of the

 EVIDENCE OF:

Dr Denis Balsom

held at

THE ROYAL WELSH SHOWGROUND, INTERNATIONAL PAVILION

 on

THURSDAY 8 MAY 2003

 LORD RICHARD: Thank you for coming back.

Dr DENIS BALSOM: Thank you for having me.

 LORD RICHARD: Why do you not just open it up?

Dr DENIS BALSOM: Thank you. I just wanted to say that the election was not a terribly long time ago, so there has not been a huge amount of detailed analysis of the results as yet. There will, of course, be some studies published by the Institute of Welsh Politics and others, based on surveys that were done on polling day and subsequently on how and why people voted and they way they did so. But it will be some time however, before these become generally available.

What I have tried to do here, is to put in a paper that is complementary to the submission you had in April. The basic thrust of my argument, or my critique of the system, remains the same, but uses more recent electoal evidence to illustrate these points. Unfortunately, I did not see the questions which you drafted for me until I arrived a few minutes ago. Some of them I think are touched upon in the paper I have circulated. Others I am very happy to come back to and discuss more generally, as I am to discuss any aspect of the election that you want to bring up.

The paper really reiterates many of the points that I raised before, put into the context of the recent election. Last time I did not discuss election turnout, and obviously that has been an important aspect of this particular election. I feel particularly sensitive on this point as I was involved in the opinion poll that was published a few days before polling which seemed to suggest that 51% of people said they were ‘certain to vote’. I am not quite sure what happened to their certainty, but that in itself is quite interesting because similar responses to past surveys has been extremely reliable. When people say they are ‘fairly likely’ or ‘very likely’ to votewe know they are not going to vote. But when they actually say they are ‘certain’ to vote, in the past that has been a good guide. It was a good guide in 1999.

I don’t really think anything happened between when we were interviewing, up until the Monday night before polling day, and polling day itself. I know some people have talked about the two peaks of Labour voting, particularly early in the morning and then after tea time. It was certainly wet in parts of South Wales after tea, but I think the issue of poor turnout is going to take some further investigation.

There is very little that I have in this paper that really addresses the question of turnout, other than to reiterate the point that generally speaking, people tend to become engaged and keen to participate when they feel their vote is important. So in some of the closer constituency races, we saw higher turnouts. This is a general rule of elections and it certainly happened this time.

What probably did not convey itself terribly well, but certainly the parties were making the point, was that the overall outcome of the election, and whether or not the Labour Party would achieve a majority, meant this was an extremely tense contest of which the outcome was uncertain. The idea of using this uncertainty as something with which to motivate the people was tried by the parties, but it clearly didn’t convey itself with any great effect.

Underlying that, there is a case to say that the present electoral system we have is still misunderstood in some ways - the distinction between the two levels in the election and the two votes people have. Again, I think this needs more detailed research of the kind that the Electoral Commission are undertaking. I was rather intrigued by the Liberal Democrat party political broadcast which emphasised over and over again ‘the big ballot’, as if this was the only way the public were going to understand the significance of their two votes when they were given their two ballot papers.

Of course the Liberal Democrats also made a great ploy really to say vote Liberal Democrat in order to deny the Conservative Party their regional seats. I sometimes felt that tone of campaigning was not very elevating.

If we look at the constituency election, the most significant thing about 2003, is the difference between the share of the voting and the share of the seats was more disproportional than it was in 1999. This is specifically in the constituency section.

Another point, because I emphasised it last time I spoke to you, was that I do fundamentally believe in the single Member/constituency relationship. I think that is important and I think that John Marek’s ability to win Wrexham as an Independent shows that perhaps the public have an interest in this relationship as well.

 LORD RICHARD: What was the turnout in Wrexham? Do you know?

EIRA DAVIES: It was 34%. It was very similar to the turnover in 1999.

Dr DENIS BALSOM: It was 34%,. exactly the same as in ’99, yes. Whether or not it would apply to a candidate who is not as well-established and as well-known as John Marek is, of course, a different point.

Moving on to the regional election, I still feel that the areas as defined do not mean very much in themselves to the voters. If you think about South Wales, whether one is one side of the line in South Wales Central or in South Wales East is not quite arbitrary, but it is not a very obvious border either. Yet these ‘virtual’ reions can have a major impact on the outcome of the election.

There has been a lot of talk post-election about representation of North Wales in the Cabinet. I can tell you the one Cabinet post I heard announced on the 12 o’clock news is from North Wales is Karen Sinclair, who has become the business manager. Whether or not there will be others, we will know later today. Even the electoral definition of North Wales does not perhaps satisfy the kind of representation that people might be looking for, either from Northeast Wales or Northwest Wales.

