 And I'm a regional director for Asia and the Pacific at International Idea Office. And very warm welcome to this very first session in our webinar series on democratic development in Melanesia. I'm so very glad that you're all able to join us. Together with our Melanesian partners we came to think that now is an opportune time to discuss the state of democracy in the region and the reforms that may be considered. And so International Idea 2 will intensify its work and attention to the region. The Pacific Islands is sprawling with democracies and a lot of knowledge and experiences can be shared with one another. And to reflect also this, our International Ideas Global State of Democracy indices is expanding its data for the Pacific. Starting with three Melanesian countries in addition to Wabunigini which is already part of the indices. But we will now expand the data gathering to Vanuatu, Fiji and Solomon Islands. And hopefully to include more of the Pacific in days to come. These indices can be found from our website in a couple of months time. This online webinar series will be held monthly and will feature democracy building topics relevant to the Melanesian region. Such as women in politics, civil society advocacy, money in politics, climate change and so on. Often using a particular country as a case study just as for today's seminar. And it is our hope that this webinar, not only the participants will not only gain knowledge, but also able to exchange views with one another during and after the webinar through various channels. This pandemic has provided us all with an opportunity for online communication and discussions. And even when at times we may feel a bit of a Zoom fatigue and miss those face to face interactions. This still provides us with an unparalleled opportunity to come together and talk of range of issues in spite of the physical distance. So I hope we can make most of it and then appreciate the physical time together even more when it happens. So we do wish that you are able to participate our monthly webinar and do send us any feedback that that you may have also in terms of topics that you think we should be should be covering. So today we have the honor of starting with one of the world's leading experts on what it comes to electoral systems, Mr Alan Wall, and this session will be facilitated by an other expert and influencer Niles Lal from Dialogue PG. So let me also extend my very special welcome to Serena Zazinjan, who is our member of International Idea Board of Advisors and joining us, I believe, from Papua New Guinea. Welcome again to everyone and hope you enjoy and participate in today's discussion. Over to you Niles. Thank you very much, Lena for the welcome note and for those insights into this webinar series. So, as has been mentioned by Lena, my name is Nilesh Lal and the Executive Director of Dialogue PG, which is one of the PG and CSO partners of International Idea. I'm very pleased to be moderating this first of the webinars under this series and I thank you all for your participation and definitely look forward to some great interactions over the next hour or so. So, before any further ado, I'm going to now invite Mr Alan Wall, who is going to be the main speaker for this webinar to deliver his presentation. Before I do that, I will briefly introduce Alan. He is, you have a disjun from the flyer that was sent. He is a senior global elections expert. Alan is an electoral management and democracy development specialist with over 35 years of management and consulting experience in electoral management and planning, electoral capacity building and training, electoral law reform, voter registration, electoral systems, qualitative research, public opinion surveys, civic and electoral education, program evaluation, electoral integrity, promotion, gender, disability and youth inclusion, advocacy, political party and political finance regulation, electoral dispute resolution and electoral observation. Alan is based in Australia at the moment and today is going to be speaking on the electoral system of PG and specifically on the electoral formula and the design. So, I'd now like to invite Alan to deliver the presentation, which I believe he has prepared for this webinar and then we will proceed on to accepting interactions from the audience. Alan. Please excuse my voice, everyone. I've got a rather bad cold and if I start coughing and spluttering, I may have to take a break for a couple of seconds. But I do have to take a break for a couple of seconds first just to get this PowerPoint presentation up on screen. So if you can bear with me for a minute, I will do that. Yeah, it's coming. And all we're going to do now is start it. My computer wants to work. Great. It's there. Okay. Now, before we start talking about the electoral system in PG and the Pacific things in terms of the electoral formula and ballot paper design, just like to give it a couple of parameters with which we can frame this discussion. We need to understand that what electoral systems can facilitate certain outcomes, they can't determine them. Another issue is, of course, that no electoral system is neutral and none of them is really inherently fair. Fairness tends to be a subjective judgment by people depending on which objectives they think are most important that the electoral system actually achieves. Now, for example, is an electoral system that has totally inclusive representation but results in weak governance, fairer to the population and one which provides more concentrated representation but effective government. In PG, representatives of parliament was one of the major critical issues, a major criteria the NCBBF used to assess fairness. However, we look at the history in PG, the move from the plurality majority system to an open list supposedly more inclusive system has not resulted in a more inclusive or representative parliament. In the 2006 election, 94% of parliamentary seats were held by two major parties that hadn't changed under the second election in 2018 under open list PR. Now, we know from other experience that this changes like this can take a while to flow through to greater representativeness. In New Zealand, it took four elections in Sri Lanka, for example, it took three. So it can be a long-term change rather than a short-term fix when you mess around with electoral systems and electoral formulae. Open list PR, what's being used in PG, it's used in 39 countries solely and another four is part of a hybrid system. The thing about open list PR is it's more of a concept than an electoral system. There's almost as many versions of it as there are countries that use it. PG has a unique combination of features and these impact particularly on ballot design. They have a single nationwide large magnitude electoral district, 51 seats up in the one district. They have a voting method by which voters vote only for a single candidate and they have relatively high barriers for party registration and no provision for non-party or you'd probably call independent lists of candidates. Now, this combination of three things is unique to PG amongst open list systems. And we also need to understand what open list systems do and don't do. They are proportional representation systems. So no matter who you vote for, the candidate vote totals do not determine which candidates or which party lists win seats. The whole impact and emphasis is on how many lists, how many votes the actual list gets, not how many votes the candidates get. So if a candidate wins a high number of personal votes, if their