Local Government and Communities Committee, 21 Nov 2007

Meeting date: Wednesday, November 21, 2007

Official Report 284KB pdf

---

“Scottish elections 2007”

I reopen the meeting. The fourth agenda item concerns the Electoral Commission report "Scottish elections 2007: The independent review of the Scottish Parliamentary and local government elections 3 May 2007".

We have with us Ron Gould, who conducted, on the Electoral Commission's behalf, the independent review of the electoral process for May's Scottish Parliament and local government elections and the problems that arose from that. We give him a warm and genuine welcome. He got up very early this morning, as he is in Ottawa, five hours behind us. He told us earlier about the first of the winter snows in Canada, so we are lucky to have him.

Mr Gould, would you like to make a short statement before we move to questions?

Yes, thank you very much. I thank you for giving me the opportunity to discuss my report of the review of the 3 May elections and my comments on options and recommendations on the local government elections.

As you may be aware, my terms of reference included the impact of the new single transferable vote system for the local government elections, as well as related matters such as the new ballot marking system and electronic counting. I had great flexibility in putting the report together, but from the outset I made two exceptions: I made it clear that the review would not explore the outcome of the election—including the validity of the results, which were already complete and had been accepted by stakeholders—nor take any actions that might threaten the secrecy of the vote.

I stress that my objectives throughout the review were to examine the background of each of the issues that were identified, describe the key related problems and suggest options and make recommendations that would minimise or avoid similar problems and would strengthen the Scottish election process for the future. In other words, I was looking ahead as opposed to looking back.

I would welcome any comments or questions that you may have and will do my best to answer them as comprehensively as possible.

Thank you for that short statement. You said in your report:

"Almost without exception, the voter was treated as an afterthought by virtually all the other stakeholders."

That is a clear criticism of political parties for putting their own interests before those of the voters. There was a great focus on that by the media and politicians. Do you regret that focus? Did it prevent the wider recommendations of your report being presented properly?

Following the release of my report, I was somewhat disturbed by the fact that, in my view, some of the media and some of the debates that occurred were looking back and trying to apportion blame and assess who was responsible for what, rather than taking the problems that arose as the starting point for resolution and avoiding them in future. I was concerned that the report would be used as a vehicle for debate about who did what to whom rather than a vehicle for positive progress. That was one of the reasons for the explanatory letter that I felt was necessary. In reality, it is impossible to assess blame because, although many people were responsible for many things, no one had ultimate responsibility, so the buck stopped nowhere.

Good morning, Mr Gould. In your report, you comment that

"as long as the responsibilities for the decisions which have an impact on the Scottish parliamentary and local government elections are divided between the Scotland Office and the Scottish Government, it cannot be guaranteed that these electoral processes will be conducted effectively, due to the fragmentation of the legislation and decision-making in this context".

In view of that, would it be more appropriate for the Scottish Parliament to have powers transferred to it so that it can take greater control of the electoral process in Scotland?

I have been trying to avoid getting into discussions about who would be responsible for the legislation. It would certainly be useful to have one focus for all legislation. However, it is worth while to consider the election process as having two major components, the first being the legislative component and the second being the operational component. In my view, and when talking about the chief returning officer, the operational component has to be in one place and run by one body or one individual. If the legislation—or the responsibility for legislation—is split between two Parliaments, it needs to be worked into and co-ordinated with the operational side. Whether elections are devolved or combined, the legislation could still be split, and the responsibility for the application of that legislation united in one location.

Would it be beneficial, in order to focus on specific elections, if those were decoupled, as was recommended by the Kerley report at the beginning of devolution?

If it is felt that local issues and local candidates are important objectives, combining elections would defeat those objectives. If you really want the public to recognise what is needed at the local level in the local elections, then the elections need to be separated by years, months or days. As soon as you combine them, the parliamentary elections take over the spotlight.

Obviously, as the Local Government and Communities Committee, local government is our central focus. It seems to have been forgotten that the number of spoiled ballot papers in the local authority elections was three times higher this year than in the previous election. According to a report from Unlock Democracy, about 26 per cent of people it recruited as volunteers to monitor elections were unsure about how the single transferable vote worked. Are you convinced that voters knew how STV worked? I understand that you have looked at some of the ballot papers. It would appear that a lot of people went into the voting booth and put only one X, not realising that they could put their first, second and third preferences. Indeed, if two or three candidates from one political party were standing, some people may have put two or three Xs, which unfortunately invalidated their votes. Will you comment on those issues?

