The joint committee is continuing its hearings on the review of the electoral system. Today it will consider a number of technical aspects of the proportional representation single transferable vote electoral system. Professor David Farrell will speak about variations in the different single transferable vote systems used across the world. We will also consider the surplus vote transfer procedures used in proportional voting systems. Professor Gary Murphy will set out the arguments for and against using by-elections to fill casual vacancies in the Dáil. Mr. Hermann Schiavone will make a presentation to the committee on the application of the proportional representation single transferable vote, STV, electoral system in Malta. I thank them for coming and they are all very welcome. Their documents have been circulated to members of the committee. Before we commence, I must inform them that while members of the committee have absolute privilege, the same privilege does not pertain to them. I invite Professor Farrell to make his presentation.
Electoral System Review: Discussion.
Professor David Farrell
I thank the Chairman. I have circulated a copy of my presentation. I propose to refer to it, as I did on my last visit to this forum. I will begin by drawing the attention of members to the first footnote on page 1 of the document. Although it is not directly relevant to what we are discussing, I thought it might not be a bad idea to include it. I understand the committee discussed the representation of women at a previous meeting, but I am not aware that this aspect of the issue has been considered. I refer to the potential additional benefit of moving towards a national quota rule, for example, requiring all registered parties to ensure at least 30% of their candidates are women. At last week's meeting at Trinity College Deputy D'Arcy spoke about a tendency associated with the bailiwicks in our system. He suggested each Deputy had his or her own patch, which is almost like a single-seat constituency. One of the additional benefits of ensuring a certain proportion of candidates were women might be diluting the idea of a geographical emphasis in candidate selection. A number of candidates in each constituency might be chosen on the basis of sex, rather than geography. A national quota rule might begin to dilute the bailiwick emphasis mentioned by Deputy D'Arcy.
The first table in the presentation summarises the main variations associated with the single transferable vote systems used in Australia, Malta and Ireland. As Mr. Schiavone will speak about the system in place in Malta, I am drawing members' attention to the table rather than dwelling on it. In the text of my presentation I summarise the five main variations for the committee. I will focus on three that will not be dealt with elsewhere.
Ireland has the lowest average district size of any STV system in use. The point made in various other presentations — it would be worth considering constituency size in Ireland — is worth bearing in mind, particularly as it would not require a constitutional amendment. I am sure members are aware that if one were to increase the constituency size, one would get a more proportional result. This point is worth considering in the context of the proposal to reduce the number of Dáil Deputies. The two factors with the biggest influence on the average proportionality of an election result are the size of the Parliament and the size of the constituencies. If one were to reduce the size of the Parliament, one would get less proportional results. This would be an opportune time to increase the size of constituencies to ensure smaller parties have a chance of winning seats. I strongly urge the joint committee to consider that other countries have much larger constituencies and that there is no reason this should not be the case under the Irish single transferable vote system.
On ballot paper design, Figures 1, 2 and 3 show how the ballot paper can vary in each of the different STV systems in use. Mr. Schiavone will discuss the case of Malta. It is noteworthy, however, that while candidates are ranked alphabetically in Malta, this is done under each of the party groupings. It may be worth considering this model as an experiment in local elections to determine the degree to which voter preferences remain within a party group. For example, would a Fianna Fáil Party voter be more inclined to keep preferences within the party? While this may also have an effect on the bailiwick emphasis, it is not a strong point. Nevertheless, I draw members' attention to it.
The issue on which I will focus is surplus transfer rules which are set out in table 1. Attention must be paid to the available options in this regard. Four main types of surplus transfer rules are in use. The Irish system involves random selection of surplus ballots at full value. The essence of this system is that when a decision is being made on which surplus transfers to pass on to the next candidate, a "random" selection is taken from the pile of the candidate who has just won the seat and has a surplus over quota. These ballot papers are then transferred to the remaining candidates.
Two problems arise with the manner in which this system is employed in Ireland. First, not all ballot papers are treated equally because ballot papers left in the pile of the person who has won the seat remain in the pile throughout the process. As a result, the ballot papers of voters who are unfortunate enough to find their votes in this pile are not used again. As such, the ballot papers of these voters are not treated in the same manner as ballot papers which are transferred to other candidates.
Second, as the count proceeds through each stage, the ballot papers which are transferred continue to come into play. However, as these ballot papers contribute to the surplus of other candidates in later counts, the ballot papers of these candidates remain unused. There are, therefore, many random effects of the current system. I do not need to inform members that in a close election, these effects can lead to the election of the wrong candidate, in other words, a candidate may be elected who may not have won the seat if a different set of ballot papers had been selected at an earlier stage.
