 Yeah, thank you for joining us either on Webex or on Facebook Live through the first lecture in our series. This lecture series is conducted by International Idea and Friends from a number of Asian and Pacific countries. This lecture is currently planned to last for the remainder of this year and it will be held every other Wednesday. So, the next one will be two Wednesdays from now on the 11th of October. And it's designed to last about one and a half hours preceded by a 20-minute presentation by an expert. And the lecture, the presentation will be followed by a Q&A session. Now, the Q&A, the questions, we ask participants to ask the questions through the Q&A facility on Webex. For those on Facebook Live, I apologize that you will not be able to ask questions unless you log on to Facebook Live. So, my name is Adyaman and I'm a Senior Program Manager at IEA's Asia and the Pacific Regional Program Office in Canberra, Australia. And I would like to thank our panelist Alan Wall for his time to speak at this first lecture. I would also like to thank International Idea and Friends who have provided their inputs and also lend their hands in the organization of this lecture. So, as I mentioned earlier, there's a pre-lecture poll that we conduct and it's still open right now. So, I would encourage you to fill the polling out and also there will be a post-polling survey. Our first post-lecture survey, about four years, towards the end of the discussion session. And for quality purposes, there will be a survey for us to gain your feedback on the quality of this lecture. So, please do participate in on those and I also inform you that during the discussion session, there may be questions asked in the polling section. So, please pay attention to those and participate. So, the poll is about to be closed. I'd like the poll to be closed at the start of Alan's presentation. So, we still have a few minutes. So, before I hand the floor to Alan, I'd like to introduce him actually. Alan Wall has over 30 years of experience in electoral administration and as a democracy advisor. From 1985 to 1994, he held various management positions with the Australian Electoral Commission. He was country director for the International Foundation for Electoral Systems, or IFIS, in Azerbaijan in 1999, 2000, in Indonesia from 2000 to 2004, Nepal 2010 to 2013, and Kosovo 2013 to 2016. He then continued as ICANN Senior Election Advisor in Timor-Leste in 2017 and 2018. He has been a senior electoral official for the United Nations in Eastern Slabonia in 1996 to 1997 and Nigeria in 1998 and an advisor from 1994 to 1996 to the South African government for the first democratic local government elections in South Africa. From 2005 to 2009, Mr. Wall was senior advisor to Democracy International and directed the election assistance programs in Indonesia, opinion polling in Indonesia, and research capacity development program in Timor-Leste. In 2007, he held a fellowship in electoral systems at the Netherlands Institute of Multiparty Democracy, or NIMD, in The Hague. Alan has also made reviews of voter registration and or election management systems in Buktan, PG, Iraq, Macedonia, Malawi, Mozambique, South Africa, and you think. Alan has published widely on electoral issues, including as co-author of International Ideas, handbook on electoral management design, on electoral systems for NIMD, on voter registration systems in Africa for the Electoral Institute of Sustainable Democracy in Africa, or ASA, and as a lead author of the ACE Project's Electro-Encyclopedia in the Fields of Electro-Management and Voting Operations. So, I would like to now invite Alan to make his presentation and title openly his proportional representation, The Good, The Bad, and The Ugly. So, we are closing the pause now and hope you find the lecture beneficial. Alan? Okay, thanks, Adi. I'm going to have to try and share the screen and hope that this software works. Great. So, if people can't hear me, please let me know. My mic's on. Hopefully it's working. Thanks, everyone, for joining. This is not a lot of time to try and talk about Open List Proportional Representation, or OLPR. And I must admit, a disclaimer, I didn't choose the title of this at all. And I think, listening to some opinions about OLPR, a more appropriate title to some people might have been the title of the first two movies in the series. The first of which was A Fist for the Dollars, and the second of which was For a Few Dollars More, which seems to be the impression some people have of OLPR. But I sort of have a confession to make. I find OLPR configured sensibly and with adequate controls on political behavior, an arrangement that can serve very well to facilitate a wide range of electoral system objectives. But first, let's get a couple of basic things out of the way. I'm not trying to teach in the English expression, people teaching your grandmother how to suck eggs, but a few things just to get out of the way. First of all, electoral systems don't determine behavior. And a lot of the commentary I see on electoral systems seems to assume that by changing the system, you don't just change the vote to seat ratio, but you also determine how political contestants behave. And they, systems may be able to facilitate certain behaviors, but changing electoral system doesn't necessarily solve perceived political problems. It's the attitudes of the participants, the controls on political behavior that are the critical issues. And secondly, that no electoral system is perfect. I mean, the system you don't have often looks a lot better than the one you do have. The issue is how well does it serve a mix of objectives rather than one particular objective. Now, OpenList PR is described by a lot of people as more of a concept than an electoral system. And it's interesting to see how people do that. Because usually electoral system operates in largely the same way in all countries that use it. There may be some differences. For example, in CloseList PR and seat allocation formula or threshold for representation, but the mechanics are generally the same. But that's always the case with OLPR. You think 40 countries through the unicameral parliaments or national houses and my rough count, there's more than 70 in significantly different ways of implementing this concept in these 40 different countries. Most of them to do with how lists are organized, how many votes voters have and how lists are calculated. And that's what I'm considering issues like thresholds for candidates or lists or single or multiple tier electoral districts within an OpenList system. So there are almost as many variations of OLPR as there are countries that use it. Some variants are good. Some don't work so well. Some have been like in Columbia pre-2003, plain ugly. But the thing about OLPR is that can be configured to satisfy a relatively wide range electoral system objective, if you get it right. So what is the concept that we're talking about? The concept basically is that voters have the choices to represent them. And that's within proportionally and potentially inclusive electoral outcome. Now, this fits with increasing trend of giving voters more power over who represents them. The second part of it is that votes are pooled. Not only do voters have a chance, but votes are pooled so that seats are allocated in accordance with the total votes for candid lists. And this could be calculated in many different ways. And after any representation threshold has been considered. So there are basic things that are the same. But then there are multiple basic differences, mainly in how much power the voter has to determine which candidate selected, whether the system is more candidate