In the new system, the parliamentary ballot seemed, for various reasons, to be better understood. Ironically, however, the percentage of rejected ballot papers for the local elections was much lower than for the parliamentary election—both regional and constituency. Obviously, people understood how to mark the ballot. There were errors, and the error rate was higher than it had been in the past. However, the Cragg Ross Dawson report "Ballot Paper Designs for Scottish Parliament Elections 2007", from the Scottish Executive, raises the question whether voters understood that they could mark 1, 2, 3, 4, 5 and so on for all candidates, whether they felt restricted or whether they had other concerns. As it was the first election of its type, it is quite possible that people did not take full advantage of the system for marking the ballots that was available to them. A lot more voter education and information could be provided in advance of future elections, so that people are more comfortable with the process.

At the count, I noticed that in cases in which there was more than one candidate from the same political party, there was a high correlation between the surname of a candidate—and hence their position on the ballot paper—and how well they did in the election. In my constituency, I noticed that, regardless of a candidate's political persuasion, the higher up the alphabet they were, the more likely they were to be elected. Even when candidates from the same party were grouped together because they had similar surnames, the person whose name was higher up the ballot paper got more votes. The greater the distance between candidates' surnames, the greater the difference in the number of votes that they got. For example, a candidate for the Scottish National Party who was called Alasdair Allan was more likely to get votes than a candidate for the SNP who was called William Wallace. How concerned are you about that? Will you expand on your recommendation to have a lottery to determine the position of candidates' names on the ballot paper?

Yes. It is recognised internationally that the higher up the ballot paper someone's name is, the greater their chance of getting more votes. In other words, the top of the ballot paper is the most preferential position. There is also a possibility that, through advertising, a candidate can obtain a preferential position at the bottom of the ballot paper. The alphabetical listing of candidates or parties tends always to ensure that there is a preference for those candidates or parties that are higher up the alphabet and therefore higher up the paper.

A number of approaches to resolving the problem can be considered. Probably the fairest approach is to hold a public lottery once a decision has been made about whether to list candidates' names or parties' names first on the ballot paper. At present, if the party name appears first on the ballot paper, it is advantageous for the party name to be higher up the alphabet. If the candidate name appears first, those candidates whose names are in the early part of the alphabet have the advantage. In the case of the most recent council election, the candidate name appeared first. If a lottery had been held to determine the position of candidates' names, every candidate would have had an equal chance to obtain a preferential position on the ballot. The same would apply to a lottery to determine the position of party names.

There are several other possibilities. The parties showed an interest in grouping their candidates together on the STV ballot. If it was agreed to group the candidates by party, each group of candidates would appear in whatever position on the ballot that team won in the lottery. In effect, the position of party names on the ballot would be drawn by lottery.

Another approach that can be followed is to randomise the printing of the ballots so that different ballots will have the candidates and/or the parties in different positions on the ballot. In my view, that is difficult politically for the parties, because it means that they cannot use advertising that tells people to vote for the second or the bottom or the top option on the ballot, for example. Both the alphabetical listing of candidates or party names and the use of a lottery mean that the parties know exactly where individual candidates or groups of candidates will appear on the ballot.

Welcome. You said that the buck stopped nowhere. If we agreed that the two elections ought not to take place on the same day, the matter would be resolved. If responsibility for the Scottish Parliament elections remained with Westminster and responsibility for local government elections remained with us, but the elections took place at separate times, we would no longer be faced with the complexities of bringing together the two elections. The issue is not that Westminster and the Scottish Parliament have different areas of responsibility, but that they have to work together to deliver their responsibilities.

Separating the two elections would minimise complexity, and many of the problems that arose this time would be avoided. On the other hand, the management of both elections involves the same responsibilities, mechanics, knowledge, skills, background and training. That is why it would be in everyone's interest to have one chief returning officer handle both elections. That officer would answer to whichever jurisdiction was responsible for the election concerned. Returning officers, too, should be more professionalised and should be responsible for all elections and all aspects of those elections, if possible. That is preferable to having as returning officers individuals whose main jobs are as executive officers and who serve incidentally as returning officers because an election has been called.

So where jurisdiction over the election lies is not the issue. It is possible to have the clarity and accountability that you seek and the structure that you have described regardless of whether Westminster remains responsible for the Scottish Parliament elections. Such a structure is not predicated on changing where responsibility for the elections lies, as has been suggested. We do not need to bring responsibility for all elections to the Scottish Parliament in order to get clarity.

Yes. There would be no question as to who had the authority, was responsible and was accountable at each level. That would be very clear, right up to legislative level. As things stand, decisions about local government elections lie with the Scottish Government and decisions about legislation and policy for Scottish Parliament elections lie with the UK Government.

