Skip to main content
Coronavirus (COVID-19): Latest updates.


And There’s More… The Rest of the Welsh Political Barometer Poll

14 December 2015

Well, I hope that you agree that last week’s Welsh Political Barometer poll was worth waiting for. (And it’s a bloody shame if you don’t think so, because it will be a while until our next Welsh poll – the next Barometer is not due until February, and I’m not aware of any other scheduled polls before then.) I’ll be assessing the standing of the parties, as revealed by that poll and other evidence, in my annual End of Year Review before long.

But before we wrap up discussion of the latest Barometer poll, there are a few findings that I haven’t yet discussed on the blog. First of all, there was our regular tracker question on the EU Referendum. This time around the figures (with changes from the September Barometer in brackets) were:


Remain a member of the European Union: 40% (-2)

Leave the European Union: 42% (+4)

Would not vote: 5% (+1)

Don’t Know: 14% (-3)


This lead for leave is the first that they have experienced in Wales since the inaugural Welsh Political Barometer poll exactly two years previously. The full run of EU referendum polls in Wales can be seen here:


Poll % Remain % Leave % DK/ NV % ‘Remain’ Lead
ITV-Wales/YouGov, February 2013 42 35 22 7
Western Mail/Beaufort, June 2013 29 37 35 -8
WGC/YouGov, July 2013 39 40 21 -1
ITV-Wales/WGC/YouGov, December 2013 38 40 22 -2
ITV-Wales/WGC/YouGov, Feb 2014 44 33 23 11
Walesonline/YouGov, June 2014 41 38 22 3
ITV-Wales/WGC/YouGov, June-July 2014 41 36 24 5
ITV-Wales/WGC/YouGov, September 2014 43 37 20 6
ITV-Wales/WGC/YouGov, December 2014 42 39 19 3
ITV-Wales/WGC/YouGov,  January 2015 44 36 20 8
ITV-Wales/WGC/YouGov,  early-March 2015 43 36 22 7
ITV-Wales/WGC/YouGov,  late-March 2015 44 38 18 6
ITV-Wales/WGC/YouGov,  May 2015 47 33 21 14
BES/YouGov, May 2015 50 33 18 17
ITV-Wales/WGC/YouGov,  June 2015 44 37 19 7
ITV-Wales/WGC/YouGov, September 2015 42 38 21 4
ITV-Wales/WGC/YouGov, December 2015 40 42 19 -2


As I have discussed previously, there has generally been a pattern in EU referendum polling that Scotland tends to be the most pro-Remain nation within the UK (or at least GB; the few polls on the subject in Northern Ireland have also shown clear margins in favour of Remain), and England the most likely to support a Leave position; Wales tends to be somewhat between Scotland and England, although closer to England. These figures are not inconsistent with that general pattern, and with the broad though modest drift in the direction of Leave in most recent polling across Britain. Those campaigning for Leave will certainly be heartened by this poll. But there is a long way to go, and doubtless a fair few fluctuations to occur, before the eventual vote which we now expect some time next year.

There were also a few additional questions which appear to have been added to the Barometer poll by YouGov, but paid for by other clients. I didn’t know about these in advance – because YouGov are a proper company who fully respect client confidentiality. However, it is clear from the published details (sample size and dates of sampling) that they were included in the same poll.

One of these questions was for J. Jones, a long-time contributor to the comments section of this blog (amongst, doubtless, many other accomplishments). The following, rather lengthy, question about educational provision in Wales was asked:

“In most local authorities in Wales, councils are required to survey parents to assess the demand for Welsh medium education in primary schools and make provisions for these demands. In local authorities where there are few or no English medium primary schools do you think local authorities should, or should not be required to survey parents to assess the demands for English medium education in primary schools and make provisions for these demands?”

Overall, 51% of the poll’s respondents chose the “Should be required to survey parents and make provisions for the demands” option; some 21% chose the equivalent “should not” option, while 28% chose the Don’t Know option. At least as interesting as the overall result, however, are some of the detailed breakdowns. Among supporters of the parties in the general election, Labour and Liberal Democrat supporters (and not, as some might have expected, Conservative and UKIP voters) are the most likely to choose the ‘should’ option; much less surprising is that Plaid Cymru supporters are the least likely to do so, and actually split very marginally (45-41) against the idea. The pattern with Labour and Liberal Democrat supporters may well be because the question, as worded, appears to invoke some fairly clear norms of fairness. Among speakers of Welsh there is a more or less even split on the issue (43$ choosing ‘should’ and 42% ‘shouldn’t’; non-speakers of Welsh split more than three to one in favour of ‘should’ – by 53% to 16% – but are also far more likely than Welsh speakers to choose Don’t Know, possibly because they will have be less likely to have thought about issues relating to the language of schooling provision.

