Finance and Constitution Committee

Meeting date: Wednesday, September 11, 2019

Official Report 634KB pdf

---

Referendums (Scotland) Bill: Stage 1

The last item on our agenda is evidence on the Referendums (Scotland) Bill. I welcome Dr Theresa Reidy, University College Cork; Dr Andrew Mycock, reader, University of Huddersfield; Professor Chris Carman, Stevenson Trust for Citizenship; and Professor Toby James, professor of politics and public policy, University of East Anglia. I thank you all for your submissions. We will go straight to questions.

The Scottish Government officials explained to us that the intention of the bill is that

“it will be possible for the framework to be used for different polls, so there needs to be some flexibility on the circumstances and the packaging.”—[Official Report, Finance and Constitution Committee, 26 June 2019; c 26.]

To what extent is there sufficient flexibility in the bill to allow for different referendums? That is a nice easy question to start.

I will have a go at that. The bill allows a lot of flexibility—much of the detail would come through statutory instruments. If I recall correctly, the bill also provides ministers with the ability to use secondary legislation to change any of its provisions.

This might echo the comments made in last week’s evidence, but it could be that there is too much flexibility, in so far as there might be concerns that it would restrict parliamentary oversight of the legislation. If the objective is flexibility, the bill achieves that objective. However, there is probably too much flexibility.

The bill certainly includes a lot of flexibility on the campaign conditions, but some of the discussion suggests that all referendums are equal and equally spirited and contentious, and that they will naturally involve a lot of high-stakes engagement. That is not necessarily the case.

If you begin to use referendums quite frequently as a tool, you will inevitably come across technical issues that will not engage the public. A bit more thought might be given to the role that the Electoral Commission could play at that point, in particular in providing more impartial information to voters. Active campaign participants might not be leading the debate in the field, and there will be a need for information to be provided to voters, and that will require the Electoral Commission to step up to the mark, and, perhaps, in hotly contested campaigns, do more than it might normally do.

On financing, if the referendum is on a technical issue, there might not be obvious well-resourced campaign participants on the ground, and you might want to provide small amounts of campaign funding to encourage and incentivise umbrella organisations to form, because you already have the structure to do that. Such approaches emerge when you begin to use referendums frequently, in order to deal with non-contentious issues.

I asked my initial question because I wondered whether, in viewing the bill through the prism of one particular referendum—in the committee’s last session, that tended to be indyref2, should it ever happen—there is a danger that we might lose sight of some of the flexibility that is needed. We could end up amending the bill to address concerns relating to a specific referendum, which might undermine the policy intent that was there in the first place to create framework legislation. Does anyone want to reflect on that? I see Chris Carman nodding.

I agree with that assertion. There is a problem in having a very specific example of a referendum in our heads when we are thinking about this bill. You might want to consider the issues on which referendums can be held. Would they be held on any devolved issue? How would the decision be made on what such an issue might be? There is always some discussion on the extent to which devolved powers reach and do not reach into different areas. What body would adjudicate that? Would that be left to an independent body? I assume that you would not want the Electoral Commission to do that, because it tends to remain a step back from those sorts of contentious political issues. So, how would you make decisions on that, and what would be the mechanism for determining how to do so? I agree that you should keep in mind—as Theresa Reidy just said—that referendums can be held on a variety of issues, from small technical constitutional issues to larger issues that affect the public mind.

It is encouraging that the bill shows a sense of keenness for policy learning from previous experience of referenda. There are two substantive questions on this matter. One is about frequency of referenda, and at the moment that seems to be rather open ended. The other issue that has not been addressed is what triggers referenda. Often, we find that the process is contentious in itself. Some parts of Europe have stipulated conventions that allow for that, such as Switzerland and the Republic of Ireland. I do not see that consideration in the content of the bill, and it is something to think about. There is tension between the political process and purpose of those referenda. Both matters need to be considered.

Before I ask my question, I declare an interest in that, like Professor Carman, I am a member of the professorial staff of the University of Glasgow. Thank you for reminding me about that, convener.

I want to continue the line of questioning. The reason why it is not a problem in UK law is that, under the UK equivalent of this bill—namely, the Political Parties, Elections and Referendums Act 2000—each referendum requires its own bespoke act of Parliament. As introduced to the Scottish Parliament, the bill does not have that stipulation. Section 1 of the bill gives to the Scottish ministers the power by regulations to provide for a referendum.

Given that, unlike PPERA, this legislation would not require a bespoke act of Parliament for each referendum, should the bill specify the subject matters, within devolved competence, on which referendums should, might or could be held, or is the solution just to omit section 1?

In a way, that refers back to what I just said about devolved competences. There are different ways in which you could handle that. You could set up an independent commission that would take a view on devolved matters. I am reminded that, over the years, petitions committees of the Scottish Parliament have taken very different views of what is and is not devolved and the extent to which devolved powers reach into, say, the health sector.

Where devolution starts and stops is contentious. To some extent, it is a political decision. There needs to be a mechanism to decide that, whether that is allowing ministers to do it as the Government of the day—that is one political decision—or setting up an independent commission. I suspect that going so far as to try to provide for that in this bill would be quite difficult. Trying to specify areas in the bill would be a minefield and could end up causing more problems, because the level of specificity that you would have to get to would be quite granular, I imagine.

Let us take an example that is wholly in devolved competence and for which there is no argument about whether it is devolved or reserved: the setting of the Scottish budget. It is an argument that I explored with the Government’s bill team at the committee meeting before the recess.