I was interested in your solution for making the ballot paper more accessible—having a public lottery to determine the position in which candidates and parties appear. Have you done any work on literacy issues and on people's ability to understand the ballot paper? People with basic literacy learn alphabetical order as it helps them to locate information. Did you do any work on the disadvantage that could be created for people who have problems understanding the ballot paper? Would grouping candidates by political party clarify the situation for people, especially where parties have put up more than one candidate? Anecdotal evidence suggested to me that a large number of spoiled ballots could have been related to parties having more than one candidate; people put down three crosses rather than one, only one of which was counted. Did you do any work on how the ballot paper should have looked with those who understand the particular needs of people with literacy problems?

That is an interesting question. You raise a fundamental issue. Every proposed change to any stage or aspect of an election process must be tested in order to determine what confusion it may cause, as the approach may not be traditional, and to determine the easiest way in which voters can understand the story. Once proposed changes have been examined, the approach that is determined must be backed up by a lot of voter information and education.

An alphabetical listing is not confusing if used in one way, but becomes confusing when it is used to vary the approach. I give as an example the use of party descriptions on the regional ballot papers. There was so much flexibility that different approaches could be taken, including the use of the party name. That caused confusion for voters. Such an approach may be useful for positioning on the ballot paper, but it certainly does not take into account the voters' interests. If candidates are grouped by party—in local government elections, for example—the determination should be that the voters' focus should be on the parties. In other words, the campaigning should be by party, and people will focus on party interests being most important as opposed to individual candidates. In that case, the parties would probably appear in alphabetical order and candidates would be grouped by party. There can be variations on the theme, but it is important to determine well in advance what the voter's focus will be, the most logical approach from the voter's point of view, and what the voter will easily recognise when they vote in the polling booth.

Is it fair to say that those who promote STV systems do so on the basis of fairness to political parties? Therefore, given that the test of the system is the extent to which it results in fairer shares of votes for the political parties, it could reasonably be expected that the ballot paper should give parties a place, as opposed to pretending that a party election was not taking place.

I am not sure of your question. In a single transferable vote system, votes are, obviously, transferred from candidate to candidate as opposed to from party to party. One issue that was raised with us during our consultations was that there was a pattern whereby, when a party had more than one candidate on the ballot paper, the candidate at the top of the list would always have the favourable result and the second candidate would pay the price. It appears that if the party focus is important in local government elections, party candidates should be grouped by party name.

Good morning, Mr Gould.

Will you say more about the Cragg Ross Dawson research or the test driving of potential ballot papers? I understand that that research was commissioned by the previous Executive in September last year and involved four trials of 25 ballot papers. That seems to me to be wholly inadequate. Perhaps the poor quality of the research was one reason why the number of spoiled papers tripled under the single transferable vote system in the new elections. Indeed, the Local Government and Transport Committee acknowledged the problem in December last year.

You said earlier that the buck stopped nowhere. On a more positive note, rather than talking about the failures of the previous Executive, we need to make sure that any future governments get it right. If we had a chief returning officer, would the buck stop with them? Would they be independent of the Scottish ministers but accountable to the Scottish Parliament?

I will deal with the Cragg Ross Dawson report first. As you are well aware, there were two reports, one of which was conducted by the Electoral Commission, and the second of which was done for the Local Government and Transport Committee and the Scottish Executive.

The Cragg Ross Dawson report was on the STV ballot and, interestingly, it suggested that the clear overall preference was for alphabetical listing by party name rather than by candidate, and for the party name to be in an additional column to the left of the candidate's name. Although the listing resulted in the party name being to the left of the candidate's name, the listing on the ballot was by candidate name.

I am not sure how that evolved after the Cragg Ross Dawson report but, given the lateness of the decisions and the technical difficulties, I assume that the contractor might not have been able to do what was asked. I am aware that, because of the late decision, the contractor was unable to group candidates across from the party names as was recommended in the report.

I have some concerns. Despite the fact that, statistically, 100 people is adequate, given a brand new system and combined ballots and so on, the work that was done did not consult enough individuals, and the results were not followed up or tested sufficiently.

On the question of accountability, if you give a manager the responsibility and authority to carry out all aspects of the operational side of the election and apply the legislation, you can focus on them and the buck will stop with them. That individual—I am talking about a chief returning officer—will have authority over the 32 returning officers who, at the moment and on paper, have responsibility for all aspects of the operation in their jurisdiction, but have no control or authority. On many issues, responsibility and power is out of their control, so it is difficult or impossible to hold them responsible. With a chief returning officer, the responsibilities of the 32 returning officers would be more clearly defined, they could be held accountable, and they would be much more comfortable in such a situation.