focused or list focused. Some of the worst of these variants have been incredibly candidate focused, like the one that used to be used in Columbia in the late 20th century, which resulted in huge numbers of single person lists. What people like Matthew Sugar have called a personal list style of election. So the concept, the voter not only has a choice of what party or list they want to support, but also which are the party's candidates. So now I come to the variations in the detail and some of the variations are common to all PR systems and need really to talk about them and district magnitude of legal thresholds for representation and the methods of turning types into seats. In many or more rather, most of the variations in how LPR are distinctive to open list PR. And they're down bound to the voting methods. The party or actually a candidate, a candidate or other party or candidate, then how many candidates can you vote for and do they have to be for the same list and can you vote against candidates for candidates or change their order on the list. And can you do that across every part of this list. And this then feeds the ballot paper design issues on the how party candidate lists is shown on the ballot. Now in most systems, fairly standard way of showing your candidates on the ballot paper, but under open list, you can have many various ways. How are the list designed out of a single or multiple party listener district. Can they contain both partisan and non partisan members. Can you have independent candidates for lists. There are vote threshold to qualify candidates for seats one by the list. So instead of the highest voted candidate necessarily winning that candidate has to have a certain proportion of the party list vote to be able to be considered to have a seat. How votes only for a party distributed to candidates is at least three different ways of doing that. And how to quotas disadvantage groups work. And this be difficult in open list PR compared to closely PR. Often they need people need to either reserve to reserve seats, which has its difficulties or a best loser formula, which will look at a bit later. So these variations when you look at the 40 countries that use open list PR, you can see that the variations are really widely spread. Most common is to vote for a list and either be able to vote for a single candidate or multiple candidates within that one list. Once you get past that you get huge variations. The nine countries where your list total votes is made up for the total votes for candidates. And that may be for voters who voted for candidates from the one list or multiple candidates from across lists or multiple candidates across lists and also be able to negatively vote for these candidates. Then you have the systems where you can either vote for a list or for candidates and those votes are totaled to get your list totals. And again, you've got five major variations within that. So yeah, there's lots of different variations just purely in the voting method and how your list totals are calculated. So what sort of issues and advantages does OLPR bring to the table? Well, firstly, it's a modern democratic reform in line with trends over the last 30 years. It gives the voter more influence and B provides a more proportional inclusive outcome than plurality majority systems and it can combine these two together if the system parameters are right. It can provide the voter with multiple choices so the voter can have not just one choice of who's going to represent them but more than one. It can strengthen intra-party democracy and limit the power of executives to direct people within the party. That has of course an opposite to it which it could promote party fragmentation as well. It can promote both local and proportional representation depending on how candidates are chosen for the list and how people tend to vote for people from their own locality or from somewhere else. It does provide opportunities for independence and non-partisan groups. We'll see this is possibly a bit overstated in a minute. And overall it can be simpler to operate than similar systems that can try and combine these things like mixed member proportional systems. Now, talking about independence just briefly, one of the strong advantages that people propose for OLPR is it allows independent candidates to compete equally which isn't necessarily the case under closed list. Well, when we look at how this is distributed, there's really 13 of the 40 countries don't allow any sort of non-political party list to compete. There's only seven of them that allow individual independence to compete whereas others allow independent lists or independent candidates and lists or some sort of non-partisan groups. When we look at the result, there's really only three countries using OLPR that have significant numbers of independence in Parliament. That's Chile, Jordan and Lebanon. Now, this is something that many in civil society seem to think is useful to have some sort of non-partisan presence. But OLPR doesn't necessarily deliver that. Then what could possibly go wrong with OLPR? Well, the main thing is and this is where it comes in for a lot of criticism is the external political environment has to be calibrated appropriately. Now, two really critical things are the ease of political party formation. So if you have very, very easy formation of political parties and you can have particularly multiple lists from the one party competing in an election, you get the situation you have in Colombia pre-2003 with single person electoral lists. Really importantly, there need to be adequate controls on candidate funding and expenditure and particularly controls on a legal vote by. Though, to be honest, without these controls, any electoral system is going to be in trouble. However, because of the nature of intra-party competition in Open List PR, it can be seen to exacerbate tendencies as can plurality majority systems to have individual candidates out trying to buy votes. There is a case that can be made that if intra-party competition does get too intense, it can lead to party fragmentation or a lack of actual influence of parties in the parliament. And this happens, say, in countries like Jordan where you can have mixed partisan and non-partisan lists. It's likely to happen, say, even where I've been looking recently at a potential new electoral system for Afghanistan where the incentives are in dependent candidates are very, very similar to those that they've had under SNTV in the past and it may not promote what they're looking at for, which is a more cohesive party structure. One issue that you have in OLPR systems where you can have a multiple number of votes, but not a fixed number of votes. And in many OLPR systems, voters have as many votes as there are vacancies to be filled in the district. So it's a floating number. If there are five vacancies to be filled, you could have between one and five votes and have a valid vote. In other countries, again, like Kosovo, you have a fixed number of votes or up to a fixed number of votes. You could vote for one, two, three, four or five candidates. The problem here being that reconciling ballot papers becomes very, very difficult. It's not possible to reconcile the number of votes for candidates with the number of voters who have voted. And that leaves room as it has done in countries like Kosovo in the past for people to try and manipulate results sheets post-election. It can have an interesting impact on ballot paper design. Now, the issue with Open List, of course, is that you have to be able to know who all the candidates are to be able to see who you're going to vote for. And if you've got particularly a large