The Scottish Parliament is a parliament of minorities, as it is designed to be. Let us suppose that the cabinet secretary for finance cannot get his or her budget through Parliament, and he or she decides to use the power under section 1 to establish a referendum on whether the budget should be passed. Given your understanding of the international best practice in the use of referendums in parliamentary democracies, is that good practice? The Government bill team confirmed that, as drafted, the bill would enable a referendum in circumstances such as that. Is that the use of referendums that we should contemplate in Scotland?

It seems that the bill would enable a potentially rapid expansion in the number of referendums that could and would be held, which could have profound consequences, both positive and negative. In many respects, it would change the nature of Scottish parliamentary democracy by making it much more direct. There are advantages to that, but there are also disadvantages. Referendums are about the will of the majority—the 52 per cent, the 55 per cent and the 60 per cent—but we must think about minority interests as well.

The committee might want to consider being cautious about allowing ministers to call referendums in that way and with that degree of speed. Globally, referendums are still relatively new territory. It is a relatively new era for them—certainly in the United Kingdom and Scotland. The committee should be wary about the situation leading to referendums being called very frequently, which would be a major change.

10:45  

If there were to be a shift towards holding referenda to resolve issues that emerge in the Scottish Parliament, we would be looking at a fundamental change in Scottish political culture.

In that context, if you are thinking about listing a range of policy areas on which referenda are to be held, you are entering into territory in which parliamentary sovereignty and popular sovereignty increasingly come into conflict. One of the experiences of the 2014 Scottish independence referendum and the 2016 European Union referendum has been the creation of a binary effect in society, which can often be in conflict with the multiparty framework of the Scottish Parliament itself. There is great need to think about the implications of increasing the number of referenda, in the context of the broader framework of political culture.

The bill as drafted appears to provide an enduring framework for the regulation of referendum campaigns. The question taps into a higher-order question: under what circumstances and for what types of issue would referendums be held? In a way, the citizens initiatives issue raises the broader issue of the kinds of referendum that would be held. Would a referendum be purely about constitutional issues or could it be about ordinary legislation? Adam Tomkins is talking about ordinary legislation, with the referendum triggered by a minister, but if there were provision for citizens initiatives, citizens could initiate legislation in particular areas.

The international experience is that budgetary issues are generally excluded from citizens initiatives, and that top-down referendums on ordinary legislation are quite uncommon, albeit that there are some examples. For example, in the United States, famously, a citizens initiative in California affected property tax and was a significant contributor to the near-bankruptcy of the state. Designers tend to leave budgetary issues, foreign policy and minority protection rights out of referendum procedures—indeed, such issues are often specifically precluded from inclusion in referendum procedures, for that very reason.

I was going to say something similar. Budgetary mechanisms can be excluded. Instead of thinking about what is included in the context of the legislation, you can think about what is excluded.

In the context of the example that Adam Tomkins gave, serious consideration of the regulated period or relevant period—whatever you are going to call it—would be needed, because the length of the referendum campaign could be problematic for certain pieces of legislation, such as budget legislation. A long regulated period would create a period of uncertainty about the budget. That, in itself, might stop the behaviour that Adam Tomkins talked about.

The flipside of the point about shifting Scottish political culture is that the approach might be regarded as extending Scottish political culture. Openness, transparency and participation were core founding principles under which the Scottish Parliament was set up. Those principles led to, for example, the petitions system and cross-party groups, which have been ways of bringing the public into the political sphere. One might think that those approaches have not shone at their brightest, so extending such mechanisms could—possibly—lead to a more participatory democracy.

Alex has the next question. Sorry, I should have said Alex Burnett; I have to remember that we now have two Alexes on the committee.

Mine is pronounced Alec.

That will help me out. Thank you. On you go, Alex.

Thank you, convener.

Last week, the committee heard strong evidence that the Electoral Commission should be consulted on question testing, even when a question has previously been proposed and tested. Will the panel members give their views on that?

It is very good practice to involve an independent body in question testing. Colleagues are more expert on this than I am. The precise wording of a question can obviously affect the result. Therefore, it follows that we have an independent body acting as a check and balance, ensuring that the question is reasonable, fair and fully tested.

It clearly matters. It may well be that there should be something that involves not just the Electoral Commission but citizens themselves. It is the kind of issue that would work well in a citizens assembly. The idea that the contours or conventions of any referenda are designed by people who are not entirely accountable to them can seem rather abstract to voters. There may well be a need to consider both the independent input from the Electoral Commission and the input from some form of citizen representation.

In general, it is good thing to have an independent Electoral Commission consult on the question. Having a degree of flexibility on the design of the question is important. I can give the committee some evidence from Ireland. In Ireland, the phrasing of the question is set down in legislation, so each referendum question asks whether you agree with the legislation to, for example, the 38th amendment to the constitution, and in brackets there will be a summary phrase, which may or may not make it apparent what the referendum is really about. We have evidence to show that that sometimes causes voter confusion and can make it difficult for people to vote. In particular, where there has been a negative built into the question, such as, “Do you agree with abolishing this house of Parliament?”, we have seen that voters find it very difficult to understand. Indeed, we have evidence that shows that some voters voted the opposite way to their clear intentions. Having flexibility in the design of the question is very useful and having an independent body involved is the right road to go down. Our system in Ireland is far too inflexible and we have evidence to show that it has not served us well.