In your experience, do STV systems of voting, such as the one that we had for the first time in our local government elections, tend to produce habitually a higher level of spoiled or wasted ballot papers than a traditional first-past-the-post system does? Is the fact that we had three times as many spoiled papers simply a function of the novelty of the system, or is it the result of design failings or lack of voter information?

I do not have statistics on that at my fingertips, but my impression and understanding is that STV systems do not necessarily have higher rates of spoiled or rejected ballots. One area in which that matter could be determined or checked quickly is Northern Ireland, which has had STV for at least two elections—I have attended two such elections there. Again, we return to the point that the problem with STV in the elections in Scotland was a multiple problem, arising from the combination of a traditional system with which people were familiar, a new system that arrived late and a ballot that was defined very late in the process, which meant that the hard-nosed training and voter information was not as comprehensive as it should have been. My expectation is that, at the next election, whether it is combined or decoupled, the rate of rejected ballots will drop significantly.

Is there a standard that we should aim for? For example, if the rejection rate is 0.6 per cent, with whatever system, does that mean that by and large voters understood the system and there were no problems with design or information? Is that the sort of level that you aim for? If the figure was 1 per cent, would that be a problem? The rate was 1.8 per cent for the local government elections and 4 per cent for the Scottish Parliament elections. What standard should we set for the results that the system should achieve, to allow us to judge the quality of the system?

There is an existing standard in Scottish elections. Looking back to 2003, the rate of rejected ballots for the local government election was about 0.64 or 0.66 per cent. In the parliamentary election, the rejection rate for the regional and constituency ballots was about 0.64, 0.65 or 0.66 per cent. Less than 1 per cent is a normal acceptable level.

I want to follow up on the questions about the Cragg Ross Dawson research findings. The findings on design aspects of the local government ballot paper—in particular, that of listing by alphabetical order or by party group—were discussed in a parliamentary committee. As a member of that committee at the time, I am not aware that spoilage was a pertinent issue in those discussions. Similarly, I am not aware that the discussions on the research on the Scottish Parliament elections focused on spoilage. However, as your report states, the research suggested that a 4 per cent spoilage rate would occur, which is in fact what happened in the Scottish Parliament elections. However, the spoilage rate did not seem to be an issue that came out of that research. Is that a fair comment?

Yes. Interestingly, the Cragg Ross Dawson study for the Scottish Parliament election, which involved 100 people, found a 4 per cent spoilage rate, which equated to the reality of the election rejection rate. However, a 4 per cent spoilage rate out of 100 people is four people. The Cragg Ross Dawson study for the local government elections commented that "only a handful" of people spoiled their ballots. In my view, a handful is four or five, which again is 4 or 5 per cent. In both cases, Cragg Ross Dawson underrated the importance of that level of spoilage, because it used such a small sample in the first place. That was not pursued by either the Scottish Executive or the Electoral Commission, because Cragg Ross Dawson did not express a high level of concern about it.

But, presumably, Cragg Ross Dawson used that size of sample because that is what it was resourced for and commissioned to do. In fairness to Cragg Ross Dawson, it might not have appreciated that the result was statistically significant, as opposed to statistically insignificant, because it did not have enough numbers to work with. Is that fair?

I would not presume to try to figure out Cragg Ross Dawson's rationale, or whether it was contracted to take a sample of only 100 people or whether that was its decision. My understanding is that, from a statistical point of view, 100 people is a reasonable sample. If that is the case, the rejection rate should be a reasonable area for concern. I cannot out-guess Cragg Ross Dawson's thoughts.

So, you think that the rejection rate, even based on the sample that was used, should have been given more prominence in the findings of the research reports and that those who received the reports should have paid more attention to it.

Certainly, in considering the papers for the Scottish Parliament elections—I know that the rejection figure was 4 per cent—the Electoral Commission, which is supposedly a professional organisation, should have said immediately, "There is something wrong here. We'd better take a closer look." From that point of view, the Electoral Commission was remiss. As I said, I do not have the exact figures for the local elections. However, it is a matter of concern in either case.

Thank you for joining us, Mr Gould. Will you comment further on the potential benefits for the local elections of decoupling local and national elections? It has been suggested in some quarters that if the two elections were decoupled, the argument for having both elections administered by the same Government would lose strength. Do you think that the other side to the argument is that, depending on what electoral cycle is chosen, the potential for one election to catch up with the other, or for extraordinary general elections—not that we plan to have any imminently—to intervene means that the two elections could clash? Does that add strength to the argument for the simplicity of having two elections administered by one Parliament or Government?