If there is a quota of 8,000 and proceedings continue until the tenth count, what is the margin of error? If the process is carried out on a random basis, do two, 20 or 200 votes go to a different candidate? Does Professor Farrell have an estimate?
Professor David Farrell
I cite in the presentation a statistical work and simulations which have been done. While I do not have the figures in my head, the work demonstrates that this system can have significant effects. My presentation features a summary chart of a count in Dublin South-Central in 1992 when Mr. Ben Briscoe won the seat.
I was present for that count.
Professor David Farrell
Deputy Woods will recall that five votes separated the final two candidates at the end of the count.
It decided the Government.
Professor David Farrell
It is conceivable that the result would have been different if a different pile of ballot papers had been chosen at an earlier stage.
The same pile of ballot papers has always been counted in recounts.
Professor David Farrell
Alternatives are available, one of which I strongly support. The Gregory method is known in Ireland as the Senate rules. As the system used for electing Senators, it is not unusual, untested or untried. The point of the Gregory method which is summarised in the presentation is that one takes account of all the ballot papers. The Chairman has referred to an election where the quota is 8,000 and a winning candidate has 10,000 votes. Using the Gregory method, one would transfer all 10,000 ballot papers at a fraction of their original value in order that the original quota technically remained with the winning candidate. The beauty of this method is that it eliminates the random effect, as occurred in the example of Ben Briscoe's election to the Dáil. The slight disadvantage of the system is that in later counts one only takes account of the original ballot papers. This means that if 2,000 ballot papers are placed in the pile of Gary Murphy who is placed over quota on the fourth count, only the ballot papers which arrived in Mr. Murphy's pile from the first person to win a seat continue to play a role and Mr. Murphy's ballot papers remain untouched. There are, however, ways of addressing this matter. While I propose to refer to these, I would not recommend them for use in Ireland.
One example is the inclusive Gregory method used in elections to the Australian Senate. A second is the weighted inclusive Gregory method used in elections in Western Australia. Members who wish to do so may read about these methods at their leisure. The reason I do not recommend either of them is that they are highly complex, as they take account of all ballot papers at all stages and the computations involved would realistically require computer counting. Since I do not expect computer counts to return in the near future, I do not regard these as options for use in Ireland. My strong recommendation to the joint committee is that it support the use of the Gregory method, which would be a great improvement.
Why is it called the Gregory method?
Professor David Farrell
It was named after the mathematician who came up with the idea. His surname was Gregory.
It has nothing to do with the late Deputy Tony Gregory.
Professor David Farrell
No, it dates back to the early part of the previous century.
I will leave specific recommendations on the table, as it were, some of which relate to presentations. I strongly urge an increase in constituency sizes. I also urge the use of the Gregory method or Senate rules for transferring surplus ballot papers. By-elections should also be replaced because they are an anachronism in a proportional system. Only two proportional systems, those in Ireland and New Zealand, use by-elections. There is no reason to hold them as alternatives are available. Members will hear about one but there are others to consider.
By-elections certainly have disadvantages.
Professor David Farrell
Are we discussing Dublin South?
To return to my first point, I suggest the joint committee consider amending the party registration law to require registered parties to nominate a national minimum number of women candidates.
I thank Professor Farrell for a highly interesting and helpful contribution to our deliberations.
Professor Gary Murphy
I thank the Chairman for the invitation to address the joint committee. I have spent much of my academic life examining interest groups and how they use elections as a vehicle to advance their ambitions. I have also done other work. I will discuss some of the disadvantages of by-elections and, notwithstanding the comments of my colleague, Professor Farrell, make some general remarks on the reason it could be worth sticking with by-elections, at least for the time being, until there is a full change of the electoral system.
I will start with the arguments against by-elections. As noted in my paper, by-elections can be arbitrary in terms of which candidate wins to the extent that the outcome does not necessarily bear any relation to the party which won the seat at the previous general election. Even if no single voter changes his or her mind, as Professor Farrell noted, due to the vagaries of PR-STV, a different candidate from a different party could be elected. To take a random year, of the 17 by-elections held since 1992, nine resulted in changes of party representation.
By-elections also have an in-built bias against the Government of the day as they become a mini-referendum. It is instructive to note that the last time a Government won a by-election was in 1982 when Mr. Noel Treacy was elected with more than 50% of the vote. Former Deputy Lee was not the only candidate to achieve more than 50% of the vote. While Mr. Shane Cassells came close to being elected for the Fianna Fáil Party in a Meath by-election some years ago, he ultimately lost out to Deputy Shane McEntee. The problem with the in-built bias against Governments in by-elections is that they can result in Governments making less than optimum decisions.