party list doesn't win many votes overall, they may not get a seat. Conversely, if candidates win few votes, but if they have a rock star or really popular candidate heading the list or somewhere on the list that ensures that that party list wins a number of seats, then candidates with few votes may actually get into parliament. Now, this is a design feature. It is proportional representation. The basis of representation is the party list, not the individual candidate votes. If people want to have their representation based on individual candidate votes, they need a totally different type of electoral system. Now, in all PR systems, the major issue is how votes for a list are transformed into seats. And I have up here on screen for two major ways in which this is done. I just remained a method and the highest average method and some examples of different variations of these and how the mathematics of these are worked out, which I'm not going to go through. I just want to mention a couple of things. One, these different mathematical formula may give different results. And this very much depends on how many lists are contesting the election and what the distribution of votes is between these lists. The G, of course, uses the dump formula. And the dump formula, as we'll see in a minute, is probably one of the least proportional of all the ways of transforming votes into seats. Overall, the largest remainder systems are the most transparent and the easiest for sort of non-technical people to understand. However, on the other hand, the highest average systems tend to have a more equal average number of votes as a cost for winning a seat. Whereas they more evenly distribute votes to seats. Now, this is an example of how dump could be not proportional. The favorite one of mine is in, and this is an extreme example. In 2011 in Tunisia, the constituent assembly was run on a hair-largest remainder system. And the party, the largest party with 37% of the vote won 41% of the seats. Now, if you'd used dump instead to distribute, to actually allocate the seats, the result would have been totally different. With 37% of the vote, the largest party would have ended up with 69% of the seats. And that's purely because of the distribution of votes and the number of small contestants or small voting power contestants contesting the election. In 2000, the Council of Europe did a tabulation of the seat allocation processes from most proportional to least proportional outcomes for largest remainder and highest average. And you can see here that they assessed dump as having probably one of the least proportional and that is least representative results. Now, dump may tend to favor dominant political party lists which have large percentage of the votes. On the other hand, the St. Largue systems, also the other family of highest average systems, tend to be more proportional than outcome and more likely to give seats at the margin to lower scoring lists and resolving in a more inclusive representative body. So you need to think of what is the objective of the electoral formula in Fiji of the electoral system. Is it to maximize inclusive representation and NCBBF was advocate? Is it to provide a short government within the PR system? Is it to promote a balance between stability and representation? I mean, each of these is a justifiable objective and each of these will gather the electoral formula that is most appropriate. Now, when we look at the distribution of these systems, we find that Daunt is easily the most popular of any of these seat allocation systems. They were 30% of countries using Open List PR using it. Now, this is also true of PR systems in general that Daunt is the most widely used seat allocation system. And there's even more dominant when you look down the bottom where we have got the lines of Hagenbach-Bischoff, which is basically gives the same result as a Daunt system. So that basically increases the dominance of Daunt. Daunts used in countries such as Austria, Belgium, Brazil, Finland, St. Largue in generally newer democracies. Bosnia, Indonesia, Kosovo, Latvia, modified St. Largue which is a result sort of generally partially between Daunt and St. Largue in Scandinavian countries and in Poland. Hair, the largest remainder hair system used in places like Bulgaria, Honduras, Lithuania, Slovakia, Sri Lanka. And let's look at the practical impact of all this in CG. So what I've done here is give a table showing the 2018 vote totals for all of the political parties at the 2008 election in CG and then distributed or done a seat allocation on each of the major seat allocation systems. And we can see what the result is from the Daunt method in CG. The Imperiali method from Italy is very rarely used. But when we look down, we find that if we use St. Largue or we use any largest remainder system, we have a slightly different distribution of seats and one of the smaller party winning seats gets one additional seat and the largest party gets one fewer seat. Now this may be seen as making parliament more representative and probably more evenly gives votes to evenly distribute votes in an average number across all the political parties. But this is only one part of the process. Other things that can affect this distribution. I think one of the big things, if we look at this, I did this again using different thresholds. 5% threshold is the CG. I did it with 3% and there was no change. But if you get rid of the threshold, the legal threshold altogether, it has an even greater effect on making the parliament more representative if you move away from Daunt and St. Largue or most of the largest remainder systems. Here we see under most of these systems and St. Largue and most of the largest remainder systems and additional party enters parliament. And the third largest party gets an additional seat and these seats may be taken from one or other of the largest two parties. This doesn't mean there isn't a threshold at all. In any system, there's a natural threshold. With 51 seats up for grabs, that natural threshold is about 1.5% of the vote for a party to get a seat. But there's no legal threshold here. What we have to be careful of is note that changing electoral system components can change voting behaviour. And we can't have a certain comparison. And also, we're trying to take a slice in time. I mean political behaviour, voting behaviour will change over time. So you can't be sure that if you change one of these components, you're going to get the result that you expect. And lots of electoral system designers have been surprised by the effect of changes to their electoral systems. Now, what other things that are important factors in representativeness if this is a key goal in Fiji? In Fiji, political party registration thresholds are high. They're in the top 10 of countries around the world. There's around 8% of people on the voter lists. And that's a very high percentage of registered voters that have to support a party for registration. There's also no provision for nonpartisan lists to contest the elections. It's much easier for a nonpartisan list to get someone elected than if independents have to stand individually. And almost half the countries using Open List PR allow a nonpartisan list to compete, though. To be honest, they're not successful in very many countries. As you can see, Lord Jordan, Lebanon are the only places where they really have had a significant representation. Moving on to ballot paper design. The three major issues for Open List, mainly because you've got to try and shell a large number of