I agree with that. International evidence shows that question testing is a good thing. Mr Burnett’s question is really about the second testing of the same question or something very close to it. If there is any deviation from the previous question, I would say that testing it would be appropriate. The question is whether the same question being run again deserves further testing. The answer to that is probably yes, because public understanding of political issues changes. We can use relatively recent examples to show that, as public opinion changes, particularly in relation to pre-legislative as opposed to post-legislative questions, even if the same question is rerun in a relatively short period, some degree of testing is desirable. One might question whether that would require the full 12-week process, but there would be a need for some degree of confirmation or other sort of testing.

Professor Carman, as a follow-up to what you said earlier about being able to look at the bill objectively and separating out future referendums from past referendums, do you think that the specific reference in the bill to previous referendums—the only example being the 2014 referendum—is helpful?

It is difficult to say. It is probably desirable for the bill to be as neutral and removed from previous experience as possible, if only because the legal precedent that is established by referencing previous examples might cause some issues down the road.

As a member of the Delegated Powers and Law Reform Committee, I explored this issue with colleagues and the cabinet secretary yesterday. One of the points that the cabinet secretary made—and I am keen to hear your views on this—was that the question that was used in the 2014 independence referendum has been used in almost every opinion poll since then and is commonly understood. I think that the question has been used in more than 50 opinion polls. Therefore, it might create unnecessary confusion to deviate from that question. What are your views on the cabinet secretary’s statement? Do you agree or disagree?

I guess that I could see that. In a way, it runs contrary to what I have just said, in the sense that you might want to retest. Even John Curtice would recognise that there is error in polls, and we get slightly different responses to questions depending on sampling methodology and how surveys are run. Furthermore, the fact that a question is used in public opinion polls does not mean that the public necessarily understand what they are responding to, although there are certainly ways of improving question wording in polls. As a member of the Scottish election study team, I note that we ran the same questions over and over again, recognising that they tend to lose their meaning to the public over time, or change their meaning.

I still think that some sort of independent evaluation would be required. If a question has been used repeatedly in polls, we might not require the full 12 weeks, or the full period that the Electoral Commission would require to test a full, unique question, but we would probably still want to have some independent experts look at it in order to certify that it was still a fair and reasonable question.

If we think about the influence of the independence question on Scottish society over the past decade, regardless of how the question is framed, many people will translate it into how they see the question, and they will simply reinforce their particular view. There is of course a constituency of don’t-knows, or those who are prepared to shift their view, but much of the evidence suggests that, regardless of how you frame the question, many people in Scotland will recourse to the position that they have on that, which is rather fixed.

Drawing from some of the evidence in the Republic of Ireland, I note that we have had multiple referendums on many different issues, and I think that it is good to think about the subject more broadly. The question wording does not necessarily dictate the direct lines that the campaigns will take. We had a referendum on divorce in 1986 and another one in 1995. The questions were almost identical, but there were different outcomes and different types of campaigns. There are underlying shifts in public opinion, and on particularly deeply rooted, cleavage-type issues such as that, opinion change happens very slowly over long periods of time.

It is important that the question is clear, that voters understand what it means and that any underpinning legislation that might be implemented on the outcome is clear and direct, but the question will not necessarily determine the types of campaign that there will be or the issues that will come up. The participants in the campaigns and the issues that are live can change over time, even with the same question. We have had that experience with fairly intractable social and moral issues, but we have also seen it in relation to the repeat of European Union referendums. We have had a couple of those, and we have had very different debates. Economic and socioeconomic circumstances can change, and that is really what delivers the final outcome. The crucial thing about the question is that it is clear and understandable to the voters.

I have a supplementary question on the idea that Professor Carman raised of having an expedited process for the Electoral Commission to engage. Whether the referendum was made through regulation or through primary legislation, it is likely that a committee—perhaps this committee—would take evidence, so there would be an opportunity for the Electoral Commission, via that process of engagement, to give at least a preliminary view or a provisional opinion. Rather than there being a full 12-week period of testing, would that be a forum where it could give its informed view?

Yes. Obviously, it is up to the committee to decide who it invites to give evidence, but one would think that the Electoral Commission could come and provide it. It could depend on when the committee hearing was held and whether the Electoral Commission had enough time to gather evidence before it was invited to provide evidence to this committee, or any other committee.

11:00  

Another layer of consideration is public perception. If the process were to be run through the Parliament, that might be seen as politicising the testing of the question. You might then like to think about whether a more independent view would need to be taken, as opposed to the matter coming through one of the committees.

The simple point that I was making was that we have an opportunity for the Electoral Commission to submit evidence, whether orally or in writing, so we would not be ignorant of what its view was.

Again, that is assuming that it would be granted enough time. The questions would then be what that amount of time would be and whether it would need to be legislated for. There is therefore a bigger issue to do with timing.

I think that we are due to have the Electoral Commission before us soon, so I look forward to asking about that then.

I have a question on the same theme of legitimacy. Is it important that referendums are seen to be legitimate and not set up by any of what we might call the governing parties? If the Government decided to hold a referendum and it determined the question that would be put, is there not a danger that that would bring such legitimacy into question?

I think that there was a second referendum in Ireland over Europe.

It has happened twice.

I would need to look at whether the situation was different there. In general, if a referendum has to be rerun but the circumstances have changed significantly, is there not a danger that, if the governing party simply insists on its question, the legitimacy of the whole process will be doubted? If a country is divided as it goes into a referendum in which even the legitimacy of the question is in doubt, is that good for democracy?