An appointment which just fulfils a rather simple sense of representation, based upon the electoral region, is perhaps not tackling what people mean when they say there should be more diverse representation in the Cabinet. Perhaps the important thing is that at least it would be somebody who was not from Cardiff.

HUW THOMAS: There is a point I just wondered, wanted to look at. I picked up, It is totally unscientific, but I’ve been surprised by the number of people who mentioned to me afterwards their cynicism about the regional list, where people who stand in the constituency and are defeated are nevertheless returned. Their comments to me have been, ‘Well then, why bother voting?’

Dr DENIS BALSOM: Yes, we had a discussion on this point when we met in Haverfordwest. Unfortunately, Huw, I don’t think you were with us. The results for Clwyd West, when all four candidates have ended up as AMs was to be anticipated. As I think I mentioned last time, somebody said to me in a similar vein, that it reminded them of a golf club election, where there were five candidates for four seats; four were elected and the other one was co-opted. Therefore, what is the point? My principle concern is how the candidates for the additional member system are put before the public, and perhaps we can come back to that in a second.

Obviously the allocation of additional members from the regional election has corrected in part the disproportionality of the constituency contest, but it still strikes me rather artificial and less effective than perhaps an all-Wales list might have been. Without picking on a particular party, but perhaps if there was an extreme example: if you think of two of the new Conservative regional members – Lisa Francis and Laura Jones – had the Tories known at the end of the count they were entitled to ten top-up seats, as they were, I cannot believethat these two people would have been their next two choices. There may have been other candidates within the party hierarchy, or whatever, who they would have preferred to see in their Assembly team.

It seems to me that this is something a party ought to have more control over. Those AMs were not voted for as individuals; they are there because the electorate voted for their party. It is almost coincidental, perhaps the Tory nominations for the list for South Wales East were not complete and somebody like Laura – who I have met and I am sure has a very successful career ahead of her – ends up number three and gets elected. In a sense the outcome goes against the grain, whereas somebody perhaps like Peter Rogers, who had been I think a fairly forceful member in the previous Assembly, fought a very vigorous constituency campaign to secure second place, which he was not necessarily expected to do but was not highly placed on the North Wales list and is therefore is out. But there we are.

The outcome I of the election reinforces the point I made last time, that the parties do not share an equal interest, and will not share an equal interest, in the two halves of the election come the next contest in 2007. The Labour Party have not returned a single list member, the Tory Party 10 out of 11, the other two parties are more balanced in this respect.

We saw the Labour Party campaign very, very explicitly on asking their supporters to give them two votes. We know that in many regions, particularly the South Wales regions, a second Labour vote is effectively wasted. It strike me as being somewhat cynical that the party would prefer to have electorates waste their votes than have them go to anyone else.

 LORD RICHARD: Why are you surprised at that? I have never heard of a Tory knocking on the door saying, ‘One vote for me and a vote for the Liberal Dems.’ It sounds very strange.

TED ROWLANDS: It would be an extraordinary thing to do.

Dr DENIS BALSOM: Yes, but presumably you would not knock on the door and say, ‘Vote for me and tear up your ballot. Don’t return the rest.’ If you have invented a system which requires the participation...

TED ROWLANDS: In the back of a party’s mind, anybody’s particularly a campaigning party’s mind, is that if in fact an elector does choose to vote something else other than the party, that might become a more regular permanent choice. It is not a question of wasted votes; it is a question of good solid thinking about trying to maintain your core vote, to not split your core vote up.

 LORD RICHARD: I think you are asking for a degree of sophistication on the part of...

Dr DENIS BALSOM: Of the parties or of the electorate?

 LORD RICHARD: Of the electorate.

Dr DENIS BALSOM: Well, we came from an Assembly where the Labour Party was in coalition. It only just avoided going into coalition again. In other parts of the world, where systems similar to this are in place, one would have expected the party to have a view. If we cannot elect anyone else, then it would be better that the party we wish to be in coalition with had the benefit of our support. I accept your point, that it is very difficult in the culture we come from to go on the doorstep and make that kind of appeal. Possibly the system we have, though it would not happen perhaps if the Assembly was larger, is still predicated on the basis that yes, you can possibly squeeze a majority out of it. Clearly that is not the case in Scotland, and it probably would not be the case in Wales if the ratio of seats between the constituency...

 LORD RICHARD: Is there any evidence that any other parties did what you think the Labour Party ought to have done; in other words, told people ‘One vote for us and one vote for someone else’?

