The Silk Commission: new evidence on public attitudes to devolution

The Silk Commission has published new evidence on public attitudes to devolution in Wales. For the most part, this evidence tells us little that is actually new. Nonetheless, I think it is important in two respects. It reinforces previous findings. And it adds some further details to the broad picture.

On the first point: the Silk research, conducted by Beaufort, offers a very similar picture of public attitudes to devolution to that produced by other studies over the last decade and more conducted by GfKNOP, ICM, NatCen and YouGov. These various studies have used different survey methods (face-to-face, telephone and internet), differing questions, and been conducted for a variety of clients. Yet they all point to essentially similar conclusions. Though none of these studies is perfect, their sheer consistency is impressive.

What picture do these studies, as a whole, reveal? They show that opposition to devolution in 1997 was substantial, while support was quite limited. (Hence the close referendum result.) Attitudes then changed quite quickly in the years to 2003: opposition to devolution fell substantially, while support for devolution – but not independence – grew. Since 2003, attitudes have moved more slowly, but further in a pro-devolution direction. These findings are robust across multiple different surveys, survey methods and survey questions.

Also striking is the lack of impact that differences in government in London have made to attitudes to devolution. The most rapid growth in support for devolution occurred during those years when there was a generally popular Labour government in London. Support has continued to grow, though more slowly, both when the Labour government in London was less popular and under the Conservative-led coalition that has fairly limited support in Wales. It would probably be unhealthy were Welsh devolution to be supported simply because of hostility to a particular government in London. But that is not the case. The findings of the Silk research, and much research before it, are not primarily about Welsh dislike of the Tories.

The main question used in the Silk research on broad attitudes to devolution is a multi-option one where respondents are asked to choose how Wales should be governed from several alternatives. The findings are:

Wales should become independent from the UK


The National Assembly for Wales should have more powers than it currently has


The powers it currently has are sufficient and it should remain as it is now


The National Assembly for Wales should have fewer powers than it currently has


The National Assembly for Wales should be abolished


Don’t know


Several answer options here are subtly different from those used in a similar question deployed in a number of surveys since 2009. While slightly annoying – it prevents direct comparisons with previous findings – this also re-assures us that previous results didn’t depend wholly on one particular wording. The results strongly reinforce previous research in showing that devolution is now the ‘settled will’ of the Welsh people. Support for independence remains limited, while support for the ‘No Devolution’ option, at 9%, is even lower than the c.15% found by most surveys in recent years. There is overwhelming (86%) public support for the National Assembly to enjoy at least its current level of powers, with the centre of gravity of public opinion being firmly located in the ‘more powers’ option.

Several other questions in the Silk survey also demonstrate broadly positive public attitudes to devolution. The Assembly and the Welsh Government are shown to be substantially more trusted than Westminster and the UK Government; they also score higher in terms of public satisfaction with their performance. And two-thirds of respondents endorse the idea that the Assembly has given Wales ‘a stronger voice in the United Kingdom’. Once again, these findings are very much in line with those of previous research.

The Silk research also probes knowledge and attitudes about the devolution of specific policy areas. Their questions about whether or not specific policy areas are devolved constitutes the most detailed such exploration of which I’m aware. It is re-assuring that in all but 2 of 12 areas asked about the majority chose the ‘correct’ answer when asked whether a policy was controlled by the Assembly or by Westminster. (Although it should perhaps be noted that, given that respondents were given only two options, the majority should get the right answer about half the time simply by random guessing).

With regard to whether policy areas should be devolved, the Silk research again uses somewhat different question wordings from previous research. But it again reinforces what previous studies, like the 2011 Welsh Referendum and Election Studies, have found. There is strong majority support for health and education to be devolved. (The interpretation placed on the findings by BBC’s Wales Today last night was bizarre, in implying a fairly even split of attitudes. Devolution of health is supported by well over a 2-1 majority; devolution of education is supported by an almost 4-1 majority). Even greater majorities support the continued devolution of powers over agriculture, roads, housing and tourism (in the case of tourism, devolution is supported by a more than 15-1 majority!). And clear majorities also support the devolution of several policy areas currently not devolved:

  • Policing (by 63% – 35%)
  • Renewable energy, including large wind farms (70% – 25%)
  • Broadcasting and media regulation (58% – 39%)

There is also a narrow majority in favour of the devolution of the ‘welfare and benefit system’ (51% – 46%). However, support for devolution is not across the board. The devolution of ‘Courts and the Criminal Justice system’ is opposed by a roughly two-to-one margin (63% – 35%); devolution of ‘defence and foreign affairs’ is also overwhelmingly rejected (by 82% – 15%).