district magnitude, that can lead to huge ballot papers. So people take a number of ways of looking at this. They could try and code the candidates or they could randomly put them on the ballot paper to try and reduce the influence of political parties. So here's a few examples of ballots under OLPR. First one, I don't like it all from Fiji, which basically randomly assigns numbers to each of the candidates. They're not in any party order. There's nothing either on the ballot or in the voting compartment to tell voters which party each of these candidates comes from. Voters have to know. So that weakens the sort of party consolidation that you may well need in a democracy. Similar way of doing things using numbers is from Kosovo, where you both vote for a party and can vote for up to five candidates. So you vote for a party name on the left and then you have each of the party's candidates has a number. And you pick that number and in each voting compartment is a list for each party of which numbers each of their candidates have. Then if you want to go for names, you get a very, very big ballot paper. This isn't the biggest when you look at, say, from DRC, you had 16 page ballot papers for candidates in an OLPR election. But here you can see that you can get a very large ballot if you are listing all the names of the candidates. If you have smaller district magnitude like here in Honduras, you can have a more compact, compact ballot paper in here. And this is one where there is pan a charge where you can vote across parties for candidates. And where you vote for the candidate rather from the party, they've taken the view that it's easy to list the candidates horizontally rather than vertically. It's a fairly unusual thing of ballot paper design, but ballot paper design becomes critical and the OLPR to be established easy links between the voter and the candidate that they want to vote for. Let's talk a bit about district magnitude issues because it does have an impact on this ballot paper design. There are only six countries that use OLPR that use a nationwide electoral district and they're listed here. They also use smaller electoral districts, which of course smaller electric district magnitudes. They make voter choice and ballot design a lot easier under OLPR, but their big disadvantage, of course, is they raise the natural representation threshold. So they ended up reducing proportionality, which is one of the things you're probably trying to do with a list system and gives you more wasted votes, is you're trying to in a list system and to give you an indication of what sort of impact this has. If the district magnitude of 15, the natural threshold for representation is around 5%. If bring it down 30 to 10, it's around 7%. To six, the threshold was just under 11% estimated, not necessarily hard and fast, but these are best estimates experience. So if you do try and make the ballot paper easier for people to understand and easy to make a choice of candidate, you are likely to raise the representation threshold significantly and cut out smaller parties. Now, to try and around this, some countries that have open list proportional representation have a two tier system. They basically have districts and they have a national compensatory district, sort of like under MMP, but you have the one electoral system, open list in both national and open list in district systems. And basically the national seats being compensated to make sure that these natural thresholds do not reduce inclusiveness in the overall parliament. Okay, what else can possibly go wrong? Ah, well, you could put limits on voter choice. You could put candidate vote thresholds and what some people call flexible OLPR. So in Croatia, for example, a candidate has to win at least 10% of the votes that the total list gets to be considered for a seat. And that then if not enough candidates get this much, many votes rather than seats are allocated or the remaining seats are allocated in order on the party list like in closed list PR. And there are other countries with lesser provisions. One thing that can happen is a public unhappiness with what we call leapfrogging. So an example in Fiji in 2018, there were 14 ruling party candidates that are elected with fewer personal votes than people from opposition parties that weren't elected. Because of what we call the coattail effect of high scoring candidates up the top of the list. Now in this OLPR can have similar issues to both single transferable vote and alternative vote where the people you don't expect to win win. But to be honest, even though you can have candidates elected here with few votes, they're not elected with zero votes as they are under closed list proportional representation. Implementation agenda and other representation quotas can be difficult as I noted earlier. Now they may need to rely on what we'd call a best loser leapfrogging. If you have a gender representation quota, then it's insufficient women are not elected. Then the least or the lowest vote meaning the lowest vote winning a man lose their seats and they leapfrogged by the highest unelected vote winning women. You could have reserved minority seats. So either way, it's a bit awkward if you want to try and work representation quotas with OLPR. You can't do it. You can do a CLPR and zip your list to make sure that you're going to get a certain proportion of disadvantaged groups elected. Another thing that can annoy people is the use of apolitical star candidates and their coattails effect. I mean, in the Netherlands now it's basically unannounced policy that the last person on every party's list is a star, star sports person, star theater actor, whatever to try and attract votes to that list. Though that person probably has absolutely no intention of taking a seat in Parliament if that party gets sufficient number of votes. So this can be seen as dragging people with few votes along on the coattails of candidates that have high visibility and high recognition. And it may be possible to through lower district magnitudes to reduce the impact of this, but it may not. If a voter votes only for a party, which candidate gets the vote? Now under some systems, no one candidate gets the vote. Under others, the first on the list gets all the votes for people who vote for the party. And then under other systems, if someone votes for a party, everyone, all the candidates, pick up that vote and it gets included towards their personal vote. This is something also that OLPR can be criticised for is that which one of these you use. There are a couple of common misconceptions. Firstly, that party nominees can't distinguish themselves by running on party platforms, but this to me is a straw person. I mean, party nominees run on their ability to implement the party platform that can distinguish themselves quite neatly through doing that. And then this major one that OLPR facilitates vote buying. And as I noted before, this isn't a function of OLPR, it's a function of poor political finance controls. And I've yet to see a study with a control group in it. So something that shows CLPR and OLPR at the same time in the same country to indicate whether OLPR actually results in higher vote buying incidences. Now, it may be over time or compared across countries, but that's not a really scientific way of comparison. It's also said that it can win parties as aggregators of opinion. But this also can be controlled through policymaking through internal party democracy and also the way in which candidates selection processes can be implemented so that you select the candidates that are actually going to push the party policies. Okay, very, very quickly because of running out of