To be published as HC 1608-v ii

House of COMMONS



Scottish Affairs Committee

the Referendum on Separation for Scotland

Wednesday 18 APRIL 2012



Evidence heard in Public Questions 815 – 979



This is an uncorrected transcript of evidence taken in public and reported to the House. The transcript has been placed on the internet on the authority of the Committee, and copies have been made available by the Vote Office for the use of Members and others.


Any public use of, or reference to, the contents should make clear that neither witnesses nor Members have had the opportunity to correct the record. The transcript is not yet an approved formal record of these proceedings.


Members who receive this for the purpose of correcting questions addressed by them to witnesses are asked to send corrections to the Committee Assistant.


Prospective witnesses may receive this in preparation for any written or oral evidence they may in due course give to the Committee.

Oral Evidence

Taken before the Scottish Affairs Committee

on Wednesday 18 April 2012

Members present:

Mr Ian Davidson (Chair)

Fiona Bruce

Mike Freer

Jim McGovern

Iain McKenzie

David Mowat

Pamela Nash

Mr Alan Reid

Lindsay Roy


Examination of Witnesses

Witnesses: Martin Boon, Head of Social & Government Research, ICM Research, and Nigel Smith, Chair of the Cross-Party Campaign for a Yes Vote in the 1997 Scottish Devolution Referendum, gave evidence.

Q815 Chair: Gentlemen, thank you very much for coming along to the Scottish Affairs Committee to give us the benefit of your experience and knowledge. First, perhaps I could ask you to state, for the record, who you are and whether you are speaking as an individual or on behalf of an organisation.

Martin Boon: Good afternoon. I am Martin Boon, a director of the opinion research company ICM. Am I speaking on behalf of myself or the company? I guess it is a bit of both, but most of the views that I will be presenting are based on my own experience of conducting high-profile social and political research over almost 20 years or so.

Nigel Smith: I am Nigel Smith speaking as an individual, but my experience was chair of the Yes Yes campaign in the referendum in Scotland in 1997.

Q816 Chair: Perhaps I may start off by asking whether you agree that any question that begins "Do you agree" is a loaded one?

Martin Boon: No, I do not agree with that, as long as it is followed by "or disagree". That is a really pertinent point. I do not have many rules when it comes to writing questions for opinion research. Obviously, referendum questions are very different from opinion research, but some general thoughts should still apply. One of those thoughts that I try to stand by categorically is that any question asked by ICM that is intended for publication in the public domain should be fair and balanced. It is as simple as that-fair and balanced.

What "fair and balanced" means is a matter of conjecture. My subjective view on what is fair and balanced may be disagreed with by people who view the question and the poll results, but both I and my predecessor Nick Sparrow held the view that no question would ever go out of the door of ICM without it being fair and balanced at least in our own minds.

The question presented beginning "Do you agree" fundamentally cannot be balanced if it excludes the words "or disagree". That is my primary starting point and fundamental objection to it. That said, it takes only a very minor tweak by the addition of the words "or disagree" to bring it into an area with which I would be entirely comfortable.

Q817 Chair: To be clear, if you say "Do you agree or disagree", do you then have a yes and a no?

Martin Boon: Not at all. You would have answer codes that said "I agree that" or "I disagree that". That would be the typical way of doing it. Indeed, we have asked an independence question for Scotsman Publications going back 15 or 20 years. It has been pretty consistent over time. The exact wording has been, "Do you agree or disagree that Scotland should become an independent country?" The two answer options have been, "I agree that Scotland should become an independent country", and, "I do not agree that Scotland should become an independent country." In market research terms, that is a pretty obvious question, which has the benefit of being simple to understand and on which respondents have clearly given us an express view, because we rarely get many don’t knows answering that question. It just works. It is always best to put something like that to the general public. Make it simple, concise and unambiguous, and it should work.

Q818 Chair: To be absolutely clear, as this is obviously a matter of some importance, would you have a preamble and then two alternatives, and you would put a tick beside whichever of the two alternatives you favoured?

Martin Boon: That is a different question in many ways.

Q819 Chair: I was trying to be clear about what you are saying.

Martin Boon: "Do you agree or disagree that Scotland should become an independent country?" is fine; that works.

Q820 Chair: How do people score it?

Martin Boon: Then we would offer them the alternatives, "I agree Scotland should become an independent country", or, "I do not agree that Scotland should become an independent country."

Q821 Chair: That is the point I am making. You have the preamble "Do you agree or disagree", and then two alternative statements, one or other of which you tick.

Martin Boon: Typically, it is a telephone interview with us, but the principle is still the same.

Q822 Iain McKenzie: You say that you apply the "disagree" to it as well, so what should you put first? "Do you agree or disagree?", or, "Do you disagree or agree"? We are back to the same question. Does the first part lead in before the second part kicks in?

Martin Boon: Probably not. It would be taking a little bit of a step too far to interpret a question that simply begins "Do you agree or disagree" as leading simply because you put the word "agree" before "disagree", but others might disagree with me on that.

Q823 Iain McKenzie: Or agree.

Martin Boon: That is the thing about being somebody who works in opinion research. There are no right answers but there is a whole host of wrong ones. More often than not, you find yourself covered with bullet holes, being kind of "messenger shot". I entirely accept that these are legitimate questions, and often there are no perfect answers. All I can tell you is that historically what ICM has done has been to put a question which begins "Do you agree or disagree".

Q824 Chair: Nigel, do you agree or disagree?

Nigel Smith: I certainly agree with what has just been said. I would want "disagree" as well. I would go a bit further. I think that "independent country" on its own is very shy of information, but that can be tested. The thing to bear in mind here is that the campaigns go a long way to filling in these answers. You can start off with, "Do you want to vote for an apple or a pear?", and nobody knows what that means. After a campaign you have given some meaning to this, but you could still be left with a bias for an apple or a pear. If you can remove the bias, it is absolutely essential that you do. I am quite happy with "disagree" being included; that goes a very big step towards balancing up the thing. The rest could be done either with a preamble, which is allowed, or by extending the question, but the question is greatly remedied by "disagree".

Martin Boon: I would like to add one point, which may be slightly unorthodox but which we have found works over time-that is, not to have any preamble at all. In the context of a referendum I see no reason why you should not simply say, "Please tick one box." You provide two boxes and simply have, "I agree that Scotland should become an independent country", however you want to phrase "independent country", or, "I do not agree that Scotland should become an independent country." Therefore, you have no question at all; you just have two possible outcomes.

Q825 Chair: In those circumstances, that is one proposition, as it were, which is either negative or positive. What about the alternative of having, say, "I believe Scotland should become a separate country", as compared with, "I believe Scotland should remain part of the United Kingdom"? Both are positive statements; it is essentially the same grouping of arguments, but, rather than simply supporting a negative, you are supporting an alternative positive. Are there any choices or preferences there?

Martin Boon: I think it would be wrong of me to state preferences because that is inherently a political debate, and as an opinion researcher that is something I avoid engaging in totally. Forgive me for a cop-out. It is inevitably true that the way in which the question reflects those two statements will have a direct impact on the likely outcome of the referendum. Words like "separation from the UK", for example, are not so much provocative as strongly perceived emotional words that some people might relate to in a different way from the way they would relate to a question that comes entirely from the initial position you talked about. That needs to be very carefully handled, and I certainly would not have the answer as to what would be a better reflection of public opinion in Scotland or elsewhere.

Q826 Chair: I was not intending to ask your view about the wording of particular options. What I meant was that, rather than having, "Do you agree such and such?" or "Do you not agree such and such?", is it better to have two positive statements as alternatives to each other? Does that affect the result or the perception of the alternatives in any way?

Martin Boon: I would like to know exactly how that could work to make both statements positive.

Q827 Chair: The two positives could be, "I wish Scotland to become an independent country or a separate state", however you describe that, and the alternative is, "I wish Scotland to remain part of the United Kingdom", both of which can be seen as positive statements that are mutually incompatible.

Martin Boon: I think the danger you would run is in confusing a certain portion of the population. It may be self-evident to people around this table what you mean by presenting two options like that, which indeed are mutually exclusive, but any opinion researcher would tell you that you need to have complete clarity in your question. I do not believe that kind of question would give you the clarity that everybody would understand.

Q828 Chair: Nigel, do you have any comments on those exchanges?

Nigel Smith: You are getting very close to the two-question issue. Of course, having seen quite a bit of the evidence given to the Committee, there has been a lot of focus on how the question is put on the paper. But I worry as a campaigner about the debate. It is the referendum debate that makes the thing a referendum and not an opinion poll. If you have what you propose, which the Scottish Government are tempted towards, then you have a multi-option debate, whether you like it or not, for three or four months. For me, that is the big issue that the pollsters have not really brought up so far.

Q829 Chair: I thought that the alternatives I was suggesting, one against the other, did not involve a third option, so to speak, and were another way of presenting a single question with two outcomes.

Nigel Smith: Yes, but in order to answer that question you are venturing on to the ground of whether we want to stay inside the Union as part of the devo-max question. To me, it is making it rather more complicated than it need be by having it phrased in the way that you want it.

Martin Boon: You may have seen some public opinion research conducted and commissioned by Lord Ashcroft, which asked three questions. The first one, which we are currently discussing, is, "Do you agree that Scotland should become an independent country?", which was yes 41 and no 59. Then he added the word "disagree" to the statement, and it moved from 41/59 to 39/61. He then presented entirely that kind of option, which was, "Should Scotland become an independent country or should it remain part of the United Kingdom?", and the numbers moved to 33/67. In those three variations of the same thing you have a significant movement of seven or eight points. That is the tangible effect of what we are discussing. These words will move the Scottish public one way or the other depending on the choice that is made.

Q830 Lindsay Roy: All the questions I had have basically been covered. To clarify it, the Chairman gave you two options. What is more complicated with the second pattern in your view?

Martin Boon: It is not so much that it is complicated. All I am giving is an opinion from the perspective of market research. The guidelines, such as they exist, and the way in which pollsters operate simply give people a consistent set of answers, such as "I do believe" or "I do not believe". A question that comes at it from the perspective, "I want Scotland to be independent", or, "I want Scotland to remain part of the UK", polarises the debate in many ways. You are giving people two ways of coming to the same thing-it is true-but they are two different ways of coming at it. That, in comparison with a simple "I do" or "I do not", is in itself inherently more confusing, and that is the simple point that I am making.

