Local Government and Transport Committee, 09 Dec 2003

Meeting date: Tuesday, December 9, 2003

Official Report 258KB pdf

---

Local Governance (Scotland) Bill: Stage 1

In today's meeting of the Local Government and Transport Committee, we will continue our evidence taking on the Local Governance (Scotland) Bill. I welcome our first witness, Professor David Farrell, who is from the department of government at the University of Manchester. Members should have a copy of Professor Farrell's submission, which sets out some of his views on the single transferable vote.

Before I invite Professor Farrell to make his introductory remarks, I put it on the record that we have received apologies from Michael McMahon, Bruce McFee and Iain Smith. Iain Smith's substitute is John Farquhar Munro—I ask John to confirm that he is here to act in that capacity.

That is a fact. I am deputising for Iain Smith.

I also welcome Tricia Marwick. Although she is not an official substitute, she is most welcome to participate in today's proceedings.

I invite Professor Farrell to make his introductory remarks.

Thank you for the invitation to give evidence to the committee—I am grateful to be involved in a small way in the process. I do not have an awful lot to add to my written document. From what I have seen from the committee's extremely detailed documentation, much of what you are trying to design is modelled closely on the Northern Ireland system, which is probably as good a version of STV as possible. As I tried to make clear in my written presentation, STV systems can vary in all sorts of great and wonderful ways. However, if one had to pick a system, the Northern Ireland one is probably the best. That is all I that have to say at this point, but I am happy to take any questions that the committee may want to throw at me.

I will start off with a specific question about the redistribution methodology that is proposed in the bill—I think that you refer to it as the Gregory method—by which only some of the votes are redistributed from second and further preferences. From our evidence last week, it seems that the system is complicated and not comprehensible to the average person. The system also seems to go against the philosophy that is used to argue for STV that everybody's vote should count in the same way. If we use the proposed method, everybody's vote will not count in the same way, because some votes will count more than others. What is your view on that method compared, for example, with the method of full redistribution?

As you can imagine, the issue may take us into some dreadfully anoraky areas of the discipline. You are correct to say that the procedure that is proposed in the bill is generally called the Gregory method, which involves the transfer of the surplus vote of candidates who have been elected. The advantage of that method over the one that is used in the elections to the lower house in the Republic of Ireland is that it takes account of all the preferences that are expressed on ballot papers. Under the system that is used in the Republic of Ireland, when the calculation is made of which ballots should be transferred, only the very next preference is taken account of, which can lead to distortions when there are very close results at the end of the race. The Gregory method is fairer.

There is an alternative, which has been used in elections to the Australian Senate since 1983. No one else has given it a title, so we have called it the "inclusive Gregory method". It would deal with the issue that you raise, because it involves transferring all the ballot papers that were originally in the pile of the candidate who was elected, not just the last parcel of ballots. The problem with that method is that it can lead to an anomaly. In certain circumstances, the value of a ballot paper can increase, because the value of a ballot may be greater when it is transferred to candidate B than it was when it came to candidate A—if the committee is interested, I can provide written documentation after the meeting to demonstrate that point. For that reason, many of us would criticise the Australian Senate rules as inappropriate. If one had to pick a version of the Gregory method, the one that is proposed in the bill is about the best available. Alternative methods are available and are theoretically possible. However, the general advice is that computer counting would be required for those methods.

Do you agree that, under the system that is proposed, a considerable number of preferences that voters express are simply ignored?

There are swings and roundabouts—we gain in one way, but we lose in another. If we were to take account of all the ballot papers, not just those that are received in the last parcel, we would arguably be adding much more weight to some ballot papers. It is difficult to describe this verbally—it is much easier to do so in writing—but such an approach can replace one potential anomaly with another. The bottom line is that the surplus has been created by the election of a candidate. One could argue that, even if only some ballot papers are transferred, the remaining ballot papers in candidate A's pile elected that candidate—the voters concerned had a say in the election of one candidate.

However, they would then have a say in the election of another candidate.

Some would. If one wanted to be a purist, one might want to adopt the weighted inclusive Gregory method or the Meek method. The advantage of the weighted inclusive Gregory method is that all the ballot papers—not just the last parcel that is received—are transferred and appropriate weights are attached to the value of the ballot papers when they came to the candidate. That is possible, but to do it by hand would greatly increase the length of the count.

To summarise, there is no definitive way of producing a pure result. It is all about judgments in relation to how ballots are transferred and whose ballots are accorded what weight.

I would prefer to say that criticisms may be made of the counting process in STV, depending on the choices that are made in determining that process. However, criticisms can be made of virtually all electoral systems. No electoral system is perfect.

On our visit to Dublin, academics told us that there was an alternative method, which was to take a truly random sample. Obviously, that is not done in the Republic of Ireland. I imagine that such an approach would overcome the problems associated with the weighted method, which involves examining all the voting papers. What are your views on the method that I have described?

You say that using the weighted method would involve moving to some sort of computerised system. E-voting might make that easier. What are your views on that issue and on the experience in Australia? As you know, a dispute is under way in the Republic of Ireland about how secure the system can ever be.

On the first question, I bow to the greater experience of the people to whom you spoke in Dublin. My lay impression is that the number of sampled ballot papers in relation to the size of the population would determine whether the random sample could adequately take account of all the subsequent preferences. I am just not sure about that.

On computer voting and counting, Ireland is blazing a trail. In Australia, there has been some movement in the direction of computer counting, but I think that I am right in saying that, apart from the Australian Capital Territory, nowhere in Australia has gone down the road of computer voting. From what I know of the experience in the Republic of Ireland, the experiment last time, which incorporated three constituencies, was seen as a great success. I understand from reading political science accounts and journalistic accounts and from speaking to people who were there that there were no major problems. As far as I understand it, computer voting will be used across the whole of the Republic of Ireland for the next general election, which suggests that it has been a success.

You are absolutely correct to suggest that once the idea of going towards some kind of computer counting is embraced, there is no reason why one cannot move towards something like the weighted inclusive Gregory method. I always get my states confused but, as far as I know, the Western Australia Parliament is debating the possibility of moving over to the weighted inclusive Gregory method, which is facilitated by the fact that the state can do computer counting.

In your written evidence, you raise the issue of by-elections. You observe that, under the proposed system, single-vacancy by-elections will use the alternative vote system. You suggest the alternatives of

"a replacement candidate from the same party or a recount of the original district election".

However, there are significant problems with both those approaches. A replacement candidate from the same party might be someone who has never been in front of the electorate and so has no demonstrable support, whereas a recount of the original election might mean that the seat is awarded to someone from a completely different party, which is against the wishes of the voters at the time. What is your preference and what is your justification for it?

I cannot give an easy answer to that question. I infer from the bill that the proposal is probably to use the alternative vote by-election method. That seems to be the implication because, if all contestable elections are to be conducted by preferential voting and one person needs to be replaced, that has to be done by the alternative vote. In essence, that is what is done in Ireland and it seems quite sensible. However, if the object of the bill is to introduce proportional representation for elections to Scottish councils, the possibility of a by-election that, by definition, will not be conducted through PR is not totally in the spirit of the bill. That suggests to me that you may want to consider other models.

I do not know the full details—perhaps you already have the information from the people in Dublin—but, from what I understand, Republic of Ireland STV elections to the European Parliament work on the principle of making available an alternate candidate in the event that somebody retires or dies in office. That is the Australian equivalent of the party picking a preferred other candidate. There could well be ways—such as having a name somewhere on the ballot paper or manoeuvring in some other way—of making clear to the voter the alternates who will be available. Certain list systems do that. It is clear to the voter that they are not electing those people, but that those alternates will be available in the event that seats need to be filled. That could get round the democratic problem of the person not facing the electorate.

As to the count-back method—the third method that is used in some parts of Australia—you are right to say that in many circumstances it means that the party whose seat is to be filled does not get that seat, because the seat goes to somebody else. If one was speaking from a party perspective, one would probably prefer the method of letting the party choose the alternate.

The other issue that I want to raise from your written submission relates to the lack of specifics in the bill on the way in which the elections would be run. You might have read the evidence that we took last week, when John Curtice suggested that the rules under which the elections are to be conducted should be included in the bill. Do you take a similar view?

I do not. I understand from the documentation that a lot of the detail can be done through secondary legislation, so I am not particularly concerned about the absence of certain details. I have seen some of the questions that have been raised about how a vote would be called invalid and so on. The one issue that raised a question mark for me was how the ballot structure would be organised. As I tried to suggest in my submission, different methods can be used. I presume that the idea is to go down the route that is used in Ireland and that you have therefore resolved the issue, but I raised the question because it is not covered in the bill.

Surely the details of how the election is organised—ward boundaries, for example—are crucial to the way in which the whole system works. What happens elsewhere? Do other countries put such details in primary legislation or secondary legislation?

To be honest, I am not sufficiently aware of that area to give a good answer. My understanding is that practice varies a lot. In some countries, the role is given to an electoral commission. In other countries, the role is taken by the relevant department—in Ireland, that is the Department of the Environment, Heritage and Local Government. I do not know, but I assume that you would have enabling legislation to cover issues such as boundary reorganisation. My concern is whether the voting method that is applied, once the boundaries are established, is the correct one.

Can you point us to a source, academic or otherwise, that gathers such information together?

One place to go, particularly on STV elections, is the relevant electoral commissions of the Australian states. I notice that you have a written submission from Michael Maley of the Australian Electoral Commission. His view was that the issue is not relevant for the Australian Senate because the Senate boundaries are determined by the states. That implies that you should go further into the six states and talk to the state electoral commissions. I am sure that their officials would be able to give you good advice on how they organise their systems.

On electronic voting and counting systems, are any safe, secure systems evolving elsewhere that could become a standard model?

I am guessing that the answer is no, because it is all so new. As I said, the Republic of Ireland is blazing a trail, particularly with regard to STV. To go down the road of e-voting as well as e-counting is a first. On the basis of what I have seen, my general advice is that, whatever the Irish do—if we assume that they get it right—will be a good system.

I wonder whether I could examine the evidence that you have submitted on the proportionality of the proposed system. Last week, we heard from two academics, Professor Curtice and Professor Miller, that the proposed system is the least proportional system of which they are aware. Are you aware of any systems that are less proportional than the proposed system?

I am not aware of any STV system that is less proportional than the one that is proposed. In some places, such as Malta, people build in ways of making the STV system less proportional. However, in terms of the effect of the size of the districts on overall proportionality, I am not aware of any STV system that is less proportional.