The framing of the question is one element of legitimacy, but we have learned—particularly from the European Union referendum of 2016 and, to a lesser extent, the Scottish independence referendum of 2014—that legitimacy can be doubted in a number of areas. For example, there could be questions about the way in which the different campaigns use facts or are being less than secure in what they claim, or about the way in which they are funded. It is a remarkable feat that both remainers and leavers have come to the conclusion that, for many people, the European Union referendum campaign process was not legitimate. In many ways, that highlighted the fact that the UK Government had rushed into it, had not thought about the conventions and had not given the electorate enough time to learn about the issues.

Many things about the set of campaigns in 2014 were appropriate, such as the two-year period of learning that allowed Scottish society to comprehend the issues concerned. Although both campaigns were febrile and rather hotly contested at times, they largely provided the electorate with substantiated cases for and against the proposition. That was not the case in the 2016 European Union referendum, in which, at times, both sides were loose with the truth.

Therefore although I agree that the question is an issue, it is the wider political approach towards how a campaign is run that often produces challenges to legitimacy.

I understand that but, in 2014, all parties had come together and agreed on the question. Is it not crucial that there is some kind of agreement on what people will be asked?

In principle, yes—I strongly support that position. It is very important that all political parties and citizens—I highlight again the idea of having their input—have a sense of having legitimised the process at the beginning. If they do not have that sense, you will find that the legitimacy of the referendum will be contended even before it has been held.

On legitimacy, an important starting point is that there should be clarity about the circumstances in which a referendum can be called and who is responsible for calling it, so that all referendums are created under the same types of structures. The instrument should be applied and used equally within the system, which gets back to questions about whether it is on constitutional matters or ordinary legislation, who can trigger it and so on.

There is a second point, which relates to the likely success of referendums. The international evidence shows that when there is cross-party consensus—particularly when there is parliamentary consultation and agreement around the referendum issue—there is a greater chance of it being successful. I was reading some work yesterday about constitutional amendments in central and eastern Europe, which said that they have a much higher chance of passing in referendums if there is some degree of cross-party consensus. It is a political issue, which translates into the campaign, but the starting point has to be clarity about what types of issue will be dealt with in referendums and how they will be triggered. That gives an enduring framework and way of delivering referendums that voters can understand and become more familiar with over time, which helps to underpin and build broader legitimacy for the referendum instrument as a general tool for making decisions.

Regarding the idea of revisiting questions, lots of countries revisit referendum questions and the same is true of citizens initiatives; the same kinds of things come up over and over again. That comes down to the much bigger question of how a country’s direct democracy intersects with its representative democracy. Again, it is about who can trigger referendums and what happens when results occur that perhaps run contrary to other policy issues or more complex ones. What, then, is the role for representative democracy to address those kinds of issues? Those are very big design questions for a system, which have to be understood and addressed in a broader sense, rather than through the prism of a particular—albeit very live—referendum issue.

Professor Fisher, who gave evidence last week, said:

“polling companies constantly review their questions because the questions rapidly go out of date in respect of people’s understanding of what they mean.”—[Official Report, Finance and Constitution Committee, 4 September 2019; c 19.]

Tom Arthur talked about the number of times that polling companies have used the question that was used in 2014. I am struggling with this in my head but, if polling companies have not altered the question over that period, can we assume that they are comfortable that people’s understanding of it has not been lost? There is a conundrum there.

Not necessarily—there is a variety of reasons why one would not change the question. In, particular, if someone is running the question for academic research or media purposes, an important factor is to have a longitudinal timeline of responses to the same question. As we have already established, small changes in question wording can lead to different responses. Therefore, even if the question is not asking exactly what they want it to ask, someone might ask the public the same question over and over again to get that timeline of responses.

Does that not create confusion if the question then changes?

Possibly. That goes back to the question of whether an independent body should look at the evidence. It could use the evidence from the polling firms and does not necessarily need to run its own surveys. It could do some statistical analysis to find out how stable responses are over time across different groups, which will give a sense of whether the public still understand the question in the same way.

If the justification for running a second referendum is a material change in circumstances, it is not the same proposition or the same question that is being asked. In those circumstances, do you not think that the Electoral Commission should be fully involved in testing the question?

I think that the Electoral Commission should be fully involved. I cannot see any advantage in limiting its role or the time that it has available to do that.

It is appropriate for every referendum—if it is repeating an issue or if the material circumstances have changed—to go through that process, even if it is simply a confirmatory process, so that you get buy-in from as many citizens as possible on the legitimacy of the particular referendum.

John Mason has a question about the Irish experience.

We have heard a lot of evidence so far, and there are some interesting points to pursue in your evidence, Dr Reidy. You have mentioned that some referenda might be on contentious issues. Clearly, there will be a lot of people on both sides of a contentious issue, but some referenda might not be on a contentious issue. Can you give us an example of such a referendum, maybe from Ireland? I do not think that we have had any here that have not been contentious, so I am interested in that point and in your suggestion that the Electoral Commission, for example, should provide objective information. Would that be only for the non-contentious referenda, or should it be for the contentious referenda as well?

I will answer your second question first. If you conceive of this as an enduring framework that would apply to all referendums, in principle, the same rules should apply to all referendums in relation to technical matters. If it is good practice for the Electoral Commission to comment on all questions, that should apply to all referendums; equally, if the Referendum Commission in Ireland or the Electoral Commission here provides objective information, it should do that over time.