candidates on the ballot paper. So people can actually find the candidate they want to vote for and the party that want to vote for. So you have to try and make the ballot paper as compact as possible while allowing each candidate to be easily identified. So this is for reasons for practicality in a voting compartment and also for reasons of cost. The voters have to be easily able to recognise the party and candidate they wish to vote for. And also that candidates should be strongly identified with the party they represent. Now, if they're not, personality issues become prominent. And you have the possibility of Open List PR fragmenting the party system of personalities are dominating rather than parties. If candidates are strongly identified with the party they represent, policy issues can be more easily the dominant factor. Now, the CG ballot at the moment with its numbers in random order certainly meets the compactness criteria to quite a good extent. But I think most people would see it as failing the other criteria. Now, there are bigger problems for ballot design than in many other Open List countries in CG. No other country using Open List PR has such a large magnitude electoral district while requiring voters to vote only for candidates and not for a list or political party, which is quite important. There's significant constraints on how you can design the ballot. When I looked at the FAO's post-election survey during the week, I found it asked questions about how easy did you find it to vote? And what did you think of the layout of the ballot paper? Now, these are far too general really to be able to get an idea about whether the ballot paper is effective or not. It would have been probably more useful to ask voters whether they found it easy to identify the party list and the candidate that they wished to vote for. And that would give probably a better idea about how effective the ballot paper design was. How can it be improved? Now, the proposed PG Law Reformancy Party affiliation shown on candidate lists but not the ballot paper. And to me, this is only a small step to improving design. It doesn't deal with issues like the fact that a party identifies on the ballot. The random order of candidates makes it really difficult to identify which party or which group of candidates represent a party. And this is important both for the voter and also for political parties. In PR systems, parties generally have some control, whether by internal primaries or their members or other means, on the order on which their candidates appear on the ballot. I mean, this is an internal party matter. It's important that there's fairness in voting between the parties, but within the party list, normally that is up to the party to control. And it helps party discipline and it helps party cohesion. Now, in some extent, then the PG ballot paper design seems to be sort of opposed to the general intent of PR to pool votes to parties. Instead, it seems to reinforce an emphasis on individuals and personalities. And to some extent, it seems to appear to favour lists dominated by big personality. So how can it be changed? It'd be great to be able to look at international exams and say, hey, we could do that. Unfortunately, it's not quite that easy. Now, mainly because in all similar district magnitude durations, with 50 or more seats up in a single district, the voter must vote first for a party and then a candidate, which means you can do different things with your ballot paper design. But let's look at some of the things that have been done. Obviously, Kosovo. Again, here, similar to PG, you've got numbers for the candidates, but you have on the left all of the parties listed. So you vote for a party and then you select the candidate numbers. And similarly to PG, there's an index of these in each voting department for that particular party. Let's look here. This is similar from Columbia for the Senate. Only this does the ballot paper horizontally. It only has as many numbers for each political party as there are candidates standing on their list for the seats available. And then, and this is probably most relevant to PG, this is the Netherlands. Now, in the Netherlands, they have up to 80 candidates per political party. They list them all by name under the party header. True, in the Netherlands, you do have to vote for a party as well, but you need not put, if you wanted to only vote for the candidate, you need not put a box beside the party and just list the candidates under the party header. It does result in a bigger ballot paper than PG, but it is not enormously oversized and it does allow clear identification of each candidate and which party they belong to on the ballot paper. Okay, a couple of others. And these are more modern design concepts showing logos, showing initials for the parties, showing symbols, showing photos of the candidates and candidate numbers. The first here is from Honduras, makes it very easy for voters to be able to identify who they want to vote for. Honduras has a maximum district magnitude of 25, so only half that of PG. So it would be a much bigger ballot paper if it was this style in PG. And secondly, reversing the way in which the ballot is shown, from Ecuador, where the candidates are listed vertically. In Ecuador, the maximum district magnitude is 15. And in both these countries, voters vote only for the candidate as they do in PG. They don't vote for the party as well. Now, there is some other off the wall or more extreme solutions. Brazil, which has a district magnitude of up to 70, uses electronic voting machines. And Slovakia, which has OLPR, Open List PR, in one single district of 150 members to be elected, uses a French ballot. That is, each party has its own ballot paper and voters get given a pack of ballot papers and select the ballot paper for the party for which they want to vote and mark the candidate for which they want to vote on that ballot. Now, this is not something that's ever been used in countries with an English tradition. Probably not appropriate for the PG in context. It would be a very radical departure. What are some other issues or some other potential ballot design solutions? Well, you could get the current style of the ballot with numbers, but group the candidates on the ballot by party or list so that voters are much more able to work out which is the party and which candidate is from that party they want to vote for. You could print the relevant party symbol beside or behind each candidate number on the ballot. And they would, each of these would require minimal change to the way in which the ballot is laid out at present. More radically, you could, and this would require major electoral system change just to get the ballot a little bit easier. You could change the multiple electoral districts to smaller magnitudes. So fewer candidates would be on each ballot, but this has significant impacts on the electoral system. It will make it harder for small parties to win seats. It will reduce inclusiveness. But on the other hand, it may limit the influence of a rock star candidate in a candidate list. So it will be limited to a particular district rather than affecting the national result. Or even you could modify the voting process that voters vote first for a party in a candidate. But these last two are major differences that would require a lot of work and a change of philosophy in terms of the electoral system in PG. Okay. Thanks. Thanks for putting up with my croaky voice at the moment. Nilesh, I'll hand back to you. And if we have any questions or