Dr DENIS BALSOM: For no other party would the second vote be wasted.

PETER PRICE: Just to pursue it for a second. Germany is the example, where the CDU had implicitly and a little bit explicitly encouraged people to vote FDP. I think that is fair, is it?

Dr DENIS BALSOM: Yes, and I think it does happen elsewhere as well. But this requires parties to talk sensibly about coalitions before the election and not afterwards, as, I think, is implicit in Scotland. But, as I say Wales offers a tantalising opportunity for a majority, which on this occasion did not quite materialise. But had it not been for Wrexham there would be a majority. I think Rhodri Morgan is quoted as saying that he felt that this system would deliver a Labour majority three elections out of four, and of course they started with election number four. But maybe it is the other way around; maybe the system will deliver a majority one out of four rather than three out of four.

I think the next point, about the balance of party nominees on lists, we have already covered in response to Huw’s question, but I think people do not quite understand how it is; the idea that people have a list nomination as a safety valve. In this respect the list is again, my view, deliberately manipulated so that a candidate, a very prominent candidate perhaps, who is almost bound to be elected anyway, appears at the head of the list, as happened in South Wales Central. In fact the Labour Party generally listed their candidates, with the exception of the six minority candidates who were then spliced into the list, on the basis that the person with the biggest majority went number one. In a sense those least likely to have any recourse to the list at all were put there as your prime list candidate, which strikes me as misinformation of a kind. But it would of course reinforce Ted’s point that parties are trying to maximise their party vote, irrespective of the consequences.

Although a number of minor parties did put up lists in each region, unlike in Scotland no candidates were elected. One of the questions you have raised with me is about what the vote threshold for this election might be, obviously it is dependent upon turnout as to what that level would be. I do not have all the calculations in front of me now, but it seems to me, recalling from the night, the threshold for election by the time you get to the fourth candidate in each region was perhaps about 12,000 votes.

If you take the Marek group in North Wales, they got 11,000, but, of course, having already elected one member they would not have been entitled to another member with this level of support. As soon as you get into the calculation their vote would have been halved. By comparison, the Greens got 30,000 votes overall but obviously not enough in any one region to get to the threshold. UKIP got just under 30,000 votes overall but again, not enough votes in any one region.

TED ROWLANDS: The UK Independent Party got how many votes?

Dr DENIS BALSOM: 29,427.

 LORD RICHARD: Were there any BNP candidates?

Dr DENIS BALSOM: Yes, there was a BNP list in South Wales East which got 3210 votes, so it was not in contention for that region, nor would it have been elected under any national list.

 LORD RICHARD: Those were the minor parties, were they?

Dr DENIS BALSOM: The big ones, Socialist Labour, Communists, Independent Wales Party, all fought as parties, as well as the usual fringe candidates, such as Captain Bean. An interesting one in Islwyn was actually: Tinkers Against The Assembly, actually got over 2,000 votes, but did not stop Labour successfully recapturing the seat.

Looking at how the seats might be distributed under an all-Wales list is obviously the main point that I have tried to bring to your attention. Unfortunately in the table on page three under the 60-seat solution the columns for Labour and Conservatives have been juxtaposed. It should be 30 seats for Labour and 11 seats for the Conservatives, though the percentages are correct.

I think there are two points here which are fairly significant. First of all, the degree of disproportionality achieved on the first ballot is so large that it is never going to be corrected by allocating just 20 additional members, whether that is done on a regional basis or on a all-Wales basis. If the number of AMs were to be increased, as some others propose – not necessarily for electoral purposes but for the purposes of how the Assembly works and so on – then clearly there are more seats to be allocated and that could help correct that disproportionality and the ratio between first past the post seats and AMS seats would be different, in this case 50/50, which would slightly exceed the balance in Scotland. One of the reasons Scotland does have a multi-party system is as a consequence of that balance.

The 80-seat share at the bottom of the table, gives an outturn of share of the seats remarkably close to the votes as they were cast, if you compare the bottom row with the top row. Of course it does not give Labour a majority. In fact, even allocating 40 additional members, the Labour Party, on the basis of votes cast last week, would only have won three more seats.

Again, just to reiterate the points I made last time, I think that if in the course of your deliberations you were to advocate that the Assembly should be made larger, I would hope that those extra seats would be alloctaed to the list rather than in some other way. It seems to me there is a case that can be made for an all-Wales list: it would negate the danger of wasted votes, it would give each party an equal incentive to compete in the AMS election, I think it would also allow minority parties to compete more equitably against the main political parties. Not necessarily that they would win, they would still need to achieve a substantial threshold to get in, but they could concentrate their resources and put forward their principal spokespersons rather than being forced to provide slates across all five regions.