The key learning from the Irish experience is that voters have become familiar with the information that is provided by the Referendum Commission. They expect to get the leaflet in the post, they listen to the advertisements and the chairperson of the Referendum Commission participates in a lot of public information sessions. A lot of the research evidence that we have shows that, as a result of that, the Referendum Commission’s information is highly valued—it is influential in shaping voters’ positions and their understanding of issues—and the Referendum Commission is very much trusted by the voters. That did not happen overnight. It happened because of the cumulative experience of running referendums. We have been having roughly one referendum a year for the past 10 to 15 years, so there has been a cumulative build-up.

Can I press you on that point? How much detail and what kind of information does the Electoral Commission give? My impression is that what we have had so far has been quite insipid—it is about which box to tick and so on—whereas, in the European referendum, there were big debates about, for example, whether the health service would get more money if we left the EU. Would the Electoral Commission comment on that sort of thing?

In Ireland, the Referendum Commission provides objective, factual information that is not disputed. There are limits to what information it can provide, because there will still be areas where there are substantive elements of contention—

So, the commission could not say anything about the future, could it?

Let us take, for example, the fiscal stability treaty, which was the topic of the most recent EU referendum that we had in Ireland. The Referendum Commission developed information to explain the contents of the fiscal stability treaty, what it was about, what the implications would be for Irish budgetary policy and the Commission’s oversight of the budgetary policy. That was all fairly objective information. The chairperson of the Referendum Commission then participated in public interviews and debates on those questions.

Previously, a persistent thread running through European referendums in the Republic of Ireland was that the new legislation or treaty would bring in abortion and supersede the provisions on abortion that were in the constitution at that point. The chairperson of the Referendum Commission was able to conclusively say, “There is nothing in the fiscal stability treaty about abortion—it will not affect abortion rights one way or the other.” It is not always that clear cut, but that is a nice, clear-cut example of how the Referendum Commission can shut down a line of argument.

Your first question was about the types of referendum. In the Republic of Ireland, referendums happen because there is a requirement to change the constitution.

Is that the only time that you have referendums?

We have legal provisions for what are called “ordinary” referendums, which are the policy referendums that were mentioned earlier. However, they have never been used. We have had referendums only on constitutional changes. The constitution is a fairly comprehensive document. As a consequence, as we have outgrown the mores of the 1930s, we have had to have a lot more referendums in recent years.

You have probably heard a lot about referendums on equal marriage and on abortion. We have had lots of questions on such issues. We have also had lots of referendums on quite technical issues to do with legal and political institutional design, which are in the constitution and relate to, for example, the houses of Parliament and the courts.

In 2011, for example, there was a referendum on whether parliamentary inquiries could be held under particular circumstances. It is hard to find members of Parliament who have strong views on parliamentary inquiries; finding citizens on the street who have views on the matter is even more difficult. That referendum was on a matter on which there was very little information and there were very few strong views among voters, so it was important that the Referendum Commission stepped into the field and provided information on what the proposed changes meant, how they would be implemented and what the consequences would be. Ultimately, the proposal was defeated.

11:15  

That brings me to a bigger point about referendums. There is an interesting piece of work by a Canadian political scientist who has classified the types of issue on which we have referendums. He talks about big, cleavage issues, on which people have fundamental views. Scottish independence is probably a big, cleavage issue, as is unification of Ireland. We expect people’s opinions on such sincere, value-type issues to change really slowly.

There are then the mid-range issues—the more technical, policy issues. The European Union referendums in Ireland fall into that category. What the political parties and civil society actors say about them is very influential.

After that, there are the more technical issues that come up over time. A referendum might relate to a specific policy issue at a point in time. Views on such issues can be quite volatile—as was the case with the parliamentary inquiries referendum in the Republic of Ireland. People really do not have any opinion on the issue three weeks before the referendum happens, but a rapid process of opinion forming happens during the campaign, with each new piece of information potentially changing a person’s mind. Today, a person hears something that makes them favour Oireachtas inquiries, but, the next day, all the attorneys general come out and say, “This is a dangerous change,” and the person changes their mind again, because they are not making the decision from a place of fundamental values on the topic.

We have referendums across the spectrum, but, in the context of the bill, the issue is largely envisaged through the prism of the more cleavage, fundamental questions.

Is that why you seem to suggest that both sides of a campaign should be financed—so that a campaign can be created if none exists? We might not need to do that if we are going to have referendums only on contentious issues.

If you are dealing only with contentious issues and the intention is not to have technical referendums, it might not be necessary for you to do that. We do not do it in Ireland, even though we have technical referendums fairly regularly, but we have done post-referendum research that shows that a consequence of that is that voters tend not to know much about the issue. Often, there are no posters and there is not much by way of debate.

A solution that is often proposed in that regard is the provision of a small amount of money to campaign groups. However, you should keep it in mind that such an approach artificially creates a no side, in a way. There is concern that money could be given to people who would use the referendum to create a platform to campaign for something else. You would have to be careful about how you went down that road.

Some referendums can take place on technical matters that do not generate much interest or intensity on the part of voters or the political parties.

Thank you. I addressed the question to Dr Reidy, but I think that other panel members want to respond.

Our conversation started with consideration of the role of the Electoral Commission. So far, we have talked about the commission in a variety of ways—I have jotted them down. We have talked about the commission providing fair, neutral and balanced information to the public on the content of a referendum. We have talked about the commission being involved in the regulatory processes of the referendum. We have talked about it administering and running referenda, and we have talked about it testing questions.