Other systems, such as list systems, might be less proportional. The question comes down to the size of the districts—a district size of three to four members is small. As a generally accepted rule of thumb, if we want a reasonably proportional result, we should have districts with five or more members. At somewhere around the five-member mark, there will be pretty good proportionality, so three to four members is rather low.

I will return to the point about Malta in a moment. The Kerley report considered the McIntosh commission's work and recommended that the districts should have three to five members. That seems to add weight to what you say. However, the bill proposes that three to four members should be elected in each ward, which academics referred to last week as a "political fix". What would you call that proposal?

I am not as aware of Scottish politics as Professor Miller and Professor Curtice are, but my best guess is that what is proposed is akin to the Jenkins proposals for the House of Commons. That process was referred to as "broad proportionality". The aim is to try to achieve a slightly better measure of proportionality than there currently is so that small parties and independents have a slightly better chance of winning seats. However, that is not proportionality.

I have just completed a study with a colleague in Australia of STV systems in Australia, Ireland and Malta. If my memory serves me correctly, the system in Ireland comes out as one of the least proportional of the STV systems. In any event, the academic literature often refers to STV as quasi-proportional. It is right to say that Ireland has an average district magnitude of three to five members, but in parts of Australia the district magnitude can increase to 21 members. There is no particular reason why the cap could not be lifted; I am a rank outsider, but there seems to me to be a deliberate attempt to keep a lid on the degree of proportionality of results.

I want to clarify matters for the Official Report. Your work was quoted at the meeting last week. You have said that, in order to achieve a balance between local representation and proportionality, there should be a minimum of five members for each district, but you are not arguing that having five members represents pure proportionality. You are saying that five members is the minimum number that is required to keep the link between local representation and proportionality.

I would prefer to say that five members is the generally accepted minimum number if a degree of proportionality is wanted. As I said, that is a rule of thumb. Trade-offs and balances are introduced in all systems. I imagine that, in far-flung rural parts of Scotland, there would be a desire to ensure that wards are not too huge as a result of the low population densities. I remember seeing in your documentation references to wards with just two members. I do not see a particular problem if there are reasons why parts of Scotland must have small wards with, say, two members, provided that there is balance. I do not understand why a ward must have three to four members. I do not see anything wrong with having two to seven members or two to eight members, if I had to pick figures.

Last week, Professor Curtice suggested something along the lines of what you are saying. He said that, in some rural areas, there might have to be wards with as few as two members, but he saw no reason why cities such as Glasgow and Edinburgh and other urban areas should have wards with fewer than five members. Do you accept that?

I would certainly agree with that.

In relation to—

I would like to come back in.

Do you recognise that the bill is not just about introducing proportional representation, but that it tries to address the five criteria for an electoral system that the McIntosh commission highlighted, which are proportionality, the councillor-ward link, a fair chance for independents, allowance to be made for geographical diversity and a close fit between council wards and communities? Perhaps that is why the range of ward sizes—in terms of the number of members per ward—is not as great as you suggest that it should be.

I am sure that that is the case. As we have heard, there are questions about how ward boundaries should be determined. I have no particular knowledge of the Scottish system, so I cannot comment on that. I do not know about the five criteria that you mention—obviously, I have not followed the debate as closely as members of the committee have—but you referred to a fair chance for independents. Larger wards would certainly lead to a greater possibility of independents being elected. Therefore, some of the criteria overlap. However, the Jenkins process that I mentioned earlier also tried to cover a set of criteria. The combination was a bit like squaring the circle and I suspect that a similar process is going on here.

I will ask another question and then come back to Tommy Sheridan. On a recent visit to Dublin with other committee members, I spoke to local government representatives and Dáil members who indicated that in the past there certainly had been Tullymandering, in which parties press for different numbers of members per ward to maximise their political advantage. Would allowing a broad range of members—say, from two to seven—in wards increase the likelihood that such practices might be attempted?

I have a different take on that question. Although the gerrymandering that produced the Tullymander debacle in 1977 was a big problem, it was caused not so much by the range in the size of constituencies that the parties were able to play with as by the fact that the politicians were in control of determining constituency boundaries. In light of what happened in 1977, the power to determine the Republic of Ireland's boundary divisions was given over to an independent commission in—if my memory serves—1979. Since then, Ireland has not had an issue with gerrymandering or Tullymandering. Again, this goes back to Mr Welsh's question. I think that it would be crucial for either the primary or secondary legislation to give some kind of independent authority outwith politicians' control the power to determine boundary redivisions.

I am glad that you interjected, convener, because that allows me to clarify something. The proposals are based on five criteria, including proportionality, giving independent members a chance and local links with ward members. However, the experts that we have heard from so far have made it clear that, even taking those criteria into account, they still think that the proposed system is less proportional than any other system in the world. Does that view ring true? After what you have heard, would you stick by your comment that the proposed system will be the least proportional in the world?

I stick by the argument that the proposed system would be the least proportional of all the current STV systems of which I am aware, although there might be a system in some rural part of the United States that I do not know about. Some PR list systems can be less proportional. However, it all comes down to the same thing. If your district magnitude dips below an average of five—which is what will happen under the proposals—you will not achieve very proportional results.

Just to anticipate the debate, can I ask you how authoritative you are in this field? Has anyone done more academic research on this subject than you and your colleague, whom we mentioned last week?

I am the author of the standard textbook in the area and the bulk of my research is on the subject. If I have not published it, I have read it.

I want to ask a dead quick question about Malta. Did Malta build a threshold into its STV system or is the system more complicated than that?

I am not very good on Malta. As far as I understand it, a party that wins more seats than any other party is awarded with a boost in its number of seats. I think that some kind of extra mechanism is built in that is deliberately designed to manufacture a majority Government, but I would have to check that. In any case, as that system has been in place since the 1980s, most of us who examine STV systems have tended to take the Maltese experience only up to that point—we do not refer to the new system as a member of the family of STV systems.

How reliable is the formula in your paper that suggests that the effective threshold in a three-seat district would be 15 per cent? I know that other factors are involved, but is that a general formula that could be applied to other situations?

The threshold would be 19 per cent in a three-seat district.

Sorry.

The formula, which is known as the effective threshold, was originally designed by the Dutch political scientist Arend Lijphart and is built on the 1989 work of two American political scientists, Rein Taagepera and Matt Shugart. Lijphart's threshold as originally conceived was debated in quite a lot of detail in the main journal of electoral studies. As a result of that debate, everyone has settled on the formula that I have provided in footnote 4 in my paper—it is the agreed formula for what is known as the effective threshold. The effective threshold is an approximation of what, on average, a candidate needs to get to pick up a seat and is generally seen as the best measure in that respect.

So if we stick to a three-member ward system, anyone who does not get 19 per cent of the votes cast does not have a chance of being elected.

Yes. All things being equal, that would be my best guess.

The more the discussion goes on, the more complicated it seems to get. I was interested in Tommy Sheridan's line of questioning, as I read with interest what your submission says on multimember constituency representation.

Elsewhere in your submission, you refer to the situation that would develop with regard to by-elections and the complications that could arise because of the variable ballot paper that might be used. How do you see that being a complication in a small electoral ward?

I guess that you are hinting at alphabetical voting. In truth, one of the virtues of having a district that will elect only three or four members is that there will be relatively short ballot papers, meaning that alphabetical voting will be that little bit less likely. There is a direct relationship between the length of the ballot paper and the extent of alphabetical voting. An extreme example of alphabetical voting occurred in New South Wales in 2000 or 2001, during what became known as the tablecloth election. If memory serves me, there were 640 candidates. As you can imagine, in that scenario there was the potential for a lot of alphabetical voting.

I assume that, under the bill, the alphabetical listing of the candidates on the ballot paper would be based on the alphabetical order of their names. That is how most British ballot papers are organised. The spectre of alphabetical voting is raised, but its incidence is likely to be very small. All that I am suggesting is that other models might be considered for a system that is being designed for the first time. The Tasmanian approach, which is known as Robson rotation, could be a model for a place of this size. The bill could provide that the rank ordering of the candidates would be varied every X number of ballot papers, so that everybody would get a turn at being at the top of the ballot paper.

So, there would be a variable list. Your submission mentions having three or four members in a district ward. How are those members selected to represent the ward? We know that they are elected, but how does the electorate identify with the individual member?

The Irish experience suggests that—to put a good spin on it, for now—the recognition factor of politicians is a lot higher than the recognition factor of British politicians in the House of Commons. One of the reasons for that is the fact that politicians in Ireland need to get good preference votes.

Let us say that a party wants to try to win three seats and has three good candidates. In an ideal world, it would probably try to distribute those three candidates according to their places of residence and areas of operation, in good areas of the region. That would mean, in effect, that the party would have three constituencies in the one region and good geographical representation. The supporters of that electoral system will say that, contrary to the argument that moving to multimember regions could reduce the level of people's contact with politicians, if anything it might help to increase that level of contact.

Do you imagine that, within such a ward system, the current boundaries would be extended to accommodate the increased representation?

In the sense that the current boundary is for only one member?

Yes.

Yes. I imagine that a ward would have to be three or four times its current size for that reason. You would not want the size of the legislature to increase as a result of the change in system.

Your submission compares Scotland with New South Wales, where there are 21 members in one district. In terms of the terrain and geographical challenges that members would face, is it fair to compare Scotland with New South Wales? A three or five-member ward in Glasgow would be different from a three or five-member ward in New South Wales. Am I correct in saying that?

Yes, indeed. With that example, I was just trying to show that 21 members is the upper limit that is currently in use.

For the Australian Senate, districts of six members would be the norm. The ideal would be to try to keep within single figures, because, once we get into larger figures, apart from anything else, the ballot structure starts to get too difficult for the average voter. Although I say that you should settle on five as a minimum for the size of your districts, once you get above nine, the system probably starts to get a bit more difficult.

Surely the challenge would be not simply proportionality, but service delivery on the part of the elected members. Your submission compares our system of three to five-member wards with the system in New South Wales. Do you accept that that is not a valid comparison?

All that I was trying to do with that comparison was to demonstrate that, in STV, it is possible to have a district magnitude—a ward size—that is bigger than three to four members. I was simply illustrating that point by saying that New South Wales has the opposite extreme of a 21-member district. I cannot remember the figures for the other states off the top of my head, but in Tasmania every district has five members—it used to be seven, but it has come down to five.

I would not suggest for a moment that the bill should go the New South Wales way and introduce 21-member districts. If it did, you would have the tablecloth-election scenario that I just mentioned. There is an upper limit at which it would get ridiculous.

But the argument in your submission is that the number of elected members would have to be increased to accommodate some of the geographical challenges that elected members would face in places like New South Wales. An elected member in New South Wales would be very different from an elected member in Glasgow. I do not think that your submission touches on the challenges in Australia, where elected members would have difficulty in delivering and where an increased number of elected members would be required compared with Glasgow, for example.