With regard to the nomination of the European Commissioner last year, it is generally accepted among all political factions that choosing a sitting Government Deputy was ruled out on the grounds that the seat would inevitably be lost. This is not to cast aspersions on Commissioner Geoghegan-Quinn who I am sure will do an excellent job and has a long distinguished career. Nevertheless, the point is that we do not pick from the correct pool if we eliminate Deputies simply on the grounds that governing parties of all hues will not submit a sitting Deputy for the post of Commissioner.
The fear of losing a seat in a by-election gives governments of all hues an incentive to leave seats vacant and I believe the record is approximately 520 days between former Deputy Pádraig Flynn's resignation in 1993 and Deputy Michael Ring's election in June 1994. This was far too long and disenfranchised——
It could be given a run for its money.
Professor Gary Murphy
It could well be.
That would not be acceptable, obviously.
Certainly not.
Professor Gary Murphy
It is a disenfranchising and unacceptable practice. I recommend at a maximum a six-month gap. One does not want a by-election within a couple of weeks but I suggest a period of between three and six months is appropriate. Otherwise the citizens of that constituency are disenfranchised. We might complain all we like about multi-member constituencies and inter-party competition, but that fact is that if a constituency has five seats, then those voters deserve five representatives. There is no excuse for what happened in 1994 when we were told that the country was too busy and did not need the hassle of a by-election. That is completely inappropriate in the system we have.
Having said that, by-elections may be better than the alternatives, of which Professor David Farrell mentioned a number and I will concentrate on one of them. My colleague, Mr. Hermann Schiavone, will discuss the countback method, of which I am not terribly convinced but Professor Farrell is. The other alternative, and perhaps the most feasible, is the method used in European Parliament elections whereby parties put forward a list of substitutes and if the MEP dies, God help us, or retires, the first name on the list takes over. This is also what is done in PR list systems. This has the advantage of there being no delay in filling the seat so citizens are not be disenfranchised. The inter-party balance in the Dáil would not be affected by this method and I will return to this point because it raises a question on whether by-elections should be seen as a test of public opinion during a Dáil sitting. However, this method is open to abuse in that a popular candidate, such as Joe Higgins MEP, could get elected partly because of his or her personal appeal. It is clear the Socialist Party would not gain a seat in a by-election in which Mr. Joe Higgins, MEP, was not a candidate. He is a especially eloquent and exceptional candidate but his party has very little public support. I believe it was 2% in the most recent opinion poll. The idea that voters would automatically have given his representative the nod is questionable at best.
Matters could get very complicated if the individuals on the party ticket fall out and that is not beyond the bounds of possibility.
It could well happen.
Professor Gary Murphy
It could well happen. People have moved parties in the past. It could be argued that by-elections are useful tests of public opinion. The election of George Lee and Deputy Maureen O'Sullivan last summer could be seen as useful tests of public opinion.
The main advantage in retaining the system we have is that it is easy to use and voters understand it. Seán Lemass's dictum expressed 50 years ago that what happens to a third preference is understood by nobody, neither by people in the Dáil nor those outside, misses the point that because of the way it is designed, PR-STV gives voters the opportunity to know full well for what they are voting, and this should be considered before any change.
It could well be argued that the most radical thing to do is to use plurality, which is first past the post, for by-elections, whereby whoever gets the most votes wins. As only one seat has to be filled, and political competition is much tougher than it was previously, a shoot-out for number one votes should be the way it is decided. The drawback is that the plurality system was rejected by the Irish voters on two occasions and its use for by-elections might well be seen as softening up the electorate for the radical alternative of introducing it nationally.
By-elections can have very significant national consequences. Deputy Howlin spoke about Eric Byrne's non-election and some examples of significant by-elections include one that had an effect on the second inter-party Government in 1957. More recently, the Labour Party's gain from Fianna Fáil in June 1976, which occurred prior to opinion polls becoming as sophisticated as they are now, seems to have resulted in the hubristic overconfidence of the then coalition Government, which was routed afterwards. Fianna Fáil has also been a victim. Jack Lynch's tragedy in 1979 in my neck of the woods led to his demise as Taoiseach and leader of Fianna Fáil. In 1994, when there was a change of Government during a Dáil session, gains by Fine Gael and Democratic Left enabled a Fine Gael, Labour and Democratic Left Government to be formed, which would not have happened if we had had a system of replacement such as that which I outlined earlier.