In his briefing paper, Alistair Clark notes that an advantage of the bill is that it separates the regulatory function from the administering function, with the latter going to the Electoral Management Board. Within the context of the bill, the committee might want to think clearly about the role of the Electoral Commission. If the commission is in charge of providing fair, balanced information, one might want to separate that from the regulatory function, which has already been separated from the administering function. It becomes quite complex.

Is there any best practice on that, or does it vary around the world?

My sense is that it is rather varied. In the United States, if we look across all 50 states, there are different processes in place for regulating information. As the committee can imagine, in the US it tends to be no holds barred and everyone goes for it, which can be quite problematic. We have seen before that it can be quite problematic if we do not regulate the sorts of information flows, particularly in relation to internet and online access.

I will take that on. In some senses there is an important requirement for an independent body to provide information for the electorate about the context, issues and consequences of any referendum. What has been proven by both the 2014 Scottish independence referendum and the 2016 European Union referendum is that a Government body that is stimulating the referenda cannot be relied upon to provide that impartial information—that was the case in both instances. Beyond that, there is a really important question about the learning period. There is strong evidence to suggest that a campaign period of anything around 10 weeks, which is the usual norm for a UK referendum, does not provide the electorate with sufficient ability to gain an appreciation of the issues concerned.

American research on political psychology suggests that, about six weeks before an election, most electors move from being objective about trying to consider both sides of any issue towards a more emotional response that tends to be less evidence and fact driven.

You are suggesting that they will have made a decision after six weeks.

They may have closed down to the idea that there are multiple perspectives on a particular issue and then turned to a more emotional response. The Scottish independence referendum had a two-year lead-in, which provided much of Scotland with that chance to learn about the issues. It is possible that it was too long. In many ways, it was all enveloping and people felt slightly exhausted by the end of the process. At the same time, the European Union referendum has proven that such a short period did not allow many segments of the electorate to fully appreciate the other side’s arguments.

That campaigning period is one of the things that needs to be considered in respect of the bill. The problems around the electorate learning about the issues must be taken into consideration in deciding what that period should be.

Before we look at how we can encourage voters to participate in a referendum, I want to ask about voter registration. Scotland’s population is at a record high, yet the Office for National Statistics produced a report in March 2019 that highlights that Scottish voter registration dropped in 65 per cent of council wards between December 2017 and December 2018. Is there a particular difficulty with voter registration in Scotland? What underlies that drop?

There is a UK-wide problem with voter registration. There has been a long-term decline in voter registration rates since the second world war. One reason for that is changing lifestyles. One intervention was the move to individual electoral registration in 2014, although it is important to note that it was introduced at a different point in England and Wales to when it was introduced in Scotland; the Scottish independence referendum, for example, was still run under household registration. It is widely thought that one of the effects of individual electoral registration has been a reduction in the completeness of the electoral register. There were some positive outcomes, including increased accuracy and reduced opportunities for electoral fraud, but research shows that young people and students in particular were negatively affected.

If we think back to 2014, the system was that one head of household—that was the phrase that was used—could complete a registration form for everyone in that property. Parents would possibly have filled in the electoral registration form on behalf of 16 or 17-year-olds who were to be included in the franchise for subsequent elections. It is estimated that of all 18 to 24-year-olds, roughly a third are missing outwith peak election season—possibly 8 million people across the UK who are either missing entirely from the electoral register or are incorrectly registered. That could be a challenge in a future referendum.

Some of what I put forward in my evidence to the committee, which draws from the “Missing Millions Still Missing” report, points to ways in which public service agencies could play a role in registering people to vote. There are possible interventions.

The situation also places considerable pressure on electoral officials, from whom the committee will hear next week. The types of pressures that we hear about include, for example, last minute spikes in registration applications. We have seen about a quarter of a million applications over the past week in the UK, just because of the mention of the words “general election”. In the run-up to the 2016 Brexit referendum, the voter registration website crashed because there was such a great volume of traffic. The electoral officials have to process every single application and check whether a person is registered. That is a potential problem.

Should we move back to head of household registration, or should we have some form of automatic compulsory registration—for example, a young person being automatically registered to vote when they get their national insurance card at 16?

It would be difficult to move all the way back to household registration, which has its advantages, although it is a Victorian system. Some automatic registration interventions certainly seem to be clear winners, including that suggestion for registration of 16-year-olds. At the moment, 16-year-olds are sent a letter that tells them their national insurance number, which is a very important moment for them. They could also at that point be at least encouraged to register to vote, but why not simply add them to the electoral register at that point, too?

We are working with electoral registration offices across the UK on that question. The national insurance approach would not work in Scotland, because the age of registration is 14 for some elections and 16 for others.

I was a member of the independent youth citizenship commission, which was convened by the UK Government and sat in 2008-09. We recommended that automatic registration be introduced. Schools could easily oversee that. It would save a considerable amount of money in the short and long terms.

There are general data protection regulation issues, because under-16-year-olds have a separate convention, so I urge the committee to think about that. In Scotland, citizenship education is a statutory subject, although it is not uniformly or universally well taught at present. In Wales, where they are considering lowering the voting age to 16 for local and national parliamentary elections, there is an opportunity for that citizenship programme to include all 14-year-olds signing on the electoral register as part of their political education. If you do not introduce automatic registration, I strongly urge you to consider that option.

You touched on political education. From looking at turnout figures for the EU referendum, it is my understanding that 64 per cent of 18 to 24-year-olds voted, and 80 per cent of 65 to 74-year-olds voted. In the independence referendum in 2014, 75 per cent of 16 to 18-year-olds voted and 92 per cent of over-55s voted. What can we do to address that discrepancy in turnout by age group?