Similarly, it has been suggested in the past that perhaps candidates outside the mainstream political parties might be attracted to contesting Assembly seats as an Independent. Someone with a fairly high personal recognition factor could secure support on a national basis perhaps more easily than on a regional basis.

How this would work? we talked about this last time. Clearly the list would remain the same, in that a person would cast a single vote for a party. The parties, conceivably, would have published their lists of their candidates, both constituency and others. But essentially, after the election, the party would know how many top-up candidates it would be entitled to nominate and it would choose them. It might choose to include one or two people who had been defeated in constituency contests. It could also choose to address any imbalance in its representation geographically, by gender or to ensure ethnic minority representation.with the party would take a view that it was presenting itself as a collection of politicians representative of Wales and who they would want to put before the electorate and the next election. The final point is put in the revision there…

TED ROWLANDS: Can I just clarify? You say that under this all-Wales list system you would have to go into the ballot box and you would have to vote for a party, but you would, could also have on that ballot paper a list of individual names. How would these national figures get on to the list?

Dr DENIS BALSOM: I do not believe there would be a list of names. There would be a vote for a party.

TED ROWLANDS: Yes, but how would your little [inaudible]? If you could see the potential for national figures from business, sport, or the media.

Dr DENIS BALSOM: They would stand as individuals, as they are able to do now.

TED ROWLANDS: You would be entitled to vote for an individual, if they stood as an individual, but not entitled to vote for an individual if they had a party banner attached to them?

Dr DENIS BALSOM: You do not have that option now; it is a closed list.

TED ROWLANDS: I am just clarifying. Thank you.

Dr DENIS BALSOM: A final point is that political scientists have interpreted the first Assembly elections as indicating that the electorate has a clearly differentiated view of the world, whereby for Welsh elections and Welsh politics it has developed a particular set of attachments and loyalties, but when inWestminster mode, and voting in a Westminster election, then other stimuli and issues would come into effect.

This theory was used to explain the very substantial swing to Plaid Cymru in 1999. Obviously that was undone to a certain extent last week. It seems to me, perhaps less obviously from the findings of the election, but perhaps in the build-up to the election, that during that month first we had the war in Iraq, there was a very general swing to the Labour Party across Britain as a whole. Certainly from some of the evidence we have seen in polls that were being done, both in constituencies and for Wales as a whole, it did seem to me that this gave Labour Party support considerably more robustness than had been the case in 1999. Now there are other factors I appreciate, but it did mean that the Labour Party retained far more of its support, particularly on the regional ballot, than had been the case in 1999. Plaid Cymru were far less successful in winning people over to support them, especially on the regional ballot, than they had been previously.

I just wonder, mea culpa whether we have overplayed this theory and the distinctions between Welsh politics and British politics are not quite so marked as we might have anticipated they were. This is not to say it will not evolve in the direction of greater distinctiveness, but clearly there are all sorts of overlaps. There are, especially, important overlaps in terms of the distribution of the population of Wales, in where people are getting their news from, in terms of media, television, press or whatever.

The First Minister certainly has made a great play that he found it extremely difficult to get any press coverage of this election at all until the last few days. Part of that was to do with the particular circumstances of the war, but we do know that, generally speaking, the print media in particular do not give Welsh politics very much attention.

 LORD RICHARD: Is there any evidence the Labour Party was better liked in Wales than it was in England?

Dr DENIS BALSOM: I am not aware of any specific evidence on that. There was some investigation by the polls as to leadership, or rather the standing of the party leaders. Of course Rhodri Morgan himself, I think, is a uniquely popular politician, no only amongst his own supporters, but amongst the supporters of all other parties – both as leader of the Labour Party and even more so as First Minister. When asked whether he had done a good job as First Minister, the approval ratings amongst his so-called opponents are extraordinarily high. I am sure that had something to do with the overall election outcome as well.

 LORD RICHARD: Nothing to do with the fact the Labour Party in Wales had some relatively old Labour policies?

Dr DENIS BALSOM: I have no even pseudo-scientific evidence for that point. Again, I am sure more detailed studies will be looking at those kinds of issues.

PETER PRICE: Can I follow through that last point then? You mentioned right at the outset that there were studies conducted about why people voted the way they did on the day and things of this sort. Are you aware of exactly whose done those, any indications of the kinds of questions they asked, and when the results will be available?