any discussion, I'm happy to see what we can answer. Thank you. Thank you, Ellen, for that very insightful presentation. So we're now opening this discussion up for questions from the attendees. So if there's any question that you'd like to pose to Ellen, or if there's a comment that you'd like to make in relation to the subject of today's webinar. Or even, you know, on any issue that is related to the electoral system design in PG in general. So you are most welcome to do that. And you need to use the raise hands feature that is available to you to indicate that you have a question that you'd like to raise. Alternatively, you could send us your questions by using the Q&A pane that you see on the right hand side of your screen where you can put in your questions and we'd be able to accommodate those. So we definitely look forward to some good questions that would ensure that we have interesting discussions on the subject matter. So whilst we were waiting for the attendees to pose the questions, Ellen, a question that an issue that I'd like you to talk a bit about is the co-tail effect. You know, in PG, this has been a contentious issue with a lot of people perceiving it is unfair that candidates with a low number of votes are getting elected into parliament on the coattails of more popular candidates in the party. Recently, there was a lot of discussion on social media when minister Faya square who had pulled 547 votes in the last election was appointed as the acting attorney general. Incidentally, the position of attorney general has come to be perceived as a very powerful position in PG. Due to the influence of the incumbent minister, yes, you can use. So it's the code to tell you if a thing, you know, does that make the system unfair. Okay, I said before unfairness is in the eye of the beholder. It's generally in the eye of the loser rather than the winner. The issue is you dealing with a proportional representation system. Proportional representation systems are based on how many votes a party list overall wins, no matter how many votes the individual components or candidates within that list win. So it said before, if you if you want votes to people to be elected on the base of their individual votes, you want a system other than proportional representation. Go back to an alternative vote system or a first pass the post system or STV. But proportional representation is based on list totals, not candidate totals. Now, okay, you have a choice between open list and closed list. In closed list, no candidate gets any votes at all. All of the seats are allocated in terms of how many votes the party gets. Now, is that any fair or less fair than having a system where people have to get some votes or be ranked highest in the number of votes within their party list to be able to win a seat? I mean, that's a matter of individual judgment. But if you're going to have a PR system, you are going and you're going to have the sort of system that PG has in voting for candidates only, you are going to have a coattail effect and it is part and parcel of a PR system. You could modify making the coattail effect by changing the system as many countries have that you vote for a party first and then vote for a candidate. And that tends to even out the candidate votes because fewer people overall will vote for the candidates than actually vote for the party. It's generally not compulsory. You have to vote for party may vote for a candidate. I mean, if you don't like the coattail effect, get another electric system, get a non PR system. Yeah, so essentially the alternative is then to have a closely system if one has issues with the coattail effect in a PR system. What you're essentially saying is that the other alternative is the closely system, which in some ways, I mean, not in some ways, which in in a very significant way, then restricts vote a choice as far as choice of candidates are concerned. It's totally up to the party as to how the how the candidates are ordered on their list on the and it's up to the party to determine who are the highest on that list and who it is that gets elected. The voter has no influence apart from perhaps if they're a party member and take part in a party primary in selecting the list as to who actually ends up in the parliament. They're purely voting for the party list. Close list does have one sort of advantage and that it is easier to get inclusiveness within the party list or within people being elected through requiring that every X number of people on the list. Ranked down be either of different genders or a different ethnicity so it can be easier than open list to be able to get a more inclusive parliament within each party list. But it does restrict voter choice and leaves it up to the party is to basically which candidates are likely to be elected. So to me it's probably better for the voter to have some influence than none at all. The other issue of course as I said is if you do change to having smaller electoral districts, you can get rid of this rock star effect to a significant extent. It's hard to find lots of rock stars to leave the lists in every electoral district. But if you're doing that you're losing on the other hand by making it harder for smaller parties to win seats in those smaller electoral districts. The problem with electoral systems and the way in which you work these different components is each of them has advantages and disadvantages. There's not one thing that is clearly better in doing everything than any other system. And the whole issue is to try and get the balance as close as possible as to what you want the objective of the electoral system to be. So the inherent disembarked disembarked as they are for the other political parties would be that they're too many. They're not enough rock stars around then I guess so. So I see Moshmi you've got your answer. I believe you have a question for him. Moshmi Bhim. Thank you, Nilesh. Yeah, I was waiting for you to unmute me. Thank you for the nice presentation, Mr. Alan Wall. Yes, and. Yeah, it's very interesting the way you have presented how the design of the electoral system affects voter outcomes. And yes, I agree the closed system seems to be worse because then the voters do not get a choice to vote for. With regards to the current system in Fiji, I believe why it is unfair that a person with the low votes is able to get into parliament when the leader of their party is able to redistribute their high votes. Is because in the other party, you have your candidates with much higher votes, which end up not having any seats is all. For example, for example, 500 votes by Fayaz and then an SDL. And then a soda candidate who got over 1500 votes, which is three times the vote that Fayaz got. So I think if, for instance, they want to have the leaders of the party who are able to amass a high number of votes. And because it's easy to campaign for a leader who people already know well, and then they're able to amass votes from all over the nation, the other candidates who are less known would have found it easier to get votes within their district if it had been in that system. Yes, so I guess the unfairness here is the leader of the party who are generally well known for decades locally and internationally. They are redistributing their votes. And if the leader of the party is the prime ministerial candidate, then their vote should not be redistributed because they are already getting the reward for the vote by becoming the prime minister. And when people vote for the leader of the party, they are making a