I cannot speak with any knowledge on that aspect. I do not cover questions concerning delivery by politicians; I look at electoral systems.

Along with your studies on proportionality, have you considered how effective members have been since they were elected?

That is a difficult question to answer. In the book that I have just completed with Ian McAllister of the Australian National University, who has also spent a lot of time in Edinburgh, we tried to examine how politicians represent their voters. We made use of survey data and, to be honest, the only survey data that were available to us for the Australian context were those that were based on the senators and the members of the House of Representatives. As you have seen from some of the other written documentation, particularly that by Michael Maley, the Australian Senate uses an unusual variant of STV, and it is therefore difficult to draw generalisable conclusions from it. It is difficult to examine New South Wales, Tasmania, Western Australia or any of the other states or territories, because the data are not available.

So there is no evidence to suggest that the introduction of STV has improved local services.

There is no evidence of that because nobody has studied it; equally, there is no evidence of the opposite, because nobody has studied that.

STV has increased proportionality slightly, but no studies have been done on whether it has improved service delivery.

Nobody has tried to address that issue, and we do not have the academic resources to do that. The argument is that, if we introduce STV, a good link between the politicians and the voters is maintained, and one can therefore expect that the degree of contact between politicians and voters will remain good. However, cohesion and discipline within the Parliament and the nature of its output are under-researched.

When you talk about a 21-member district, is that in state Government or local government?

State Government.

Do you have any evidence or statistics from local elections—in the areas that you have been talking about with five-member wards, or three or four-member wards—that show how many elected members there are per head of population? Do you have any evidence on the population sizes for five or six-member wards?

I do not have such information to hand, although it would be relatively easy for somebody to research that. I may be picking you up wrongly, but, when a change from a non-proportional, single-member system to a proportional multimember system is being considered, people often imply that the number of elected representatives per voter will be reduced. However, there is no particular reason why that has to be the case. There is no reason why increasing the size of wards to form multimember wards necessarily means fewer representatives per voter; it simply means that those representatives are now organised into multimember regions. There should be no other effect. The number of elected members per voter should, in principle, be able to remain the same.

I accept that. We have talked about how many elected members there would be per ward; we have not talked about how many people they would represent, whether the ward is a three, five or eight-member ward. Nor have we talked about the geographical size of the areas that members might have to represent. Are there examples from local government elsewhere in the world that show how many people councillors are expected to serve?

I am not aware of figures on that. It would be difficult to measure the degree of contact between individual councillors and a set of voters, and to measure the areas that councillors claim to cover and seem to cover.

That last issue is crucial. We tried to obtain evidence on it during the previous session of the Parliament. As you suggest, very little research is being done.

Is it true to say that the Australian context is quite different from the Republic of Ireland context? It seems that the Australian system is based on party, which is completely different from the Irish system in which the individual in a community counts for more. To go back to what Paul Martin said, is it fair to take information from one context and apply it to another? I think caution would be required before doing that.

I got the distinct impression that competition is much more heightened in the Republic of Ireland than it is in first-past-the-post systems. Candidates identify clearly where their vote is and stick to that. That must have implications for how widely the electorate is covered. Because councils are sometimes hung, I got the impression that there had to be more co-operation between parties. If there was an unpleasant result in your ward, you might well be able to work with another councillor to make things not quite as bad.

All those considerations are going on in the background, but has any research been done—especially within parties—into the selection of councillors? There could be far fewer than are being selected in big urban areas. All that friction is going on. Can we get our hands on evidence from anywhere that would address some of those questions?

Those are great questions. There were a lot of them, though, and I am not sure that I will be able to remember them all. Please come back to me if I forget any.

On your first question, I did not quite follow where Paul Martin was going with his line of questioning, but I take the point. There is no such thing as a bespoke electoral system. For that reason, it would be incorrect to suggest that, if a system works in a certain way in one cultural context, it will work in the same way in another. That does not happen, for obvious reasons that we all know about. The way in which a system operates in Australia or Ireland might have a lot more to do with the features of party politics and the institutions of government in those countries than it has to do with the STV system. Your general description of the Australian system being more party based and the Irish system being more candidate based is spot on.

If I understood you correctly, there was a second strain to your questions. You seemed to be asking about a constituency that had a big urban area and a large rural expanse.

No. In the Republic of Ireland, I got the impression that because there will be a candidate with a close electorate within an area, candidates might not look for votes right across the multimember ward and that they seem to be very aware of the number of people who need to vote for them. The situation can get almost incestuous. The members seemed to be saying that it was more important to go to funerals than to community council meetings, for example.

My question was to do with whether we will get a better system of local governance than we have at the moment.

There are certain traits of the system in Ireland that can better be explained by the fact that the system is operating in Ireland than by anything else. After all, one of the best exports from Ireland was the big boss politicians in the cities of the United States. That is what the Irish do; that is the kind of politics that exists there. It is arguable that that would result in a kind of focus on church gates and funerals almost regardless of what the electoral system was. That is why, as I said in response to your first question, I would not blame STV for those traits. Rather, I would blame the nature of Irish political culture, and I am not the only one who would say that. STV might facilitate the situation that you are talking about, but other systems would do so equally.

More generally—again, I might not have followed you exactly—even if a candidate is focusing on the core number 1 vote, he or she will not rely solely on that but will think about the second, third and fourth votes because, ultimately, they need to pick up preference votes in order to get that seat. That means that the rational candidate, if they are not in a party, will be in some degree of contact with the other candidates to try to find out how best to play with the other candidates to try to sweep up lower preference votes from them. If the rational candidate is in a party, they will be liaising with their running mates to ensure that, while focusing on their number 1 votes in one area, they get second and third votes for their running mates.

That system has become a bit of an art in the Republic of Ireland, where it is referred to as vote management. People try to ensure that all the candidates from the same party end up with the same proportion of first preference votes to ensure that they survive through the count long enough to pick up the seats. Things can reach the point where a candidate will go to certain parts of the constituency and say to the voters, "Give me your number 2 vote, don't give me your number 1 vote."

That might sound like a game, but the virtue of the system is that it enriches the representative process for the voter because they are thinking not only of one candidate but of the other candidates that they might want to relate to. Equally, the candidate is aware that, if he or she does not do right by that voter, somebody from around the corner will get that vote.

I want to ask about the simultaneous operation of multiple voting systems, which is what we are contemplating in Scotland.

We will discuss voter education with other witnesses. Between now and the date when the next local government election is due to be held, two other elections will be held under completely different rules. We plan to hold two elections on the same day, in which people will get two ballot papers for two types of electoral system and will have to put an X on one ballot paper and 1, 2 and 3 on the other. Is there experience elsewhere of the simultaneous operation of multiple systems and of the impact that that has on the electorate?

The best example of that is probably the United Kingdom, not least Scotland, given that you already have different electoral systems for different levels of election. There is experience of that in some Australian states, but I do not know whether any of the local council, state and federal elections are simultaneous there. The state and federal elections would not be simultaneous, at any rate. The short answer is that I am not aware of any such evidence.

Having read the committee's documentation before today, I noticed that some individuals have expressed concern about the dangers of holding the local government and Scottish parliamentary elections on the same day. My question would be whether it is necessary to hold them on the same day. I understand that that does not happen in Wales, where the two elections are still held on separate days. I suppose that there is a danger of overload if both council and parliamentary elections are held on the same day. At least the counts could be separated by a certain number of days, so as to give the officials a chance to rest. In that sense, a problem might be being created where there does not have to be one.

There is no evidence of which I have been made aware that the number of invalid votes is significantly higher under STV systems than it is under other electoral systems. The extent of voter confusion is therefore not significantly greater under STV systems.

If STV is so great, why have countries—aside from the examples of Malta, Australia and Ireland—not flocked to take it up?

I know—that is a terrible conundrum. For those of us who tend to find the system fun to look at, it is a pity that there are not more cases of it. There is no simple answer; that is the reality.

I am in the process of conducting a survey among experts in electoral systems. I have a list of 640 experts from around the world, and we are surveying them to find out what their preferred systems are. I would hazard a guess that most of them will say the mixed member systems, which are in fashion. You are right, however: it seems strange that STV is not as popular as one might expect.

Which voting system will you use to tally up all those views?

That was a difficult question. We will be using a preferential system.

That brings our questioning to a close. Thank you for taking the time to give evidence to the committee, Professor Farrell. That was a very useful session.

Our next evidence-taking session is with David Green, who is speaking today in his capacity as chair of the single transferable vote working group. I welcome David to the committee and invite him to make some introductory remarks.

Thank you, convener, for the opportunity to make a presentation to the committee on a subject that is now becoming topical and interesting. I never in my wildest dreams thought that I would be speaking about such interesting subject matter; my wife now calls me a crofter with a disproportionate interest in proportionality.

I will outline the background to the group's role. It was set up by the Executive in September with—I like to think—an independent ethos and it reports to ministers. Its remit is to examine the procedures needed to facilitate local government elections using the STV system, consider how multimember wards will work in practice and commission research as needed. In particular, the group is considering the practical implications of the introduction of STV throughout Scotland, taking into account the geography of rural and urban areas and the need to retain the councillor-ward link.

We have met just three times; we are at an early stage in our discussions. We are in liaison with the remuneration progress group and the widening access to council membership progress group to ensure that we do not duplicate or cross over into the work that they are doing. We have split the discussion into three themed areas. The first is ward parity and geography. The second is multimember wards in practice, including access to councillors and allocation of ward business. Lastly there are the practicalities of holding elections under STV. We have concentrated more on the first two themes than on the last. We have set up a website with a discussion forum; I do not think that it has had an awful lot of hits yet, but I am sure that the interest is out there. We intend to hold meetings throughout the country to take evidence from the north, south, east and west, which will give us an opportunity to listen to views. We are happy to consider other issues that are brought up today. In the meantime, I will expand on any of the themes if you would like me to, before we start questions.

Thank you. We will just move to questions. You may not have had many hits on your website, but I am sure that when Paul Martin gets back to his office he will put that right.

Thank you for coming to speak to us. You said in your submission that the group is

"giving early consideration to aspects of the implementation of STV which may affect the content of the Bill and will report its findings to Ministers at an earlier date in order to inform Parliamentary consideration of the Bill."

We are already considering the bill and I wonder what you have to tell us at this stage. Are ministers and the Parliament not going to know about your considerations until we move on to a further stage of the bill?

I am happy to tell you what stage we are at with the deliberations just now. I will tell you about the group's findings on ward parity and geography and the number of members per ward.