Notwithstanding the relative national importance of by-elections, I am not particularly convinced that tinkering with the by-election system for casual vacancies without full electoral change is the way to go. It may be that without a full-scale change of the system, the filling of vacancies via PR-STV may be as good a method as any.
I thank Professor Murphy, whose last words were interesting. I welcome Mr. Hermann Schiavone and thank him from travelling from Malta to assist us. We appreciate it very much.
Mr. Hermann Schiavone
I thank the Chairman. I am very privileged to have been invited to give evidence before the committee. As Professor Farrell stated, there are a number of divergences between the two systems and I would like to focus on some aspects of the Maltese system which are different from the Irish system. I will begin with an overview of the Maltese system by highlighting some of its salient features. The system was introduced in 1921 and the islands are divided into 13 districts, each returning five members of parliament. The average quota is 3,700 votes. In the most recent general election the number of eligible voters was one tenth of the number in Ireland, at approximately 315,000. The system is applied in general, local and European elections. Malta has moved from a multi-party system to being almost a two-party system. Support for the Maltese political parties is not only sharply divided along partisan lines but also between the two major parties. Only 1,500 votes separated the two major parties in the most recent general election. Voter turnout is almost universal in Malta. The 93% participation rate in the last election was the lowest since 1971. As there is no legal obligation for voters to cast their votes, the high turnout reflects the electorate's genuine motivation.
Since 1987, the first preference vote has had a dual preference in terms of electing the Government as well as choosing among candidates. Parties have an incentive to nominate multiple candidates because a few extra first preferences may decide the balance of power. One consequence of this system is the occurrence of intra-party competition and friction among candidates. There is also a repercussion in regard to the number of counts required to determine elected members. Rarely has an election been determined in the first ten counts and normally between 12 and 25 counts are needed to determine the five elected members. It also has an impact on the ballot paper. The longest ballot paper in the last general election listed 28 candidates. Normally voters allocate their top two or three preferences judiciously but then they simply go down the list. This has an effect on the outcome for individual candidates.
This would particularly affect someone whose name started with "S".
Mr. Hermann Schiavone
Yes, my interest in the electoral system arose after I blamed my father for giving me the wrong surname.
Another characteristic of the Maltese system is that our electoral law permits candidates to stand in two districts simultaneously. Candidates may stand in two constituencies as a sign of status, a kind of insurance policy or, in the case of lesser known candidates, to boost their chances of picking up the crumbs when a vacancy arises or because redistribution has split their base. The consequence of dual candidacies is a higher number of casual vacancies than would be expected were candidates to confine themselves to one district. Some candidates are successful in both districts and as a result have to vacate one of their seats.
In contrast to the by-elections held in Ireland, casual vacancies in Malta are not put before the voters but are filled by means of a recount of the vacating candidate's ballot papers. Any candidate who stood unsuccessfully for election in the district is eligible for the seat provided he or she can win 50% plus one of the votes. As cross-party voting is negligible, at 1.5%, eligible candidates who contested on another party's ticket do not bother to compete in casual elections because they have no chance of getting elected.
The procedure begins with the opening of the parcel of ballots credited to the vacating MP and transferring the votes to the first available ballot preference, skipping excluded candidates. All candidates other than those nominated for the vacant seat are deemed to be excluded. Where the number of nominees is two or higher, the candidate with the lowest number of votes is excluded and his or her votes are transferred to the next available valid candidate. This is the same procedure as that used in general elections. The process of excluding the candidate with the least votes and transferring his or her votes to the next available preference continues until one candidate reaches the quota.
According to a survey I have conducted on the opinions of the PR-STV system, 67% of Maltese MPs are either extremely or somewhat satisfied with the system and only 18% are dissatisfied. In response to a question on whether the system should be changed, 58% said "No".
Since its inception in 1921, the system has generally worked well for Malta. When its mechanics failed, politicians managed to address perceived or real shortcomings by means of creative adaptations. I do not have time to go into further detail now but I would be happy to answer members' questions.
That was a very informative presentation on the electoral system in Malta. It is interesting to note that the proportion of Maltese MPs who are satisfied with the present system is exactly the same as among Irish public representatives. Similarly, 57% of Oireachtas Members do not want the system to be changed, compared with 58% among Maltese MPs. I do not know whether that is a coincidence or a case of great minds thinking alike.