time. These are some countries where OLPR has been relatively recently introduced or modified. And the notes say in some places like Honduras, Panashas are the introduced now in most of these places the idea of OLPR has been to increase voter choice. And that has been seen as the most important priority objective for the electoral system. So many of the electoral systems have to satisfy a mix of objectives and people have to basically identify because many of them are contradictory, which is the one that's most important to them. So in these instances, it's been introduced as an increase in voter choice. Okay, about the end of this now. I think I'll get within time summary. It's OLPR can be seen more of an electoral concept with multiple variants rather than a rigid electoral system, something that you can play around with to get the best, the mix of components for your particular environment. Secondly, none of these variances perfect, but then not as any electoral system. Third, with a carefully calibrated framework for OLPR, you can put into a simple package, both voter influence and proportionality, if that's what you're looking for in your electoral outcomes. And it's the simplest way probably of doing that. It's a flexible electoral concept. You can't reinforce this enough. You can be configured for specific objectives. It can be wrongly blamed for non-system factors that could be present no matter what the electoral system. And it does need to be configured carefully to maximize its positive impact. So like any electoral system, it has disadvantages. I tend to see that it's more, it can be a cue into the good than the ugly, but there certainly have been some ugly vary events of it in existence over the past 20 years. Okay, thanks very much for listing. I don't talk anymore. This is about you guys discussing things, hopefully amongst yourselves. I don't want this to turn into a Q&A directed at me and Artie, but this is something really that you guys are in the field. You guys are having to deal with your electoral systems discussing between yourselves how OLPR does or doesn't fit into your sort of environment. And if it brings advantages or the disadvantages are too great for you. Okay, thanks very much for listening. And Artie, can I hand this back over to you now, please? Sure. Thank you very much, Alan, for the presentation. I'm sure it's insightful to everyone. Unfortunately, this lecture is meant to be a two-way communication between you and the participants, although participants can of course discuss them with each other through the chat function. So we've got a few questions already, Alan. The first one is from Michael May Lee. How might an OLPR ballot structure work in a country like Timor-Leste, which votes as a single electorate with therefore a large number of candidates? Go ahead. Hi, Michael. Not all that well, as you can see. Can I, Artie, can I pull this back to sharing screen again? Yes, of course. Okay, let me just... Okay, so these are the sorts of choices you have. The Fiji example where you code everyone, and then you have something, hopefully, that tells people what each of those codes means. Finally, the Netherlands example in the middle where you have a huge ballot paper with all the names listed on it. And in Timor, that's going to be a significant number of names on a ballot paper, not as bad as DRC, but a large number. Kosovo variant, where you have the parties, and I would imagine, given the history in Timor, you'd still probably want to vote for a party and vote for a candidate, where you have the parties listed and then you have a coded number of candidates, or coded numbers for each of the candidates in each voting compartment, has a book, basically, that says these are the candidates and these are their codes. So you have choices, but as I said, for huge numbers of candidates, OLPR can be messy to get both a ballot paper design that is succinct, and also to have a link between the voter and the candidate established, which is why some countries go to two-level OLPR and have national compensatory seats for smaller and smaller electoral constituencies, both running OLPR. Yep, I'm here. Okay, so there's another question. Sorry about this. There's a second. I can't see. Eric Larson, would you mind reposting your question? Okay. So Eric Larson asked, I appreciate the point that OLPR can allow party nominees to distinguish themselves. OLPR would allow voters to have say both on parties and on particular party priorities. Yet in some places, especially with a single national constituency, media parties and hence voters may see their task as voting for national leaders rather than for a representative parliament. What can be done to decrease the focus on selecting the national leader? I'm not sure that as long as you have dominant personalities in politics, that's a fact of life. I'm not sure that you're going to be able to get away from that. It could, and you can try perhaps by reducing the or making multiple electoral constituencies, in which case this dominant national leader can only win in one of those constituencies. Now, one of the things I wanted to look at before here was to see if this really did have an impact. And I wasn't really able to get hold of any sufficient data to be able to see whether it did or not. And in countries say like Finland, even when you reduce the number of the district magnitudes and have multiple electoral districts, you still have dominant vote winners in each party. And I'm not sure that this is something you're ever going to get away from no matter what the level of what the level of district is going to be under OLPR. But it probably will be significantly less if you have multiple electoral districts. Then of course you have your proportionality and whether you want to have some sort of overall compensation within the OLPR system. But certainly say in the case of countries say like Fiji, there's been a dominance of a particular personality for a particular party that has dragged that party's vote along on his coattails. But political parties would say, I'm sorry, this is reasonable and natural if we have a dominant personality is super popular. Why shouldn't we be able to take advantage of that? Okay, next question is from Dean Crabb. What Australia benefit by moving to OLPR for its Senate, where it is virtually a closed list PR? I'm not sure everyone's going to agree it's closed list PR in the Senate in Australia. Look, this comes down to history of things. STV is it's been taken away from its initial individual candidate focused and made a lot, a lot more this focus through what people have called reforms over the years, which has also made it easier for voters to vote. I'm not sure that there's going to be any real advantage in Australia to move to OLPR from a system that people are now used to and seem happy with. Next question, Alan is from Awang Ilham. Can OLPR help in reducing the incidence of MPs elected switching sides as what happened recently in Malaysia? I'm not sure it can reduce the incidence of that. I mean, what you need is proper is is clear anti defection laws. I mean, you can build an anti defection law as the Indians have into a first pass the post system. You can build it into an MMP system. You can build that into any electoral system. I'm not sure the electoral system is going to, you can build it into closed list PR like you can in South Africa or tried to do in South Africa in the past. But I think the thing here is the strength of the anti defection laws rather than the electoral system itself. They can be two separate things. They don't necessarily have to be in lockstep with each other. Anti defection laws can work under any sort of