Dr DENIS BALSOM: No, I am personally not involved in any of those. There has been an ongoing commitment to election studies under the banner of the British Election Study, which I believe Aberystwyth have had the Welsh end of, and I think the ESRC will have funded a study which will appear in due course. I cannot tell you when, though I would have thought some preliminary information would be available relatively soon. The Electoral Commission have also undertaken a certain amount of work, which is not to do with policies or personalities, but about people’s perception of the Assembly, of the election and the election method. Some of that work was done before polling day and there has also been some qualitative work done. I think there is also a post-election survey to be done.

PETER PRICE: Those are the two main bodies?

Dr DENIS BALSOM: That I am aware of.

PETER PRICE: The other question I had was just statistical. You have drawn attention to the very close voting correlation this time during the first and second votes. One can see a pattern if you look at the different constituency votes and region that each party got a little less in the regional vote than in the constituency vote. I assume that is because extra parties come in – like the Greens – and therefore that is why you have got each party depressed by a little. It is fairly consistent across the board that they are so depressed.

Dr DENIS BALSOM: I think that is mostly an illusion. If you look at the 1999 results there were not substantial gaps between the parties on the regional and the constituency ballots. But when more detailed studies looked at the behaviour of individuals, there was a great deal of churn between the two parts of the election; they just happened to largely cancel each other out. So I think the data is slightly misleading here.

Of course the large number of extra party lists which were there in the regional elections did attract overall a fairly high proportion of the vote. If there is any evidence at all that I was aware of from the polls, it appears that the Liberal Democrats were the probably the biggest losers, in that they retained the lowest proportion of their supporters on the regional ballot. I think that is where the dispersment of votes went. The Greens and others did disproportionately well from the second votes of Lib Democrats, even though some other parties’ supporters were also giving their regional vote to the Lib Dems.

PETER PRICE: You mentioned the ’99 figures for the regional vote, but they are not actually on your table. Do you have them handy that one can annotate? The ’99 share of vote.

Dr DENIS BALSOM: Yes, I can give you those. In the regional election, the Labour Party got 35.5, Plaid Cymru got 30.6, Conservatives 16.5, and the Liberal Democrats 12.5. In every case that was within 1.5, or 2% at the most, of what the parties got on the constituency ballot. As I say, that does disguise a great deal of churn.

SIR MICHAEL WHEELER BOOTH: May I ask you a question about the table? On the table on page two of your paper, what seems to be the indication is that you are really recommending going up to 80 on a 50/50 proportion. But one looks at your figures and that means that the present that instead of being able to have a voting majority, 30 out of the 60, the Labour Party would only have 33 and they would have to go into coalition with somebody or be a minority government. It is quite a big thing to ask, is it not? I just ask you whether you think it is not sort of asking them to be rather selfless?

Dr DENIS BALSOM: I think there are a number of points here. One is that the 40 seat/20 seat division that we have, 2:1, is perhaps what Ron Davies intended as an element of proportionate representation, but the outcome is not proportionate. I do not necessarily think absolute proportionality is the ultimate goal in itself and we certainly had this discussion with Dr McAllister who was here last time.. But I think it is extremely valuable that all the principal political parties in Wales are represented in the Assembly. And that the political dialogue that goes on there does represent other people, and we do not have a position where a political party can be excluded, as it can be in Westminster under a first past the post system.

I think one is looking for some model whereby the predominance of the Labour Party in Wales is somehow moderated by a system to provide a House, or a chamber, which represents all parties. I only used 80 because other people I know have been before you arguing the case there should be 80, though probably for other reasons. That would give a seat ratio of 50/50, which even exceeds the ratio in Scotland.

Clearly in Scotland, the balance is such that one would not anticipate there ever being a majority administration. The system is predicated to create a coalition. I would not necessarily say that the system in Wales should be similarly based, or have that as its prime objective, but it seems to me thatto have more seats to undo the greater degree of disproportionality would give the people a greater sense of engagement with the system.

SIR MICHAEL WHEELER BOOTH: Absolutely, I see that, but what arguments could we put to the present Labour administration of the National Assembly that would make them so to speak ‘buy’ your bottom set of figures in that chart on page two?

Dr DENIS BALSOM: They probably would not want to buy them at all. Other than the one we have got, it is whether you want to set out to define a new electoral system. On the basis of political expediency, which I understand, this may be behind closed doors, or whether in putting proposals before the public, one is trying to put forward a system which has embedded within it, fundamentals of equity of some kind.