clear choice that they are voting for that person to lead the nation, I believe. And they are not voting for another person to represent the leader of the party. Well, that's my view. Of course, other people's views will differ. So I think, like in the presidential system, the leader's election is conducted, their votes are voting is done separate from the rest of the members of parliament. The leader's vote is not redistributed to the other party members. And I think that would be fairer, but that's my view. Do you have any response? Yes, I think response is necessary. Yeah, one way around this is to move to a presidential system. So you have the super popular people competing in the presidential election. Yeah, and taking a lot of votes at the top of their list, but I mean. I have difficulty with the idea that the votes are redistributed to other people on the list. They're not. They're distributed to the party. And then the party votes determines how many seats the party gets. And that's the essence of proportional representation. As I said before, if you don't like that, move to an individual candidate based system where individual candidates have to fight it out amongst each other. You're not redistributing votes to other people, you're distributing your votes or adding your votes to the party, which is a quite different concept. But yeah, a presidential system would help get rid of that as would, I agree, having smaller districts would make it more difficult. But then the downside of that is that if you have smaller districts, for example, your 50 districts were within the 5 of 10, say 5 districts of 10, your natural threshold for representation would increase significantly. And it would make it very, very difficult for smaller parties to be able to win seats in the parliament. So that's the downside of that. You might have a more even distribution of votes amongst the candidates from the bigger parties, but smaller parties may find it a lot harder to win a seat at all. So the representativeness overall of the parliament might be a problem. Excuse me, but presidential system may be a way that you think is fairer, but basically what you're saying is that I'm going to deny the votes of everyone who voted for the leading candidate for that party. They shouldn't be counted towards what that party wins in the way of seats. And that's not fair to those people who voted for the highest rank or the biggest vote winning or the rockstar candidate in that party. That's basically denying their votes. I think there's a lot of misconception on this issue and I think it is important that it needs to be emphasized that the PR system is based upon the proportionality between political parties and not proportionality between candidates. So you do not need to see what is the proportionality of votes between the distribution of votes between a Sodelepa candidate and a VGFES candidate for that matter. Rather what is the proportionality of the vote share of the two parties involved in the poor two parties get allocated a number of seats based on their aggregate number of votes that they receive. So thank you for the question. This is a lot easier to see in open list PR systems where voters have to vote for the party first and then vote for a candidate. It's much easier to see that that's the party vote that determines how many seats each party gets. Whereas it's more difficult to both see and to understand and recognize if the party vote is a total of all the individual candidate votes as it is in Fiji. So, but again, it'll be a matter of changing the nature of the voting method within the LPR system. Probably not as big a change is changing to a presidential system, but that would probably solve the problem. Okay. All right. So there was a question that was sent in earlier. And it is in relation to the, to the electoral formula. So what do you think are the consequences should be changed its electoral formula. What? Well, it is pretty much open ended. So, in case there's a change in formula. Then what would be the consequences? I suppose that this would tie in also. And you've pressed on that. You know, what are the electoral objectives in Fiji? What are some desired electoral outcomes that we want to see. As I noted, the don't formula is probably the least proportional of the highest average for formula that can be used and does tend to concentrate seats in the largest political political party with the largest number of votes. Whether that's a good or not a good thing is a matter of the, of the objectives for the system. Whether you want to stable government with few parties in coalition or none in coalition or whether you want to have total inclusiveness as a matter for people to determine that what's best for their own country. As you saw from, let's see if I can cancel that up easily again at the moment. You saw from the slide, if you get, if you change from don't to sent like way in any form or to any largest remainder formula. You get seats moving from the largest party to the smallest party under the voting patterns that were there in 2018. Now that may not hold if voting has changed in a future election. It may be a bigger strength or as we saw as sometimes happens. If you have lots of small parties that fall below the 5% threshold, that's it all because there is lots of parties falling behind under the 5% threshold would mean that there would be lots of wasted votes and that the votes again will be concentrated in the larger parties. So it depends on distribution of votes and number of contestants what the impact of a change in the electoral formula would be. It's a little difficult to predict but all we can do is look at the past data and say if this voting pattern was the same again there would be in Fiji a slight move away from the largest party. But again, as I know, the bigger change would come from getting rid of the 5% threshold. That would make more of a difference changing on 2018 voting patterns than just changing the electoral formula change the formula and get rid of the 5% threshold. If what people want is broader representative this that would be the more effective thing to do. Okay, well, you've done a simulation of the results of 2018. And to see what the city allocation would have been like under the different systems or methodologies for allotment. But, you know, there was in fact a change in voting behavior after the experience of 2014 elections. So it would be very interesting also to do a similar simulation for 2014 elections. You know, in which the smaller parties received a greater share of the of the votes. Because, you know, by 2018, a lot of people had given up voting for the smaller parties and independent candidates. In fact, by 2018 there were no independent candidates that were participating in elections anymore. And then essentially, you know, this smaller parties, which are absolutely critical to sustaining a multi party system in Fiji and bringing in some sort of a moderation in Fijian politics, you know, from this preoccupation with race that has been there all along. So this parties have been inherently caught in a cycle of inclusion. We are by people do want to vote for them, but they fear that their votes will be wasted just because of the excessively high threshold. High by international standards, but you know the dynamics of the Fijian situation and the desired electoral objectives here. It was very interesting also to do a simulation and see, you know, how many parties would have got inclusion into the parliament. I would have got a seat in parliament had the same method being used in 2014 to one more party that one seats. But this is the issue. Each set of votes is going to give you a different distribution, which we cannot predict for either 2018 or 2014 data. What the impact of changing the electoral formula would be in a 2022 election? We can say, well, it's probably more likely there will be more representative, but we don't know. And there's terms of independent candidates, independent candidates in PR systems. I don't think I'm struggling to recall an instance ever wearing it. Now, where they can win seats is if you allow independent candidates to group together in a list. And that way, all the independent votes can be added together. And maybe one of those candidates has got a chance of winning a seat. But standing alone, it's almost impossible for an independent to win a seat in a PR system. So that again is something to consider should an independent list be able to compete in the election. But the threshold is a critical issue. 