I shall put my first questions to Professor Farrell, who has made a significant input into our deliberations over the past several months. A clear recommendation appears to be emerging from today's discussion on the Gregory method of surplus transfers. I ask the professor to pursue the idea further. I acknowledge that it works in the Seanad because one is dealing with manageable parcels of votes in the election for that House. However, if one is dealing with 90,000 votes in a Dáil constituency, the logistics are more complicated.
This is amplified by the professor's suggestion on increasing the size of a constituency.
If members' questions are kept brief, we may be able to get immediate replies.
That is the only question I wish to ask. I am seriously interested in the recommendation and believe the committee could take it on board
Professor David Farrell
To be honest, I have never had to conduct a count. For all I know, the people in the count centres would be tearing their hair out if we went for that option.
I suggest that if the committee still has time that it could call a witness who has managed a count and ask his or her advice on it. As members will see from the table in my presentation, the bulk of countries that use STV do not use the Irish surplus transfer system. If they can do it there is no particular reason why we cannot do it. Perhaps it would slow the count down but I would hazard the argument that it probably would not slow it down quite as much as one might fear.
The Clerk of the Seanad, Ms Deirdre Lane, could do those things in her sleep. She might have a useful input on the practicalities.
We will ask the clerk to the committee to contact her.
I have two questions. I do not think I have attended a European count but I have attended the last number of Leinster counts. It is almost surreal. One has people tallying on the basis of shopping baskets. There might be 15 baskets from Dunnes and the concern is whether they are bigger than Tesco baskets. People are trying to estimate tens of thousands of ballot papers. If all of the votes remained in play for the entirety of an election, I do not know if that would be feasible without electronic voting, although it would be fairer and better. With all due respect, Ms Deirdre Lane's handling of up to 200 votes is not comparable to a European election where one has half a million ballot papers that would constantly be in play for the duration of the count. People would shoot themselves.
One of the reasons we are looking at the system is because of the waste of time in the representational role of Deputies. There are five Deputies in Wexford, which is the largest of the constituencies we currently have. If an issue arises in north Wexford, five of us turn up. The same happens if an issue arises in south-west Wexford. Five of us generate representations. Five of us are expected to attend events. If one had a nine seat constituency, one would have a football team arriving for every constituency event and generating the representations.
Plus county councillors and an MEP.
It is important to get away from that. The argument has been made that the system we are discussing empowers people but it is not our system. No matter what people say, the electorate wants Members of Parliament to be available to their need. If one does not have that system, and people are excluded from it, my final comment is that one would have what a great colleague of ours, Michael Moynihan, used to call the strong curate. If a Deputy does not attend events then his or her county councillor will attend events and he or she will replace the Deputy at the next election.
I will not ask for a response from Professor Farrell now, he can respond later. Mr. Schiavone would also like to respond.
I have a question for each of the witnesses. The only alternative to the use of the Gregory method is to recount the votes every time. The only pure votes are the first counts or the transfer of the lower candidates. If one wants to avoid fractions one would have to recount the votes every time a quota was passed and reset the quota. That is the only way of keeping all the votes in play. As Deputy Howlin indicated, it depends on scale.
In terms of by-elections, one of the arguments against the use of a named candidate as a replacement on the ballot paper, as was outlined is if there is a falling out between the individual and the party or between the individuals concerned. The situation could arise whereby an individual named as a replacement switches political allegiance even before it is necessary to make a replacement. That is one of the bigger complications that could occur in that situation. It would be important to consider how one might avoid it and whether parties would have an ability to replace people in such circumstances.
I have a question for our Maltese guest, which I raised previously. The sample ballot paper that has been circulated shows a unique element of the Maltese system, the fact that both the National Party and the Labour Party run more candidates in each constituency than there are seats available. It seems to be a kind of primary process as well as an election process for both those parties. They do not have a pre-selection process. Whoever wants to run on a party ticket runs and the electorate makes the choice subsequently. Perhaps Mr. Schiavone would comment on that.
Does Mr. Schiavone wish to comment on that point?
Mr. Hermann Schiavone
Parties have an incentive to over-nominate because as far as the parties are concerned at the end of the first count they would know whether they won the election. They rarely turn away any candidates.
The voters normally allocate all the preferences so parties are not worried about losing out because of over-nominating. They would win a few extra first preferences if they nominated someone close to a community such as the village doctor. That is why lists are normally of that nature.