Finally, there were a series of questions about Syria, bombing and refugees, asked for WalesOnline. There has already been some coverage of the answers to these questions here. There were a number of questions included on this matter, and the findings are of sufficient interest as to deserve more detailed attention than I can give them in this blog post. The findings deserve a blog post of their own, so I’ll be back fairly soon with that.


  1. J.Jones

    Yes, there was a history to the question and the whole experience was quite interesting. Firstly there’s cost; I wouldn’t have asked if I had to stump up all the money and so I was a conduit for several people who consider that Welsh language politics are not fairly discussed (or at all) by political parties. I have to say that Cymdeithas Yr Iaith and “Yes Wales” started it; they used a YouGov poll to ask a rather unfair question which gave this:

    “The results found that 56 per cent wanted to see all pupils taught to communicate effectively in Welsh with only 33 per cent against the move, while 11 per cent didn’t know.”

    I just thought that this was a dodgy question; It actually asks about the effectiveness of teaching and I would expect that any question couched in this way would come up with a majority in favour of effectiveness in teaching any subject.

    It was used to suggest that Wales wants more WM schools.

    ITV then asked a question about compulsory Welsh in schools which you can see in the June Barometer poll. The headline for that was that 64% of people who responded were of the opinion that compulsory Welsh lessons in English medium schools should stop at the end of key stage 3, age 14. Welsh is compulsory up to age 16 at present.
    Clearly these two polls tell a conflicting story to a degree. Roger points out that the ITV poll may be influenced by the idea of “compulsion”….we don’t like it… anything.

    At first we were just wondering how much people cared about Welsh in education. Was it a vote winner or loser? The technical difficulties of the question rather stopped us there; asking if someone was more or less likely to vote for a party which wanted all Welsh medium schools or wanted no compulsory Welsh was unlikely to be anything other than messy and inconclusive.
    The eventual question also risked having a large number of “don’t knows”. Everyone in the Fro Cymraeg knows that primary education is almost exclusively through Welsh Medium schools. Many agree with this, some don’t. Few people in the rest of Wales know that local authorities are obliged to survey for demand for Welsh medium schools and provide those schools but the LAs of the Fro Cymraeg have no duty to similarly survey parents to see if English medium schools are desired. So the question is one about equality and discrimination at a certain level.
    As Roger points out, Welsh speakers know what the question is about and are evenly split. Plaid voters do not think that the rights given in, say, Newport to request WM schooling should be paralleled by an equal right to request English medium schooling in Anglesey. Not a surprise.

    As I said, an interesting experiment but my feeling is that, in a country where half the people don’t know who has responsibility for the Welsh NHS, asking a question about the finer points of education policy is likely to find a lot of people who know nothing about it.

  2. J.Jones

    Is the Beaufort the only ‘phone poll? It seems to be out of step with YouGov. I wonder about the recent analysis of “what went wrong” amongst the polling companies. Partly they blamed their online panel saying that they appeared to be a more politically aware group (more likely to vote, more knowledgeable). This should be OK when it comes to voting since they will carry disproportionate weight in any vote but not THAT much weight.

    I wonder if YouGov has a more “Liberal Left” panel than is actually present in the population as a whole. Beaufort phone polls might find a truer cross section.

    • Roger Scully

      Yes, it is the only phone poll in Wales thus far. A slightly odd result, because the few GB-wide phone polls thus far have given bigger leads for Remain. Recent YG polls, by the way, have been weighted by 2015 vote, so really should not have the sort of biases you suggest, JJ.

  3. J.Jones

    Two new UK online polls published today with polling dates 10th-12th December:-

    ICM Remain: 42% Leave: 41% DK: 17%
    Survation: Remain: 40% Leave: 42% DK: 18%
    YouGov Wales: Remain: 40% Leave 42% DK: 19%

    Really speaking Neck and Neck with no difference between Welsh opinion and UK opinion. The “Wales benefits by a huge amount” message is not really having an impact and the notion that Wales thinks very differently to the UK doesn’t look well founded.