First and foremost, when the voting age was lowered in 2014 for the independence referendum, and subsequently for Scottish local and national parliament elections, the issue of youth democratic socialisation was not considered in a developed manner at the same time. I urge that that situation be addressed.

Local authorities in Scotland provide different levels of citizenship education. About one third of young Scots take the modern studies curriculum, so they get a good level of political education, but there is clear evidence that sizable numbers of young Scots do not receive appropriate political education. So, regardless of the issue, they are uncertain about that territory.

11:30  

There is also a second question, which is not simply about 16 and 17-year-olds. The biggest drop in turnout is actually among 18 to 24-year-olds. Something problematic is happening there. Although there is not a particular issue with one group of 18 to 24-year-olds, it is true to say that it seems to be those who do not go to university who do not turn out. In that case, it is not just that there is a need for statutory political education in schools and colleges before young people are enfranchised; there is also a need for political parties to think about the spaces and places where those young voters congregate, particularly those who are not in some form of education after the age of 18, and to ensure not only that they understand the issues but that they are listened to and represented.

One of the things that has not yet really affected Scottish political culture is the fact that lowering the voting age has not caused any real change in the way in which political parties appeal to younger voters. We have not seen policy making change radically, and young people’s voices are not yet voluble enough.

Furthermore, the age profile of representatives in this august institution and in local authorities has not changed. In fact, since the voting age was lowered in 2016, the average age of members in this institution has gone up. There is something to think about there with regard to the broader way in which Scottish politics resonates with and reflects younger voters.

I agree with all that. Obviously, a political education should begin in schools. It is interesting that the bill encourages statutory bodies to promote participation. I also add that we should exercise some caution with regard to how effective they can be in doing that. Obviously, public organisations have many skills and have well-trained staff, but there might be limits to their ability to capture the grass roots. A positive contribution to reaching hard-to-reach communities could be made by having a board that includes charities and schools, or by having grants or competitions through which money could be given to small organisations so that they could have democracy champions who could encourage everyone to vote.

Our evidence is from a study that we undertook following the 2014 referendum that surveyed teachers and students in the west of Scotland. One of the clear findings was that teachers felt that the guidance that came from local councils and headteachers was quite varied across council areas and schools. Some teachers felt that they were strongly discouraged from discussing controversial political issues. Therefore, unless there is some sort of mechanism, relying on schools to take care of the sort of political education that we are discussing is highly problematic.

Some 20 per cent of secondary schools in Scotland do not offer modern studies, which means that there is a limit to the extent to which students have access to that subject. The curriculum for excellence says that citizenship education should be a part of every school year and go across all subjects. However, it was pretty clear from our survey that most teachers who responded to our survey who are not modern studies teachers felt that that was an area that only modern studies classes were supposed to handle. For example, I think that only 26 per cent of teachers felt that citizenship was something that should be talked about in the context of a history class, even though it is a good example of a subject that you might imagine citizenship could be integrated into.

It is fairly clear that political literacy is not integrated across the entirety of the curriculum, which means that you have to be careful about how you think about the issue. That is why, in our recommendations, we say that you would need to think about mechanisms—whether they are dealt with in secondary legislation subsequent to the bill or in some other way later on—that can measure the role of schools. At the moment, pupils across Scotland are exposed to different levels of information and inconsistent information, and there are obvious questions about fairness in that regard.

I accept what you have said. However, I am most concerned about 18 to 24-year-olds, whose turnout level is only 54 per cent. Obviously, there is an issue there that needs to be addressed.

I am curious about the Irish situation. Does Ireland witness differential turnouts by age?

Yes, we have a particularly acute difference between younger voters and older voters—in European data, it is second only to that in the United Kingdom. It is worth noting that by having voluntary procedures, the United Kingdom and Ireland are outliers in terms of voter registration procedures—most other European countries have automatic registration procedures, which simply removes a barrier to voting.

That said, the international research tells us that things that we do to make it easier for voters to get to the polls, such as early voting and postal voting, help just a small percentage of voters. The overall impact on turnout tends to be quite modest—just a couple of percentage points. The fundamental things that mobilise people to vote are an interest in politics and a belief that politics affects them. Unfortunately, that is a much more difficult conundrum to address.

In the United States, various experiments have been tried. Polling stations being put in shopping malls is one of the few examples that have had any sort of effect on the age profile of voters—it also usually makes a difference of just a few per cent, but it is a start. If that were to be combined with automatic registration, we would start to see slight changes to the age profile of voters.

I can add another American example. About 25 years ago, the US introduced the National Voter Registration Act, which required particular public agencies to ask people to register to vote when they came into contact with them. That is another option that could be implemented here, given that the Scottish Government has responsibility for a wide variety of public agencies. The agencies could say, “You’ve come to see us about this today: by the way, do you want to register to vote?” Such everyday conversations can make a big difference.

Thank you.

There have been a few side comments on regulation of the digital space around donations and accountability of money as it flows through the political system, and around the accuracy of information. I want to draw some of those threads together.

Dr Reidy’s submission says:

“As concerns about mis-information grow, it is likely that the role of independent electoral commissions will become more crucial in providing comprehensive information which can be trusted by voters.”