Overall, the Silk Commission’s research reinforces what we have known for some time. People in Wales are not regularly taking to the streets en masse to demand a reserved powers model of devolution, or the devolution of policing and broadcasting regulation powers. But if political leaders decided to substantially strengthen Welsh devolution, they would have a broadly supportive public behind them.To borrow a term formerly used in studies of the EU, there is a ‘permissive consensus’ behind substantial Welsh devolution. For advocates of substantial Welsh governmental autonomy, the main problem is no longer one of convincing a reluctant public. The problems, and the people to be convinced, lie elsewhere.


  • J Jones

    Unlike you I was quite amazed at the level of ignorance about the present devolution settlement. I was also quite sceptical about the survey overall; for instance, how valid is it to ask someone how well they think the Assembly is dealing with a devolved issue such as the NHS when the respondent didn’t know that it was a devolved responsibility until the researcher told her so? I thought that there was a lot of confused thinking evidenced and evidence of a badly designed poll…..although you can see the report struggles to remedy the situation.

    The danger is that only the bare figures will be quoted.
    Perhaps you, RS, could bring some light to bear; my impression is that the percentage of people in the survey who claim to have definitely voted in the last GE (take away all those under 18 now and then) either indicates a particularly politically engaged sample or a lot of wishful thinking. Time and again polls actually find that people respond with what they perceive as the socially acceptable answer……ask people how much they give to charity on average in a year and few say nothing at all while many inflate the figure.
    Then there is the break down of the figures. Let’s say for instance that we work from the 24% who don’t want any further change in the devolution settlement. We can add to this the fairly strong contingent that wants the NHS returned to Westminster control, 27%,. We must also consider that some of the 20% who want Education returned to Westminster aren’t the same people who wanted the NHS taken away. We have a range between 27% and 47% who want a reduction of devolution in the most significant areas where Cardiff has control. A minimum of 51% of the survey want a significant reduction of powers or the status quo.
    I can do the same sums for some of the popular areas where people want control brought to Wales….major renewable projects for instance, recently on the news, many people incensed at the rape of the fair country…..a gut reaction that WE must have control.
    What Silk has done is to discover too much about soft Nationalism and party tribalism….and abject ignorance.

    There is quite clearly a confusion around the definition of “more devolution”. If a person would like to see control over policing moved to Wales, it isn’t at present, therefore that person wants “More” devolution. The same person might want Education taken away from Cardiff…..he wants “Less” devolution but he might be quite happy that the sky hasn’t fallen in over the last 10 years and is broadly happy about the status quo…..what he understands of it.

    The poll allows each respondent to be ignorant, enlightened, wanting more or less and contented……all simultaneously.

    How not to do a survey.

    • Roger Scully

      Dear J Jones,

      Thanks for the interest. However, I think I have to disagree with some of your points. I’ll address a few of them here:

      – On people’s level of knowledge about devolution: I think this study supports previous ones, which suggest that most people in Wales have a fair idea about the basics of devolution; they struggle more with the details, except where it concerns things that directly touch their own lives (e.g. bus passes, their kids’ education). Of course, we should remember that it is not only Welsh devolution about which public knowledge is substantially less than perfect. At the UK level, how many people know about (for instance) Statutory Instruments (which are hugely important)? And the EU – I used to study that, and public knowledge about even the basics there is appallingly low. A final point on public knowledge: to the extent that it is limited, to where or whom should the blame be directed? Some, I think, to much of the media which fails adequately to cover the realities of a devolved UK. Some, also, probably to the fact that the devolution ‘settlements’ we have had in Wales thus far have been chronically unclear, and difficult for people to engage with.

      – On the overall conduct of the recent Silk survey: I probably wouldn’t have done everything exactly as they did, but I think your dismissive verdict goes much too far. Moreover, whatever shortcomings this study may have had, the overall picture it paints is very similar to that from ALL major surveys that have asked questions on devolution over the last decade and more: multiple Welsh Election Studies, the 2011 Welsh Referendum Study, as well as individual surveys conducted for the BBC, ITV-Wales, the Electoral Commission, National Assembly Commission and the All-Wales Convention; surveys conducted by a variety of world-leading survey agencies. Some of these studies, including the Silk one, have also conducted focus group work that has supported the conclusions of the surveys. As I said in my original post, none of these studies are perfect, but when they all come to very similar conclusions it starts to become silly to dispute their broad conclusions.

      – On social desirability in survey responses: there is a voluminous academic literature on this. In general, we seem to find that this is most prominent where surveys are conducted face-to-face by an interviewer; a little less so when interviews are conducted by telephone; and even less so when they are conducted via the relative anonymity of the internet. Voter turnout is consistently over-reported in surveys, but this seems to be only partly about social desirability; it is also prompted by the fact that the sort of people who agree to do surveys are also the sort of people more likely to vote… I don’t think there is a magic way around this, beyond doing studies using different methods to try to ensure that you cover as many types of people as possible.