  4. Nospin


    your last comment is important and I agree that on many major UK issues Wales thinks very much like the rest of the UK.

    It would be interesting if at some point the Barometer asks do we support this idea of a regional veto, ie all 4 nations must vote the same or result nullified for a UK wide referendum utterly daft idea. I believe this is supported by Welsh labour Plaid, not sure about Libs, wonder if these parties will agree to re-run devolution referendum with same stipulation for each constituency or county.

    Re your comments on the language poll what concerns me is that the don’t knows are don’t cares, ie doesn’t affect me so I have no opinion, but the result showed a broad nation wide support for the idea and was only narrowly beaten with Plaid voters and Welsh speakers a virtual draw, both groups having the lowest don’t knows.

    • leigh richards

      when people wake up’ to the tens of thousands of jobs in wales dependent on eu membership we’ll see the polls moving in a different direction.

      • john workman

        your in a dream world sir, the 55 million a day paid to this corrupt club would keep you going in wales ,you still have no valid response try again

        • Roger Scully

          Much though I am grateful for your interest in the blog, I’d prefer it if you found a more suitable forum in which to continue this debate.

  5. J.Jones

    With regard to the poll on equality of language provision I would tend to agree with you that few people are aware of government policy apart from Welsh speakers who usually feel more strongly (and are better informed) about language issues. Do “don’t knows” not care? It’s hard to decide; they obviously have had no occasion to consider it.
    In Wales we have grown to accept that government policy on language goes unquestioned and certainly unopposed but there comes a point where one man’s affirmative action becomes another man’s discrimination. With several years of data showing that pupils from English speaking homes perform less well in Welsh medium schools than similar pupils in similar English medium schools parents are rightly concerned that government is willing to turn a blind eye to that harm being imposed in the Welsh speaking areas.
    In the South East of course WM schools are increasingly popular but freedom to choose remains for the time being.

  6. Dylan Foster Evans

    Thanks to J. Jones for the background to the question on WM education in the Welsh Political Barometer poll. But as education is a complicated and many-faceted field it’s difficult to come up with a truly useful yes/no question.

    The problem with the question as it has been phrased is that it seems to assume that WM and EM are mirror images of each other. But they are different from each other in a significant way. It’s expected that pupils at the end of their time in a WM primary have a native competence in both English and Welsh. But children in EM education are not expected to have native competence in Welsh.

    So WM schools in mainly English-speaking areas produce pupils who have native competence in the main language of those areas (English). But an EM school in a Welsh-speaking area would produce pupils with a relatively low level of skills in the main language of the area. That could have significant implications for them later on should they wish to enter the local job market , and be able to play a full part in local civic society.

    So it would be very difficult for a local authority in a Welsh-speaking area to justify introducing EM schools in the knowledge that the children from those schools would be set at a disadvantage. It is important to consider parental choice as the poll question suggests, but as the current system safeguards the development of English-language skills it seems to me that WM schools in Welsh-speaking areas meet the needs of the children. Certainly, a meaningful yes/no question on this would have to be very complicated!

  7. J.Jones

    Thanks for engaging in the question in a meaningful way Dylan. Firstly the assumption that Welsh is the majority first language in each of the LAs of the Fro Cymraeg. You can track home language for pupils on the Welsh statistics site; Primary pupils who are fluent Welsh speakers and speak it as their home language are:-
    Anglesey 37%
    Gwynedd 56%
    Ceredigion 27%
    Carmarthenshire 21.5%

    All 2015 figures. So, with the exception of Gwynedd the area first language is English and, as the Welsh language use survey tells us (page 48), even amongst those who speak Welsh fluently a proportion (20%) prefer to speak English. The further problem is that analysis of Key stage 2 results in WM schools by language of the home shows pupils from Welsh L1 homes doing well at the end of Primary and pupils from English L1 homes doing less well than, not only their Welsh home language class mates, but also pupils in similar EM schools.

    Parents recognise the disadvantage. The largest secondary school in Gwynedd is the only EM school and it draws pupils from as much as 30 miles away as well as pupils from Anglesey and Western Conwy.