5%, as you say, is not high by international standards. Probably around the top of the range for what people like the Council of Europe would think is acceptable. But it's, it's, it is a barrier. As is the high number of people that are required to support the registration of a political party. That's another barrier to people being able to contest the election. Okay, so we just like to remind attendees that you are very much welcome to pose questions and to contribute to the deliberations that we're having. You can indicate your intent to pose a question by using the raise hands feature that you have. Okay, so coming forward or proceeding further, there's another question that I had for you, Ella. And this relates, you know, to one of the contenders issue in Fiji, which is about the separation between the two electoral men and women. A lot of people in Fiji, all the perception that the Electoral Commission and the Fijian elections office need to be independent of each other. In fact, I had posed this question to the electoral expert in the multinational observer group of 2018. And he told me that there is no requirement internationally that, and you know, that is definitely not the practice internationally that EMBs need to be independent of each other. So what is the view on that? Do you think that there should be a separation between the Fijian elections and the Fijian electoral commission? Okay, as a first comment on this, let me just emphasize that the structure of an EMB is not the determinant of its behavior. It's the behavior, it's the attitudes of the EMB staff and members that determine the behavior. There are lots of election commissions called independence that are anything but independent in their action. There are election commissions that are in two parts that perform well, there are others that don't perform well. So the structure really is not the determining factor. There is no requirement that they be normal, independent of each other. The more normal practice would be that an independent body in name or in reality actually is a single body that manages the election. There aren't as many bodies like in the Fiji model that has an independent commission and then a bureaucratic, a separate bureaucratic sort of implementation arm, if you like. And when you have that, you have all sorts of issues. Now, I've worked a lot more, they have that there. It works quite well because each of the bodies knows their respective powers and knows how to relate to each other. I've been in places like Mozambique where it works particularly badly because it's become incredibly politicized in the bureaucratic level and there's no clear lines of authority. One of the things you have to do if you've got this bifurcated model is work out what it is the elected commission actually does. Does it just think the administrative arm, bureaucratic arm, to get on and implement it? What sort of monetary oversight powers does it have over that arm? Do the members of the commission take particular responsibility for particular areas of operations or activity in the separate administrative arm? To me, this separate model is a more difficult one to get to work effectively than a model where there is a single structurally and in reality independent body. It can be useful in transitions where you want to have either a politically representative body as the commission and then bureaucrats doing the actual work. Or if you want the great and good of civil society to form the commission, you don't actually have to do any administration work but just set vague policy goals and let people get on with managing the electoral process. But it's a more difficult problem to make work effectively. There's certainly no requirement to be like that or not be like that. And I'd have to go and sort through which I haven't got in front of me now. I'd have data somewhere on how many commissions have these various structures. But let me emphasize again that the important thing for an election management body is that it has autonomy of action, not its particular structure. And you can have autonomy of action no matter what your structure. It depends on the behaviour of the people who are the election managers. Yes, Alan, thanks for that. In Fiji at the moment they are proposing electoral reforms and part of the reforms includes increasing the powers of the electoral commission here. So having autonomy of power or action is okay as long as that body is actually also controlled by regulations. That preventive from making decisions that will harm the other political parties that are not part of the ruling party. Because in the past elections that we've had in Fiji, the courts during election time have made decisions, some of their decisions, which have favoured the ruling party members and have actually been acted against the interest of the other candidates that were not part of the ruling party. So that's why in that regard, the exception whether members of the electoral commission are independent or are they known to have been supporters in some way of the ruling party can actually, well, can have an effect or in the way elections are done because it's some of the decisions of the members, I mean, or the director of the electoral commission can affect the election outcome as well. It's not the structure that counts, it's the behaviour of the people in the positions that determines how independent that management is. Now, in terms of political party supporters there, I mean, some election commissions work on a balance of care. Basically, we acknowledge representatives of people with links to the various political parties sitting within the commission and fighting with each other, not necessarily a great model. But you can structure, if you wanted to, the criteria for eligibility to be a member of the commission to make sure that you can limit, at least as far as it's practically possible, the appointment of people who are going to be influenced by political party to those jobs. And you can't prevent it because, again, it's a matter of individual behaviour and people who have been members of parties or people who are known supporters of political parties becoming members of election management bodies. You can put codes of conduct in place for the election management body members that they have to adhere to that are justiciable. But, I mean, it's a matter of ethics and behaviour. You can't totally, they just like to prevent political influence in an election management body. You can have one instruction. Okay, yeah, so there's another question, Ellen, that I