We are aware that there are strong arguments in favour of greater flexibility in the number of members per ward. The majority of the group thinks that the number of members per ward should be between three and five, and two in exceptional circumstances of remoteness and sparseness. We believe that that would provide a better fit in terms of parity and sustainability for natural communities. Although that is not in line with the coalition agreement of three to four-member wards—I am well aware of that as chairman of the working group—it meets the concerns expressed in the recent consultation on the white paper "Renewing Local Democracy: The Next Steps". It meets the criteria of the Kerley report, which recommended three to five members and two in exceptional circumstances. That number of members per ward takes account of proportionality, the councillor-ward link, fair provision for independents, allowance for geographic diversity and a close fit between council wards and natural communities.

Three to five-member wards work in Ireland; regardless of whether that is good or bad, there is a proven track record. The group argues that three to five is in line with the Executive criteria in the white paper, which are: retaining the councillor-ward link; giving a clear reflection of voter preferences; being useable throughout Scotland; being responsive to the views of voters in both rural and urban Scotland; and not favouring unduly either larger or smaller parties or acting against the interests of independent candidates.

You said that the majority of the group were agreed on between three and five members per ward. Where do you anticipate each number working? What would apply in rural areas and in urban areas? If you are proposing variances within a council area, would that not produce a distortion of the result, given that there would be a higher degree of proportionality in one part of the area than in another? Would some parties benefit from such a split?

There is always the possibility of political interference in the setting of wards. The early indications are that setting the number of members between three and five would allow a better fit with the natural community. We have not worked out what that means for voting on the ground, but we believe that having three to five members per ward would provide the basis for proportionality.

What would be the best fit for a rural area as opposed to an urban area? Do you suggest, as others have done, that urban areas should have more members per ward and that rural areas should have fewer members? Would that not favour parties that have stronger support in rural areas?

We have not made a specific recommendation for rural and urban areas. On the catch of allowing two members per ward in exceptional circumstances, my personal view is that, because of the geography of the Argyll islands, if there were three-member wards, some of the islands would need to be linked to the mainland. There is an argument for two-member wards on islands such as Jura, Islay and Colonsay. However, we have made no specific recommendation that urban areas should have five-member wards and rural areas should have three-member wards. We must wait to see how everything comes out in the wash when the wards are merged.

I want to return to the remit of the working group. The group was set up in September 2003 and is expected to take a year to report to the Scottish ministers. As I mentioned, your submission states that one aim is

"to inform Parliamentary consideration of the Bill."

If you will not report to ministers until September 2004, how will your considerations affect the bill? How will the committee find out what your views are and whether those views have been agreed to by ministers? I am speaking particularly about your views on three to five-member wards. Can we expect an amendment from the Executive on that issue? How many other bits and pieces are likely to arise during the passage of the bill?

I have asked for an early meeting with the Minister for Finance and Public Services in the middle of January to put the group's views to him. At that stage we will recommend having three to five-member wards, but it will be up to the Executive to feed that back through the parliamentary system. We will have an early meeting, rather than waiting until the end of the year.

I hope that you are still in post in January to have your meeting with the minister because, given what you have said today, it seems that your recommendation will affect the outcome of the bill. Are you saying that, after three meetings, the majority opinion of your working group is that three to five-member wards should be used, not three to four?

That is right.

We have heard evidence from the Executive that, in creating the new wards, there is no plan to affect the present local authority boundaries and that the only change will be within local authority boundaries. Will your working group make recommendations for each local authority as to which wards should have three members and which should have five?

The group has discussed that, although the decision will not be one for the group; it will either be for the Local Government Boundary Commission for Scotland or for local authorities. There are a number of options. We can either merge existing wards to get to three, four or five-member wards—or two, as the case may be—or we can start from the beginning with the Lego blocks. The group did not reach a view on that. Some members say that it would be simpler to merge the existing wards into three, four and five-member wards and to ask the local authorities to bring their proposals to the Local Government Boundary Commission, whereas others feel that the commission, in consultation with local authorities and others, should come up with the proposals. The time scale is fairly tight and both courses would present a number of challenges.

Will you clarify for the record the basis on which the members of your group were appointed? What was the qualification? Are you well known in the field? What qualifies you to be a member of the group? Will you clarify whether ministers appointed you?

I was the convener of Highland Council until May this year. I stood as an independent convener and my political view was that I was totally opposed to proportional representation.

Did the other members have similar qualifications?

The views of Jim McCabe from the Convention of Scottish Local Authorities on proportional representation are well known. Leslie Evans is a civil servant from the Scottish Executive. Jeff Hawkins is from the Society of Local Authority Lawyers and Administrators in Scotland; you will hear from him later on. Vicki Nash is the chief executive of the Society of Local Authority Chief Executives and Senior Managers. The other members are Billy Pollock, who is from the Association of Electoral Administrators; Ken Ritchie, who is from the Electoral Reform Society; and Bob Smith of the Local Government Boundary Commission for Scotland.

Andrew Welsh has a supplementary on the previous question. Can we take that first?

I am anxious not to miss my follow-up point, convener. Would it be fair to say that you must be quite expert in the area of local government and STV—otherwise you would not have been on the group?

I would hesitate to call myself an expert.

What would you call yourself?

I reckon that I have a fair bit of experience of local government and of working in large wards. In particular, I have worked in the representation and service delivery roles that are associated with local government and I have awareness of the general issues around the debate. I would not call myself an expert on STV, although I am fast becoming one.

It is the majority view of your group that the recommended number of members per ward is three to five rather than three to four.

That is right.

You are clear on the number of councillors for each ward, but you are not clear on how ward boundaries should be determined; you are saying that that should be left to the boundary commission or to local authorities. In considering that, should your group not be looking for some fair general principles that could apply to the whole country? Is your group not looking to give advice to the boundary commission on how the ward boundaries should be made up?

We will be giving that more consideration in due course. At this stage, we have not managed to come to a view; we are just examining the options that are open to us, which fall into two camps. One can either merge the existing boundaries or one can leave it to the boundary commission to come up with views on three-to-five member wards or on three-to-four member wards, if that is what the Executive decides on.

You will be giving detailed consideration to whether there should be a merger or a wholesale re-examination and you will be providing advice or guidance.

We intend to do that.

Do you believe that that advice or guidance should be in primary legislation or in secondary legislation? In other words, should it appear in the bill?

We have not come to a view on whether that should be in the bill. It may be something that can be done anyway.

Surely it is of fundamental importance that the mechanism for running the elections is fair and democratic and that rules and principles are agreed on a neutral basis. For example, will you be making suggestions on rules about the shape of the ballot paper? Should not that also be in the bill?

I think that I mentioned that we have not had a great deal of discussion on the practicality of running the elections. We will deal with that at a later stage. We have views on decombining elections and whether, because of the complexity of the elections, there need to be greater transparency and robustness and more secure systems. We have not come to a decision on those implementation issues and we have not discussed the design of the ballot paper.

I am worried that you have made a decision on having three to five members per ward when all that you have mentioned to us is that you based that decision on the idea of natural communities. We have heard nothing about what Tricia Marwick mentioned in a previous question, which dealt with whether you were thinking about the population that a councillor would have. As I understand it, existing wards could be put together to form multimember wards, which might mean having three or four councillors per ward. You do not know whether there will be five councillors per ward. You seem to have made a decision without considering any of the boundary arguments, such as those that Andrew Welsh has made. I do not understand how you have made your decision or what you have based it on.

The boundary commission sits on our group and it has given us an early indication that including five-member wards would offer more options to fit natural communities together.

I am sorry—are you saying that having five councillors would allow natural communities to be fitted together?

If there were three, four and five-member wards, there would be more options for keeping natural communities together.

However, you are not saying definitely that there would be five councillors per ward. You will not know that until you have examined the boundary implications of such an approach.

We believe that it would create more options.

The Local Government Boundary Commission for Scotland is represented on the working group and recommends that five-member wards would keep local communities together. That is to assume that the present boundaries match local communities, as the commission would claim.

We are not aware of evidence that the present ward boundaries do not match natural communities. There is bound to be the odd instance in Scotland of boundaries not matching communities, but we do not know the extent to which that is the case. We have asked the Local Government Boundary Commission and COSLA to provide examples of the present boundary set-up not fitting with natural communities.

I return to the key point—what is the status of the activities of the single transferable vote working group? Will you publish a report that states your conclusions and is submitted to the Executive before or after your meeting with the minister? I have a fair idea of what the minister will tell you if you present him with a proposal for five-member wards. I see that people are alarmed—not necessarily for the reasons stated—but the partnership agreement states that there will be three and four-member wards. What status do you have to interfere with that?

I do not wish to interfere with the views of Scottish ministers. The single transferable vote working group has been set up to make recommendations and that is what we will do. Whether a report is published after that is in the gift of ministers. In the first instance, we are reporting to Scottish ministers.

So after today, we may never know what you have concluded.

Members are hearing what we have concluded on the issue of ward size. They will not know the result of our discussion with the minister until he shares it with the committee.

You have indicated that you have not reached a conclusion on a number of the issues that are within your remit. You have said that you have reached a view on ward size. The next that we will hear is whether the minister has accepted your view.

That matter is outwith my control. I report to the minister. It is up to him to decide how he shares that information. I can say no more than that.

So much for transparency.

The process is about as transparent as the STV system.

I am sorry to keep labouring this point. The working group was set up by ministers and will report to them. Do I take it that you are giving us your conclusions before ministers know what they are? If your conclusions are genuinely to affect the bill, that can happen only at stages 2 and 3. If we are not aware until stage 2 of any conclusions that you reach, we cannot take evidence on them then as we can at stage 1.

I am greatly concerned by the timetable and the process. I am not sure that I understand what point we have reached. We are considering a bill that states that there will be three and four-member wards. You are suggesting that you will tell ministers that there should be three, four and five-member wards. Does that mean that from now on at stage 1 we should discuss having three, four and five-member wards, or should we continue to discuss having three and four-member wards—as we ought to do, given that that is the proposal in the bill? Are amendments likely to be lodged?

I would like to answer that question. I am trying to be as open as possible, but how we proceed is in the gift of the minister. I can make recommendations and the minister can choose to say, "I am sorry, but I am not accepting that recommendation." The group has been set up to make recommendations and suggestions.

I understand your position but, as you are indicating, it is not terribly satisfactory. Your work runs in parallel with that of the committee and with the information that we are receiving. Are you saying that your recommendations will never be published?

A report will be made but it is up to the minister to decide whether it is made public or not.

You will be making a report to him and he will decide what he does with that report.

That is as I understand it.