It seems to me that the position I am trying to outline has to be consistent with the broader British political tradition. I feel quite strongly that the individual Member/constituency is very important. Therefore, if we are going to have a moderating factor to create a more inclusive Assembly, then whether it is 40/30, 40/40, or whatever is something we can discuss. However, what we have seen in two elections is that 40/20 is not going to undo disproportionality and is not necessarily going to deliver the major party a majority either.

SIR MICHAEL WHEELER BOOTH: I have one more question. It is really about the arguments for the list members to be on a single list but putting in the power of the parties the choice of members. What are the arguments for that? It would seem to me that if you would have a single list, you could have the names given and it could be open. The electorate would be able to say they did not like Alun Michael, they did like Rhodri Morgan, just for the sake of taking two names.

PAULVALERIO: Can I just add to that, because that was the basis of my question? Do you not think it bad for democracy when we have a low turnout in an election, so that means fewer people are interested in it, there are less party activists across the board, therefore we are increasing tremendously the power of a small caucus of party activists in creating these politicians. Is that not generally a bad thing?

Dr DENIS BALSOM: Do you mean now or under the proposals that I am suggesting?

PAULVALERIO: The proposals.

Dr DENIS BALSOM: Actually, most of what you are saying would apply now, as well.

PAULVALERIO: At least people have the power to rank candidates in virtually most of the parties.

Dr DENIS BALSOM: Only within the party. And they would have presumably, have a party choice to exercise their ability to nominate say ten top-up members, it would be for their own purposes. It might actually encourage people to get more involved.

PAULVALERIO: Would the open list not...

Dr DENIS BALSOM: I am not sure how an open list would work in this context. First of all, presumably people on the open list could not be included at all on the constituency list. So that would force a difficult choice for many people. I think I do accept that parties do not have the ability to put up a discrete list of 40 candidates in 40 seats, and a list of some length and accept that, in the case of the Conservative Party, 39 of the people they put up for seats could go to the Assembly in this way. I think there does have to be some means of combining them, but in a less crude way than the point that Huw raised, where people get soundly beaten in their individual contest and yet they end up being elected anyway. It would be much more explicit if it was a party nominee; in other words, the party’s put them there, not the electors of South Wales East or whatever.

 LORD RICHARD: I do not quite follow actually; it deals with Huw’s point. You have somebody standing in the constituency against people. There is then somebody who the party actually wants in the Assembly. You then got the elections list. They are going to have to tell you in advance who their candidates are going to be, but they do not have to. If you have somebody who is beaten in the constituency, they are perfectly entitled under your system to switch over to the list.

Dr DENIS BALSOM: The point is they would not be voted in from the list. You vote for the party on the list.

 LORD RICHARD: It is effectively exactly the same.

Dr DENIS BALSOM: Not quite. If you take the Clywd West example, where all four principle candidates have ended up being elected; three candidates have gone to the Assembly in that way. Which hat are they wearing? Yes they are there because of the list election obviously, but there is a sense surely amongst the voters of Clwyd West that they had been rejected.

They often define themselves. The point I think I made about Nick Bourne last time, is that when Nick Bourne talks about his constituency, is he talking about Brecon or is he talking about Mid- and West Wales? He is there as a party nominee. I think if these people became national members for their respective party, one gets round some of that confusion.

 LORD RICHARD: You would still end up with the situation in which defeated people in Clwyd, the parties could put them on their respective lists after the election, and they get in.

Dr DENIS BALSOM: They would be part of the party’s pool of candidates, I accept that point. They need not necessarily all be put on the same slate.

VIVIENNE SUGAR: I am struggling to understand how this would work as well. If we had an all-Wales list for the present 20 top-up Members, do you really think that somebody is going to stand in a polling station and put 20 crosses on a piece of paper which would have 250-300 names or parties on it?

Dr DENIS BALSOM: They would not do that; they would vote for a party, as they do now.

VIVIENNE SUGAR: Sorry, your paper is suggesting that both are possible, that you could have individual names, say I want to vote for Cerys Matthews and Colin Jackson, but then I want to give my other 18 votes to one political party.

Dr DENIS BALSOM: You do not have 20 votes. You only have one.

VIVIENNE SUGAR: You only have one, even though there are 20 seats? Keep going, I am trying to understand it.

Dr DENIS BALSOM: It is just like in your region, you would have elected four members this time; you only had one vote.

VIVIENNE SUGAR: But that is for part of Wales.

Dr DENIS BALSOM: Whether it is part of Wales or all of Wales. When you vote in the European election you have one vote in an all-Wales election.