It is all thought provoking. One could have much discussion on the matter. The question of having 30% women candidates seems reasonable. Most parties are trying to do that, but it does not always work. The party has to win seats as well and the seats make the Government. The individual constituency does not have all the freedoms one might think. At the end of the day the people are looking for a Government and one has to ensure one can achieve that. In a three seat constituency with a number of parties and independent candidates also running how does a party have 30% women candidates it has only two candidates? One would need to have constituencies of at least four or five seats, which are better anyway. I accept we have a body that decides on the number of seats in constituencies since the time of Jack Lynch. The number of seats in constituencies is being increased. It is much easier to achieve a quota when one has at least a four seat constituency. One has to take into account the practicalities in terms of winning a seat in a constituency at a particular time. One might say forget the practicalities and take whatever turns up. That is fair enough. That might be necessary, but one is talking about building it into the Constitution. Parties are trying to get 40% women candidates. There are many reasons why so many of the women who have been involved have left. It is a tough area vis-à-vis family and for many other reasons. The number of women coming forward is increasing. Constitutional difficulties and difficulties in terms of practicalities could arise in small constituencies.
People frequently talk about reducing the size of Parliament. One should not forget that in reducing the size of Parliament one is reducing the level of service available to people and the involvement with them. My party sought to introduce single seat constituencies on two occasions. One is a representative of the people. The fact that the media does not recognise that is nothing to do with the Constitution or the people. One is there to represent the people. One can do that easily by giving the resource that is needed. I mentioned previously that when I was first elected we had one secretary for eleven Members. Everything had to be supplied from outside, which is not a great idea. When I went to America to discuss matters with Ted Kennedy, the Lord have mercy on him, he had an office where he dealt with people. He was able to deal with them because he had the minimal resource. It is necessary to provides resources for Deputies in the future and to continue to improve the public service. I have been working to improve the public service for a long time, even before I became a Deputy. We have done a great deal to do that.
People still fall between the boards and will not be picked up and but these cases take time because it takes personal time. They get into complicated situations and the service does not do what is should. A Deputy, on the other hand, tries to look after the person and get the service to respond in different ways. I am not that mad about reducing the number. The number can be reduced but if Deputies are supplied with the resources to do the job and quibbling about this stops. We must remember it is all about having active democracy on the ground and we have a good system at the moment.
I agree with Professor Farrell about changing the system. One would have to use a computer to analyse the preferences and transfers. The other submissions explore other aspects that are relevant to our deliberations. Malta has a two-party system but we will never have such a system. There are active and high levels of participation in Ireland and people take politics seriously. I thank the delegates for their contributions.
I apologise for my late attendance. I had other commitments and I have only had a brief opportunity to read the documentation. I would like to put a contrary view to Deputy Woods, which Deputy Howlin articulated earlier. More representatives does not necessarily mean better representation for the public. If that is the argument, there would be no limit to the number of Deputies that might be appropriate, other than cost. One might say we should have 200, 300, 500, 1,000 and people will have even more access but it could become ridiculous.
I am putting the contrary view to the notion of multiplying the number of Deputies available to people. What is the option the person is exercising when he or she enters the polling station to vote? Is it who got to Enniscorthy, for example, first on a given night and who got in touch first the following day with an e-mail? Is that how the person has to decide who to vote for? The dilemma is what is the basis for the system. The reason we are having the debate is to see whether we can improve the system of representation. Everybody knows his or her politician. My knowledge of other countries in this area is limited but everybody knows politicians in this State. Irish politicians have the closest contact with the public. Few people do not know a Deputy or councillor. Until two years ago, we had a Taoiseach who had met practically everybody in the country. This presents a paradox where people feel they know and like their Deputies and they can meet them in pubs and restaurants but the quality of representation is an issue.
I did not suggest an increase in the number of Members but the way in which they are elected should be reorganised in order that, for example, more women are on the ticket and so on.
The Deputy intimated the number should not be decreased.
Not hugely.
Professor Farrell stated the weighted, inclusive Gregory method is the fairest but this can only be done on computer. Could it not be done manually by recounting all the votes and then, on a proportional basis, assigning them to the candidates as they come up for elimination?
There is an eternal debate about constituency work versus legislative work. I am not aware of any serious academic work that defines what is constituency work and what is legislative work. One merges into the other because many Bills emanates from constituents raising problems either on a personal or national level. If work has been done in this regard, could the other academics direct the committee on where we might find it? It is increasingly becoming a defining issue as to whether we do too much constituency work and too little legislative work for the committee and the media. We should have clear definitions about what each constitutes if we are to discuss this problem in depth.
Senator Boyle asked a specific question of Professor Murphy. Deputy Howlin put a question to both Professor Farrell and Mr. Schiavone about constituency size. The witnesses can reply in general to the other questions.