Q831 Lindsay Roy: Is there not an inevitable polarisation about the whole debate anyway?

Martin Boon: You are probably right, but when it comes to referendum wording the general principle should be to make it as simple and concise as possible. My view is that if the question were phrased that way it would not be the most simple and concise way of doing it.

Q832 Lindsay Roy: Intelligibility is critical.

Martin Boon: Is it not in everything? You want people to be able to understand what you are asking them and give a well-formed view in return. That is a fundamental premise in everything I do when it comes to talking democratically to the public.

Q833 Lindsay Roy: Therefore, you feel that the first option the Chairman presented to you is the simpler and more concise one.

Martin Boon: I think the second option, "Do you agree or disagree that Scotland should become an independent country", is the best question wording I have seen to date, although I do not think it is perfect. There are easy ways to improve it further, but I would also say that the more information you give people, the greater the chance of loading or leading the question in a particular direction, and on that basis the shorter and simpler it should be.

Q834 Chair: I understand your point about there being a single issue, as it were, which is "agree or disagree". The alternative way of looking at this is, "Should Scotland remain part of the United Kingdom?"-"I agree" or "I disagree." If you pose that pair of questions, is it likely to make any difference as compared with posing it around the question of independence?

Martin Boon: It will make a difference, but we are entering into the realms of speculation. My view is probably no different from anybody else’s. I would guess that, if you asked the question that way, there would be a bigger majority in favour of Scotland remaining within the UK because it is a more comforting way of presenting the issue, but that is just my opinion.

Q835 Iain McKenzie: Should someone independent construct the question? If so, can you suggest anyone you think would be independent enough to pose the question?

Martin Boon: I probably could. We have the Electoral Commission in this country and it is an independent Executive body. I think it does some great work, but I would say that. For the purposes of clarity, it spends quite a lot of money with ICM, but it is an organisation that is there to administer elections. It has a huge amount of experience testing previous referendum questions. I do not see why on earth it should not be considered as a body that would be capable of objectively, intellectually and independently assessing question frameworks.

Q836 Iain McKenzie: Is it essential that the question be taken by, say, the Electoral Commission and tested?

Martin Boon: I would absolutely concur that it is essential. On a separate but, I suppose, related point, the Electoral Commission warned that the ballot papers used in the election in Scotland four or five years ago, which had difficulties, were not completely understood by people. I think that is as good an example as I can give about the value of looking at these things very closely.

Nigel Smith: On the actual writing of the question, I was not quite clear whether you were suggesting the Electoral Commission formulate it as opposed to commenting on it. I have been very critical of the SNP for departing from the generic rules in PPERA, as it is known, even though these rules are not complete, and if they were extended they would still allow the authority calling the referendum to choose the date and propose the question. As I want the PPERA rules to be used throughout and am critical of the SNP, I personally would not support a change in what in the old trade union days we called custom and practice; in other words, in the UK’s referendums the questions have always come from the Government. All I would ask here is that the Electoral Commission comment on it and that the Government take note of them.

Q837 Iain McKenzie: When you say "the Government", do you mean that the Westminster Government constructs the question and the Electoral Commission takes it out for a test?

Nigel Smith: No. I mean that the Scottish Government, the authority calling the referendum, constructs it. You arrive at a legal problem here because of the way the whole debate between the two Governments is going. The idea is that the generic rules would leave only the date and the question in the hands of the Government calling the referendum. In this case I would say that is the Scottish Government, but I recognise there is a nuance to that debate. I would not want to change that, for the same reason I object to the SNP changing some of the other rules under PPERA, which I think are quite significant changes.

Q838 Lindsay Roy: Should the ballot paper be logo-free?

Nigel Smith: I would have the ballot paper logo-free for the simple reason that a referendum is directed towards an issue. Parties may be split on an issue and usually are in a referendum. For example, we would expect a significant minority of SNP voters to be voting no to independence. From my point of view, anything that de-party-politicises the ballot paper and referendum is moving in a direction of what a referendum is about.

Q839 Lindsay Roy: You would agree with Ruth Stevenson that the Saltire should not be on the ballot paper.

Martin Boon: Yes.

Q840 Chair: Nigel, to come back to the point about who should determine the question and the issue of legality, the evidence we have shows pretty clearly that the Scottish Government do not have the legal power to call a referendum. If they tried to do so it would end up in court, and the appropriate avenue is a section 30 order from Westminster. In those circumstances, that would probably mean, would it not, that the Westminster Government propose a question that the Electoral Commission reviews and is then passed to the Scottish Government for their observations?

Nigel Smith: Yes; I accept that. I am just making the point that all of the nine major referendums we have had in this country, as it so happens, have come from the British Government and, therefore, it has been cut and dried, but I entirely accept what you say. If the political interpretation is that this is a referendum called by the generosity and legal position of the British Government, it would fall to them to propose the question; if not, or the Government take another view, clearly it is for the Scottish Government.

Q841 Chair: I do not know whether you have had the opportunity to see all the evidence we have had and the various seminars up and down the land. It seems to be overwhelmingly clear that the Scottish Government do not unequivocally have the power to hold a referendum. The best the protagonists argue is that, yes, it would end up in court, which is undesirable. There is no doubt that a section 30 order could manage it. That would mean the UK Government putting forward the question to be tested. Presumably, there is no reason why they should not put forward a proposal to be tested and also ask the Electoral Commission to test the Scottish Government’s proposal as well.

Nigel Smith: If it is settled that they are the authority, I have absolutely no problem with it. Somebody has to propose the question. Up to now it has been the authority calling the referendum, and, as we know, that is the issue right now. When you have decided that, I will fall one way or the other.

Q842 Chair: We are pretty satisfied on the basis of the overwhelming majority of the evidence we have heard that the responsibility lies with the UK Government. Obviously, there are some who are not happy about that.

Nigel Smith: However, one caution would be that this referendum is creating new precedents for British referendum democracy, which is in its infancy anyway but not completely; it is not a blank sheet. Therefore, whatever we do here, it needs to be realised that it will be written into the record in some way.

Chair: Unless there are any other points on the issue of wording, perhaps we may now move on to multi-option issues.

Q843 Pamela Nash: Do you think it is possible to have a fair referendum that has more than two possible options on the ballot paper?

Martin Boon: It is probably possible. I would worry about it somewhat. You would need clarity, of course, on both questions. I guess that the extent to which you can judge what the settled view of the Scottish public would be in terms of their final constitutional preference would be open to question depending on the way in which the votes were calculated and the system used. You are doubling up the problems. We talked about this as an opinion poll. You would ask one question after another and you might have 20 political questions on a poll and not think twice about it. In this case, you could worry considerably about the "questionnaire" effect of the first question followed by what impact that has on the second question. It is probably doable. You would need to include minds immeasurably cleverer than mine to ensure the questions are fair and balanced, but you would struggle.

Q844 Pamela Nash: Can you expand on the "questionnaire" effect in a referendum?

Martin Boon: When you write a questionnaire, the order in which you structure the questions can have an important influence on the cognitive path that people go down when they go through an interview. You might ask a question at the beginning and the same question at the end and compare the results to see what the effect has been on the other questions of that critical first and last one. You see if the numbers have moved-among the same people, remember. If the numbers move from first to last question as a result of the others they have been asked, then you know you have some form of "questionnaire" effect; you have led people to a particular view that may be different from the view they had at the commencement of the interview. That is a "questionnaire" effect.

Nigel Smith: I think it is possible to have a fair multi-option referendum, but you have to ask why, when so many political questions have more than one answer, there are not more multi-option questions. It is very difficult and it is more luck than good design that leads to an outcome that is fair. For example, I do not think anybody would say the multi-option referendum on nuclear power stations in Sweden in 1980 was unfair, but it was inconclusive. That is the pattern. If you look perhaps at 10 major multi-option referendums since world war two, you have got that sort of problem. Has it settled the issue and is it considered fair? If you take the first New Zealand referendum on PR, that was greatly helped by the fact that there was already something of a consensus around MMP. The debate was held only among those who had selected to be part of the debate; in other words, it was a stand-alone referendum with a 55% turnout. Because one thing already had a kind of established position it got away in the first referendum, but everybody knew that it was to be followed by another referendum.

Eighteen months later there was another referendum, which this time was a binary one. The binary referendum was between the winner in the first one, in this case MMP, and the status quo, and was held at the same time as a general election, which completely swamped the referendum. I think there were 420 new no votes for every new yes vote. What happened was that the general election brought in a whole load of voters who did not have a view about this and said no, which is a very common pattern in referendums. That is a rather complicating example. You asked me about "fair". I think it could be fair but still be inconclusive, and that is why there are so few of them for major issues.

Q845 Lindsay Roy: I understand that, of the 200-odd constitutional referendums that have taken place, 98% have been binary. Is that probably the reason?

Nigel Smith: Yes. It is actually far more than that. I do not think I could find 200 multi-option referendums of any kind, let alone constitutional ones.

Q846 Lindsay Roy: But in terms of the constitutional referendums 98% have been binary.

Nigel Smith: Off the top of my head, I could quote maybe eight or nine constitutional referendums that have not been binary out of hundreds.

Q847 Lindsay Roy: What do you think is the reason for that?

Nigel Smith: The one I quote in my paper is Puerto Rico, which has been wrestling with whether to become a state of the union for 50 years. It has always gone down the multi-option route. Finally, a presidential commission recommended that it try a binary this time round. I do not know whether that will produce an answer, because they are now talking about having two binaries on the same day, which is another aspect of the problem.

Q848 Jim McGovern: I think it was in 2007 when the Scottish Parliament elections and local government elections took place on the same day, and there were multiple options on both ballot papers. I think it is fair to say that there have never been so many spoiled papers in an election in history. Do you agree that the best possible option in a referendum is just one question with yes or no?