I will go back to one or two of your specific recommendations. When the convener was asking questions, I did not quite grasp whether you had a firm view on different-sized wards in the same local authority. There might be three-member wards and five-member wards in the same local authority.

Yes, that is perfectly possible.

You have not come to the view that all wards in an authority should be the same.

No, and neither have we come to a view on the urban-rural question that was asked earlier.

Do you intend to come to a view on that?

That will become clear when the Local Government Boundary Commission for Scotland or the local authorities go through the consultation process to decide how the wards will be made up; that is, whether they will have three to four members or three to five. I cannot answer your question at this stage because it has to be solved by either the boundary commission or the local authorities as they make up the wards.

The evidence that the committee has heard, as Sylvia Jackson said, suggests that determining whether all the wards should be the same will be fundamental to the process, and more so than determining whether they are three-member or five-member wards will be. How much work has been done on your view of ward business and multimember wards in practice?

We have had a fair bit of discussion about the evidence. We asked that all the desk evidence be brought together and that any gaps in it be identified. Several issues are coming through. There exists the potential for politicisation of case work and there is a need for a protocol for opening schools, swimming pools and so on. The group has also agreed that there might be a need for guiding principles along those lines that could be adapted at local level. There are other issues about duplication of resources of council staff and duplication of effort by councillors in a multimember ward situation. There are several issues on which we could call for more research to find out how multimember wards work in practice.

Last week, COSLA's evidence said that it could not find any examples of protocols for multimember wards that operate elsewhere.

We did not find evidence of formal protocols, but there appear to be many informal protocols through which councillors sort matters out among themselves and take turns to open schools or whatever needs to be done. It is very informal, but some guiding principles might be needed.

Are you working towards a mid-January deadline?

That will be when we give to the minister the interim report on what we have found out up to that stage.

That is quite a tight time scale to get around the practicalities of conducting elections. We have heard a lot of challenging evidence about that.

I am aware of the time scale.

Several of the issues that you are considering have fundamental implications for the bill and you expect to send the interim report to the Executive by mid-January. Is there a deadline, or a time by which you expect to publish your final conclusions?

I was given an early indication that the final conclusions will be required in September of next year.

But the bill will have been passed—or not—by that stage. The target for the conclusion of consideration of the bill is June 2004.

I suppose that that is why the interim report to the minister is so important.

I would say that that is why it is so important to get the final report before we consider the bill.

It is in the gift of ministers to know what they want: it is up to them to do what they will with the final report because they have set up the working group and they will be getting a report from me and the group.

It will be absolutely crucial to your work that the voters understand the system and how STV is to be implemented. I note that you have said that there should be three to five-member wards. What kind of consultation has your working group carried out with the electorate and communities? The only elected member on the group is Councillor Jim McCabe. Did you consult tenants associations and community councils?

If I may be frank, we have not yet consulted anyone. We have taken account of some of the comments that were submitted in responses to the consultation on the white paper and we have done a bit of research into what happens in other countries. We have looked at the Kerley report and we have come to an interim conclusion that there should be three to five members per ward. As I mentioned earlier, we intend to go round areas and meet people who are concerned about the issue.

I would like to know whom you will consult. The experience is that some such documents receive responses only from the usual suspects and from those who are consulted. It would be interesting if you could confirm that. Is it not wrong that you should decide that there should be three to five members per ward prior to consulting the people who will take part in the process? Surely the electorate who take part in the process will need to understand that. You also mentioned the need to clarify some of the issues concerning casework prior to confirming the position.

Our working group was set up with a certain remit. We have come to an interim decision, which we are sticking with. We will feed in what we get from our various public meetings, but we have come to a fairly early conclusion on that issue.

Does not that put the cart before the horse? Should you not have consulted communities first, so that you could then clarify whether we should have three to five members per ward? Is that not the process that you should have used to ensure that you could deliver that to local communities? Surely the most crucial aspect of your work will be in implementing STV. The most effective way in which to do that would be to ensure that communities understand the system. How can you make a proposal for three to five members per ward prior to that?

As I said earlier, we believe that the early indications are that we can get a better fit with natural communities by allowing for five-member wards as an option as well as for three or four-member wards.

I do not want to labour the point, but it is important. How can you possibly say that five-member wards fit better with communities without having first consulted those communities? To be effective, I would expect the working group not only to issue consultation documents but to get out there into communities to experience some of the concerns, or the lack of concerns, that people might have. Who will you consult as part of the process?

I cannot give you a list of names, but our consultation will be as wide as possible. In particular, it will include those that have made representations on the white paper. There has been—

Sorry, will you clarify which organisations you are talking about? Which organisations will you consult?

We will consult local authorities that have made representations on the bill. We are also keen to meet the Scottish Youth Parliament. We will also consult academics who have fed in representations.

What about community councils and residents associations?

Yes, we will consult community councils, the voluntary sector, business associations—the normal cross-section.

If those people say that they are unhappy with having three to five members per ward and want three to four members per ward, in what position will that leave your working group?

We will have decided prior to that and we will have to justify our position. There has already been a fair bit of discussion on the issue. The Kerley report, for which there was a fair bit of consultation, proposed that there should be three to five members per ward with an exception for remote areas where there could be two. There is some provenance for three to five members per ward—it works in Ireland.

To be fair, however, the bill mentions that there will be three to four members per ward. My understanding is that any consultation exercise should not have pre-empted the decision prior to the consultation's taking place. What is the purpose of the working group's meeting community organisations if it has already decided one part of its report before meeting them?

We will take on board the views that are expressed—

How can you take those views on board?

Paul, let David Green answer.

We have come to an interim position that there should be three to five members per ward. That is a recommendation at the moment.

Tommy Sheridan has a supplementary on that point.

For the working group's remit, did the Executive detail how many public meetings it had held and how many community councils and tenants associations it consulted before arriving at the proposal for three to four members per ward? Did the Executive say how many meetings it had held?

I was not aware of how many meetings there had been to discuss having three to four members.

Do you know of any meetings that the Executive held to discuss the matter?

I am not aware of any.

So you do not know how they arrived at the recommendation of three to four members per ward.

No.

Professor Farrell made the point that there is very little evidence on how multimember wards operate in practice. I am interested to know where you will look for that evidence. The second thing that he mentioned—I think that this is what he was saying—was that the matter is context specific. The situation here differs widely from that in Australia, and the Republic of Ireland has a very particular context, so we should be very wary about transplanting ideas from there. How will you go about getting evidence and how will you take context into account?

My second question relates to the system itself and to distribution of surplus votes. I wonder how you are looking at that issue, because there is obviously a range of ways of doing that. We have been saying that one of the fairest ways to distribute surplus votes seems to be to do it by computer. That would be the nearest thing to doing it randomly—indeed, it could be said that it would be better than that. Are you going to look at the possibility of using a computer system and, if you are, what are your views about e-voting and time scale?

You are correct in that the one thing that one cannot do after reading the research, which I did over the weekend, is to make generalisations about STV. It appears that there is not much research on how multimember wards work in practice. We have the power to commission such research, so if it appears from the desktop work that there is not a lot of research, we intend to commission research from wherever.

The experience in Australia differs to that in Ireland. There is talk of a greater social worker role in Ireland because of the nature of Ireland. That is just one of the differences in context, so we would have to weight research accordingly.

You also asked about e-voting and e-counting. The working group is taking a precautionary stance on the issue, because we feel that the introduction of STV is a big step in itself and that strain might be put on the system if we were to introduce e-counting or e-voting at the same time. There are issues around the transparency and security of the count and around the available technology, which is only now being piloted in some areas. Although we have not had a lot of detailed discussion on the subject until now, we are taking the precautionary approach.

Although members of the working group have different views on whether it is a good thing to combine elections, the realistic view of the group so far is that the public and politicians expect the elections to be held on the same day. Given that, I am minded to recommend a staggering of the counts to avoid undue pressure on staff. At this stage, it is difficult to imagine that proper e-counting or e-voting will be in place.

I have two quick supplementary questions. First, how will the redistribution of second votes be done? Secondly, what is the time scale for the commissioning of research, if there will have to be new research because there is no research on the matter?

Research will be taken forward as soon as we have full information on the desktop research that is available: it will be done as soon as possible. As I said, the working group has the powers to go ahead and commission research.

I am not sure whether the question about redistribution relates to the distinction between Northern Ireland and the Republic of Ireland.

Yes. Which method should we be going for in terms of fairness and so forth?

I have to say that we have not debated that in detail, but the advice that we have received is that the system in Northern Ireland is better in as much as it is less random in the allocation of the second votes, which makes it more transparent to those who were looking at the system.

I am sorry. Did you say that it was more random?

It is more transparent. If there was a recount, you would get the same result. That would not happen with the system in the Republic of Ireland.

You said "less random", but you meant "more random".

Yes. I am sorry.

The content and timing bother me. I have seen so much bad legislation pass through Westminster. The work of your group is absolutely crucial because you will be taking evidence on a host of issues that will affect our democratic system; I hope that you will be able to evolve principles that will make that system better. However, it looks as though your recommendations will be produced after the bill becomes law, which bothers me. For example, if you consult local authorities, what will be the cost of implementing the system? Will there be electronic counting or voting? Those and all of the other issues that we have heard about will have a massive impact. I am sure that you will have thought those issues through, but the bill might become law before anybody hears about the results. I find that incredible.

I intend to take up the matter with the minister when I meet him in January.

Thank you.

I will say first that although you might perceive that you have had a hard time from some committee members, it is not you that is getting the hard time—

It just feels like it.

We all appreciate your honesty in presenting your evidence to the committee. We invited you and you have given us an honest assessment. Whether the ministers have thought the situation through is another matter, but you should not take the blame for that. However, are you going to arrive at a detailed position on whether the Scottish Parliament and local elections should be decoupled? Is that part of your remit? Do you have an opinion on that yet?

That is related to the practicalities of conducting the elections, which is part of our remit, so we will come to a view. Currently, views seem to be somewhat polarised, but we are anxious to come to one view when we have more information on e-counting and e-voting.

Will you take into account any of the evidence that the committee has heard? We have taken a great deal of evidence on the matter and I am concerned that our evidence might be deemed irrelevant in the light of the same information that you seem to be taking. We have taken a lot of evidence about the size of wards, much of which supports five-member wards: that is in my opinion as somebody who is objective about that because I am totally against the introduction of STV.

We have heard lots of evidence on redistribution from the academics. Will you pay any heed to the evidence that people have given to the committee or do you see the two processes as being completely parallel and without connection?