    As for employment; you won’t find many pupils whose home language is English employed in Welsh essential jobs in Gwynedd. Even those who get level 2 at GCSE get a “C” rather than an A or B. In 2015 84% of Welsh home language speakers attained A*-C in Welsh at GCSE and 62% of English home language speakers attained A*-C.
    Looking at A*-B:
    Welsh home language:- 50.2%
    English home language:- 22.4%

    English first language speakers made up only 24.3% of the entry cohort.
    There’s a lot of mythology about Welsh in Wales and Welsh medium teaching in the Fro Cymraeg. The biggest myth is that English home language pupils in WM schools do as well in Welsh by the end of school as the Welsh first language speakers. They don’t by a mile. The second dangerous myth is that other subjects don’t suffer by being taught through a pupil’s second language….they do.

  8. Dylan Foster Evans

    Thanks for the reply J.Jones.

    I don’t think I set out to define ‘Y Fro Gymraeg’ or a ‘Welsh-speaking area’ so I’m sorry if that was misleading. I do realise that English is the most common home language everywhere but Gwynedd. But thanks for the statistics.

    Other than that it’s difficult to respond as there are lots of assertions with little real, verifiable evidence (or links).

    For instance, you say ‘you won’t find many pupils whose home language is English employed in Welsh essential jobs in Gwynedd’. On what basis? And if it were true, would we know why? You also say that overall those who speak Welsh at home do better at Welsh. If so, is that any way surprising (given the sociolinguistics of modern Wales)? Isn’t the obvious comparison between those from English-speaking homes in different kinds of schools?

    But comparing schools and pupils is extremely difficult and complex, given all the socio-economic variables involved. As such, the information presented here is insufficent and unreferenced. And even if it were shown that children from certain backgrounds did less well in certain tests, why assume that the answer would be to stop them from acquiring fluency in Welsh by removing them from WM education, and so reducing their opportunities on their own doorsteps? Wouldn’t thinking carefully about the pedagogy used in the schools for different kinds of pupils be a better starting point? I woudn’t agree that a child’s educational attainment is predetermined by his or her home language.

    In fairness to Roger and his blog, though, I think that this is a different debate so I won’t take it further. And there are numerous blogs and discussions on the advantages of bilingual education.

    But I think it all does go to show that it is very difficult to think of a yes/no question that would do justice to this issue.

  9. J.Jones

    Well yes to part of that Dylan…Roger’s tolerance of his blog being hijacked will wear thin.
    I have in the past taken pains to compare only non FSM pupils since, where job opportunities favour Welsh fluent adults, it is likely that non Welsh speaking parents are more likely to be unemployed or employed in poorly paid jobs. That is English L1 pupils would be more likely to be eFSM. In Gwynedd the most logical move for children of non Welsh speakers is to leave Gwynedd and of course Hywel Jones study of migration by ability to speak Welsh shows young non Welsh speaking people doing just that.

    I got these figures for KS2 in 2014 covering Wales as a whole:-

    The 2015 figures are part of a data set from Gwynedd LA so linking is a problem.

    It would be interesting for Chris Taylor to do a targeted study of his MCS cohort of 11 year olds. I have tried to interest him in doing that…maybe you could have a word?

  10. Dylan Foster Evans

    Thanks for the links J.Jones – plenty of number crunching for someone. It’s hard to imagine though how stopping children from becoming bilingual would be the best way forward for them. There are numerous pedagogic interventions that can be considered should various groups (socio-economic, linguistic etc) show evidence of underperforming. That’s one reason why the yes/now question is insufficient. But that’s for another blog!

  11. J.Jones

    Quite Dylan. I don’t for a minute dismiss the idea that those WM schools could improve outcomes for pupils from English L1 homes. Personally, when, several years ago, I had a dialogue with Prof Bialystok I was struck with two things that she said; there’s no such thing as compulsory immersion schooling…it doesn’t work and there’s no “one size fits all” with children. She also admitted that her samples of pupils in immersion settings actually learned the mechanics of core skills through their home language.

    In Wales there plainly is compulsory immersion schooling and it plainly doesn’t work and we are obsessed with Welsh language immersion in the foundation stage rather than skill acquisition in literacy and numeracy.

    The prerequisite to dealing with failings in education is an acknowledgement of the problem. We are politically incapable of admitting that WM schools under-perform.

Leave a reply

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