Nigel Smith: I have absolutely no doubt, even though I am a devo-maxer, that I want a binary referendum. I do not care whether it is about devo-max or independence, but I want a binary referendum. I accept that the paper could be designed. You have heard quite a bit of evidence along these lines. It is more difficult to design a multi-option or double referendum ballot paper, but it can be done. My argument is a democratic one about the referendum debate, which becomes enormously complicated. In Britain, along with a very few countries in the world, broadcasters are required to be balanced 50/50 in a referendum, and that gives a fair chance that the voter can hear both sides of an argument. If you make it multi-option, the broadcasters immediately have to put three politicians into the studio. It becomes 33/33/33, and it is a nightmare. In a referendum 60% of the population know how they will vote. They could vote tomorrow probably the same way, but the rest will decide the matter. We are likely to have a big turnout in this referendum. It is hard to believe but a lot of people come to a referendum in the last eight or 10 weeks-as late as that-so all these years of argument we have been having are stirring only the top of the pot. The debate is incredibly important. If you are a democrat, you want that debate to be as straightforward as it can possibly be. Therefore, as a devo-maxer I would want two referendums separated in time, and I do not care which way round you put them. There would not be a second one if they win independence. Anyway, that is the way I would do it.

Q849 Jim McGovern: I certainly don’t want to sound patronising, but in canvassing recently for the local government elections when I said it is a multi-option ballot paper and you put one, two, three and four, a lot of people said to me, "Why don’t I just put a cross in the box for the person I want?" The point I am making about a referendum is: should we not try to make it as simple as possible so that people understand exactly the question they are being asked to answer?

Nigel Smith: I agree with you. You have heard evidence that says you can design the ballot paper to do the multi-options. We know that in New Zealand, for example, they did it. We have one or two examples, but my criticism of this line of argument is that it focuses the argument only on the ballot paper, whereas as a democrat my argument is about the referendum. That is what makes a referendum. That is why you get bigger movements in public opinion during a referendum than in electoral campaigns-at least very often.

Q850 Chair: Surely, there is a difficulty about obtaining clarity. This applies to both of you. Since presumably we will not have all the issues resolved by the time of a ballot, we will have negotiating positions outlined by those who are in favour of separation with an assumption that the UK Government will simply roll over and have its tummy tickled, but I suspect that is not likely to be the case. Until we have these negotiations we are not clear what the eventual turnout will be. Therefore, people are being asked in these circumstances to vote for a pig in a poke. Absolute clarity on the outcome of a decision to go for separation is impossible without it being a vote post negotiations.

Nigel Smith: I could not agree more. This is shit or bust for both sides. This may be unparliamentary language, but I personally think it is the wrong way to use a referendum. If you are trying to develop a broad principle, you would not be doing it this way. In Northern Ireland we had 18 months of multi-party talks in which the issues that mattered to people were being fed out to the public, and so a debate went on for those 18 months. Then people were asked to vote on the final deal. It was not a pig in a poke, but we have a situation where quite clearly there will be an opening White Paper. I would be very surprised if the Scottish Government did not put their best foot forward in the White Paper. There may be a counter-balancing White Paper from the British Government, but it is only the opening of negotiations and you are asking the public to vote on that. This is second best. Like you, I have to deal with where we are now.

Q851 Chair: So what is the best option?

Nigel Smith: The best option would have been for the SNP to have had a win in an enabling referendum, like de Klerk did in South Africa. He went to the electorate and asked, "Do you give me the power to negotiate with the British Government? I will come back to you with whatever deal I have got, and we will put it to a ratification vote." That, of course, raises the question that, if he wins the first and loses the second, you have to deal with that.

The other situation is that the British Government could say, "All right; we’ll negotiate with you on the condition that the referendum is changed from being pre to post negotiation-a single referendum." But that is not likely to happen either. I am quite clear that what we are doing is not the best way of tackling the issue.

Q852 David Mowat: Why should not the first referendum that we are going to have in two years be an enabling referendum in the way you describe, which would effectively be a mandate to negotiate, and, after that negotiation was done, there would have to be effectively a ratification? Maybe that is the path we are on.

Nigel Smith: Yes. The interesting thing I posit to people is what happens if there is a close result in the first referendum. You are likely to have a situation where the bolder spirits say, "Let’s vote SNP and see what comes out of the tree when we shake it." Let’s say the SNP won 52/48 or 54/48. Then, during negotiations, quite a considerable amount of detail is revealed by the negotiations, perhaps things that both Governments had not fully clocked. This comes out into the public sphere and public opinion begins to say, "Hold on a minute. I didn’t really get that." You then find there is a political demand for a second referendum, which is completely unplanned but to which politicians then have to respond.

Q853 David Mowat: Would not the ratification referendum be that anyway, so that if stuff did come out of that process you have that backstop?

Nigel Smith: Yes, but that is if you have planned it. What I am saying is, suppose you just do it with one referendum and you have a narrow win. Public opinion swings during the course of, say, 12 or 14 months of negotiation, and people say, "I don’t know; I’m not so sure about this." The British Government suddenly say, "Well, we respond to public opinion. Let’s have a ratification referendum."

Q854 David Mowat: But the issues that require detailed negotiation are so immense in this case that it could well sway public opinion; in fact it would be natural that it would.

Nigel Smith: Yes.

Q855 David Mowat: That implies that the process has to be like you have described. You called the first referendum "enabling". It could be a mandate to negotiate or something like that.

Nigel Smith: Yes.

Q856 David Mowat: Therefore, there has to be a further part of the process, because the issues are so big.

Nigel Smith: Yes. Of course, what worries the British Government is that, as with trade unions, as we all know as experienced people, people will say, "I’ll vote for it the first time because I can always withdraw my vote."

Q857 David Mowat: But there is no harm in that because what they are saying is, "I’m voting to see what deal the Scottish Government can get out of the British Government, and then we’ll look at it again after that negotiation."

Nigel Smith: Yes.

Q858 Chair: A double system makes it more likely that you will get a positive result in the first ballot, if you know there is a second ballot.

Nigel Smith: Yes.

Q859 Chair: Given that there is a general election due in 2015, if there is a referendum in 2013 or 2014, to some extent people who are unhappy about, say, a yes vote as the terms emerge will effectively have the opportunity to cast their vote either for a party that seeks a mandate for another referendum or one that favours maintaining the Union at that subsequent general election, so in effect there will be another referendum. Just as the SNP see every election as a referendum, similarly those who are in favour of the status quo, devolution, or whatever the alternative is, will have the opportunity to express their view.

Nigel Smith: Of course that is what I do not want. I am not now talking about this particular issue, but in terms of referendums in general we are stumbling towards some kind of use of referendums in the UK. Mixing them up with general elections is not the way to go about it, but that is precisely what would happen in the situation you describe. I think we would have been far better anticipating this and designing a procedure that tried to cope with this, but we have not.

Q860 Mr Reid: Is it possible to have a fair and decisive referendum with three options?

Nigel Smith: You could have a fair referendum, but I do not personally think it could be called decisive if people feel they have not heard all the arguments, or the arguments are mixed up in their mind and they have not been able to discriminate between them because they overlap. It is a loose use of the word "fair", but I regard the debate as very important. I do not underrate it. It is crude, rough and imperfect, but it adds to or detracts from the public’s knowledge for one side or the other; it does matter.

Martin Boon: I would suggest that you could design a fair referendum but you would not get a decisive outcome. In any question with three options, unless you have a specific set of pleasant circumstances, you will fail to get a majority position. In a referendum that, for argument’s sake, gets a turnout of 50% with a winning option of 40% of those who voted, it implies that only 20% of the population voted for the preferred option, and you are entering into a world stuffed full of cans of worms.

Q861 Mr Reid: Take the three options of status quo, devo-max and independence. How would you organise the ballot paper? Would you have three separate boxes, or some other more complicated option?

Martin Boon: You would certainly have to have three specific mutually exclusive answers, which of course they must be, but, to reiterate what I said earlier, they should attempt to be as consistent as possible in the way the three options are presented to people. In my world probably many would make the argument that people would adopt the middle position as it is the most uncontroversial, if you like, so those people who would support devo-max probably might like the idea of that, and opponents of it would struggle to ensure that their preferred outcome won, but that is a political argument that is outside my remit.

Q862 Mr Reid: What is your view on the Scottish Government’s proposal to have two separate questions-one in favour of independence and one in favour of devo-max-on the same ballot paper?

Martin Boon: It would not be my preferred option. I would be one of those who view a two-option referendum as the most simple, concise and understandable. Coming from the perspective of the general public, I would always argue that their ability completely to get the options that are presented to them is fundamentally important when it comes to having outcomes that are generally accepted.

Q863 David Mowat: To follow up your observations on devo-max, it strikes me a little that separation or non-separation are two discrete out-turns and at least you know where you are. With one you have got no Scottish MPs here; with the other you have 57, whereas devo-max is a continuum; it means everything to everybody. Is it possible to use the words "a decisive out-turn" in the context of a vote for devo-max, because it could be so wide?

Martin Boon: If the people proposing devo-max are not capable of defining exactly what it is, why should anybody expect the public to understand what it is about?

Q864 David Mowat: But we have just said that we are not even going to be able to define very easily what independence is without a great deal of negotiation on things like the national debt and all that goes with that, but at least it is a fairly discrete outcome and you know the direction you are trying to achieve. Devo-max is everything else; it is everything in between. It seems to me to be intellectually wrong to have it on the same ballot paper; it is a little like apples and pears without getting that discrete thing sorted out first.

Martin Boon: I think you have a case, although I suggest it could be managed in some way. I would reiterate that the two-option question is a preferred thing for me in terms of it being a simple and accepted referendum question. I honestly think that if it went down the three-option route, enormous difficulties would be encountered and you would need to do it again.

Q865 David Mowat: It seems to me there is a clear mandate for this referendum; there is no ambiguity about that, because that was the Scottish election result last time. But is there a clear mandate for a devo-max part of this referendum, which seems to have come in as a likely compromise answer just because that is what people often tend to do when faced with tricky questions? Do you believe there is a mandate for it?

Martin Boon: I have seen some opinion polls that suggest people like the devo-max option, but that does not equate to a mandate in any sense.

Q866 David Mowat: Quite. I could do an opinion poll about whether we would like a bit more power in Cheshire. We really would, but that is not a mandate.

Martin Boon: No. You are probably asking the wrong person.

David Mowat: I am asking Nigel. I am a devo-maxer.

Chair: Hear, hear.

Nigel Smith: I do not think so. There are several reasons. First, I share your view that the two issues are different in nature, but, more importantly, there is no definition of devo-max. In order to get a definition, there will have to be the consent in some way of the people of Cheshire, Cornwall, Wales and the rest of Britain in defining what devo-max might be, and we are nowhere near that.