      – Finally, regarding public attitudes to devolution, there is some fairly obvious double (and maybe even treble) counting in your statement that “A minimum of 51% of the survey want a significant reduction of powers or the status quo.” The Silk study, I think, simply reinforces the substantial body of previous evidence in suggesting that the vast majority of people in Wales wish neither independence nor the abolition of devolution. Devolution is supported by a very clear and stable majority; and the bulk of the evidence suggests that the centre of gravity of public attitudes is now supportive of devolution probably going rather further than at present.

  • J. Jones

    We will have to agree to disagree about some things but I wasn’t disputing your overall assertion (or that of the survey) that the population is either content or wishing more specific devolution. What I was pointing out is that the survey certainly reveals that the respondents may want more in some areas and less in others, that is, each individual may want a different settlement. The survey is not internally consistent and would have benefitted from restricting itself to fewer, better targetted questions.

    On the matter of ignorance of political responsibility obviously the question is far harder to answer for people in Wales but there has to be a greater understanding before acting on the opinion of the populace.
    For instance…..devolution of responsibility for major renewable energy projects is popular. Devolution of responsibility for foreign affairs is not. The UK government signs agreements with the EU on targets for renewable energy generation and on greenhouse gas emissions and Wales is happy to see the Westminster government do this. To comply with UK responsibilities Westminster must site renewable energy projects in appropriate locations within the UK…..Wales is not happy for those sites to be within Wales.

    Quite minor seeming questions of responsibility keep coming back to how the UK operates. If we are happy to see re-distribution of wealth (since we want more from the Barnett formula I assume that we are) then can we keep chipping away at our responsibility to the wider UK?
    More and more I’m becoming disenchanted with non-stop devolution. The rhetoric of all political parties in Wales is now Nationalist……the way to the voters hearts is to emphasise committment to Wales in narrow terms. There is no voice that dare speak of responsibility to the UK.
    On the matter of the media I can only agree that we do not have a diverse enough media nor are the Welsh public particularly focussed on Welsh politics….amazing when you consider that the Assembly controls the most vital aspects of day to day life. Up to a year or so ago there was a vibrant and active “Blogosphere” in Wales with very active forums…….these have mostly been curtailed or shut down completely. Free comment is in short supply; information and opinion increasingly the preserve of an elite.

  • Jon Jones

    As I said in another place, the Beaufort Surveys are a bit dubious on the percentage of Welsh speakers and I maintain that Welsh speakers are substantially more likely to support further devolution and independence. You will remember that I said that I would ask the Silk Commission about the make up of their 2000 person survey and how many were Welsh speakers, Fluent Welsh speakers etc.
    I recall that you disagreed with me on the likelyhood of Welsh speakers disproportionately supporting devolution or independence…..any way, you are an academic and well able to test that theory if you can get the survey results from Beaufort…….and in the interests of truth you should do that. This is my reply from Silk:-

    “Dear Mr Jones,
    Thank you very much for your interest in the Commission’s work.
    The Commission is not actually subject to the Freedom of Information Act as we are not listed as a public authority under that Act. The Commission is however committed to being as transparent as they can in their work.
    I’ve checked with Beaufort Research, and they have told me that 3% of the respondents to our poll did so in Welsh, which I am told is typical of surveys of the Welsh general public. The demographic data that they collected at the end of the poll showed that 14% were fluent Welsh speakers, 26% said they were non-fluent Welsh speakers, and the remaining 60& did not speak Welsh.
    We did not commission Beaufort Research to analysis the information collected according to Welsh-speaking ability, and so do not have the answers to the other questions you have.
    I hope you find this information of interest, and thank you once again for your interest in the work of the Commission.

    So, by whatever means, Beaufort Research interviewed 800 Welsh speakers in a sample of 2000 people. The definitive estimate of Welsh speaking ability in over 16s is the 2011 census, which finds 16%. Face to face surveys usually find more people who will say that they can speak a few words of Welsh so that the National Survey of Wales 2012-2013 found 24%.

    Beaufort found 40% able to speak Welsh!!
    What is usually less open to variation is the percentage of the adult population who consider themselves to be fluent in Welsh….12% in 2004-2006 Welsh Language use survey, 11% National survey 2008 and Beaufort “Attitudes to Welsh” 2008 and in the 2012-2013 National survey of Wales 10% fluent in Welsh amongst the adult population.
    Beaufort had a sample where 14% were fluent in Welsh.

    My feeling is that this is a very distorted sample that should be corrected back to 16% Welsh speakers 10% Fluent. Only then can it give a true picture of attitudes to Devolution.

Leave a Reply

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