electoral system and cannot work under other electoral systems if they're not there. One of the things with OLPR say compared in common with majority plurality systems and unlike closed list PR is that there's no question. In many ways that the person has been elected on the strength of a personal vote, no matter how small that personal vote may be. Now, Venice Commission and others would say that no matter what the electoral system, there is no. There is no imperative vote. There is no way you should take a seat away from someone who has won it. But certainly if you have a personal vote, you have a stronger argument for retaining that seat with you no matter what party you belong to. But look, it's up to the strength of the anti defection laws rather than the electoral system itself. I can't hear you, I'm sorry. There's a comment from Carlito Nunes. Hello Carlito. I'm sorry, you're disappearing. So in case of, he made a comment in case of Timor-Leste's OLPR system is good, but technically very challenging to the EME, the ElectroMesh employee. I'd also like to add a follow with a question from once again. I can't hear a thing. I cannot hear anything. Okay. How about now? Yep, now I can. This is from Wang Xinhua. Did the introduction of OLPR to replace FPTP or block vote in Sri Lanka and APV or alternative vote in Fiji encourage inter-communal moderation? Sorry, I didn't hear what the question was in that. Did the introduction of OLPR to replace first pass the post of block vote in Sri Lanka and alternative vote in Fiji encourage inter-communal moderation? Look, Sri Lanka, I don't know Wang Xinhua. So I can't really comment on that. It's nothing I've ever had anything to do with Fiji. It doesn't seem to have been much in the way of moderate parties evolving. Though on the other hand, OLPR has not resulted in a coup either. Unlike the introduction of AV, to that extent, things have moderated. But certainly there are still in Fiji very, very few parties. It hasn't seemed to encourage moderates to form their own parties. And certainly the view of many of those I talked to in Fiji and that the parties still tend to be dominated at the extremes. So to that extent, probably not. Yeah, I'm here. There's another question. There's a question from Paul Thornton Smith. How does OLPR differ from STV in terms of political culture and election results? Good to do. In terms of election results, it's probably hard for people with very small fractions of the vote to be able to get elected so that it is not as fragmentary as isn't necessarily the fragmentaries of STV. It's certainly easier for people to understand than STV can be. It doesn't give the exact proportionality that some people love of STV. In terms of political culture, it does create, it does, depending on how STV is configured. And if it's the Irish version where everyone is sort of listed indiscriminately on the ballot, it possibly doesn't have the same party cohesion effects as it does somewhere in the parties and people can vote for parties within a supposed STV system. So politically, so party-wise, possibly not as can be not as fragmentary. Voter-wise can be easier understandable than STV. Political culture-wise, it probably has equal opportunities for candidates to compete against each other. Next question is from Eric Lovier. What are instances in the adoption of the game, game, system, and managed allocation of STV? Sorry, Adi, I couldn't hear a word of that. I don't know what's happening, but you're breaking up really badly. Yeah, I'm probably too far from the mic. What are instances when the adoption of OLPR gamed GAMED, the system, and manage the allocation of seats? It depends again on how you configure OLPR as to whether you can do that. Are you doing it in a very complex system, something like Jordan, where you have basically few political parties? Yes, it should be gamed really easily. If you have it somewhere like Columbia where your incentives were zero for vote pooling and you're basically incentivising people to have their own electoral list, yes, the system was totally gamed. It became something more like an SNTV system than an OLPR system. Where it's been adopted recently? I'm not sure if there are other places where it's been adopted recently where it has actually resulted in people gaming the system. Not in the same way as you can game the system. It's not the same way of gaming, you say, under MMP, which is aiming for a similar sort of result. Above having proportional and individuals being voted for where it can be easier to gain the system through splitting your vote. In OLPR, you can gain the system in countries where you can vote against as well as for. If you can organise people to vote against candidates they don't like, you can gain the system in that way. Particularly if there is panachage available and you can vote against candidates from a party that you have absolutely no intention of voting for. As in places like Switzerland or Luxembourg. So yes, you can game it to that extent by negatively affecting other parties' candidates that you particularly don't like. Yeah. I apologise to those following on Facebook. The live feed has been interrupted a couple of times so apologies for that. But we try to get the live stream back as soon as that happens. So Alan, next question is from Ronil Kumar. How do you come to a threshold where it does not discriminate using the term loosely? How does one come to a threshold where it does not discriminate using the term loosely? Smaller parties and independent candidates as is the case is for Fiji. Threshold is 5% because when these smaller parties don't make it to the threshold. A large number of vote is wasted or we can say transferred to the parties that have qualified. Yeah, I'm just trying to think back to Fiji 2018. I think the issue to me more with Fiji is the lack of competitors. Sorry, why do I stop sharing this screen and look for some data for a moment? The issue to me is more with Fiji is, as I said, the fact that there are a few competitors rather than the threshold is stopping people. The 5% threshold would make a difference and simulations I've run recently show that if you got rid of that, if you had a 0% threshold in 2018 that you would have had. Depending on which allocation, seed allocation system you use the government could have either just had a one seat majority or had been in a one seat minority, depending on what seed allocation method you use. So yeah, it would make a difference there. One of the other things you've got to think of is what are the incentives to parties to actually participate as well? Even if you have a 0 threshold, what are the party registration requirements? And in Fiji they're fairly stiff and they're fairly stiff for reasons such as countries as Indonesia to make sure that there are regional and fragmentary parties contesting. But they are in Fiji. It's a very high barrier in terms of percentage of voter population that a party has to reach to be able to have enough supporters to be able to register the party. Independence, there are very few OLPR systems as I noted before where significant numbers of independence get elected. And of course they tend to be more easily elected where the threshold is lower, either naturally through a big electoral district or through having no legal threshold there. But hey, the Fiji situation, the getting rid of the threshold would make a difference. If you've got a 0% threshold, you would have had more parties represented in parliament. And I'm just still trying to find where my data is on this because it's not coming up because I just worked on this in the OK. So with a 0% threshold, no threshold wouldn't have, even having no threshold would