Q867 David Mowat: I agree completely with that, but doesn’t that mean that devo-max, which may be the right thing to do for Scotland-that could be fine-is not the kind of issue that could be sorted out in a referendum just in Scotland?

Nigel Smith: Absolutely.

Q868 David Mowat: That is why you have Parliament to sort out things like that.

Nigel Smith: Somebody has to work it out, and that is why I would have a referendum on the issue. For example, the British Government could say, "Two or three years down the line we will have a referendum on devo-max, but prior to that we will have a convention process that sorts this out with the rest of Britain", but that is something completely separate.

Q869 Chair: But by that logic we ought to have had a referendum on the Scotland Bill. Surely, there are mechanisms other than a referendum to have devolution evolve; it is an iterative process.

Nigel Smith: Yes.

Q870 Chair: Therefore, not to use your delicate phrase, it is not an either or nothing option; you either have it or do not have it. You can have a degree of Fabian gradualism.

Nigel Smith: I would make the point that the country spent 80 years discussing symmetrical devolution and failed to get an answer. In 1997 Scotland got asymmetrical devolution and the rest of Britain said, "Okay, we’ll go along with it." But Scotland cannot expect to get devo-max out of Britain without some kind of political process with the rest of the UK.

Q871 Chair: Absolutely. That need not be a referendum. That is almost in a sense an iterative dialogue, is it not, where discussion takes place; maybe something is conceded either way about the practicalities; and it is not an all-or-nothing proposal in those circumstances?

Nigel Smith: No. I tend to attach referendums to a lot of political decisions that are constitutional. In this case, because it has become bound up with the debate, we might find it is politically easier to say, "Look, vote down independence and we will have a referendum on a British negotiated thing at some future time." I agree with you.

Q872 Chair: That is the Douglas-Home danger, is it not?

Nigel Smith: That is why I suggest a referendum. It was not just Douglas-Home who did it; John Major did it in the general election of 1992. You will remember Taking Stock. It is not enough for the British Government to say, "We will look at some way of improving devolution."

Q873 Chair: But Douglas-Home and the others were Tories and, therefore, bad people and not to be trusted. Presumably, it is possible to have an agreement, as we seem to have evolving, that other forms of devolution will be worked out, and therefore progress will continue. As David said, separation or non-separation is a clear and decisive result, even though there is a lot of muddle about it, and these other processes can take place in another forum.

Nigel Smith: The only problem I was raising was with reference to your Douglas-Home argument. Because the ploy, which happens a lot in referendums, that "Something better will come along in a minute so don’t vote for this one" has happened twice in Scotland, the Conservatives or the British Government-the coalition-might find themselves having to say "and we’ll give you a referendum on this" to solidify the promise. That is all right.

Q874 Pamela Nash: Nigel, to take that argument a little further, earlier you said that if we had further devolution to Scotland it would have to be negotiated with other parts of the United Kingdom. If that had to be decided by a later referendum and not through Parliament, would you envisage that referendum would be UK-wide and not just in Scotland if it needed the permission of other areas of the UK?

Nigel Smith: I would not rule that out, and I do not think it is impossible to win it either if it has been properly negotiated. As the Chairman says, if this has come via a unanimous view in Parliament, you could equally say, "Well, all we’re looking for is an affirmative vote in Scotland to accept a deal."

Q875 Pamela Nash: Why would we need that affirmative vote in Scotland?

Nigel Smith: Because you have got yourself into a position of promising it. I am only saying you are promising it because of the history of failed promises in Scotland before this-the Alec Douglas-Home argument-and, therefore, one might say, "Vote this down. We’ll give you a referendum after a convention on this", but it does not need to be that way. I just think it sounds better to the public in trying to sell the proposition. Otherwise, what do you say when Alec Douglas-Home did that? This is the third time it has been suggested. I have to convince you, so I would say, "Well, there will be a convention and a referendum."

Q876 David Mowat: I suppose the only point I would make, just to think through where the Chairman was going, is that a discrete decision is very well suited to a referendum; it is a good use of a referendum. With a yes or no we know where we are, whereas a continuum decision like devo-max is particularly unsuited to a referendum given the complexity, nuances, shades, the fact that negotiation is multiple-layered and all the rest of it. Therefore, I just wonder whether we have allowed something that had a mandate-i.e. this discrete decision for the referendum-to be somehow almost adulterated by the new concept of devo-max. I do not resist it; I just think it is the wrong mechanism to sort it out.

Nigel Smith: I think it is very important that you try to settle the Scottish issue, and one way of doing that would be a referendum.

Q877 David Mowat: Because devo-max is a journey, by its nature it will never be settled; it will always move, will it not?

Nigel Smith: I am not sure I agree with that.

Q878 Jim McGovern: Nigel, I think you have made it clear that you are in favour of devo-max, but probably everyone in the room wonders what devo-max means. Are you in a position to define what your opinion of devo-max is? What does it mean?

Chair: In seven words or less.

Nigel Smith: I cannot answer that question in any helpful way, because it just means more devolution to Scotland.

Q879 Jim McGovern: I think they are with us now.

Nigel Smith: Central to it would be Scotland becoming responsible for raising the bulk of the money that it contracts to dish out on our welfare-all our public spending in Scotland-which I think was a hole in the devolution of 1997. I would like to see that corrected in some way. That would have to be a key element of it.

Q880 Jim McGovern: But Scotland voted yes in favour of tax-raising powers, did it not?

Nigel Smith: It was 3p in the pound.

Q881 Jim McGovern: It was a yes, yes vote, was it not?

Nigel Smith: Yes.

Q882 Chair: It has never been used.

Nigel Smith: It has never been used.

Q883 Jim McGovern: Correct, but it is there.

Nigel Smith: It is there, but it is not a serious power. All it would allow, if the Scottish Government came asking for money, is that London would say, "Sorry. Use your power first, sonny boy." It is a little discipline.

Q884 Jim McGovern: How will devo-max differ from that?

Nigel Smith: Devo-max would be all about taxation and the grip of the Treasury. Macro-economic policy would be in London and the details of tax raising would be in Edinburgh, but the point of your asking this question is that it is not defined. I cannot tell you what it is; I would just say I am a devo-maxer.

Q885 Jim McGovern: You are in favour of it.

Nigel Smith: As a general direction. It is not defined, and here we are; it is standing up there with independence. It is not true-

Q886 Chair: We have just about covered all the points. That has been immensely helpful. Are there any answers that you have prepared for questions we have not asked? Is there anything that you feel we should get from you in addition?

Martin Boon: I have not prepared any answers, but I would make one point that I think the public would like me to make. My job is about being a mouthpiece for their views in many ways. The simple point is: give the public something that is simple, clear and they can understand and report back to you with an answer. We have talked at length about the complexities of the situation. Complexity is not good in a referendum. The public will not get it if it is complex. I understand there are constitutional settlements and democracy at stake in all of this, but a referendum that fails to appreciate that the public may struggle to understand what is part and parcel of the question they are being asked to answer will not do justice to this matter.

Q887 Chair: Can I respond to that? Quite often we have been told that anybody who suggests anything like that-that complexity cannot be grasped by the public-is told either that we are talking Scotland down or that we are anti-Scottish in some way; Scottish voters in total are so intelligent that they can understand anything at all, and this is just an excuse not to give them the sort of options they want. How would you respond to that?

Martin Boon: I never pre-empt what the great British public say or do, and, bless ’em, my job is dependent upon them, but they do not walk around thinking about pertinent political issues, irrespective of how important they are. God knows this is one that is right up there. They walk around thinking about how they pay their mortgage; whether they get treated well if they go to hospital; and whether their kids are getting a decent education. It is only when a campaign that is seriously important, and well managed by both sides taking part in it, conveys their views that they form an impression. It is all right to say that this is an entirely patronising position for somebody to take and the Scottish public are above all that and are able to answer, but it is true that on any political question that we ask there will be a section who are not interested; they will not care; they will form an opinion based on the flimsiest of understanding, and that is what we all have to deal with, whether it is about Scottish independence or anything else.

Nigel Smith: I wanted to raise spending levels with you. I referred to the generic rules of a referendum. The SNP are tailoring the rules, as you will know. That is their word, not mine. One thing they have done is to withdraw the public grant to the two sides. It is important you understand how that public grant came about in the first place. In 1997, as the chair of the yes side, we did not have enough money to send a leaflet on our case to every household in Scotland. I went to Donald Dewar, the Secretary of State for Scotland, and asked him three times that we use public money in some way to send a leaflet to the public so that at least they had some basic information about what the Scottish Parliament would be about.

Eventually, that was done. There was an eight-page leaflet; it was a bit anodyne because there was no mention of the no case. There were supporting videos and so on. This exercise cost in current money about £1 million. Alex Salmond’s proposals to repeat that are based on that, but when Lord Neill heard about this he rejected it as unfair. Instead, he had the grants. He said, no, Governments should not be doing this; each side should be given a grant to make its views known. So we got the grant because of the way we behaved in 1997, but this has been withdrawn. We are back to 1997, except that this time it is the other side who gets to spend £1 million of public money. To be quite honest, I think we should not be doing this; we should be sticking to the rules in PPERA, which have been set up by Parliament and stand for all referendums in the UK.

Q888 Chair: Presumably, the logic of that position is that that proposal would be part of the section 30 order, and, if Westminster gave permission to the Scottish Government to have a referendum, it would determine not only the question but also things like spending limits and making sure that it was going to be equal and balanced.

Nigel Smith: Yes.

Q889 Chair: And all these individual decisions that the Scottish Government are taking to try to skew the result would be undone, as it were, to make sure that it was conducted as fairly as possible, presumably with the Electoral Commission identifying where attempts were being made to rig the result.

Nigel Smith: No. This is what the SNP have proposed to do in their tailoring. Their argument is that they will take PPERA and tailor it, and this is one of the tailorings. My argument would be that we should stick to PPERA completely with no tailoring.

Q890 Chair: What they describe as tailoring, normal people would describe as rigging.

Nigel Smith: Those are your words, not mine.

Chair: You have used some fairly colourful language earlier on, so you are hardly in a position to object now.

Nigel Smith: Please excuse me.

Chair: Thank you very much for coming along. This has been an immensely helpful session to us.