I am worried that if we rely on that kind of argument we will miss the effect of the way that digital campaigning is happening at the moment, in that it is not only having the effect of undermining voters’ trust in objective information, but is designed to do just that—it simply offers voters a choice of information to subscribe to, rather than recognising that anything is objective.

What is your view on how well the bill innovates in respect of regulating online activity? How much more scope is there for innovating in that area? Where is the balance between respecting freedom of speech and holding people to account for telling objective lies? There is a difference between expressing a contentious or contested opinion and saying,

“Turkey (population 76 million) is joining the EU”,

which is a barefaced lie—or was during the campaign. Another example is:

“Now the EU wants to ban tea kettles”.

Those are objective lies. Once they are out there in the digital space, they may be being shared by individuals, rather than by companies that can be held to a digital imprint. If all that money goes into the campaign in the last few days or weeks, the effect will have happened and it will be too late. What potential is there to engage with, and regulate effectively, the new form of campaigning?

I am afraid that there is no answer to that question.

That was my fear.

The conversation on that is being had in Parliaments across Europe. It was a particularly acute conversation in the run-up to the European Parliament elections earlier this year. The question is how to challenge misinformation and deal with the microtargeting of voters—as members will have heard a lot about in relation to Cambridge Analytica.

Online platforms pose significant challenges because of the nature of the medium. There is also the secondary issue, which is that such platforms are transnational by their very nature. That is where the real regulatory challenge emerges, because in effect you are trying to regulate companies and platforms that may exist outside the state, which is a fundamentally difficult thing to do.

Other European countries are getting at that by working collaboratively with the companies involved. Let us just gloss over that particular point. In advance of the European Parliament elections, a code of conduct was agreed between the social media platforms and the European Commission, with the quid pro quo being that if the platforms did not engage with the code of conduct, ultimately the Commission would legislate. While individual states may have particular difficulties—even the Republic of Ireland, where most of these companies are located, faces that challenge—it is easier to address such issues transnationally, and to have policies of good practice specifically in relation to the disinformation, to ensure that companies respond quickly to take down clear examples of disinformation, to deal with hate speech online and to have greater clarity around microtargeting and the ways in which companies can use people’s data.

At the moment, there is the code of practice, but European legislation will bring the issue into the legislative arena. Fundamentally, though, there will still be problems. National regulations can be created, for example on financial transparency and advertising, and it can be made illegal for somebody to fund particular types of advertising campaign. For example, we ban broadcast advertising in the Republic of Ireland. However, all kinds of broadcast ads on the referendum issue can be run on YouTube.

We are relying on online platforms to come together and decide that they will not take advertising revenue or facilitate those types of contribution. That actually happened in the abortion referendum in the Republic of Ireland. About 10 days out from the referendum, the companies came together and decided that they would not take advertising revenue. Although that was seen as disadvantaging the no side, in the end the margin of victory was so large that that was a bit of a moot point. Ultimately, you will have to have direct co-operation with online platforms, and you will have to rely on those platforms adhering to or complying with any regulations, in full awareness that they are transnational by their very nature.

I take your point that Europe-wide political institutions would be quite helpful here—that is something that I would have a lot of respect for.

Sorry about that.

No, I totally agree.

In terms of the bill, it seems to me that digital imprints almost treat digital campaigning just like a form of leaflet, which seems inadequate. Are there any other views about what scope we have in the bill to ensure that the framework for future referendums takes account of those new challenges?

What Theresa Reidy said is true. You are trying to address a moving target. The growth of artificial intelligence will make these things even more difficult. It may well be that this is less a question of regulation and more a question of education. Digital education of young people, and of citizens more widely, has been largely overlooked. In many ways, the problem with regulation is that it addresses the symptoms of the problem rather than the causes, which are that citizens themselves struggle to navigate an increasingly complex digital world. In recommending that you think about political education for citizens across Scotland, particularly young people, I would also say that digital education is very important.

I would urge the committee to consider connecting with the House of Lords select committee that is looking at the issue of digital democracy, because it may well be able to inform some of your work.

One issue that we have not covered, which we covered last week, is the Gould six-month rule. Toby James made particular reference to that in his written submission, when he said:

“The establishment of this framework would make that goal”—

the six-month rule—

“more realisable and put Scottish referendums onto a more surer ground.”

Will you expand on what you mean by “more realisable” and “more surer ground”?

11:45  

Sure. At the start, you stated that one of the aims of the bill was to provide flexibility, but it will also provide certainty about the legal framework, not just for electoral officials but for candidates and citizens. Research has shown that one of the challenges that electoral officials face has been the variety and complexity of electoral legislation and legislation on referendums. Also, in some cases, the legislation can arrive very late on. The Gould principles arose from problems in 2007, when legislation was late. The bill is an important step forward because it provides foundations. That is why I mentioned that flexibility can be a problem in how referendums are run.

Obviously, the question here is whether six months is enough. What about any statutory instruments or regulations that follow from the bill? You will hear evidence from electoral officials in due course, but the research that I undertook with Dr Clark indicated that they always prefer certainty—they prefer to know the date of the referendum and what budget is available. If anyone has ever tried to organise a wedding, they will know that it is very difficult to get the venue in place. It is exactly the same for elections. They tend to be the most complex logistical task to undertake in peacetime. The more information that can be provided to electoral officials, the better.

The similarity between weddings and elections is going through my head. It is probably the same. I should probably not ask the question. I will avoid it.

Just leave it.

Thank you, deputy convener, for counselling me properly.

If there are no further questions, I thank the witnesses for a helpful, interesting session.

Meeting closed at 11:47.