had received. So how can you mitigate the notoriously large size of OLPR system ballot should the party symbols and or candidate names are included? I mean, the number system that Fiji is using is probably the way of mitigating the size. However, the problem in Fiji is it totally destroys any link to any political party. It doesn't group candidates by party and it gives no indication, it doesn't list, it gives no indication of which party it's to. I mean, as in Kosovo, as an example, I showed you can use the numbers and still allow people to identify which party and not have a large ballot paper. You can have a group of numbers by political party on the ballot. And again, you can the party identified at the top of each column or row of numbers as in the Columbia example before you could then basically vote for the candidate and you would minimize the size of the ballot. It's not great ballot design international standards. International standards are moving towards making it as clear as possible who the candidates are on the ballot paper itself, but it's a mitigation method that could be used with. I mean, the other thing as we talked about before is reduce the size of the electoral distance and you're going to have fewer candidates to fit on each ballot paper. But that has other impacts that are probably going to be more negative than you're going to get benefits for in ballot in ballot paper design. So you couldn't work with current ballot and group by party have a party symbol at the end of the end of each column list the numbers and still have in the voting compartment a index of numbers of each candidate against their number on the ballot paper. I think the intent behind the complete randomization that we see on our paper is to ensure that, you know, the ballot design does not accord any a priori advantage to any candidate. It was even if we were to have all the party candidates listed with, you know, every candidate, you know, pulling, you know, within a column for the party. So that is how it was going to be arranged. Then I suppose the one the candidate at the top, you know, who whose number would be closest to the party symbol would actually receive an advantage. Which can, you know, in some ways also exacerbate the complete effect. Political party leader is put right on top next to the party symbol. So I think, you know, in that case, there would need to be two levels of pros to ensure complete randomization. Because then questions would then be asked on, you know, who decides the order in which political parties are listed. So who falls on the left then who is, you know, is the next one next party to fall on the ballot paper. So these are the questions that would need to be to be to be considered there too. In fact, we welcome the change in the bill number 52. Sorry, bill number 50 of 2020, which has proposed changes to the law where, you know, the national candidates list is now allowed to be brought into the voting booth. So then voters do get a chance to verify whether they're voting for the candidate of their choice should they forget the number. So these are some of the issues that would need to be considered. Given that one of the intense is to have complete randomization and to eliminate any sort of a priori advantage to any candidate or political party. And I would see that as as not being in line with the proportional representation system. Well, you're looking for a candidate, you're looking for a ballot paper that's more at home in a majority plurality, a plurality majoritarian system rather than the PR system. Can we get back to this thing PR is about parties. It's not about candidates. Right. But what we have in Fiji is that people have a problem with the code tail effect. And if we actually have a random, I mean, a ballot design that gives, you know, further advantage to certain candidates, then it would be perceived to be a problem. Unless of course you know we, we've got a lot of hope to do obviously civil society organizations, you know, command greater trust and confidence across, you know, the political spectrum that you know the PR system is not necessarily about candidates. It is about political parties. Yeah. And there seems to be a lack of understanding of that. And I mean, from my point of view, and it may not always be shared. The political party has an absolute right to determine what order its candidates are going to be on the ballot. These candidates are representing the party. They are not representing that an individual in themselves. They're there as a representative of a party. So obviously the party, you have to have things in place to make sure that parties don't get an advantage. And you can do that through randomization or double randomization of where parties appear of determining the order of parties on the ballot. You can do things like they do in Tasmanian Australia and have what they call a Robson rotation of parties on the ballot. So each party appears, parties appear in different positions on an equal number of ballot papers so that none gets an overall advantage. But in terms of PR, to me, parties have an absolute right to determine where their candidates are shown on their list. That's not a matter for the EMB to do that. I could perhaps come up with a situation where they're grouped by parties, but the EMB holds a lottery amongst each candidate as to where that candidate appears on the list. And one of the issues also, I mean, some countries get trying to get around this coattail effect by saying if you don't get X percent of your party's vote, you don't come into the seat allocation. Which can create problems to actually then wins that seat if only few people actually pass that second representation quota. But there are a number of countries that have opened this PR that actually do that. I mean PR is about parties. It's not about candidates. That's the basis of the system. If Fiji wants a PR system, there's going to have to be work done to educate people as to what the PR system actually means. Okay, so they're not used to it. I mean, it's not usual in English-speaking people in English heritage. They're much more used to plurality majority systems where candidates are the focus. And that, of course, was Fiji's history as well. So it's a major change. And it requires a lot of work to get people to understand what this difference is. It's at the moment they're trying to maintain as many as possible of the elements of a majority PR system. Two things don't gel together. Okay, I'm mindful that time has caught up with us. But there's one last question that we're going to take from Elam Prasad. Because I see that she had a hand up for quite some time. So we'll quickly take that question from Milam. Sorry, it's not Milam. We have short of time. So we logged in through whoever registered. This is the supervisor of elections. Okay. Thank you, Alan, for an excellent presentation. I'm very grateful that someone with an electoral experience of my age has been able to literally spill the beans on the system we have in Fiji. So that it becomes a bit more simpler compared to how I tried to explain it all this while. So heads off to you, Mike. I just wanted to run two things that I think you have endorsed quite well. There is a big misconception in Fiji that the Open List PR system is a mixture of the post system and the first preference system that we have. I've tried to do this on numerous occasions and this is where a lot of the people that made mistakes with this to assume that it's a mixture of the other two systems, whereas it's a completely