Examination of Witnesses

Witnesses: Andrew Hawkins, Chairman, ComRes, and Mark Diffley, Research Director Ipsos Mori Scotland, gave evidence.

Chair: Gentlemen, I apologise for being slightly late in starting. Our previous session ran slightly longer than we expected. We understand that a vote is anticipated at 3.49, so we will be obliged to make a start and then adjourn. I thought that was better than simply waiting until the end of the vote. I am not sure how many votes there are. We do not know whether or not there is one vote or more, but either way we shall be back before seven o’clock so we can carry on then. Let me start by asking you to introduce yourselves. Which organisations are you from, and what is your area of expertise?

Andrew Hawkins: I am Andrew Hawkins, chairman of ComRes. I have worked in polling for probably the last 20 years. I have done a lot of work with political parties, Governments, companies and so on, including some polling on the issue of Scottish independence.

Mark Diffley: I am Mark Diffley. I work for Ipsos Mori. I am a director in our Edinburgh office, so I am in the thick of it all. We do quarterly polls, which include attitudes to independence as well as a range of other related issues.

Q891 Chair: Do you agree that any question that begins "Do you agree" is a loaded question and therefore the question that has been put forward by the Scottish Government for the referendum is a biased one?

Andrew Hawkins: In the way it is construed, I do agree it is a biased one because it does not present any of the alternatives. It is notable that in other referendums, including, for example, the AV referendum, the alternative was put even though in market research terms it was not a fantastic question. One can say, as we did with the AV referendum, "At present this is the case. Should the alternative vote be used instead?" It is not what I would have chosen, but it is a lot better than a straightforward "Do you agree".

Mark Diffley: Certainly I agree that it is not a question we would use in a poll. However, as you have heard earlier, it is important to recognise the difference between a poll and a referendum. In a poll we simply would not use it because it would be leading and, as Andrew said, it would not present both sides of an argument. We would not use it; it is as simple as that. The 1997 parliamentary referendum that established the Scottish Parliament just got round the issue by having two statements: "I agree that there should be a Parliament"; and, "I do not agree that there should be a Parliament." That strikes me as a reasonable way of approaching this.

Q892 Chair: Andrew, you were here during the previous session; I do not think Mark was here.

Mark Diffley: I was towards the end.

Q893 Chair: We sought to clarify with the two earlier witnesses what they thought was an appropriate arrangement, and you touched on it with the question "Do you agree" and "Do you not agree". The alternative that we were exploring with them was whether or not you had one negative and one positive statement; instead, you had two positive statements that are mutually incompatible, and therefore that gives the same sort of choice but expressed in a positive way. From your perspective, is there any great advantage to having one structure rather than the other?

Andrew Hawkins: In a sense, it is almost coming at it from the wrong angle, which is to say, "What does it look like?", instead of, "What are the criteria we should use to judge it?" Those criteria have been well rehearsed. I do not have a problem with either approach. I think the public are sensible enough. In the wake of the carpet bombing that will take place before a referendum in terms of media coverage, debate, discussion and all the rest of it, they will get it either way. Perhaps I may by analogy talk about another example, which I did not hear earlier this afternoon but which has a couple of lessons for us. That was the example of the two referendums in Quebec between 1980 and 1995. I will not read out either question, but the first one was 106 words long. The second question was 42 words long. The Scottish independence question as currently composed has the great advantage of brevity in being 10 words long. Turnout in the first referendum in Quebec was in the 80s-80%; it was 95% for the second referendum. If the Québécois can answer sensibly a question that is 106 words long, I do not see a problem in asking a question that is a bit more complicated or has a bit more of a preamble in the Scottish referendum.

Q894 Chair: But, surely, there was considerable argument in Quebec that a lot of people had voted not understanding the question, and therefore it was misleading, which is why it was shorter.

Andrew Hawkins: I take the point. I am not advocating 106 words; I quite agree that is ridiculously long, but it goes back to the point about the criteria used. The Electoral Commission’s own criteria talk about the question being accessible to all voters, being focused and factual and having a lack of bias, which ensures voters do not consider one response more fairly than another. The current Scottish referendum question falls down on that third one. It is accessible and focused, but it has to be all three.

Q895 Chair: It has brevity, clarity and bias, which is what we believe the Scottish Government’s question is. I want to press you on the point about whether or not both of you believe that it is better to have a "Do you agree/do you disagree" alternative or to have two statements that effectively are the same thing. Is there any issue of bias in either of those? It could be argued that to say, "I do not agree" is simply being negative, whereas having two positive statements allows both sides to put forward a positive alternative.

Mark Diffley: In reality, a positive alternative will be put forward by both sides anyway. By the time it comes to the referendum question itself I am pretty relaxed about either of the alternatives you have put forward in the way that Andrew described earlier.

Andrew Hawkins: The point I made at the start about the referendum questions is that in a sense they put both sides in a way that the Scottish referendum question currently does not. They state the status quo, which in Scottish terms might be, "Would you support the continuation of the Union as it is?", but then go on to present the alternative as well. As long as it does present both sides, I do not think the format matters too much. If we are fretful about the order in which the alternatives appear, there are ways round that; they are not pretty, but it would be possible, for example, to randomise the alternatives and then distribute them. I readily accept that you may-

Chair: I had not thought of that.

Andrew Hawkins: But there might be other ways round that. It might be possible even to have a modest bit of colour coding on the things so that when it comes to the count they can be separated and so on. All I am saying is that, if people are worried and if a question, once it has been tested, throws up a problem about the order effect in which the statements are put, then there are ways round it.

Mark Diffley: The wording of the question is clearly important, and that is why we are having this discussion. We have used the new or proposed Scottish Government wording in a poll and, to be honest, have not found the answers to be massively different from when we had a completely different question previously. That is one poll, and clearly any question would need to be cognitively tested over a long period of time.

Q896 Chair: You said it is not massively different.

Mark Diffley: It was 1%.

Q897 Chair: I would rather have 1% on my side than 1% on somebody else’s side. The fact is that it does make a difference.

Mark Diffley: As to the extent to which the question itself makes a difference, clearly there is a time period between the two polls and there will be other factors at play. The point I was trying to make is that there was a percentage point difference in support for independence in two questions-[Interruption.]

Chair: We will adjourn. If there is more than one vote, we will be more than 20 minutes.

Sitting suspended for a Division in the House.

On resuming-

Q898 Chair: To go back to the writing of the question, do you have a view about who ought to be doing the question to make sure that it is fair and reasonable?

Mark Diffley: Clearly, the Electoral Commission will have a key role. The question will need to be tested in polls. There would be some cognitive testing, by which we mean that we are not just testing the question and looking at the attitudinal data that arise from it but asking people, "What do you understand by the words in this question? How do those words help you make up your mind? What process do you go through when you read differently worded questions in order to reach a conclusion about how you would vote?" We use these techniques. For instance, we did it before the census last year when we looked at the thorny issue of the wording of the national identity question in the Scottish census. We carried out a series of cognitive tests that helped the authorities to clarify what that wording should be. We would be very much in favour of that type of approach, as well as polling evidence that may suggest what biases are at play and, of course, listening to the views of the Electoral Commission. We know from the AV referendum last year that the Electoral Commission, particularly on the issue of clarity, were very clear in rewording the initial question that was put forward, which I think you and I would have found perfectly reasonable, but they considered that it needed to be further clarified.

Q899 Chair: What words of those that are around at the moment do you believe should be cognitively tested?

Mark Diffley: You used the word "separation", which is very emotive and has certain connotations.

Q900 Chair: Is it more emotive than "independence"?

Mark Diffley: I was going to refer to "independence" as well. I note that you have had discussions not just today but looking back through previous evidence that has been given about whether we really know what independence means. Clearly, we are in a process whereby that is beginning to be defined. At the moment it is not absolutely clear what we mean by "independence". "Independence" and "separation" would be the two obvious words.

Q901 Chair: What about "country" as compared with "state"? It seems to me that separation and independence are two sides of the same coin, but one can be seen to be prejudicial one way and the other is prejudicial possibly the other way.

Mark Diffley: Yes.

Q902 Chair: But "country" and "state", for example, have different meanings, with "independence" and "country" being the cuddlier versions, so there is a bias inherent in the use of either of those and a bias in the other direction in the use of "separation" and "state".

Mark Diffley: I think that is right. "State", when used not necessarily in the Scottish context but more widely, is often used to denote state interference, and so on. It is often the context in which the word "state" is used. Yes, I quite agree that there is a difference between "country" and "state".

Q903 Chair: Scotland is already a country. Therefore, to include it in a question can be misleading.

Mark Diffley: Yes.

Andrew Hawkins: There are two separate issues here, are there not? One is: who should be responsible for initiating the question? My answer to that is that it does not really matter, as long as the second part is in place, which is that there is a body accountable to Westminster and/or Holyrood. That is for others to decide, but it has to be an accountable body with a clear set of criteria and expertise, which will be held accountable for the advice it produces. I entirely understand why there is a question out there in the ether that has been proposed. Politically, it is a useful opening shot to the whole discussion and debate around the question. It seems to me that, as a pragmatic response to the situation, the Electoral Commission is the obvious body to scrutinise and test that question, but it is to do with its expertise, the fact it has done it countless times before and it is respected. It will help the process and be seen to be fair, reasonable and effective if it is involved and it is the body tasked with that.

Q904 Chair: This is possibly a more complex procedure than I had realised. What is the time scale? If the Electoral Commission was being asked by, say, Westminster to consider something that could be put in a section 30 notice-a Westminster option or something coming from the Secretary of State for Scotland-and to consider the Scottish Government’s biased proposal, how long would it take to assess two of them? Are we talking about a week, a month or six months?

Mark Diffley: I think that to do it properly would take quite long. The answer is that I do not know exactly how long it would take, but, if you used some of the techniques that I described earlier and, in addition to the Electoral Commission giving its authoritative view, you went out to test the question in some of the ways that I described earlier, that would take months.

Q905 Chair: Would it be 24 months or maybe six months?

Mark Diffley: A few months-three or six months. It would not take for ever, but you would need to go through a process where you cognitively tested the question with enough people and in different ways to give you definitive answers about how people perceive some of the words and phrases that we discussed earlier.

