary, I have also computed Ratio Swing seat projections for the latest Barometer poll. These are:

Labour: 28 seats (keeping the 26 they won in 2010, and gaining both Cardiff Central from the Liberal Democrats and Cardiff North from the Conservatives);

Conservatives: 8 seats (losing Cardiff North to Labour, but gaining Brecon & Radnor from the Liberal Democrats);

Plaid Cymru: 4 seats (keeping all three seats that they currently hold, and gaining Ceredigion from the Liberal Democrats);

Liberal Democrats: 0 seats (losing Cardiff Central, Brecon & Radnor, and Ceredigion).

As I’ve explained previously, Ratio Swing tends to produce numbers that are particularly unfavourable  to a party that has lost a significant proportion of support since the last election – just the position that the Lib-Dems are in.


I’ll be back later in the week with further results from and analysis of the poll.

(The poll for ITV and the Wales Governance Centre had a sample of 1189 Welsh adults and was carried out by YouGov on 24-27 March 2015.)

Where We Work We Win?

One of the more remarkable statistics about the last general election in Wales is that the Liberal Democrats gained more than 10% of the vote in every Welsh seat bar Ynys Môn. This statistic speaks to a time that was only a few years ago but, politically, feels almost like another century. Since entering coalition in London, times have been unremittingly tough for the Lib Dems. Although the 2011 devolved elections saw the Welsh party mostly escape the heavy losses inflicted on their Scottish counterparts, last year’s European elections were particularly grim for the party here in Wales: their 3.9% share of the Welsh vote, which put them in sixth place, was by some way the Lib-Dems’ lowest vote share anywhere in Britain.

Nor are things looking obviously brighter for the forthcoming general election. The party’s showing in the Britain-wide opinion polls remains stubbornly low, at around one-third of their vote share in 2010. The picture of the Welsh polls is similar – indeed, possibly even worse. The most recent Welsh Political Barometer had the party in sixth place, on five percent of the vote: almost exactly one-quarter of the 20.1% that they won here in 2010.

One thing that has kept Lib-Dem spirits up, however, is the belief that while the party may suffer substantial vote losses in seats that they don’t hold, they will be able to put up a much stronger fight in many seats they do hold. Well-known and hard-working incumbent Liberal Democrats MPs, many in the party contend, will be much more electorally resilient than the national polls would suggest.

Is there any evidence to support this claim? Lord Ashcroft’s constituency polls have provided a mixed picture. In some of their seats the Lib-Dem vote seems to be holding up well, but in others much less so. Two such polls conducted in Wales illustrate this mixed picture well. A poll conducted in Cardiff Central in September 2014 suggested that the Lib-Dems’ vote share was down by around two-fifths on 2010 and that they were well behind Labour in the seat; however, another poll done two months later in Brecon and Radnor had their vote holding up rather better and the party clinging onto a narrow lead from the Conservatives.

Is there any other evidence we can bring to bear? Lacking a spare few tens of thousands of pounds to conduct my own series of constituency polls – I had a look down the back of the sofa, but no luck – I’ve been trying to see what clues we might be able to draw from the polling evidence that we do have. The detailed data supplied by YouGov for each Welsh Political Barometer poll contains the parliamentary constituency of each survey respondent. We also, crucially, have a measure of how each respondent voted – if they did vote – in the 2010 general election. With this information to hand, we are then able to compare directly how the same individuals, across different types of constituencies, voted in 2010 with how they say they intend to vote now.

What I’ve done is combine the last three Welsh Political Barometer polls: those conducted in December, January and March. (Between them they give us 3,446 total respondents; 604 of whom are recorded by YouGov as having voted for the Liberal Democrats in the 2010 general election. This is slightly below the percentage of people in Wales who actually voted Lib-Dem in 2010, mainly because the polls include some people who didn’t vote in 2010).

These three Barometer polls showed the Lib-Dems on 5% (December), 6% (January) and 5% (March). The average of 5.33% across the three polls is just barely above one-quarter of the support level the party was winning in 2010. But to what extent is this decline uniform? Or do the detailed Barometer figures show the party doing notably better in their four key seats? To assess this, I’ve split those respondents into two categories:

  • Those respondents in the three parliamentary seats that the Lib-Dems currently hold (Brecon & Radnor, Cardiff Central, and Ceredigion), plus the one other seat that they appear to be seriously targeting this time around, Montgomery. (I did ponder whether or not to include Montgomery; however, the Lib-Dems own recent Welsh conference talked very positively about their prospects in this seat, so I have decided to take them at their word. It is maybe worth adding that I checked, and the results I report below do not differ substantially if Montgomery is removed from this category).
  • The respondents in the other 36 seats.

The data is all taken from national Welsh polls and so they are not sampled to the demographics of each of the specific constituencies. For this reason the findings should be seen as indicative rather than definitive. Still, I think they are interesting.

In the 36 non-core seats, we see the Lib-Dems’ support falling by more than three-quarters on that gained in the 2010 election. That is much as we would expect to see based on the polls. What about the four key seats? Well here we do indeed see clear evidence that the Lib-Dems are doing better in those seats. But when you are losing about three-quarters of your support nationally, better is very much a relative term. What it means in the case of the Lib-Dems is that, according to the figures from these recent polls, they have lost slightly more than three-fifths of their 2010 support among our sample of respondents in these four seats. Yes, that is better than elsewhere. But it is not very much better.

If they do lose three-fifths of their 2010 support even in their four key seats, then the Liberal Democrats would come nowhere near to regaining Montgomery. They would also be very unlikely to hold any of Brecon & Radnor, Cardiff Central, and Ceredigion. Is the game up, then, for Roger Williams, Jenny Willott and Mark Williams? Not necessarily. There are several reasons for being cautious in interpreting these figures.

  • First, these figures are taken from polls conducted between December 2014 and March 2015. As all readers of this blog should know by now, a poll is not a prediction of a future event but an attempt to measure attitudes at the time it is taken. Several weeks of remaining campaigning could well improve the Lib-Dems’ position in general.
  • Second, the Lib-Dems will be campaigning hard in their key seats over the next few weeks. As with campaigners for all parties in marginal seats, they will be trying to get voters to focus on the local dynamics of their particular constituency. (Think ‘Focus’ leaflets with slightly dubious bar-charts, and horse-racing graphics alongside the message that ‘It’s a Two Horse Race’…). As this message gets across, one would expect the Lib-Dems’ position in their key seats to improve somewhat.
  • Third, these polls used a generic voting intention question. Yet Lord Ashcroft’s constituency polls have shown that one can get rather different results when asking people to think specifically about their own constituency.

In short, the evidence at present is that the Welsh Lib-Dems will do rather better than the national swings in their four key seats. In the rest of Wales they are likely to see their vote slip back very considerably from 2010. However, with more than 10% of vote in every seat but one last time, the Lib-Dems have a lot of votes they can afford to lose. What matters to them this year is their vote in their key strongholds. Here we must, I think, expect them to be considerably more resilient. The key question for the Liberal Democrats in 2015 is whether better will be good enough.