VIVIENNE SUGAR: Why is STV not a simpler system? You kind of dismiss it at the end of your paper.

Dr DENIS BALSOM: Because STV destroys the Member/constituency link.

VIVIENNE SUGAR: Not necessarily.

Dr DENIS BALSOM: In my view, absolutely.

VIVIENNE SUGAR: No, we are talking about a national list now.

Dr DENIS BALSOM: But STV would not work in a national list.

VIVIENNE SUGAR: Right, why not? It would depend on how many seats you were electing on a national list.

Dr DENIS BALSOM: If the nation is one constituency, then it is an all-Wales list whether it is run as a list or however else you wanted to run it.

VIVIENNE SUGAR: I am still not clear. You are saying that you could not have an STV system at all in Wales unless you broke the constituencies link?

Dr DENIS BALSOM: I am sorry, I have gotten slightly ahead of you. If one was advocating STV, then presumably you would not have the two-tier election. If the regional election was organised on an all-Wales ballot, yes it would be conceivable that you could run an STV system, but I think the evidence of most STV systems is that the optimum is to go for units which would elect four or five Members, not 20. If you were breaking it down, then you would be left with some sub-units of Wales, which takes us more or less back to where we are now.

What I would propose is that these additional members would be all-Wales Members. Again, if the seat ratios were different and again, Peter mentioned Germany – it would be quite normal for the party leadership to head those lists, and to differentiate their tasks as party leaders and spokespersons in the chamber from those members who have become constituency members. Obviously, at the moment, the particular character of politics in Wales would probably not facilitate that.

VIVIENNE SUGAR: How close would we get to what might be the best of both worlds of having a constituency link and proportionality by basing the constituencies around local authority areas, not parliamentary constituencies, and having three seats on an STV system in each?

Dr DENIS BALSOM: Of course there are very large population differences between the local authorities. You have 22 local authorities, ranging from whatever it is, 300,000 or so in Cardiff to 35,000 inMerthyr, or something of that kind. You would be working with units of that size, it could not apply across the board. I can see in parts of urban Wales STV can work reasonably well and the idea of the constituent/Member link can probably be retained fairly comfortably. In somewhere like Swansea, if one has got three members or so, then it is not a disproportionately large area and one might gain something from people feeling they perhaps had a member of their own political party or whatever.

I think in rural areas however, such as where we are now and such as where I live, the idea of constructing even a three-member constituency is going to be huge, it is a huge geographic area. There are local loyalties and local factors that are very important. Is it more important that you are represented by a Conservative, even though your neighbourhood might be Labour, or that it is somebody from Montgomeryshire or South Pembs. Those identities are very important. It seems to me, particularly in places like Ceredigion, the local member is very important. I think we would all lose something if we lost that basic relationship.

TED ROWLANDS: Would you exaggerate actually in some ways oddly the distinction between the two by creating an Assembly that is 50/50? Only 50% which is elected by the constituency and 50% by the list. I think that is a recipe as much for more tension between those who would believe they drew their authority from a particular community and the other 50% who would wander around the place.

Dr DENIS BALSOM: Yes, it is just that there is the huge disproportionality that the first constituency contest generates. The Labour Party had 40% of the vote for 75% of the constituency seats. If it had 75% of the vote, then obviously it would be just fine.

EIRA DAVIES: Whether you have a regional or a national list member, the electorate is certainly very confused about the roles of the constituency AMs, and also seem to cause problems as well, between the additional members and the AMs. Is there not a case for some more clearly defined roles for them, so that the electorate has a clearer definition of who is doing what, and therefore perhaps engaging in the electoral process?

Dr DENIS BALSOM: Yes, and I would expect the parties to make very clear nominations as to who they see as speaking on behalf of particular areas or communities in the way that they would issue policy portfolios to their members. It has already been mentioned that membership and the number of activists in parties are diminishing. It might be one of the things that might invigorate party politics somewhat if the parties had to become much more proactive in this way and were consciously appointing their spokesperson for Ceredigion Northeast Wales, or wherever it was. They would have offices and all the rest of it, as they do now.

PETER PRICE: If you look at the national list that you are proposing, presumably one of the two reasons why you are advocating post-election naming, not names on the ballot paper, is because you end up with a huge number of names on the ballot paper otherwise. Therefore, what you are proposing have a series of party names and then any individual who wishes to stand as an individual will be named in the same way as the parties. Does that not give a huge advantage to those standing as individuals, as compared with the parties, because this election has shown that there are a lot of people who do not wish to vote for the party man and are very keen to vote for an individual. If only the individuals were standing on their own were shown on the ballot paper, would that not add to that tendency?