not have helped hope or Fiji Labor get a seat. But you would have increased the seats to four parties with a 0 threshold and using some form of central lag or a largest remainder method of allocating the seats. So yes, that could have had, you would have had unity Fiji with a seat and you would have had an extra seat to the National Federation Party. So yes, it would have made a difference in Fiji. But thresholds are there, I mean, often thresholds are there as they are in places like Timor-Leste to try and help party consolidation and stop party fragmentation. And it's a deliberate attempt on the part of the system to limit the influence of small or independence in the parliament. Now, that's a rational system objective. It depends on whether that is the objective of the majority of the people who make decisions on what the electoral system is going to be. There are arguments you can make for consolidating parties or for forcing small parties into coalitions so that they have a better chance of winning seats and thresholds are one way of doing that. On the other hand, if your objective is to have as many different voices as possible represented in the parliament having a 0 threshold and most countries or the majority of countries using OLPR have a 0 threshold, that is the way to go. So, Kamito followed with another comment that in countries like Timor-Leste where the older generation and charismatic leaders are still in politics, they would like to maintain closed list PR system because it would give them more power in the politics as they decide who will be on the list. Do you have any comment on that? Well, there's two different issues I think here. One is, well, there's three. What sort of OLPR system do you have? Do you have where the one voter has total control and votes for the candidate only and those candidate votes are pooled to make the candidate list. In that case, you move to OLPR and the party leaders can lose control because while they can determine who is on the candidate list, they can't determine who people vote for. The majority of OLPR systems still have allow for a vote for a party. And while you do that, you can still buy if you have either candidate thresholds or even where you put parties or candidates on the list, you can have some control over who people vote for. I don't have data, but being up the top of the list, and this is why places like Fiji have random order of candidates to try and break this up, is if you're up the top of the list, it is worth a small percentage of the vote to you. The people voting for the candidate who is at the top of the list. The CLPR does maintain party control over who gets elected, but the varying degrees of party control that can be exerted over different forms of OLPR system as well. To Sham Raju, I think Alan has addressed the question on independent candidates in OLPR. But if you are not, if you have a follow up question, feel free to write it down. Next question Alan is from Yolanda Panjaitan. Are there indications that an implementation of any variations of OLPR is better suited in a parliamentary or a presidential system of government? Don't know, haven't looked. I'm sorry. I'll have a look at that, but that's not something I've ever looked at. So I'd be totally talking off, as we would say in my country, talking off the top of my head or talking through my ass or whatever. So I'd rather leave that. And if you don't mind, I'll have a look at that and get back to you. Okay, so that's parked Yolanda. Next question is Eric Larsen again. To what extent are intra-party debates made explicit in OLPR? For example, are there good examples in OLPR systems of candidates advocating for particularly legislative priorities? Again, I don't know. Sorry, not something I've looked at. So I'm not sure there's going to be any data. Sorry, in the countries I've worked with, where they've used OLPR, that hasn't really been all that evident. So no, it's not something I have an opinion on. Sorry, Artie. Yep. So that's any more questions? That's it. That's all I can see here. One of the things to do with thresholds that I was looking at recently that I found interesting is that there's no real correlation between percentage threshold and the number of parties in parliament, which I found quite interesting. The vote that thresholds don't seem to be a determinant of vote spread. There are countries with 0% thresholds with few parties in parliament and countries with 5% thresholds with, say, like Algeria, with 22 parties in parliament. So thresholds don't really determine how many parties can be in a parliament. It can influence, but it's not the direct correlation. Okay. Artie, I've lost you. Okay. Sorry, the internet connection was intermittent, so it disappeared for some time. So how about, I know, okay, there are more questions now from Wang Qinhua. Alan, you list vote buying as non-system weakness. You list vote buying as a non-system weakness. What would you consider as a non-system weakness? So vote buying as a non-system weakness. Yeah, yeah. What would you consider as systematic errors? In OLPR, as issues of the systematic, have I got that correct? Have I understood that correctly, Artie? What would you consider as systematic errors? In what? In OLPR? Yeah. The major issues are things like party fragmentation, ensuring electoral integrity, and the whole issue of connecting voters and candidates and doing that effectively through ballot paper design are three potential big weaknesses in OLPR. And they're inherent in the system, and you can mitigate them through design functions, but you have to have fairly interesting workarounds to do it. The other system weakness is the inability to, in somewhat some people would say, fairly implement quotas for disadvantaged people. And that also is a weakness within the system that to get representation quotas to work, you have to bounce out people who actually want enough votes to be elected. So they're four fairly significant system weaknesses. But then every electoral system has weaknesses. You pick your weakness and you pick your strengths as to which ones are more important to you. No, there's no perfect electoral system, is there? No, what amuses me is the fact that the electoral system you don't have is so often better than the one you do have, meaning for the fact that it's different. Now, I mean, Afghanistan and Iraq at the moment are really fascinating. Afghanistan with SMTV looking to move to a fairly mild version or fairly weak version of OLPR to empower parties because parties don't have enough power. Next door in Iraq, they're thinking of moving from PR to a majoritarian system because they don't like the fact that parties have power and they're trying to break that. So whatever you don't have is always better for you. Okay, there's a question from TP Angraini. Yep. Of electoral justice scheme, we can offer, can we offer to reduce the vote buying phenomena? For example, in Indonesia, vote buying becomes a big enemy of OLPR. Yeah, but there's no, it's a big enemy of OLPR because there aren't any restraints on people buying votes. No one ever gets punished for it. That's not a matter of an electoral justice system. That's a matter of the willingness to actually take punitive measures against those who conduct vote buying. Again, it's a political culture weakness. It's a political will weakness rather than a system weakness. I'm not sure you could have any number of systems and it wouldn't make any difference if it wasn't the intention to actually implement the penalties. The pile has a similar issue and that has a really strong codes of conduct on everyone that says this monster code of conduct for everyone who has anything to do with