Andrew Hawkins: I would tend to shrink the time scale a little, perhaps to two to three months. The challenge to bear in mind here is that it is not the easily accessible demographic groups that need the testing; it is people who are by definition hard to reach. The Electoral Commission is impressively scrupulous in ensuring that in these situations question design and consultation is done among those hard-to-reach groups.

Q906 Chair: I had not quite appreciated how long that might take. Is there any advantage to the process in having the question cleared by the Electoral Commission sooner rather than later? Does it matter much that the question has to be out there for quite a long time before the vote takes place, or could it be produced relatively close? I am thinking that there are proposals that it should be in 2013 or 2014. Does it make much difference whether or not the actual wording of the question is produced, say, a couple of months before the referendum if everybody knows what the issues are?

Andrew Hawkins: That is the key thing, isn’t it? With the AV referendum it was cut and dried; with this it is quite a different concept, because, if independence today means one thing and in three months it means something else, there is no point in testing the question today based on that. The fundamental problem with the whole issue, which has been well rehearsed elsewhere, is what independence and devo-max mean, and what they mean to respondents ranging from the very well briefed on the issues through to people in the street.

Q907 Chair: That is right. I understand the point about independence or separation being to some extent a pig in a poke and we will not know until after; I understand that. But what I am focusing on now is the question of the wording. If the wording is available, say, a year or six weeks in advance and there is still lack of clarity, there is still a lack of clarity. I am just unclear about whether or not your advice is that the exact wording of the question should be available as soon as possible or you would say it does not make much difference.

Mark Diffley: To the extent that it will affect the outcome of the referendum, I do not think it makes a huge amount of difference. To the extent it might move us past arguing about the process and start engaging with the issues, from the public’s point of view, and we can put all these arguments behind us and start to consider whether it is a good or bad idea-clearly, we are not there yet-it would be a good idea to have the question resolved sooner rather than later.

Q908 Chair: That is helpful. Would you agree with that?

Andrew Hawkins: Yes, I would.

Q909 Chair: The question should be resolved sooner rather than later.

Mark Diffley: Since this became a hot issue in January, in all the ways of which we are aware, we have focused almost entirely on the process, and in Scotland as well. Our organisation is living and breathing this in Scotland. Of course, there are forums in which we are talking about whether independence is a good thing, but primarily in the media and public discourse it is all about when it should be. Should we have to wait two and a half years? Should it be one question or two? What should the wording of the question be? Should we allow 16 and 17-year-olds to vote? All those types of process issues are dominating the discussion, and the issues have not begun to cut through yet in any meaningful way.

Andrew Hawkins: For example, if the question concerned not the detail of what an eventual settlement would look like but the process-for example, a mandate to negotiate with Westminster to deliver such and such-there is nothing to stop that being done as early as possible in the process. As soon as you get beyond that and rely on someone sensibly to give an answer based on their own understanding of independence or any other concept, that is when time is against you. You need to take time before drafting or finalising the question.

Q910 Chair: I thought you were saying possibly the opposite of that and that the difficulty was not about the clarity of the wording but the substance.

Andrew Hawkins: Yes.

Q911 Chair: Therefore, if you get the process issue of the question out of the way, you can then focus on the issue of the content, and then words like "separation", "independence", "country" or "state" all get interpreted so that the emphasis is on deciding the question faster to give you more time to discuss the eventual results. Is that a fair synopsis?

Andrew Hawkins: Yes, that is right; that is correct.

Q912 Pamela Nash: You said that three to six months would be time enough to formulate the question and also test it.

Mark Diffley: Yes, depending on how many questions you wanted to test. So far, there is only one question out there. We do not yet know the results of either Government’s consultation processes and whether they will lead to further questions being proposed; whether indeed there will be a second question on what is now commonly called devo-max or devo-plus, which would clearly need to be tested as well for its clarity, bias and so forth.

Q913 Pamela Nash: We will get on to that.

Mark Diffley: Yes, I am sure. All those things would impact on the amount of time any testing took. It is months rather than weeks, but it is not years.

Q914 Pamela Nash: You think six months is a reasonable amount of time.

Mark Diffley: Yes.

Q915 Pamela Nash: Therefore, in terms of the question, we do not need to wait until the autumn of 2014 to make sure we have a properly formulated and tested question.

Mark Diffley: No. You may need to wait that long for other reasons, but as far as the question is concerned, no, I do not think so.

Q916 Jim McGovern: Possibly, my question is for Andrew. You mentioned demography. Which part of the electorate do you think would find difficulty with the question?

Andrew Hawkins: People for whom English is not their first language and those with learning difficulties.

Q917 Jim McGovern: That sounds like geography rather than demography.

Andrew Hawkins: However you slice and dice it, there are sections of the voting population who will need to be able to take part in the referendum, and the question, in my view correctly and appropriately, has to be designed in such a way as to maximise the prospects of everyone taking part in it.

Q918 Jim McGovern: But which part of the electorate do you think might find difficulty?

Andrew Hawkins: It depends on the wording of the question. By way of illustration, the AV question was changed because it was thought it would be too complicated for some people to give a reasonable answer; so it was simplified. Until you test it you do not know. We have an idea about who might find it difficult.

Q919 Jim McGovern: Are you prepared to share that idea?

Andrew Hawkins: People with learning difficulties and those for whom English is a second language. There might be words used that mean one thing to one group and something quite different to another. It depends on whether any technical words need to be used. We do not know. What does "devo-max" mean to someone for whom English is not their first language?

Q920 Jim McGovern: I do not know.

Andrew Hawkins: I just give that as an illustration of the sort of things to bear in mind, which the Electoral Commission is expert at considering.

Q921 Chair: And the fact that the Electoral Commission employs you to do these things has absolutely nothing to do with it.

Mark Diffley: It has nothing to do with it.

Chair: Perhaps we may turn now to the whole area of multiple option referendums. Jim, do you want to pick up question 17?

Q922 Jim McGovern: Should the referendum pose only one question that would give two options? When you move to two questions, possibly there are three options and so on and so on. What do you think the referendum question or questions should be? Should there be one or more?

Andrew Hawkins: I can see the elegance of having a single question, but I think the public is perfectly capable of answering more than one question and has done so in the past-if that is what you are asking-rather than whether in this instance it should be more than one question, which is a different question.

Q923 Jim McGovern: Confused-yes, I am. For example, in 2007 the elections for the Scottish Parliament took place on the same day as the local council elections. It is a matter of record that there were more spoiled papers on that day than in any other election in British history. I do not want to sound patronising and say that the Scottish public will not understand a multi-option question, but there is a case for keeping it as simple as possible.

Andrew Hawkins: Presumably, on that day there were also other factors at work, such as high turnout possibly. I do not know the figures, but there might have been greater participation overall because of the coincidence of different sets of elections on the same day, so it is not necessarily all in one direction. It also comes back to the point that, where you do have multiple elections on the same day, often you are faced with sets of ballot papers that look very different. I forget the exact numbers, but I believe that in the last London election there were multiple sets of papers all with different electoral systems being used. It was very confusing for people. That needs to be borne in mind when we think about the complexity that people can cope with at one time, how we want to maximise participation by coinciding different sets of elections and so on. I think that, if it were a single referendum, especially on a single day, given the focus and attention this is receiving and the importance of it, as long as the questions themselves are properly constructed they can be multi-dimensional.

Q924 Jim McGovern: Would you agree that a single question would probably be the preferred option by the electorate?

Andrew Hawkins: It is Ockham’s razor. It is always better to have fewer words in these things than more, and fewer options than more, yes.

Mark Diffley: I agree. There is an argument that we are quite used to casting more than one vote in any election. In the additional Member system for the Scottish Parliament we vote for a constituency MSP and the list MSP.

Q925 Jim McGovern: That is one way of having the most spoiled papers in history.

Mark Diffley: That was when we had the council elections as well on the same day, but for Holyrood-any Scottish Parliament election-we are still ticking two boxes or putting crosses in two boxes, which effectively is what we may be doing depending on a number of different options. It could be one option for the referendum. From the public’s point of view, could people cope with it and understand it? I think the answer is yes. There are other issues about the clarity of the results and so on, but as far as public understanding is concerned I think it would be fine. In our latest poll we asked, "Do you think there should be one question or two?" Actually, we asked, "Do you think devo-max should be on the ballot paper?", and 59% said they thought it should and 37% thought it should not. That does not necessarily mean there is a second question.

Q926 Chair: If you ask people, "Do you want to have more choices or fewer choices?", am I not right in thinking that disproportionately they are likely to say, "I want more choices"?

Mark Diffley: I accept that.

Q927 Chair: Therefore, that is essentially a meaningless statistic.

Mark Diffley: By itself, but maybe we will come on to talk about devo-max and the popularity of it. When you look at it with other evidence about how popular it is, how many people have heard of it and all that kind of stuff, in the round it makes some sense. I am not arguing that there should be two questions, but that as far as public understanding is concerned it is feasible. There are other issues about how you count the vote, decide the winner and all that stuff.

Q928 Chair: That is right, and that is what we want to come on to. If people are asked whether they want two hard choices or a nice cuddly one in the middle, that is how the Liberals used to prosper sometimes. Therefore, that is not an entirely fair way of looking at it and extrapolating results from those conclusions.

Mark Diffley: But one of the arguments for having a second question or another option, whether in the same question or as a separate one, is that at the moment it appears to be where the public is at, so to speak. It seems to be the most popular choice if you look at the three options. If you look at the status quo, full independence and the bit in between, it is the bit in between where the public is at.

Q929 Chair: To what extent do you think that is because it is the most nebulous? It is difficult to be against it because nobody knows what it is.

Mark Diffley: It is not defined, and we know that. There may be some of that at play. I think it is the most popular option because at the moment the majority of the public do not want to take the step towards independence, separation or whatever you call it, but there is an appetite for the Scottish Parliament and Scottish Government to have more powers.

Q930 Jim McGovern: Probably every elected Member in this room has been canvassing on doorsteps recently because of the local government elections that are coming up. When I try to explain to people that it is a multi-option ballot paper and they can put one, two, three, four, a number of people say to me, "Why don’t you just keep it simple? I used to just put an X in the box." The number or people who do not seem to understand this new system is unbelievable. I am not being patronising when I say that a lot of people will not understand a multi-question referendum. You seem to be saying that everybody will understand it.