Mr. Schiavone mentioned that he is analysing 65 MPs who represent 300,000 voters. On that basis, we would have 650 Deputies. Has he been criticised for using too many MPs relative to the population he is using? He is standing in the next general election and I wish him the best of luck.
Professor Gary Murphy
With regard to Senator Boyle's question about party replacement, he will recall during the European Parliament that sat between 1979 and 1984, it was an option for parties to replace a candidate with anyone they wanted. We had the surreal situation where the four Labour Party seats were held by no less than 11 people and this came to the attention of the European Parliament's credentials committee for obvious reasons. That is the danger of not putting in place a specific candidate. The risk remains that there might still be a tremendous falling out for one reason or another but if by-elections are to continue, it is incumbent on parties to put forward to the voters who the alternate would be because one does want a scenario where he or she is appointed at the whim of the political party.
What is the question on constituency size?
It relates to the conflict about representation. A football team of Deputies could arrive at events, for example. Either the local representative role that people want is maintained or Deputies have a constituency of manageable size.
Professor David Farrell
The issue about electoral system design is there are trade-offs. Almost any decision taken will have an impact on something else. I would never suggest this is the big thing that has to be done. I would only see this in the context of other changes under consideration. The Deputy is correct and he knows much better than I do that by increasing the size of the constituency, the likelihood is if all else remains the same, busloads of Deputies will be going around from one event to the next. I would see this happening in the context of other developments, particularly in the context of the possibility of a proposal to reduce the size of the Dáil. If we reduce the size of the Dáil, parties such as the Green Party will be devastated because proportionality will plummet. The only way to protect the chances of small parties would be to increase the size of the constituencies at the same time. There is a rule of thumb——
It would be a disparate Dáil, with small numbers and many minorities. That would be the impact.
Fianna Fáil is very good at coalitions.
Professor David Farrell
We would have to go a long way before we got to that level.
Core principles and all that.
Professor David Farrell
From the studies that have been done on what is required for a proportional system, the generally accepted rule of thumb is that one needs an average constituency size of five Members, which we are nowhere near. We could increase the size of the constituencies a little, to try to bring us up to something like five, without necessarily having a drastic impact. However, I may not persuade members.
Could Mr. Schiavone respond on constituency size? Is a constituency size of five appropriate in Malta and is there any move to increase sizes?
Mr. Hermann Schiavone
We have a top-up system as well. Actually the size of the Parliament at the moment is 69 MPs because we have a constitutional amendment that provides that parties are propped up with seats to reflect the proportion of votes they received. This happened after a perverse result in 1981 in which the party gaining the absolute majority of votes received a minority of seats. What happens is that the aggregate number of first preference votes determines the number of seats in Parliament. The winning party will top up the seats to reflect the proportion of votes. That is about it.
The constituencies have five members?
Mr. Hermann Schiavone
Yes.
Is there any move to increase that size from five to six or seven or to reduce it?
Mr. Hermann Schiavone
No.
Professor Gary Murphy
I would like to reply explicitly to Deputy Howlin's point about a disparate Dáil. I do not think it would necessarily be a disparate Dáil, because proportionality would reduce that chance, but it would give us the opportunity for that proportionality to be suggested in terms of representation through Dáil seats. Voters will still have the final choice not to vote for any Green Party members or Labour Party members at all if they so wish.
Professor Farrell, would you like to respond to the other matters?
Professor David Farrell
I will dwell on just two matters and will be as quick as I can. Senator Boyle and Deputy Devins were addressing similar points about an alternative to the Gregory method being to recount all the ballots. That is probably the Meek method, which I did not bother to put in my table because we would be getting to levels of complexity which are wonderful for social choice theorists. If the committee wishes it can get someone from New Zealand, where the system is used for local elections, to talk about that. Computers are definitely required. Bearing in mind what Deputy Howlin said about shopping baskets, if one is to recount every single ballot paper at every stage, the levels of complexity are huge.
The Gregory system is used for Northern Ireland Assembly elections, which are on a scale that is not so different from Dáil elections. If the committee intends to meet with somebody from the Seanad, it could also arrange a meeting with somebody from Northern Ireland to talk about how the Gregory system could be used. It is being managed in Northern Ireland without computers.
Deputy Woods was talking about a national quota for women. The point that has been made by others — not just by me — is that we could have a change to the party registration legislation. The legislation could be amended, without the requirement for constitutional change, to say that a proportion of a registered party's candidates nationally must be female, and it would then be up to the party to distribute these across the constituencies. If the figure of 30% is seen as a leap too far, which is understandable for all the reasons that have been explained, why not pick a lower percentage to start? We could pick 20%, see how it goes and build it up later on. It is a relatively simple change that would avoid the necessity of changing the Constitution.