If there is anything that we, as a working group, can learn from the evidence that has been given to the committee, we will be pleased to take it on. I take it as a hit that the time scale is skewed and I will be taking that matter up. However, if there is evidence that the committee thinks the working group should be looking at, I am happy to take that back and put it to the group.

That brings us to the end of our questions to David Green. I reiterate what Tommy Sheridan said earlier about our concerns being more about the process than they are about your actions or the actions of other members of the working group. We will address that issue in due course with the Executive. Thank you for your evidence this afternoon.

I welcome to the committee the third and final panel of witnesses to give evidence on the Local Governance (Scotland) Bill. We have with us Douglas Sinclair, the chief executive of Fife Council and chair of the Society of Local Authority Chief Executives and Senior Managers. We also have Vicki Nash, chief executive of East Dunbartonshire Council, who also represents SOLACE, and we have Jeff Hawkins, who is director of central services at East Renfrewshire Council and chair of the Society of Local Authority Lawyers and Administrators in Scotland election working group. I understand that we are to hear introductory remarks from both organisations.

Thank you, convener. I will begin on behalf of SOLACE. It is fair to say that there is a fair degree of consistency between SOLACE's submission and that of SOLAR on the practical issues surrounding the implementation of STV. Our shared view is that, if we are to continue with combined elections, the earliest time for the count would be the Saturday after the elections.

Secondly, there is the issue of whether the two elections should be decoupled—the committee has touched on that. SOLAR's unanimous view is that that should be the case; the same view is held by the majority in SOLACE. There are arguments either way: it can be argued that there would be a higher turnout if the two elections were held on the same day. The counter-argument is that local government elections would not have parity of esteem—a phrase that we do not hear much now in discussion of the two tiers of government. There is also some evidence that there would be voter confusion and there exists the question whether voters are voting on national issues or local issues. Nevertheless, the overall view is that, if STV is to be introduced, there is a powerful argument for decoupling the elections. Thirdly, there are some big issues in relation to voter awareness raising and training, and resources will need to be expended.

I would like to highlight a couple of other points in the SOLACE submission. The first is on the need to widen access to councils, which is a point that we have made before in discussing the importance of having some kind of citizenship legislation. There is much anecdotal evidence that suggests that many councillors find it difficult to get jobs while they are councillors, and that employers can make things difficult for those who are working and who aspire to becoming councillors. We have legislation that allows people to participate in the children's panel system, so we argue for a similar arrangement that will allow people to work as councillors.

The second point relates to our concern about the remuneration committee. We do not see the need for a timetable of 15 months to come up with results on an issue that has already been well exercised by Kerley. There is broad consensus that councillors are underpaid and we believe that the sooner changes are made, the better. The Executive has given no indication of when that will happen—we may have to wait until 2007, although we believe that the changes could be introduced earlier, probably at the beginning of 2006. Let us hope that that can be done, so that councillors will be properly remunerated from the earliest possible date.

There is a fair degree of alignment between SOLAR and SOLACE; as far as STV is concerned, I do not have much to add to what Douglas Sinclair said. I emphasise that the unanimous view of SOLAR is that a combined poll using three different voting systems with a manual count's being done on the same basis as those in 1999 and 2003 is untenable. For that reason, we advocate decoupling the elections. However, we recognise that there are other views that favour retention of the combined poll.

If there is to be no decoupling, we would push for some form of computerisation—certainly of the counting side for an STV system. I attended the elections for the Northern Ireland Assembly 10 days ago. The phrase that was used by an academic commentator to refer to the STV count—that it was "epic and labyrinthine"—certainly rang true. The count seems to take an inordinate amount of time, so if our elections were not decoupled, we would be keen to push for some form of electronic counting. Between now and 2007, there is surely time to develop some form of scanning software that could be used. There may also be an argument for computerising the poll, although SOLAR recognises that that might be a step too far.

Thank you. I will begin the questioning. I note the view of both organisations that the elections should be decoupled. When should the next local authority elections be held? Should they be held before or after the Scottish Parliament elections, and should they be set at a fixed point in relation to the Scottish Parliament elections?

We have not come to a firm view on that. I think that, given the amount of work that might be necessary to redraw the boundaries, there is an argument for local government elections' being deferred by a year until 2008. That relates to my earlier point about the earliest possible introduction of revisions to councillors' allowances, in that existing councillors would at least get some benefits from increased allowances.

I echo that. It makes sense.

I have questions on costs and timing. You said that you would look to some form of computerisation. Do you have anything in mind? Could it be done within four years? How feasible is it that you could have an electronic voting system up and running by 2007, and who would pay for it?

It would make sense to fund such a system centrally, so that every authority uses the same system. The Republic of Ireland is making headway with computerisation, albeit that that has caused some furore. The Irish are trying to refine their system. There is evidence that New Zealand is moving towards electronic counting and voting for STV elections next year, which may be informative.

I sit on the Office of the Deputy Prime Minister e-voting programme board on behalf of COSLA. The Government has stated its intention to have an e-voting general election some time after 2006. My concern for Scotland is that we have few opportunities between now and then to practice e-government.

I would not be as gung-ho on behalf of SOLACE as Jeff Hawkins is on behalf of SOLAR in suggesting that in 2007 we should combine Scottish Parliament and local government elections on an e-counting or e-voting basis. There is a good degree of nervousness about e-voting and e-counting and, coupled with the nervousness about STV, that would be too much for returning officers, the electorate and political parties. An element of risk management comes into elections. As returning officers, we are accountable to the courts, but we try to retain the confidence of our councils and the voting public. I would be careful about coupling everything together at one time.

SOLAR's main concern is that the count process for STV be computerised.

Have you examined the cost of computerisation? What if it was not funded centrally?

We have not examined the cost. If computerised counting was not funded centrally, it would be up to each local authority and returning officer to try to justify it.

Is that feasible? What difficulties would that pose?

The obvious difficulty would be in trying to persuade one's council that investment in a computerised counting system was more important than other important political priorities for councillors. I guess that it would be a fairly low priority. The local government argument would be that, given the Government's e-government agenda, there would be a strong and reasonable expectation that such a system be funded centrally.

I was struck by SOLAR's submission on the decoupling of elections, in which you use the phrase,

"The firm and unanimous view of the Society",

and you make the point that that view was submitted in September, before you visited Northern Ireland and experienced the practicalities of STV in operation. Would it be fair to say that, from a completely non-political point of view, you and SOLACE are saying to the committee and the Executive that if the elections are to be conducted under STV and they are not decoupled, serious problems could undermine the democratic process?

There would be serious problems if the elections were conducted on the same basis, but the Scottish Parliament count took place immediately after the poll on a Thursday and we then proceeded to a manual local government count on the Friday. All the evidence suggests that the manual count will probably last for between one and a half and two days. In effect, you would be asking returning officers and their staff to operate at full tilt for probably four days without sleep.

We must remember that postal voting by right means that on the day before an election, most returning officers spend the day opening thousands of postal votes, whereas previously, they might open only a handful of such votes on the afternoon of the poll or even after the poll. It would be logistically difficult to have, with STV, a combined poll on the same basis as the elections in 1999 and 2003.

SOLACE shares Jeff Hawkins's concerns. We have mentioned risk management. Concern has been expressed about our ability to recruit enough staff to manage the long process that will accompany STV.

Transparency and ease of voting for the electorate have been referred to. The Electoral Commission report "Voting for change" said that the prime concern in making any changes to the voting system should be the interests of the voter. I question whether combining all three processes—the two systems for the Scottish Parliament plus STV for local government—is in the best interests of the voter. I am thinking about how transparent and understandable the process would be, in particular for elderly voters and voters with disabilities. I share Jeff Hawkins's concerns, which do not apply only to electoral staff.

Douglas Sinclair suggested starting the local government count on a Saturday. Another idea is that we could start the process earlier and have the elections on a Tuesday. I reiterate SOLACE's previous comment that it would be nice if we could get away from the obsession with overnight counting, because that is the killer, as it keeps people up until 4 or 5 in the morning and puts pressure on regional returning officers, who must count votes until 6 in the morning then start the local government count. Risks go all round—they involve voters and electoral staff.

Another issue is electorate confusion, because the electorate will have to cope with three systems. Some evidence shows that they cope reasonably well with two systems, but to cope with the first-past-the-post system, the list system and STV is asking a lot of electors.

For political reasons, I would argue that the elections should be decoupled, because local government does not have the focus that it deserves when its elections are coupled with those for the Scottish Parliament. For the record, is your evidence as professionals and as the overseers of the workers who do the counting that, for the bill to work, local government and Scottish Parliament elections must be decoupled?

I would not go so far as to say that the elections must be decoupled.

How far would you go?

The role of returning officers is to make the system work. We flag up the risks of things going wrong and of voter confusion. I am not aware of any country that has operated three electoral systems at the same time in one count. That is bound to lead to confusion, particularly for elderly people, as Vicki Nash said. Some elderly people found the previous combined local government and Parliament elections complicated.

An interesting aspect of the Northern Ireland Assembly vote 10 days ago is that virtually every constituency had about 500 to 600 rejected ballot papers, most of which were rejected because people had used crosses instead of ascribing a numerical preference. That was a single election, but it took place in a country that is used to the first-past-the-post system for parliamentary elections. More than 500 votes per constituency were disallowed because people had used crosses. If such a figure were applied to Scotland, the votes of about 40,000 to 50,000 people would not count.

I do not want to lose the point. The bill or secondary legislation could say that crosses are acceptable, but that it depends on the number of crosses. My point is that the number of systems and the pressure on those who count the votes are your major concern, which is a dispassionate and non-political concern, rather than a political concern that somebody such as me would express.

The concern is certainly non-political.

I will add one rider: the application of more e-technology—be it e-counting or e-voting—would make the situation simpler, but we are not at that stage yet.

One great benefit of the paper-based system is that it enables people to scrutinise ballot papers. I appreciate that we could count papers with an electronic system, but there is even talk of internet voting or voting by mobile phone. The other benefit of what is a dated system is that we are able to challenge the count. How would we challenge an e-vote?

I do not have any personal experience of e-voting and I have not been to visit any of the pilots down south, but I share your concerns. As a returning officer, a prime duty is to ensure that candidates, political parties and the public have confidence in the system, and to maintain that it is an open and transparent system. I think that some reasonably successful e-voting and e-counting pilots have been run down south, but it has to be said that the information that I have received from them gives me cause for concern. There needs to be an awful lot more confidence among political parties that the system is open and transparent and that it is not just a question of putting bits of paper in and getting a result out 10 minutes later. I share your concerns. Much more work needs to be done before I would feel happy, as a returning officer, in applying that system.