Dr DENIS BALSOM: In fact, from the election last week I do not think there is any evidence to suggest that the one-man parties, which is effectively what they become, poll very well at all.

PETER PRICE: Marek? The Scottish...

Dr DENIS BALSOM: The Marek group got 11,000 votes and Captain Beanie got 1,200 or something of that order.

 LORD RICHARD: Can I sort of ask you a difficult question, if I may, which is you told us what happen, now just try and tell us why?

Dr DENIS BALSOM: I think we probably touched on this more in terms of the discussion we had a few weeks ago. Firstly, it seems to me if you have two parallel elections going on, when the parties should have an interest in both of them. I think we have got this positionwhereby one party has no interest in the outcome of the other and the other parties are entirely dependent on the outcome of the additional member system.

If you could alter it, as I said I think an all-Wales list is more likely to give all parties an interest, but given the level of imbalance as currently applies in Wales – it is not always going to be the case – the Labour Party would perhaps still not have much interest in the regional election. I think it also tackles, in part, the kind of cynicism that Huw brought up, where people who appear to have been defeated still end up getting elected from the same area. I take your point that in a sense they would end up there anyway, but it would be a much more explicit process I think, that they were nominated by their party as all-Wales members.

TED ROWLANDS: You have been against the concept of the regional lists. This is the second time you have come before us and you have said that, yet curiously didn’t the electorate actually vote on a regional basis? Your paper does not discuss the very considerable swings that took place between the valley communities, which swung 10-15% towards the Labour Party, and Southwest Wales, where the swings were 0.5% either way. And Swansea swung differently. Within this overall pattern it was in fact like great traditional Butler elections, where there was a uniform swing of 3-4%, there were very considerable differences of swing taking part in the election, in some ways almost reflecting the regional view of life. Was that not true? Your paper does not discuss the swings within regions.

Dr DENIS BALSOM: There is a great deal of minutiae about the election that one not been able to go into.

TED ROWLANDS: But crudely, there were very considerable variations of swing.

Dr DENIS BALSOM: No, I think you are picking on particularly the seats of Rhondda and Islwyn in which...

TED ROWLANDS: And Merthyr, and Penllyn Llanelli[?], and Aberdare, the Cynon Valley; they all swung hugely between 10-15%. There was a very curious uniform swing across most of the valleys which was not reflected in other parts.

Dr DENIS BALSOM: No, the swings of that size were certainly not reflected, but in a sense voters were I think, responding to what happened last time. In some of those seats last time there were a large number of independent candidates; for example, in Torfaen were effectively independent Labour candidates came second and third. 31% of the vote in ’99 was denied the Labour Party because of internal strife. These issues have been corrected, in part, and we saw a swing against Plaid Cymru and to the Labour Party. Certainly the Labour Party put huge effort into recapturing those two seats, which I understand. The other seat, which you told us you were involved in coincidentally, was a much, much closer thing. There was a swing to Labour, but not of the magnitude you are talking about.

TED ROWLANDS: And Carmarthen and South Pembs. The Carmarthen seat was equally closely fought in a very different pattern.

Dr DENIS BALSOM: Yes, but I do not necessarily think it was regional. One would have to track that through and see whether it actually applied on the regional ballot. I am sorry, I have not done that analysis.

VIVIENNE SUGAR: Can I ask you about the timing of the election? What do you think are the arguments for keeping the constituency list and the regional or national elections at the same time, or whether there is any merit in splitting them? Once you know who has been elected on a constituency basis, you can then be in a position where the parties could choose their list members. It would be clearer about who people were voting for. In your recommendations, people would not know which individuals they were voting for; they would just vote for a party.

Dr DENIS BALSOM: I am not quite sure how that would work, because if we had had the election last week and today, or Thursday, we were voting again, we would have an election where the Labour Party would not be entitled to elect anybody. I think if you want people to go to the polls a second time and there are no Labour candidates, you just would not get people out. It is not going to work.

 LORD RICHARD: Has any analysis yet been done on why Plaid Cymru did as poorly as it did?

Dr DENIS BALSOM: No, just looking at the overall aggregate figure that I think I mentioned earlier, the thing that became clear, certainly in the run-up to the election – was that Plaid Cymru’s ability to attract regional votes had diminished in itself and was not picking up support from people from other parties switching, or splitting in that way. Last time it went from something like 28% on the seat election to over 30%, where in fact it dropped between the two elections this time.

 LORD RICHARD: Thank you very much indeed. It is always nice to have a clear view of what happened.