Let me give an answer in two segments. First, the main reason that I recommended decoupling the elections was to give deserved recognition to the importance of local government elections. Technically and operationally, decoupling is not necessary; with some modifications to approach and management, the two elections could be run smoothly and without the major problems that were encountered this time. There are options for that. However, holding an election is not just about getting the vote out of the way; it is about getting the message across. I do not believe that the local government message can get across when the local government elections are competing with the parliamentary elections.

Secondly, the running and management of an election has to be done locally. A Scottish election cannot be run from Northern Ireland, Wales and London; it has to be managed and run in Scotland. Whether the elections are to Parliament or to local government, the election process must be managed in Scotland.

Who manages that process and how is that done? The returning officer has always been responsible. Sometimes that role has different titles but, often, those titles are held by the same person. We need people who know how to run an election. Why should two different groups of people—two bodies—manage an election process? That makes no sense and is costly and unwieldy. If we are talking only about running an election, which means applying the law of whichever jurisdiction is responsible, it does not matter whether the same person or group is responsible. What is important is that those people are professionals and can manage the process.

Even with decoupling, the management of elections should rest with and in Scotland. The direction of the groups that are involved and the financing of an election would rest with the body that was responsible for the legislation for that election and which had the legislative authority. In a nutshell, that is how I see the situation.

You say that Scottish elections cannot be run from London. Does that apply to legislation as much as to administration? In principle, would it help if returning officers did not attempt to serve two masters?

I do not want to get into the political aspects. What is important in electoral management and operations is communication and consultation, as was the case for the 2007 elections, on which close consultation and co-operation took place between the two Governments that were involved. That is essential for the future. Devolution—the transfer of legislative authority—is for negotiators on both sides to work out in deciding what is in Scotland's best interests.

I seek further clarification of points that have been made. Decoupling could have unintended consequences. When you made the recommendation on decoupling, was its impact on voter turnout considered? I understand that we do not know exactly what would happen, but we know that local government elections had a lower turnout historically and that they have benefited from being combined with an election that attracts a higher turnout.

Was any consideration given to voter fatigue as a result of the Scottish Parliament, local government, European Parliament and UK Parliament elections? How would we arrange the cycle of elections? I have a couple more questions, but I will ask them after your response to those questions.

The voter turnout question dogs almost every jurisdiction—voter turnout has dropped pretty well worldwide. My feeling about voter turnout concerns quality versus quantity. If people do not want to vote, and if they do not feel that the issues, the candidates or the parties are important, how important is it that the number of voters keeps going up?

There is an easy solution in the Australian approach. It is mandatory for Australians to turn up at the polls. They do not have to vote, but they have to turn up on polling day or exercise their votes in other ways, otherwise they get fined. If high turnout is important, that is a simple solution. However, I think that it is more important to stimulate and encourage voters to understand the importance of the issues, the candidates and the election, and to get them to turn out as much as possible.

There is another interesting element. When we think of voter turnout, we think of the turnout at the polls. However, as I mentioned, there has been a decline in the turnout at the polls as well as in total voter turnout. If you look carefully at the turnout figures in Scotland, you will see that the turnout for the postal vote increased to 11 per cent of the total number of voters. If the total number of voters has declined, that means that the turnout at the polls is declining whereas advance polling is increasing. In my view, if turnout is a concern, the focus should be on advance polling. We made a couple of recommendations on that in our report. The first was to have advance voting at the office of the returning officer; the second was to have voting in advance polls in shopping centres, and so on. In those ways, voter turnout will be increased.

I considered the question of voter fatigue and suggested that, if the elections were decoupled, there should be a two-year gap between the local government election and the parliamentary election. The issue of voter fatigue also applies to the interest of the voters in the issues at any particular election and in the candidates, who may be popular or controversial. Some of the onus is on the parties to encourage the voters to turn out. It is not unrelated to the question of voter turnout and the initiatives that can be pursued, such as advance polling and voting at the office of the returning officer, while continuing with postal voting. Any such initiative that facilitates voting for the voter will increase turnout, or at least maintain the present voter turnout.

Thank you—we will have an interesting debate on compulsory voting and how a two-year gap between elections could be ensured. Currently, that would allow local government a six-year term, which would be unprecedented. Is there any way in which we could manage that?