If we were talking only about a computerised count for the STV part, it could be possible to have a large screen with technology explaining each of the stages of the process—the calculation of the quota, the transfer surpluses and so on—so that people could follow what was happening. With the manual system, the counting process is not intelligible or transparent. There are large parts of the process in Northern Ireland that people, including some of the staff, do not follow.

I appreciate that the process could be described, but the benefit of a paper ballot is that it can be challenged, if someone involved in the counting process has misplaced a paper in someone else's count. I am completely ignorant of how the e-counting process would work, how the systems would work, whether errors could be detected and whether people would have the ability to challenge the result. Ballot papers have to be retained for a certain number of weeks or months after the election, but what happens with an e-vote? Do we have a Microsoft system that tells us?

I am talking only about the count. If we had scanning software, it would identify ballot papers that were defective, for whatever reason, and they would be shown to the candidates as normal, so that they could agree with the returning officer that they ought to be rejected. There are obviously greater difficulties with e-counting, and there is no doubt about that, but it does seem to work well in some other countries.

My question relates to what was said about crosses being used rather than numbers showing the order of preference. I suppose that Professor Farrell has not had that information from Northern Ireland, because he said that there was no real difficulty with STV systems. Has that happened anywhere else?

When we went to Dublin, we talked about how Ireland was moving over to e-voting. I know that we are talking only about e-counting, but the Irish seemed to indicate that there could be a staged process, so that not only could people be kept more aware of what was happening but there could be stage points in recount situations. I believe that there was a case in which a well-known person realised in a split second that they could no longer win a seat in the election—that hits you very quickly. If for no other reason, consideration was being given to staging the process for the candidates, because it is a wee bit brutal. Are you aware of what is going on in the Republic of Ireland to refine the process to overcome some of the problems that have been mentioned?

I am not aware of specific details of the Irish process, but I hope to go to southern Ireland through the Electoral Commission to see how the Irish are refining their electronic system. As far as the number of spoiled papers is concerned, I spoke to Professor Farrell when he left, because I heard him mention that. He had picked up on my paper and was surprised at the level of spoiled papers, because 500 to 600 is very high indeed.

I was fortunate enough to be in Dublin on the day of the election there, and I saw e-voting and e-counting at work. One of the complaints about it was that it seemed brutal, because the end result came suddenly. From the perspective of the parties there—and this relates to what Paul Martin was saying earlier about scrutiny—the level of printout is far too large at the moment. Ballot boxes are used for much smaller areas here. If 90 per cent of the votes in a particular ward in Methil had gone to the Conservatives, I would know immediately that there was something not quite right. The problem with the system that we saw in Dublin was that it covered a much larger area and, although the actual count came far too quickly, the breakdown in the votes did not come until some weeks later.

To summarise, there are issues around e-voting, the level of information that is returned and e-counting. However, I am more interested in whether you have any evidence about the scanning machines that are used for e-counting—as opposed to people having to put something into a wee machine and rely on a chip. That would decouple the questions of e-voting and e-counting.

We have no experience of that. We advocate that that be looked into as a matter of priority.

There is a feeling that e-voting systems are not yet quite secure enough for us to go down that route. In fact, SOLACE said that in its evidence, which stated:

"SOLACE (Scotland) also believes it is fundamentally important that electronic voting or counting should not be introduced until totally secure systems are in place and have proven themselves to be effective and robust."

Are the systems that are currently available sufficiently effective and robust to be introduced?

The short answer to that is no.

Why, then, do you think that the Republic of Ireland is going to move wholesale to that approach?

Presumably, the Irish have a confidence in the system that we do not have. It may also be a familiarity issue. I am sure that the authorities in England and Wales that have done e-voting and e-counting pilots are entirely confident about using them again. I have seen some of the feedback and have attended presentations on the subject, but for me the jury is still out on my using those systems in my role as a returning officer who interacts with the electorate and the political parties in East Dunbartonshire. Much more work has to be done on the matter. E-counting does get over some of the problems that David Farrell highlighted in relation to the order of candidates on a ballot paper. With e-counting, that may be randomised, and a ballot paper may be presented to somebody on a screen in whatever order is chosen. However, in essence, the system provides a box. People go into it, tap it and disappear again. At the end, the chip is recovered, but how does a member of a political party or a member of the public know that somebody has not tampered with it? There are big issues of security. The electronic system is great, assuming that everybody is happy with it, but I think that we are a long way off from that stage.

There is a further point about the wisdom of adopting such a change. The election process is a fairly conservative one, which has evolved over many years. There is a question about whether it is wise to make two significant changes—the introduction of both STV and e-voting—at the same time. Our sense is that that is not a wise thing to do.

The question was asked in Northern Ireland why the group of people who are most likely to vote—older people—should be alienated through the introduction of electronic voting, and that argument seemed to be holding back the authorities from making that change. I am pleased that Jeff Hawkins set out objectively what we saw in Northern Ireland, as I am not objective about the matter. Frankly, I was amazed at the way in which the count was counted.

First, there was the lack of technology, in that people were using small, hand-held calculators to work out the results. Secondly, there was the length of time involved. Thirdly, and sadly, the fact that three or four times as many people spoil their ballot under STV is dismissed by academics as not significant for some reason. It was telling to see all those hundreds of ballots on which people had put more than one X. That was particularly remarkable in west Belfast, where the final margin was just over 80, with more than 700 spoiled ballots. It is important to have that on the record.

I will ask about the idea of a voter education programme, which is mentioned in the SOLACE paper. How do the witnesses envisage that being run in parallel with other elections in which people are being asked to do something different? It is possible that, between the passing of the bill and implementation, we will have the European election under one system and a United Kingdom general election—one presumes—under another, as well as our additional member system and constituency votes for the Scottish Parliament. What are the practicalities of running a voter education programme against the background of people being asked to vote in a completely different way in another election?

That is an enormous challenge. Whatever we do before the event—such as using television advertising or presentations to community councils—presiding officers and clerks will still be faced with an enormous number of questions on the day. How the election is managed in the polling stations on the day is crucial to minimising the number of spoiled papers or papers that are put in the wrong box. We should not underestimate how many people will have missed much of the publicity, however much money we put into an education programme before polling day. To go back to an earlier point about funding elections, we in local government feel that we often subsidise Scottish and Westminster elections; the local election has to be properly funded and properly thought out. I would not underestimate the difficulties on the day for the staff in the polling stations.

There is some experience in explaining the different systems, because there were campaigns in 1999. They were not particularly effective, and many lessons were learned, so the publicity in 2003 was a bit better—it certainly was not perfect, but it was a bit better. There is no doubt that the trick is to try to explain three different systems to people, and it will take a lot of money and a lot of effort to do that.

In parallel with that, resources must be devoted to training returning officers, their staff—the system will be new to most of the staff—and the people who are employed on the day in the polling stations and counting stations. There will have to be a lot more training than has historically been necessary.

Would there have to be any more staff in the polling stations than there are under the current system? I am thinking about how we would operate the system practically.

We should certainly have no fewer than we had this year, for example. Even this year, there were times when we had lengthy queues at some polling stations in my area. We had a floating clerk, but we might need to have more than one floating clerk, because we might want to give voters with disabilities or elderly people one ballot paper at a time. In East Dunbartonshire this year, we tended to give people all three ballot papers together, but on a polling day on which there are three different systems of voting, that could be more difficult for the voters to understand, and we might want to give them one ballot paper at a time, so additional staff might be required to manage the process.

I am sure that we would all welcome lengthy queues at all polling stations.

I am concerned about the fact that the legislation allows additional polling clerks to be appointed, but not additional presiding officers. Because of some of the concerns that we had in 2003, a number of us appointed additional presiding officers—although we knew that the Scotland Office would not meet the cost of that—because we felt that it was so important. In discussions with Scottish Executive colleagues, I would push for recognition of the need for additional presiding officers as well as additional polling clerks. There could be many occasions on which it would be beneficial to have multiple presiding officers in a polling station.

An additional presiding officer might not necessarily do the normal work of a presiding officer, but could be at a helpdesk to give people advice when they come in and are not sure what the process is. There would be real benefit in that, particularly in really busy polling stations.

I have one final point. Although the matter was not touched on in your submission, I would have thought that it might land on your desk. I am thinking of the resolution of issues between councillors who represent the same ward. If, in the next three weeks, Mr Green is unable to find a solution to that problem, how do you think that such issues can be resolved within a council without the situation becoming destructive to council business?

To an extent, a similar issue is that of the boundaries between the roles of the councillor, the MSP and the list MSP, all of whom could deal with the same matter. Some constituents who are unhappy with the response that they get from the councillor go to the MSP, the list MSP or the MP—MPs are still writing to me fairly often about devolved matters. The issue is not only one for councillors in a multimember system; there is a general issue about the protocols that are needed across government in that respect.

I suspect that a lot will depend on the make-up of the multimember ward. I take the point: it would be unlikely that all councillors in a ward will be of the same persuasion. However, if a majority of them were of the same persuasion, it would be possible that their working arrangements would be reasonable. If one of the wards in north-east Fife had four Liberals and one Conservative, it might be difficult to get a degree of co-operation. It is a fact of life that you cannot take politics out of the system. We will have to try to operate the system in the best possible way. As David Green indicated, we hope that informal arrangements will emerge over time. If people want to do the best for their constituents, they might not get over-involved in party politics. It is tricky territory.

I want to try to draw the questions on voting systems to a close so that we can move on to ask questions on part 2 of the bill. Before I do so, there are one or two issues that I want to pursue. The SOLACE submission raises the issue of three or four-member wards and the implications that they could have for rural areas in particular. What is your proposed solution? The intention of the bill is to address the McIntosh criteria. Does the bill meet the five criteria of proportionality, councillor-ward link, fair chance for independents, geographical diversity and the close fit between council wards and natural communities?

Given that a great deal of the work in that area has not been completed—the boundary commission has not initiated the work yet—it is difficult to give a complete or final answer. SOLACE highlighted two principal points from the list that you mentioned simply to make the point that, in some parts of Scotland, a three or four-member ward would be extremely large. The member-constituency link is important.

SOLACE has a long-held view that we should get the maximum possible alignment between electoral boundaries and natural community boundaries. The boundary commission is perceived to be more obsessed with numbers than with natural community boundaries, to which it does not pay full regard. The commission has always laboured under that difficulty.

I am sorry if I was one of the people who got David Green into trouble but, as a member of the working group, I was trying to reflect the thoughts of some of our fellow returning officers. Obviously, not all returning officers share the same views on every matter; indeed, they may not share the views of their appointing councils. However, the view has been expressed that two-member wards might be better in some of the larger rural areas. I thought it only fair to reflect that view.