different system. It's based on political parties and it's a party list that comes for nomination and the candidates hold under the party. And it's further strengthened by the constitution to say there are no by elections. In fact, if a party candidate loses his seat in parliament, he's simply replaced by the person next in the list. And it could even go to an extent where there may be someone who received 100 votes and the person is able to get into parliament because there's nobody else left in the party list. So I know the new word co-tail effect and all these things. I think it's going to be even greater problem for people to understand when someone with 100 votes get into parliament simply because there's nobody else left in the list. So I think that's one misconception you've cleared. The other one that I wanted to agree with you is in terms of the changes in the rules, I mean, which formula do you use? St. Laguier or the Behaunt or the Hair Rule or whatever. At the end of the day, if a party does not reach the five-person threshold, you will not have a greater distribution of seats. You might just move one seat from this party to that and then one seat from the party to this. In fact, the challenge is for the party to first meet the five-person threshold. Even if you complete the 2014 election or the 2018 election, it's all to do with one seat, going out of PG first to NFP in 2018, or one seat going out of Sudolpa to NFP in 2014, using either the St. Laguier or the Behaunt. I think the competition in PG is more about what rule is better or what rule is better. But I agree with you, it's just been two elections. What is the, in the next election, you go with the St. Laguier rule and you have what Nilesh likes to call a hung parliament. And just before I close, I thought it would be a wise for you to also a little bit explain about what is a hung parliament. Right now we have 27 to 21 to 3, whereas I would assume a hung parliament is when you have 25, 25 or 26, 25. Not really, but I think it can be a worthy point to address. And anyway, thank you very much, Alan. I appreciate your presentation is very thorough and I look forward to receiving a copy so that I can use it for my future presentation. Okay, I'll send you one. Thank you for the comment. Can I just get back to this issue of people looking back into the past is to the past system. Yeah, it's not fbtp. It's not alternative vote. It's something totally different. And I think if we're talking about a hung parliament, we're going back into the past again. Hung parliament is a term you use with majority plurality systems where you have a obvious winners and losers in systems. It is more likely than not that you will have a coalition government. It's not going to hung parliament. The idea of trying basically to promote negotiation and conciliation between different political forces. I don't have the data sitting in front of me now, but I can I can find it if you like. But the vast majority of governments that come out of PR systems are coalition governments. They're not hung parliaments. They PR forces parties to negotiate to form a government to try and get a broad based government. Now it has the downside that it could be less stable than a majoritarian or plurality based parliament. But they don't have hung parliaments in PR. They have negotiations to form a coalition. Okay. I think there's a point that was raised per centium that I need to comment on which relates to the fact that you know parties need to first cross the threshold and then only can they get representation and then the fact that the city allocation method used only comes into play. Or it is only going to have an effect only after the threshold is met. So, one of the reasons why we have been advocating for a departure towards adoption of the same log method. It's because the dog system has, you know, from the experience of elections actually a marginal benefit to the larger parties. So, given the pages dynamics, we do not want to move towards a two party system, which has essentially been the case. And which has been the outcome under the 1997 constitution, which led to two rights based parties. You know, being the only place that gain representation in parliament and it had a polarizing effect on the on the electorate. And on the general nature of politics, which led to a race based politics in Fiji. So we want a system that actually reports smaller parties and which, you know, does not eventually lead to a part two party system, which can very well happen in Fiji's case because as I've said earlier on, the smaller parties are sort of getting caught in this perpetual cycle of exclusion. And, you know, in the future, there will come a time when there will be no longer smaller parties that would be contesting elections. So in order to sustain the multi party system, we need to have a system that rewards moderate smaller parties. So, unless there's any further comments or questions. I think my mic is from you. So I can speak. If you don't mind. Okay, of course. With all due respect from 2014. There has been an argument in Fiji and I suppose Ellen can run the numbers for you. That's going to a two party system. But if you look at the distribution in the 2018 election and use that as a baseline, you will note that it is now expanding into a three or four party system. After the transition election in PG, the incumbents came in with the majority vote. But following that in the second election, the trend will continue in the third election is that the number of votes is going to get distributed as more and more parties become popular and come up with more central views on issues. So, whilst it's a it's not proven 2014 that he is moving to a two party system. Simply look at the votes and if he got it was more than 2014. So, essentially you are looking at in what Ellen said coalition heading forward. Into the next few elections. So, in terms of. I think the votes are proving it wrong. I think there's some evidence if you consider the voting behavior in the two elections that have been held that people have actually shied away from voting for the smaller parties. In the second of the two elections, which means that people do think that if they vote for parties, which don't have a reasonable chance of crossing the threshold. The vote will be wasted. So, eventually we believe this will impact on the on the sustainability of a multi party system in. But, well, time will tell. And, but, but there's some evidence to a test to the view. Analyze the results of the two elections as a PR system that have been so far. Okay, so I believe there's no further comments or questions. That brings us to the end. I'd like to once again, thank you, Ellen, for your very insightful presentation. And I think the responses that you made to different questions that were posted in the course of the webinar. They've also allied a number of concerns and clarified a number of issues. The election, the election administration as well. So, thank you once again to everyone that joined us this afternoon. And as mentioned earlier by Lena, this is only the first of the webinars that we would be holding under the democratic development in Malaysia. We've been a series. So we definitely look forward to your participation and contribution in the upcoming webinars on behalf of international idea. Thank you once again. And we look forward to your continued support. Thank you. So just before you go, Lena, would you like to. Okay, that's all right. Okay, so thank you Ellen once again. I'm from here in PG. We some more day.