I suggested the two-year gap to avoid voter fatigue, but the gap could be one year, three years, a month or two days. The present situation of local government needs to be taken into account. Will it have a two-year term or a six-year term, or a seven-year term or a three-year term? There are all sorts of permutations and combinations regarding the transition. As I understand it, the parliamentary elections are pretty well fixed at the moment. In the report, I have thrown out a number of considerations and options recognising that the reality must be dealt with locally, not in a review report.

There is a concern that, in terms of the STV vote for local government, the threefold increase in failed ballot papers could be the tip of the iceberg. Some of us are concerned that the system might have masked the level of failure in that ballot. You and your inquiry team had access to those ballot papers. Has any work been done to establish patterns that would reveal underlying problems to do with an unidentified failure arising from people not understanding what to do? Papers with one X went through the system, but that would not be publicly indicated as a failure.

On the service that is provided by the chief returning officer and 32 returning officers across Scotland—the service is carried out at the moment by the chief executives of local authorities—do you think that running elections should be a full-time job? How far would you go down the road of professionalisation of that service?

On the STV ballot, the review examined only the face of the rejected ballot images. We have no statistics or information concerning any valid ballot. Of course, any voter who placed one X on an STV ballot had a valid ballot paper, so that would not appear in our analysis.

I think that it can be expected that a number of the valid ballots were marked with an X or, perhaps, only one number. As I mentioned earlier, the system and approach were brand new. Voters were unfamiliar with what to do and might not have taken full advantage of the system. That might have been one of the factors that affected the results of the STV ballot, and I suggest that a lot more voter education and information should have been provided for that count. Serious consideration should be given to making much more active use of the information officer for the next round of STV elections, to reinforce the approach to marking the STV ballot and ensure that full advantage can be taken of that system.

On the question of full-time returning officers, I do not want to jump to conclusions. However, if there were a review and the responsibilities of the returning officer were to be expanded to include all electoral functions, including the work of the registration officer, responsibility for polling stations, polling divisions and so on, there would probably be a strong argument for having at least one individual in each council who was a full-time returning officer, who would be responsible only for that role.

I think that the actual level of confusion in both elections was masked by the fact that there were officers from each local authority in the polling places to advise people. I am concerned that, if those officials had not been there, the level of confusion among voters would have been significantly higher. My understanding is that the percentage of spoiled ballot papers in STV elections in Northern Ireland, which has had the system for a number of years, is routinely in the order of 3 per cent.

You talked about turnout, Mr Gould. I am interested in what you say about that. Australia has more elections than we do and compulsory elections ensure that people there take more of an interest in politics.

I want to talk about the confusion that occurred before voters got to the polling stations. We have all seen the ballot papers with umpteen names on them—up to 23 in some areas. In Scotland, three people can register as a political party, they get a free leaflet delivery from the Post Office and they do not have to pay a deposit. That means that voters get perhaps 15, 20 or 25 leaflets through their letterboxes. Does that encourage voter fatigue and confusion and perhaps stop people going to the polling station in the first place?

I believe that the involvement of council staff was highly important and that they played a key role in supporting the returning officers in virtually every case. I continue to argue that council involvement with and support of the returning officer, whether he or she be full time or otherwise, is essential.

The problem is that, in some cases, the returning officer is the chief executive of the council, who has a full-time job and must either delegate that job to someone and abandon it for a period of time, or delegate the returning officer functions to somebody else. It is much more important to respect the process by having the chief executive of the council do the job for which he or she was hired and having a professionalised returning officer, preferably full time, as part of the support to that executive officer and the council.

You mentioned the number of candidates. That brings the criteria for being officially registered as a political party or a candidate into the discussion. On the one hand, we must give credible new parties the opportunity to emerge; on the other hand, opportunities for the system to be taken advantage of for personal gain or interest, as opposed to valid political development, should be restricted. Most countries face that dilemma.

It would be advantageous to take a hard look at the current legislation on being officially registered and on the ballot for local government elections. It would also be advantageous to explore some of the alternatives, such as refundable deposits, a certain level of verified signatures and other ways of presenting the party's mandate and platform. It is not easy to make a determination on the matter, but all sorts of options can be explored and there are ways of tightening things up without prejudicing the emergence of new and valid parties or candidates.

Several months before May's ballot, several other members and I took part in a trial for electronic counting. As the system was new to Scotland, we queried certain issues, but the trial, on the whole, seemed to be successful. However, despite assurances that it had worked well elsewhere, the system in most areas ground to a halt on the night and did not cope at all well with the volume of votes that it had to deal with. In your view, what were the main problems with the e-counting process, how should they be addressed, and should votes in future be carried out electronically?