We must overcome difficulties of that kind that might arise.
Mr. Hermann Schiavone
Deputy Jim O'Keeffe asked whether there are too many MPs in Malta and whether there is criticism of the system for this reason. The answer is "No." Actually, there was talk at one stage of whether we should increase the number because of the work associated with our EU membership. There are not too many MPs.
The MPs are part-time, are they not?
Mr. Hermann Schiavone
Yes, they are part-time. Their remuneration is around €80,000 and they spend a third of their working week on constituency work, a third on legislative work and a third doing their professional work.
Is there a system of local government in Malta?
Mr. Hermann Schiavone
Yes, there is. It has been in place for 15 years, since 1995. The strengthening of local government has resulted in a considerable reduction in the constituency workload of MPs.
That is what Deputy Howlin is afraid of really.
Professor Gary Murphy
In response to Deputy Devins's question about legislation versus constituency work, there is the question of strengthening the local government system and putting in place county or city councillors specifically for such constituency work. The dual mandate was abolished.
One would have to bar a person from standing for the Dáil.
Professor Gary Murphy
I know that, but what we could do is to strengthen local government substantially so that all county councillors would not be traipsing after Deputies because they want their jobs.
I would not say that.
Has research been done to define what is constituency work and what is legislative work? The media has a certain image of it, but I am trying to find out whether there are any hard, statistically proven results in this regard. People use the term "constituency work" to imply that somebody is looking after potholes and so on, while legislative work is seen as spending time in the Dáil. However, a Deputy might be up here talking about the National Roads Authority, which is governed by legislation. Has anybody gone into the office of a Deputy and monitored his or her work to determine whether it is constituency or legislative work?
Professor David Farrell
Such work has been done. It is mostly the classic literature on parliamentary studies by authors such as Bruce Cain and Dick Fenno. There are a couple of other classic books. The way it was done was not so much through survey work but through working with members of parliament and observing their work and what they see as being one type of work or the other. There is a good classic literature out there, but it does not produce firm delineations of what is legislative work and what is constituency work.
Deputy Michael D. Higgins has published on clientelism as well.
Professor David Farrell
Yes, and Lee Komito. There have been some studies but nothing recent. It would be wonderful to get a few PhD students to spend time with Deputies if we could do that. However, that might be for the future; I do not know whether it could be done in time for this committee's deliberations.
It is a field that is open for exploration.
One of the points Deputy Devins was making was that one gets ideas of what kind of work is done. One can read all one likes but one must put one's foot on the ground to find out what is really happening. If one does not do that one will not really know and will not have the conviction to implement changes. A connection must be maintained. It will be much more sophisticated as time goes on. As I said earlier, if the public service does its job better — although it has been improving all the way along — not as many cases will fall through the cracks.
Much depends on the system of management. I always believed in walking the factory floor. One will hear that anywhere one studies management: if one does not walk the factory floor, one does not know what is going on. Deputies need to know what is going on. They do not have to be at every meeting. I know they are always at meetings, but they do not have to do that. However, they do need to know what is actually happening to people; otherwise, they become distant from them.
My personal assistant is doing a PhD in this area, and she places much importance on the distinction between clientelism and brokerage. Would Professor Farrell see that as an important distinction?
Professor David Farrell
It is not an area on which I have focused, so I will leave it for others to comment.
We could wait until she produces her paper.
Professor Gary Murphy
There is an element of political culture. In many ways, Deputy Devins has answered his own question. As with practically every other Deputy, he would see himself as part legislator, part——
Local ombudsman.
Professor Gary Murphy
Exactly. As far as I can recall, Deputies are not dissimilar to their counterparts in France or Britain in terms of the amount of so-called constituency work. In other systems, such as those in Scandinavia in particular, there is a distinction. Irish political culture relates to the issues of multi-member constituencies and competition between parties. Unless there is a significant root and branch change of the entire electoral system, the question is moot in one way. When a Deputy or councillor puts himself or herself forward for re-election, he or she is not simply standing as a local ombudsman. Potentially, he or she could be standing as a future Minister in a potential Government. The surveys of ambition that we have examined tend to suggest that most people who enter public life would like to see themselves in the Cabinet at some stage.
Some have instant ambition.
Professor Gary Murphy
I will not comment.
I thank the gentlemen for their presentations. There is no other business.