elections, but the problem is actually enforcing it. And as long as there isn't the enforcement there, it doesn't matter what system in place you have to control it. So enforcement is key. Enforcement is key. Enforcement is not just is political will plus enforcement mechanism, not just enforcement mechanism. Okay. Mercedes Robert has a question. How would you compare in terms of the election of women between open lists and closed lists? Well, as I've just said, it's a lot, lot harder under open list because it's much, much harder to be able to implement quotas under quotas for representation under open list or even, well, okay, let's start from the beginning. Under both open list and closed list, you can have quotas for candidacy. Under open and closed list, you can require zipped candidate lists. However, whereas under closed list, that will result definitely in a number of women getting elected. Under open list, it depends which of those women people will actually vote for. So it can, so there's that problem with candidacy. It can mean that women have to have incentives to be able to campaign. Either more state funding to be out because we all know that women have difficulties in raising funds for campaigning. Or other assistive measures for their campaigns in terms of representation. It's another double whammy or double barrier there in that. If it's harder to put a candidate, sorry, a representation quota on open list PR without upsetting the will of the voter. The voters have said these are the orders that we have voted for candidates. So hang on, there's not 30% women there. What do we do about it? We get rid of men that have been voted for and replace them with women who have not been voted. We have not received as many votes if you want to have a representation quota. Or you have a reserve section of reserved seats for women, which I'm not really in favour of. That can produce more second-class representatives. My Osovo, which was under OLPR, is that I had a quota system for women and it had a best loser system. In the first couple of elections, few women got elected in their own right. By the third election, over 50% of the women who were elected were elected through the number of votes they received. And this has been increasing. So the quota was actually, even though it was a best loser quota, was leading to women being able to show that they could perform as parliamentarians and leading to them, more of them receiving sufficient votes to be elected in their own right under OLPR. But dealing with disadvantaged groups and minorities is harder under OLPR than Closed List. And again, it's something you trade off with a voter choice or a shortness of disadvantaged group representation. Hello. It took some time to unmute, sorry. Wong Chin-Wat has another question. Wong Chin-Wat has too many hard questions. Well, he's got a lot on his mind about this. In comparison to FPTP and CLPR, how far does OLPR or which of its variants encourage policy-based parties, or in other words, programmatic parties? I'm going to stop Chin-Wat asking hard questions. Again, I'm not sure that the system itself does that as much as the attitudes of the people who are in the political parties. CLPR, because it's a unified party, can possibly be better at encouraging policy-based parties. Because you have individual competition within an OLPR system, and similarly as you have within FPTP, you can have more of a personality-based politics if that is the culture within the country. And yeah, that can be another weakness within it. And that leads again to this issue of fragmenting of ideas within a political party under OLPR. That can be mitigated by the way in which you develop policy within the party as to how much IPD there is in actually developing policy positions and in selecting the candidates. But yes, it can encourage a more personality-based rather than a policy-based environment because it's up to each individual candidate in some OLPR systems to garner the votes. Now in systems where you vote for a party and then may vote for a candidate, that tendency is likely to be less because it is still a fairly list and party-based system with the option then of voting for a candidate. That is likely to be exacerbated, this personality-based competition, where all you vote for, all the voter votes for is the candidate, does not vote for a party at all. So yeah, on the variant which you vote for a candidate only, it can be less encouraging of strict party policy development. On the variants that you vote for a party and then optionally vote for a candidate, it may be very, very similar to CLPR and the way the parties operate in trying to get votes for themselves as a policy-based organization rather than for their individual candidates. Okay, Michael Maley's got a couple of more questions as well. You haven't discussed counting very much, especially counting at polling places. How can that work under OLPR where voting is both for parties and candidates? I'm not sure there's going to be any difference that under CLPR you need to amalgamate all the votes for an electoral district before you can do anything about striking a result. Certainly, many OLPR systems vote at the polling place level and then amalgamate to constituency or amalgamate to nationally to be able to do it. There's nothing to stop you counting at the polling place level. The issue, as I said before, is an issue where you have multiple votes and where you have votes, particularly if you have panacharge and how you actually reconcile your votes at any level to make sure that the results are a fair and accurate result of how the voters have actually voted. And that's close to impossible. And that's another which I forgot to mention or may not have mentioned before is another weakness with the system is if you can have multiple votes, how you assure election integrity. But under OLPR is it normal to count at the polling place level? Michael's next question is what would be the challenges of implementing OLPR against the background of a COVID-19 like pandemic? Again, something I haven't thought about specifically in terms of OLPR, Michael, and you've done a lot more work on this than I, a lot more looking at OLPR or anything in COVID than I have, which is basically zero. In terms of voting, in terms of the actual voting process, it's going to be similar. It's going to be more difficult in some ways for parties to be over candidates to be able to campaign for themselves as against getting a party policy message across. To save CLPR versus OLPR, the candidates in their entirety to be able to get through to electors is going to be more difficult. So the voters choice might not be as informed between candidates in a COVID-19 situation. In terms of ballot papers, it might in fact be easier in some ways to have remote voting on a ballot paper than for OLPR. It's a large one than actually having to do it in person in the polling place in a hurry with more time to be able to make up a decision as to which of the candidates you want to vote for. In terms of counting, the issues are still going to be the same in terms of bringing ballot papers or bringing results together to be able to determine election results. Now campaigning and also in terms of completing ballot papers, I can see differences, but on the top of my head, it's going to be different from any other electoral system in a COVID-19 environment. Right. Dahli Agumar has a question. Do you have best examples of countries whose election system is OLPR and are using electronic voting or electronic counting? She's asking because e-voting and e-counting may be a way to reduce potential election fraud in a PR system, especially OLPR, because of its