As Kerley said, three to five-member wards are a better fit. If that has got the STV working group into some difficulty, all I can say is that we are where we are. The view that three to four-member wards are the best thing since sliced bread is not unanimous. There are a variety of views on the subject. I am sure that under David Green's chairmanship, the STV working group will hear many other views that there should be more or, perhaps, fewer members in each ward.

As officers who are responsible for running the system, do you prefer the detail of how the elections should be run—including definitions of spoiled votes and allowable votes—to be written into the bill, or are you comfortable with the Executive's proposal that much of the detail will follow in secondary legislation?

It would be unusual to have that level of detail in primary legislation; normally it would be in secondary legislation. There is no legislation on spoiled papers, but little examples based on case law have built up over the past 100 years.

I take your point on spoiled papers. However, Professor Curtice told the committee that it is fairly unusual in British history not to have a degree of definition in bills that apply to elections, representation of the people or similar issues. He would expect to see a lot more detail in the primary legislation.

That is not my experience.

For returning officers, it is important that systems are as clear as possible. As Jeff Hawkins suggests, local conventions develop. The key point about local conventions is that all the political parties understand not only what you are doing but the fact that you are consistent. However, I do not feel competent to comment on whether such issues should be covered in primary or secondary legislation. That is not my area of expertise. The important thing is that the will of the Parliament is clear.

I agree with Jeff Hawkins. Election procedure changes often. That is why most changes to election law are made through regulations in secondary legislation rather than through people having to go back to change primary legislation

What are your views on holding different elections on the same day? If STV is introduced alongside the present system of elections to the Scottish Parliament, should the two elections be decoupled because of the difficulty of administering the elections on the same day, voter confusion and the difficulties of counting? Even if the elections were decoupled after we had brought in an STV election, it is your view that the counting—and therefore the declaration—would be extremely difficult. You would like to move towards a system of e-counting to allow the process to be managed better.

The e-way forward is the way forward for—

Do you mean e-voting or e-counting, or both?

Both, ultimately, will make the system easier, but voter acceptability is a long way off. With STV, e-counting would make the process much easier to administer.

It was interesting that, the second time round, the electorate were a bit more comfortable with the Scottish Parliament list system. The first time round, there was a lot of confusion. I would therefore have thought that, if we were to introduce e-voting or e-counting, we would do a couple of runs with the normal electoral arrangements.

I have a tiny question on the electoral system, and then a tiny wee add-on, which may be a wee bit cheeky. Is it fair to say that you will implement whatever electoral system is asked for, and that, although you have commented on STV, that is not really your professional concern?

The choice of electoral system is for politicians. Our job, as returning officers, is to implement that decision—but, equally, we can point out what we may see as practical difficulties in any proposed decision.

I just wanted it to be clear that you are not here to argue for or against STV, or to argue for any particular form of STV. I will ask the cheeky question now, if the convener does not mind. I was quite upset by the level of expenditure by the Electoral Commission on explaining the two different votes that took place in May 2003 for local government and the Scottish Parliament. Do you share those concerns? Should more have been done to highlight that two different votes were taking place?

As I said earlier, the message was more effective the second time round. The first time, in 1999, the wrong message went out. People thought that they had only two votes because that was what the then Scottish Office had concentrated on saying. As the local government vote was highlighted only in the last week or two before the combined election, people thought that they had only two votes rather than three. That did not happen this time. The Electoral Commission is still finding its feet but I think that, in the long run, it will be good for electoral management in Scotland.

I will give members the opportunity to ask questions about the views that have been expressed about part 2 of the bill. Let me start off by asking about SOLACE's desire, as noted in the submission, for the new systems of payment for pensions and so on to be introduced as early as possible. For the next election only, the Executive proposes to offer what will in effect be a golden handshake for councillors who choose to step down. However, the offer will not be available to councillors who choose to fight the election even if they lose that election. Indeed, it will not be available to councillors at subsequent elections. Have you any views on that issue?

We support the Executive on that issue. To an extent, the offer is regarded as a one-off for councillors who choose not to participate in an STV system. That would follow the system that was introduced in Wales. However, we point out that some councillors serve for a period of time, then stand down for personal, business or family reasons, and then return to being a councillor. It is important that all service as a councillor is recognised in the calculation of a severance payment.

My concern with the proposal is that, if someone who has been a councillor for 20 years chooses to fight the election and loses, they would get no recompense for the ending of their career in local government, whereas if someone who has been a councillor for just one term decides not to stand, they would get recompense. Is that fair?

It is a matter of individual choice. If someone decides to stop being a councillor, they will get the severance payment. If they decide that they would like to continue being a councillor, they can take their chances and, if they do not get in, they will have to accept the consequences of that. The issue is about trying to get fresh blood into local government. The recent report by the Scottish local government information unit showed that the average age of councillors is well into the middle 50s. There is also an issue about the under-representation of younger people, women and ethnic minorities on councils. I think that that is part of the thinking behind the Executive's proposal. The rule may seem hard, but the severance payment is being promoted as a one-off. If people understand the rules behind the payment, they can make their individual choice.

I have two quick questions. First, I want to pick up on the point that Bristow Muldoon made. On councillors' remuneration, your submission states:

"SOLACE, however, notes the reference to allow only ‘future' service to account for pension purposes. This seems perverse … why should not previous service as a Councillor also count?"

That is why I am confused by your argument in response to Bristow Muldoon's point. People who stand for election and are defeated will get nothing, whereas those who choose not to stand for election will get the golden handshake in recognition of their previous service. I am confused by your arguments about when future and past service should be recognised. You seem to have one argument for pensions and another for remuneration.

I do not think so. For the severance payment, one has to make the conscious decision to stop being a councillor. The councillor has to make that decision before standing for election that they will sever their link with the council. For the pension, the current proposal is that the scheme would not be retrospective. It might apply from, say, 2007. Our point is that the scheme should not ignore a councillor's previous service and that councillors should have the opportunity to buy back previous service as a councillor. A scheme would have to be developed for that. It would be unfair to discriminate against councillors who had all that previous unpensionable service.

So, when someone who has been a councillor for the past 20-odd years stands for election and is defeated at that election, you hope that they will get the opportunity to buy into a pension scheme. Alternatively, do you expect that only those who will be elected next time will be able to have their pension backdated?

I understand the point that you are making but continue to found on the point that councillors make a choice about severance. The scheme is quite clear: councillors get the severance payment only if they decide that they no longer want to be a councillor for the reasons that I have given.

You make an interesting point as to whether someone who has stood for election, is defeated and is not entitled to a severance payment should be able to buy back pensionable service. Pension experts would have to deal with that. However, I understand the principle that there should be some form of recompense for that individual to buy back. A scheme would have to be developed, and it might involve the individual making contributions as well as the Executive or the council making employer contributions.

You suggested that there should be legislation to allow people time off to participate in local government as councillors. Will you expand on that?

In Fife—and I am sure that some of my colleagues have heard similar stories—there is evidence of people who are councillors and who try to get a job having interviews that go swimmingly until the person discloses that they are a councillor, when the job offer is taken off the table. To an extent, that is understandable if the employer is small. However, given the importance of serving as a councillor, we argue that if we want the best possible mix and cross-section of councillors, the citizenship legislation that is already in place for the children's panel system should be extended to cover participation in public life as a councillor. I suspect that that would be UK rather than Scotland legislation.

Douglas Sinclair is correct; it would require UK legislation to pursue that idea.

Would such legislation include compensation for the employer?

A scheme would have to be developed. Perhaps you remember that there used to be a financial loss system in which the employer was remunerated. It would be in the interests of employers—particularly larger employers—if they had employees who were members of the local authority, as that could form part of their contribution to public life in Scotland.

What are your further thoughts on that? I can see that you want fresh blood to be introduced into local government.

I had better be careful what I say about that.

I think I know what you mean. Being a councillor is onerous and time consuming because of councillors' involvement in a massive range of daily essential services. I would like to be proved wrong but it seems difficult to attract people who want to become councillors. Does the bill do enough to encourage people to stand for election? If not, what provisions should it contain?

We welcome many of the initiatives in the bill. The introduction of a pension scheme is an important step forward. For example, I know of a trade unionist who operates in Fife and who is quite keen to stand for election as a councillor. He is prevented by the fact that councillors have no pensionable service. If that changed, it would make the proposition more attractive. Citizenship legislation and the proposals for reducing the qualifying age would also help to attract a few more people.

I still believe that the Executive has not got right the prohibition on appointment of councillors to local authority posts. If I remember correctly, the initial consultation paper said that any councillor involved in the appointment of senior staff would have to serve a 12-month disqualification, not three months. That has been amended to include councillors who have been directly involved in the appointment process for council officers. That makes me wonder how often a councillor has to be directly involved in appointments in order to be disqualified. Is it once or 20 times? As we have suggested, the only way to deal with that would be to have a class of councillors, such as those who are the most senior holders of special responsibility allowances, to whom the 12-month disqualification rule would apply. Everyone else would be subject to a three-month disqualification. The provision is worded too loosely at the moment.

I realise that much of this will come within the scope of the remuneration working group. Does your organisation have a view on the balance or proportion of councillors in a local authority area who would be required to be full-time councillors because of their responsibilities?

That is tricky territory and I would be interested in my colleagues' views. I know that the Executive takes the view that it should be a part-time job.

If you speak to councillors who have 9-to-5 commitments about their commitments to attend meetings of school boards and community councils, they will tell you that it is very difficult to get out of them because community councils and school boards expect councillors to attend. There are no easy answers.

The fact that two thirds of councillors currently receive SRAs is a comment on the inadequacy of the basic allowance system. The system has to be changed so that a smaller number of councillors receive the right level of allowances. I have never been convinced by Kerley's argument for linking the salary of the leader of the council to that of an MSP. The leader's job should be evaluated on its own merits.

I can best demonstrate the inadequacy of the allowance system by telling you that one of our councillors—the only communist councillor in Scotland—made the point that in 1974, he got £10 per day attendance allowance and £5 subsistence, which amounted to £75 per week. Now, councillors get £125 per week from the basic allowance of £6,600. Inflation during the past 29 years indicates that the basic allowance should be in the order of £20,000.

The leader of my council works full time as well as being leader of the council. Many councillors who work also do a full-time job as a councillor in the evenings. We have evening committee and council meetings. Those councillors take on the additional duties willingly on behalf of the community. They go to many community and school board meetings. We have to try to make being a councillor a more attractive proposition in its own right so that councillors do not have to work around the clock, which is what they are doing.

That brings us to the end of questions for the panel. I thank Vicki Nash, Douglas Sinclair and Jeff Hawkins.

Meeting closed at 16:32.