Mark Diffley: It depends on how it is conducted. It may be that it is two entirely separate questions. You have two options here and you put an X in there; you have another question here and you put an X there. It may not be similar to the local election system in terms of rankings.

Q931 Mr Reid: If you were going to have three options, how would you design the ballot paper? Would it be three separate options and you tick the box or would it be two separate questions, as the Scottish Government have decided? What would your advice be?

Andrew Hawkins: This is where I am very pleased I do not have the task of designing the question.

Chair: But if you were paid you would.

Q932 Mr Reid: If the Electoral Commission asked for your advice, what would you say?

Andrew Hawkins: If I was paid, the only way it could work would be if we knew what the options are and what they actually mean-what devo-max means. In theory, it would be possible to design a question that set out three separate alternatives and asked people which one of those they wanted most, but we know that is fraught with difficulty because it may well produce a third, a third and a third, as currently it is constituted. It may be necessary, therefore, to consider having a minimum threshold-it does not need to be 50%; it could even be 40%-and, if there was a failure to reach it, it would be assumed that the status quo is the outcome. To move beyond that in question design terms requires more content. It is not possible at this stage to say that it should be designed like this or that without knowing what it is we are asking people about.

Q933 Mr Reid: You mentioned 40%. Is there any particular reason you suggested 40%?

Andrew Hawkins: No. I am aware that has been used in other circumstances. About 33% say they are in favour of independence, whatever that means, which suggests there is clearly an appetite for change of some sort. The fact that, even if it is not the perfect question because we cannot make proper definitions, a substantial proportion of people would support devo-max, however they interpreted it, suggests there is an appetite for change. To make 50% the hurdle to get over seems to me to court an unpopular result, because we know we will not get 50%.

Q934 Mr Reid: Say, for example, independence got 40% and the other two options got 30% each. That means 60% are against independence, and yet you are saying that would be recorded as a victory for independence.

Andrew Hawkins: Speaking on my own account, I would support the two-stage process where there is an early non-binding referendum on permission to negotiate a further settlement and, as part of that, to explore whether that is independence or devo-max, whatever that means, and to revisit it with a binding referendum in due course. I know these options have been discussed. I say that only because I think anything else becomes meaningless. I cannot see a way ahead through this without first having a longer period to negotiate on the detail of the package to put before the Scottish people.

Q935 Mr Reid: To be clear, are you suggesting that we have a ballot paper with three options on it; the elector ticks the preferred option, and, if either devo-max or independence gets more votes than any other but less than 50%, there would be some form of negotiations, followed by a second binding referendum with that option against the status quo? Is that what you are suggesting?

Andrew Hawkins: That is what I am suggesting, subject to possibly a threshold being introduced, because we know there is a danger of one of those alternatives being as close over the line as the winner of the Grand National was last Saturday, and therefore we are no further forward than we are today.

Q936 Chair: Are you saying it is not possible to have a decision made between three alternative outcomes on the one day?

Andrew Hawkins: Not unless the electorate knows what it is voting for.

Q937 Chair: But nobody who voted Liberal, for example, in the last general election realised they would be going into coalition with the forces of darkness. In those circumstances it could be argued that people then did not know what they were voting for.

Andrew Hawkins: That was a foreseen but unintended consequence of the election. We are going into this with our eyes as open as they can be, so we want to try to anticipate the likely outcome. This is also a really important issue; it is all about statehood, nationhood and identity.

Q938 Chair: If, for the sake of argument, we have complete clarity about separation and devo-max, are you saying it is still not possible to have elections or voting on one day that gives a clear and desirable outcome?

Andrew Hawkins: Of course it would be possible to hold it on one day.

Q939 Chair: How would you do it?

Andrew Hawkins: You predicated your question with "if we had certainty of outcome". If we had certainty of outcome, that is fine; we can go ahead and draft the question.

Q940 Chair: How would you draft the question?

Andrew Hawkins: If we have certainty of outcome, we can draft the question in the straightforward manner in which we would draft any referendum question, which is to ask people which of the three alternatives they would most support.

Q941 Chair: So you would have three alternatives and tick your choice, and it is first past the post.

Andrew Hawkins: Yes, but that assumes the "clarity" point and is also subject to the risk that we have at the moment. I can answer yes in theory, but it is meaningless because in practice we could have a situation where there is 1% more in favour of independence than anything else, and that, I believe, would be an unconscionable result for the United Kingdom as a whole, probably for a lot of the business community in Scotland and so on. I would not support it if it was likely to result in a very close outcome.

Q942 Chair: If the three alternatives are all perfectly clear, either you can devise a question that produces a result or it can be dependent on what the result will be. If you give people the choice of three alternatives, you cannot say, "This is the method and we will accept it unless we get a result we don’t like." You have to assume it might be 34/33/33-if that adds up-if that is the methodology you are suggesting.

Andrew Hawkins: But there are other ways of counting the votes. I am aware, for example, of the dead Frenchman’s alternative. That also seems to me to make very good sense. As Mark has made mention, support for devo-max is there. It does seem to me to make eminently good sense that the people who are in favour of independence will also in large part be in favour of devo-max. It also seems to me that there is an obvious split in the polling done on this between what we used to call high and low politics in regard to defence, foreign policy and so on and sticking with Westminster, and there is a good case to be made for that. My slight concern is that we could risk the ballot paper becoming over-complicated if we are asking people whether they prefer option A over option B, option A over option C and option B over option C. That would be one concern, but by all means consult on it. All I want is some certainty in the outcome and to see a result that people can rally around, and this issue to go away.

The one thing we must, surely, avoid at all costs is the Quebec outcome, which coined the terrible phrase "the never-endum referendum". It all ended acrimoniously with one election candidate standing on a platform promising not to have another referendum for the next 10 years. Out of all of this we want to avoid the political equivalent of planning blight where there is one referendum after another. It would be better to design a system that produces a decisive outcome, even if it is in five years, than to have it hanging over us for the next 20 years.

Chair: I am conscious that you want to get away in five minutes.

Q943 Pamela Nash: Isn’t the point that a multi-option referendum is not ever going to provide clarity or be beyond someone challenging the result legally? If we look at the poll results at the moment and if it continues that way and the percentages are so close, does that not mean that to provide that clarity we need a two-option referendum, at least in the first instance?

Chair: Can Andrew answer because I am conscious that he has to get away?

Andrew Hawkins: I think there is a less of a problem with that if the question is one about negotiating for a further outcome and there is some provision put in place to stop us going on an irrevocable path towards independence based on margins of 1%, 2%, 3%, which I do not think is healthy or will produce a happy outcome.

Q944 Chair: If it was a straight yes/no to separation, you would not accept the validity of 51/49.

Andrew Hawkins: I do not think that goes for the reasons that Martin Boon outlined earlier. If you are given the opportunity to vote, 50% turn out and 40% support a particular option, you have only got 20%-one in five-voting for something. I can live with 51%; it is a fair democratic outcome. It is where you have a plurality that we start getting into outcomes that are questionable.

Q945 Chair: I am conscious that there is a train waiting for you. Is there any additional point that you want to raise with us before you go?

Andrew Hawkins: I covered the point about Quebec, which is the one I wanted to make, and about uncertainty and coming up with a process that leads to a definite conclusion, hopefully with a big mandate for whatever the outcome is.

Q946 Chair: Everybody wants to win with a big mandate. It is a question of working it out. I am conscious that you have to go. Thank you very much. If upon reflection there are any points you want to raise, drop us a note. The same applies to Mark. If once you see the evidence that Mark gives in the next three hours you wish you had been asked that as well and want to give us a submission, then do so, if you would not mind. Thank you very much for coming.

Mark, we will come back to apply the Chinese burns to you. Andrew touched on a number of points in response to the last couple of questions. Are there any responses you want to make to these points?

Mark Diffley: There are a number of ways in which we can do this. It could be designed in one question, which was what Andrew was concentrating on. Basically, you have a single question and three separate outcomes. You can also do it with two separate questions, which may have the advantage of clarity to some extent.

Q947 Mr Reid: If you have two separate questions how do you decide which winner has won?

Mark Diffley: Quite. I was coming on to that. It has some advantages but it has another clear disadvantage in that, as has been very well trailed, if 60% of people support independence but 80% of people support devo-max, which one is the winner? You are in the same position. Who has won? The 50% threshold has been beaten by both options, one more heavily than the other, and I am afraid that gets you into similar arguments about the outcome. These things need to be made clear in advance, and we are miles away from that at the moment.

Q948 Chair: Coming back to the question of the three options, we heard Andrew’s explanation as to how he would do it. We heard reference to the dead Frenchman and all the rest of it. If you were asked to devise a questioning system that determined between three options, would you pose them against each other? Would you have a single question? Would the first two be posed against each other?

Mark Diffley: There is no simple answer. Each option is fraught with difficulties from the punter’s point of view of interpreting how to fill in the ballot paper in the way you described earlier. Do I have to rank these or choose just one, or whatever it might be? The question is deciding which option has won the day. From the point of view of clarity, I quite like the idea of two separate questions because you can answer them both; you can say "Agree" or "Disagree" to each of the two propositions.

Q949 Chair: There are two separate questions but there are three options.

Mark Diffley: You could have, "Independence: agree or disagree. Devo-max"-or whatever the other option is-"agree or disagree."

Q950 Chair: And then whether they agree or disagree with the status quo, presumably.

Mark Diffley: Possibly.

Q951 Chair: Then who wins? This comes back to the issue that, if independence gets 51% and devo-max gets 80%, which one wins.

Mark Diffley: There are arguments both ways. This question has been raised within Scotland. You will not be surprised to hear that, even in the scenario I described earlier where independence gets, say, 60% and devo-max gets 80%, the Scottish Government’s position is that independence still wins the day. I think the argument would be that there is still a majority in favour of independence, if you look at these things in that way.

Q952 Chair: Do you agree that that position is patently absurd?

Mark Diffley: It is difficult to justify in the scenario that I described.

Q953 Chair: That is a yes.

Mark Diffley: Yes.

Q954 Jim McGovern: Does that not prove or promote the case for two separate referendums?

Mark Diffley: There is an argument, which Andrew rehearsed earlier, with which I would agree. Yes, it does. It is not my area of specialism, so I do not have a very strong opinion on it, but I agree with what Andrew said.