On your last question, under STV, applying the formulas for allotting seats is a very complex business and you would not want to have the kind of manual counting system that they have in Northern Ireland, where the form of STV that is used is much less complex than the one that is used in Scotland. As far as STV is concerned, electronic counting is important as it can allow results to be arrived at much more smoothly and quickly than any manual count can manage.

However, most of the problems with the electronic counting system in the Scottish Parliament elections can be attributed to the lateness of getting it into place, which resulted in last-minute decisions and alterations to the count—for example, in the parliamentary ballot in Glasgow and the Lothians—that affected the system. Moreover, the link between the ballots cast at polling stations and the postal ballots was out of sync. The postal ballot packs also slowed down the whole process because the papers had to be folded. The scanners could not accommodate the fold and, in some cases, the papers jammed the system.

The count went down in some—though not all—locations but, overall, the electronic count had enough successful aspects to warrant its continuation at the next election. However, that does not mean that the same contractor would be ideal for the job or that the same approach or type of scanner would be appropriate. The system needs to be examined very closely and tested much more.

Other electronic counting systems have had the same difficulties and caused the same headaches the first time they were used, but those issues have been resolved. I know, for example, that some time ago London Elects experienced many of the problems that were experienced in the Scottish elections. However, many of those have now been resolved. As I said, I believe that the electronic count is important and valuable and can work smoothly. Now that there is enough time, the systems can be tested thoroughly based on known ballots and known structures, including the postal ballot.

Another element that I want to inject into the discussion—this was not in your question—is that I have a serious problem with an electronic count that requires the secrecy of the vote to be violated. An electronic count that requires voters to carry their ballot open from the place where they mark it to the ballot box raises a serious problem. Even though people were perhaps encouraged to turn the ballot paper upside down, I gather that they had to turn it right side up to put it in the box for the scanner. It should be possible to bring the ballot paper to the ballot box without its being visible, regardless of what the individual does. In my view, there are solutions to that problem even without requiring people to fold the ballot paper. That whole aspect needs to be examined without abandoning the electronic count.

In our examination of the process, we have not so far acknowledged the obvious disadvantage that arose from the fact that one X on the local election ballot paper did not constitute a spoiled ballot whereas three Xs did. It seems to me straightforward that that resulted in a disadvantage to those parties that put up more than one candidate in any ward. Will further work be done on that issue? Do you accept that more problems arose than have been revealed by the number of spoiled ballot papers because candidates were at a disadvantage if their party put up more than one candidate? I would welcome your comments on that.

A second disadvantage concerns the information officers and the education programme. From my recollection, there was a correlation between the level of spoiled ballots and the level of deprivation in communities. There were higher levels of spoiled ballot papers in places such as Glasgow than in other parts of Scotland. If that is the case, what recommendations would you make about the nature of voter information, given that it is not possible to assume the same level of education across the board or to assume that everyone can absorb information in the same way? Do you accept that such a disadvantage arose? Did you do any work with information officers? Anecdotally, I heard from my constituents both that the information officers were helpful and that they were singularly unhelpful. I do not know what the definition of the information officer's role was, but it certainly did not seem to be about providing information to voters. Did you do any kind of sampling on that issue and speak to those who carried out that function?

I said that voters were an afterthought, but so were the information officers. They were a last-minute desperate attempt to try to resolve the problem of lack of voter information and education, especially on how the ballots should be marked. As a result, voter information officers were not appointed according to standard criteria. That is one of the recommendations in the report. As a result of their wide variety in age and competence, the role that information officers played varied from invaluable to non-existent, according to the reports that we received. The information officer can definitely play an extremely important role in remedying the problems that arose this time round.

The University of Strathclyde analysis indicated that, in disadvantaged areas of Glasgow, the rejection rate for ballots was much higher. Our findings were similar. That higher rejection rate was perhaps partially due to disadvantage and obviously partially due to the fact that, in Glasgow and the Lothians, the arrows had been removed from the parliamentary ballots. The challenges in those areas can now be recognised, certainly with regard to training. The standard voter information programmes and the roles of the information officer could be enhanced.

One rather interesting thing that a few people said in interviews was that the Scottish voter is not accustomed to having people tell them how to mark their ballots and vote. A number of people did not take advantage of the information officer and perhaps made errors on their ballots that they would not have made if they had done so. There could perhaps be a focus on support for voters and on encouraging them to seek that support and clarification, to ensure that what they do does not result in their losing their vote.