counting and recapitulation complexity. Yeah, Estonia probably. Do you have any examples? Yeah, I said Estonia. They probably have more experience with it, and I think they're probably more confident that they ought to be about their electronic and e-voting system, and it's probably the major example of e-voting and OLPR in the world. There's lots of materials on it on the Estonian Electoral Commission site. Now they were going to do it in Finland, but they delayed it due to security, and I think it's still delayed due to security concerns. But that was going to be the second big one in that part of the world using e-voting and open list. Okay. I'm encouraged looking at the Estonian example. Okay. The other places that use it are Brazil, which has also had relatively good experiences with electronic voting, mainly because they introduced it gradually. That's probably the most important one. There are people who use open list who stopped it, like Netherlands, or who were using it before and didn't use it. There were people who looked into it like Norway and haven't really taken it up, who use open list. But yeah, Estonia is probably the major example. The other place is Switzerland as well, whereas in some cantons, there have been tests of using internet-based voting also. But yeah, Estonia is the main example of OLPR and i-voting, and also Brazil and terminal-based e-voting. We're almost at the end of time, but I'd like to ask you a question. It's probably not directly related to OLPR, but to list proportional representation systems in general. In terms of formula, do different formulas truly determine the number of parties getting elected into Parliament? Assuming there are no thresholds. Assuming there are no thresholds? Yeah. Can I share a screen with you again? Just let me see if I can. And I'll show you an example of how they make a difference. Now I've got to work out how to share a screen again. Share application. I want to share Excel. Which one do I want to... Have you got an Excel spreadsheet that says 2018 Fiji election data there? Yes. Okay. So if we look at this, the way in which don't is applied in Fiji, you get a split. Oh, this is using 2018 election results. The way in which don't is currently used there, you get a split of 27, 21 and 3. Current system between the parties. If on those same votes I used the RLE version that was used in Italy, I get the same. If I use a basic... Sorry, Alan. Alan, can you zoom in, please? Can I zoom in? Yeah, make the numbers larger. Yeah. Hang on. I've got a problem that my... I've got a... It's hidden. Whoops. Let's try. This. Is that better? Yes. Okay. It's about as big as I can make it. So under don't, as is used in Fiji, you have this 27, 21, 3 split. Using the same figures under Imperiali, you get a 27, 21, 3 split. Using a basic St. Larg method, you get a 26, 24 and 1 split. Which you also get under modified St. Larg. And you also get under the... As done in Sweden and also as done in Denmark. There is another version, a rather weird version of St. Larg used in the early 20th century in parts of Scandinavia where you would get on these same votes a 25, 21, 4, 1 split. And then there's another version that's been used for internal elections where you go back to 24 and 1. And I just remained a methods group and hair quotas give you the same results. The 26, 24 and 1. And the Hagenberg Bischoff gives you a slightly different version, 26, 21, 4. So yes, on that same set of votes with no threshold, you can come up with a number of election results. With Daunt being the one that is more favourable to larger parties. And probably this weird version of St. Larg used in the early 20th century, Scandinavia comes up with a version more conducive to smaller parties. Or less large parties, medium and smaller parties. So yeah, it can make a difference. Now, my contention usually is that the election, the seat allocation method probably is not, is generally, and it depends on vote distributions, is generally not as important as getting rid of thresholds if you want to increase the number of parties represented. For example, if I put a 3% threshold on this same set of data. Sorry, let me, I went to the wrong sheet. If I use a 3% threshold, we're still getting a little bit more compressed. We're not getting this last party, the Unity 3G party picking up any seats as long as you have a 3% threshold, no matter what seat allocation method you use. If you retain a 5% threshold, again, it's not making any difference. You basically restrict it to the three parties, no matter which of the seat allocation method you use on this particular vote distribution. So the threshold is important, but when you've got zero threshold, certainly, and seat on vote distributions like this, it can make a significant difference into, well, yes, a significant difference as to how many seats parties win. That makes sense, Adi. Yes, it does. And it doesn't seem to show much difference, Alan. It's a relatively small number of seats, and it's a very skewed vote distribution towards the two biggest parties. Right. Now, it would have a bigger difference if the vote distribution was not as heavy. So when you look at here, the three smallest parties here, 10, 13, 450, have got less than 3% of the vote between them. Right. So if you even that distribution out, yeah, I mean, the seat allocation formula will react differently depending on your vote distributions. And I don't have one handy here with a more interesting to show you. Well, I do have one, but it'll take me a few minutes to find it. No worries. No worries. Okay. Thank you very much, Alan. There have been a whole lot of questions. It's been nonstop for about an hour. Some of which I could answer properly. Some of which I think I've skated over the surface of. Yeah, and there's a question that you talked as well. The question that I've batted away that I'll have to think about. I don't have an answer off the top of my head. Yeah. So thank you very much once again, Alan, for your time and your willingness to participate in this first lecture. And to participants, there is a post-lecture poll open at the moment. Please do participate. For those who did the first one, the pre-one, yes, the questions are the same. So what we're looking at is whether or not there's been a change in your views and understanding before and after the lecture. And after the event is ended, you will also see a post-event survey. This is where we ask to you about feedback on how we did today. So apologies, and this is our first time. So apologies for some technical glitches that happened. And also to Facebook viewers, apologies for having cut off twice. But it's been going on quite well for the last 30 minutes. So that's good news. And yeah, so thank you very much for your participation. And our next lecture will be on the 14th of October. Same time. So wherever you are, please mark your time. The topic is how does distant and online election campaigning affect critical freedoms? So we all know that in the COVID era, campaigning is having to resort to social distancing and therefore also online campaigns. So we're looking at how does that affect critical freedoms in terms of the candidates, the political parties in campaigning, how does that affect the leader? Thank you once again for your participation and looking forward to your participation next time. And thank you once again to international ideas friends. And if any institutions would like to join the circle of friends, please let us know. We are open for the next lectures. If you find this useful, please join in and also let others know in your network of this lecture and encourage them to participate. Thank you and until next time. Bye-bye.