Q955 Chair: But you are here as a "complicated questions to ask people to find the answers" guy rather than a "politics" guy. In terms of how to get an answer out of this multiplicity of options, I think you are saying to us that the best option is twin referendums separated by time.

Mark Diffley: Not necessarily. I am saying that I am attracted to the idea of having two separate questions, which could be asked at the same time.

Q956 Chair: Remind me again what the two questions would be.

Mark Diffley: Independence: agree or disagree. Devo-max: agree or disagree.

Q957 Chair: How do you count that?

Mark Diffley: Ultimately, you would have a proportion of people who agree with independence and a proportion of people who agree with devo-max. The problem, as I described earlier, is that they are unlikely to be identical. So which has won? I think that is quite difficult to decide.

Q958 Jim McGovern: Does that not make that ballot paper untenable?

Mark Diffley: Yes. I am attracted to it from the point of view of simplicity and the voter, but it is easy for me to understand the process.

Q959 Chair: But who gets clarity of result?

Mark Diffley: It would need to be made clear in advance. People would need to know how the results would be counted and who would be the winner in that scenario. That would need to be transparent in advance, but purely from the point of view of, "I’m going to cast a vote"-some of the issues we talked about earlier-"and how do I go through the process of doing it?", I am attracted to it from that point of view. I think it is fraught with difficulties and it is not too difficult to see where they are. But if you want a third option and you are thinking whether it should be in one or two questions, there are issues of clarity and how you decide the outcome. The two questions appeal to me from the perspective of clarity.

Q960 Chair: I completely understand that from the point of view of the question it is clear, but from the point of view of the results it need not be. Therefore, you cannot judge the clarity of the question solely by the question itself; you have to judge it by the outcome as well. Under those circumstances it is entirely unclear how you would devise it.

Mark Diffley: Yes. Rules would need to be written that said under these circumstances this position would be the winner. Everyone would need to be clear about that.

Chair: My God.

Q961 Iain McKenzie: Surely, complexity has crept its way in. With respect, you want clarity of result, but equally clarity of what originally triggered this whole debate. It was not devo-plus. Devo-max did not come in until a very late stage. The original referendum was triggered by a party winning power and saying, "In our manifesto we will call for a referendum on independence", not a referendum on independence or devo-max. To keep that clarity of result and make it as straightforward and as clear as possible, the straightforward question should be, "Do you agree with independence or not?"

Mark Diffley: I do not think you would find anyone who would argue that the result would be most clear under those circumstances.

Q962 Iain McKenzie: I think we need to get that question out of the way and then ask the electorate, once they know what devo-max or devo-plus is and what that entails, "Do you wish to have the status quo or devo-max?"-not to have the three and the one.

Mark Diffley: That is one option. It is patently obvious that as far as the clarity of the result is concerned the clearest way to do it would be to have a single question with yes or no, agree or disagree, or however you word it.

Chair: Let us move on slightly to the question of polling evidence at the moment. Pamela, will you pick up question 22?

Q963 Pamela Nash: We covered polling evidence earlier. You mentioned the move towards devo-max, although I am tempted to say that is just because they have heard the term. Unfortunately, as it has never been defined, you cannot really say that people are moving towards devo-max. Can you share with us your latest polling figures for public opinion in Scotland on this issue?

Mark Diffley: You are right to say that of course it is not defined. I think in a way we have moved on a little from devo-max to devo-plus, at the risk of just throwing all these terms out there wildly. Reform Scotland is a group that includes former MSPs from each of the pro-Union parties, not the SNP, and it has now begun to define exactly what a middle option would be.

Q964 Chair: That is the devo-plus gang.

Mark Diffley: That is devo-plus, yes. This is about what taxes would be raised and exactly in which Parliament the big issues would be decided. That is one point and that is important. How much traction that will get we do not know, but attempts are being made to try to define this middle position. So devo-max may or may not become yesterday’s phrase; it may not. However, when we test opinion on it in a poll and try to define it as substantial new powers being transferred from the UK Parliament to the Scottish Parliament-because that is the best definition you can get at the moment-we find that about 70% of people are in favour of devo-max.

Q965 Chair: If you compare devo-max with devo-plus, what are the figures?

Mark Diffley: We have not asked. It is pretty difficult to compare it because devo-max is not defined. We have not asked about devo-plus yet.

Q966 Pamela Nash: From my experience in speaking to my constituents, what we regard as a definition of devo-plus would be what most people would think of as devo-max, so it is quite an important question. The question to ask is, "Do you understand these terms?"

Mark Diffley: It is very interesting. Our polling reveals that more people support devo-max than have heard of devo-max.

Pamela Nash: I am not surprised by that.

Q967 Chair: People support it but have not heard of it.

Mark Diffley: It is small numbers. There are people who tell us they have not heard of it and yet when you ask them whether or not they support it they say, "Oh, yes; that’s a great idea", which comes back to some of the points you tried to make earlier-for example, that it sounds like a good thing; it is the middle ground.

Q968 Chair: If people are asked, "Are you in favour of more powers for Glasgow?" or "poor Glasgow", generally they will tend to say yes, will they not?

Mark Diffley: Yes.

Q969 Chair: If you say even within Glasgow, "Are you in favour of more powers for Pollok?", people will tend to say yes. If you are asked anything like that, the default position is almost always, "I am in favour of more rather than less." Whether or not that means people are genuinely in favour of a particular set of concepts is not clear from those sorts of anodyne questions, is it?

Mark Diffley: I am afraid it is the only way we can ask it at the moment because it is ill-defined. We know it is not defined, so you can only ask it in an anodyne way. When we start to get more clarity-maybe the devo-plus movement will start to provide it-hopefully we can start asking, "What about corporation tax?" and the substantial propositions that those people are putting forward.

Q970 Chair: It comes back to the two options. If you say, "Are you in favour of cutting corporation tax?", most people will probably say yes, but if you ask, "Are you in favour of corporation tax being cut and this will mean a reduction in public services?", most people will maybe give you a different answer. Therefore, it seems to me that how you ask the question determines the result, and a lot of this is not particularly helpful.

Mark Diffley: What I mean is, "Would you prefer that the Scottish Government set the rate of corporation tax?", or, "What is your opinion of that particular power being transferred from the UK Government to the Scottish Government?" It is not about at what level that tax should be set but which Parliament should have jurisdiction.

Q971 Chair: But if you said, "Would you prefer it to be transferred to Glasgow rather than the Scottish Government?", or, "Would you rather have it done by anonymous bureaucrats in Edinburgh or by your local councillor who knows your community in Glasgow?", I can tell you what the answer would be.

Mark Diffley: People’s attitudes to councils is a whole other subject.

Q972 Chair: Yes; it is the issue of proximity.

Mark Diffley: I agree, although polling evidence from other organisations will show-Andrew touched on this earlier-that, when you get to issues of defence, the Armed Forces and so on, there is no real appetite for bringing those closer. The majority of people would still want those powers to be controlled in Westminster.

Q973 Chair: Would that include nuclear weapons?

Mark Diffley: I would need to check. I do not think to the same extent but probably, yes.

Q974 Chair: That is helpful. Are there any other points you are bursting to answer that we have not asked you?

Mark Diffley: Not especially. As I said in the written submission, the question is very important, but I would argue that the extent to which the wording of the question is going to affect the result is pretty marginal, for reasons that have been touched on both by ourselves and in earlier evidence. I think the significance of the wording of the question is that it clearly sends out a signal as to which side is in control, who has an early victory about the wording of the question and so on and so forth, but the extent to which it has an effect on the referendum outcome itself is marginal.

Q975 Chair: To overcome both of those difficulties in a sense, one being the question of the early victory, presumably passing that to the Electoral Commission sooner rather than later to make a determination would blur that signal. Secondly, you said that the wording of the question does not make all that much difference. In your poll it did. We have had previous evidence from a number of experts, which you may or may not have seen, that said that, particularly in multiple choice referendums, how the questions are asked and how the results are counted can determine the result completely.

Mark Diffley: I would argue that how the results are counted is probably much more important than the wording of the question itself. You will have heard this from others, but when we design polls it is vitally important that the questions are balanced. That is important in a referendum, but it is not as important as it is in a poll. There are obvious reasons for that. There are obvious differences between a poll and a referendum, which means that the wording of a question is more important in a poll. Asking someone off the top of their head when they are not prepared for it is different from someone who has had months, years, whatever-however long this is going to go on-to consider their position. When they get into the voting booth to make their decision, the actual wording in front of them will have a marginal effect on the result. I am not saying it is unimportant.

Q976 Jim McGovern: I disagree with you on the wording. I think it is very important. For example, if the question was, "Would you like Scotland to be a free and independent country?", it sounds quite positive; it is about Braveheart and all that stuff; whereas if you want Scotland to remain within the UK and you ask, "Are you in favour of Scotland being torn out of the UK?", it would have a wholly different meaning. So the wording is very important.

Mark Diffley: It is. I think you have chosen two very extreme examples.

Q977 Jim McGovern: They are extremes. I did so to make the point.

Mark Diffley: I do not think anyone would argue that any of the words you have just used in those two examples should be included in the referendum question for those reasons. My argument is that you avoid obviously biased, leading and ambiguous terms.

Q978 Jim McGovern: I go back to what Mr Davidson said at the very start. He asked, "Do you agree that a question starting with the words ‘Do you agree?’ is a leading question?" That is what the SNP intend to put.

Mark Diffley: But I think that is self-evidently much less problematic than, "Do you think Scotland should go henceforth and become a brilliant independent country?" or whatever.

Chair: Many a mickle makes a muckle, as the saying goes.

Q979 Jim McGovern: You say the wording is marginal in terms of outcome. I do not believe that is the case. The wording is very important.

Mark Diffley: I believe it is important as well, but I do not think we should get so hung up on the wording of the question. Once we are in the ballpark of what is acceptable, I do not think we should spend endless amounts of time worrying about it. That is the kind of point I am trying to make. I am not trying to be flippant and say it is not important.

Chair: One of the very interesting things about today is that it has impressed upon us the importance of trying to get some of these process issues resolved sooner rather than later so that we can start to move on to issues of substance. We will reflect upon that and probably try to get some of the issues relating to the question resolved sooner rather than later.

If there are no other questions, thank you very much for coming along, for waiting for us while we discussed things with others and for breaking off for the questions. We very much appreciate the advice you have given us today.

Prepared 25th April 2012