You have spoken about challenges that we have to deliver on. You spoke earlier about having a chief returning officer for Scotland and the professionalisation of the 32 returning officers. You have said that the system cannot be run from Northern Ireland, Wales or London; you said that it should be run from Scotland. You were careful not to delve into politics, and I want to ask you your opinion as an independent, non-partisan academic.

On page 111 of your report, in section 9, "Conclusions and recommendations", you state:

"it cannot be guaranteed that these electoral processes will be conducted effectively, due to the fragmentation of the legislation and decision-making in this context. As a result, we would recommend that exploratory discussions take place with a view toward assigning responsibility for both elections to one jurisdictional entity. In our view, the Scottish Government would be the logical institution."

Could you clarify for us that that is the view that you reached as an independent academic?

Yes. Let me move back for a moment. Earlier, I separated out the management of the election from the legislative responsibility for it. On the management of the election, I would argue strongly, for operational, academic, political or whatever other reasons, that one management team should be located in Scotland and should manage the election from there.

Issues around responsibility for the legislation then arise. When I mentioned the "fragmentation of the legislation" in my conclusions, I was referring to the fact that the Westminster legislation—although I say this without regard to where the legislation is—is so fragmented, as is described in the legislative section of the report, that it needs to be consolidated. The fragmentation that I was referring to was not necessarily fragmentation between the Scottish Government's responsibility for certain legislation and Westminster's responsibility for other legislation. That is not what I was focusing on or referring to; the fragmentation that I was referring to was in the context of all the Westminster and Scottish Government legislation that now exists. It needs to be pulled together and rationalised for the purposes of running an election. The micromanagement that tends to occur, and which occurred at the last election, ought to be minimised.

You said at the beginning that you did not want blame to be attached. I think that that is appropriate. What do you believe the balance of responsibility between ministers and political parties should be for future elections? Do you accept the analogy that my colleagues have drawn that it is about as sensible to run a Scottish parliamentary election from London as it is to run a Canadian election from Washington?

For many years, while I was deputy head of elections for Elections Canada, I also chaired a political parties committee. Election managers are often accused of a lack of political sensitivity, and I agree with that. It is critical that there is a continuing input and a flow of communication between the parties' representatives, wherever they may be located, and the chief returning officer or whoever else is managing the elections. When I talked about not micromanaging, I was referring more to the legislation. With regard to the decisions by election managers that have an impact on political parties, the parties need to bring the political reality to the administrators and say, "If you make that decision, this is what the impact is going to be—it is not a realistic decision." The parties are critical in every step of the process—I will leave it at that.

As the convener, I have the privilege of asking the last question—I think that I have just established that precedent. It is 16 minutes past 2 now, and we assured you that we would be finishing around that time. I opened the session by reading out your words, and it is important that I also close on that. There is a big challenge for politicians in Scotland. You reminded us of that by stating:

"Almost without exception, the voter was treated as an afterthought by virtually all the other stakeholders".

I believe both that it will be difficult for politicians and that our future electoral systems and processes are too important to leave to our politicians. In taking forward your recommendations, how do we ensure that the needs of voters are met? How should we consult beyond the normal channels of communication? How do we achieve the objectives that you have given us through your recommendations concerning the voters?

I hope that there will be some sort of a mechanism—a committee or round-table—to examine each of the aspects of the election. For example, there was a huge debate over the design of both the local election and the parliamentary ballots. It is important that the parties play a role in that, but the various options for party preferences need to be clearly defined, as does the rationale for why the party feels that candidates should be grouped this way or that. That needs to be linked to whether it is technically possible—if we assume that there is an electronic count—to do that on a ballot. If it is, the impact on the electorate must also be considered—does the voter find it easy to understand, or do they find it more logical to approach the paper by candidates and not groupings? That is where the intensive work needs to be done.

There should be three stages for the consideration of the various positions and approaches of the parties, or other interests—such as an election manager—concerning, for example, the design of postal ballot packs. The first is party preferences, the second is the technical and managerial realities of carrying out each of those preferences and the third is finding out what the voter thinks—how it impacts on the voter and how easy it is for the voter to understand. That research needs to be very comprehensive and not just cover 100 people.

Thank you for all the time that you have given us today. We ask you to pass our thanks to the University of Ottawa, which has helped us to facilitate the sessions. We found the session worth while, and we hope that you have a safe journey home.

Thank you—it has been a pleasure for me. If there is anything that I can do to assist the committee or to support its work, please do not hesitate to let me know.

We welcome that offer.

Meeting closed at 14:20.