Re: [EM] My summary of the recent discussion
I think Plurality can be claimed to be the ideal method for the single-member districts of a two-party system, but then one should maybe also think that third parties should not be allowed to run, and we should stick to the same two parties forever. I don't get it. of course, if there are only two candidates, there is no problem with Plurality (because it's also a Majority). so how is Plurality so flawed if we accept that a two-party system is fine and dandy? if not Third parties, for Independents? what is the scenario with two parties where FPTP is so flawed? I think you already said it. If you want a system that allows also third parties and independents take part in the election, then Plurality is flawed. Only if you think that third parties and independents should nor run, and there should be only two parties, then Plurality is fine. These contributions to this discussion take an extremely narrow view of representation of voters as it is clear this discussion is not about a single-winner election (state governor, city mayor) but about electing representative assemblies like state legislatures and city councils. There can be major problems of representation if such representative assemblies are elected by FPTP from single-member districts even when there are only two parties. Even when the electorates of the single-member districts are as near equal as possible and even when the turnouts are near equal, FPTP in single-member districts can deliver highly unrepresentative results if the support for the two parties is concentrated in particular districts - as it is in most electorates. Thus party A that wins 51 of the 100 seats in the assembly may win those seats by small margins (say 550 votes to 450 votes) but party B that looses the election with 49 of the 100 seats may have won its seats by overwhelming margins (say 700 votes to 300 votes). Thus the 51 A to 49 B result is highly unrepresentative of those who actually voted: 42,750 for party A but 57,250 for party B. These numbers are deliberately exaggerated to show the point, but here in the UK we see this effect in every UK General Election since 1945, where the win small, loose big effect of FPTP has consistently benefited the Labour Party at the expense of the Conservative Party. And where such vote concentration exists - at is does everywhere - the result can be greatly influenced by where the boundaries of the single-member districts are drawn. Move the boundary, change the result. These are fundamental defects of FPTP in single-member districts that must be addressed if you want your elected assemblies to be properly representative of those who vote. Both of these defects has greater effect if the electorates are not so equal or if the turnouts vary with party support (as they certainly do in the UK). So even when there are only two parties, FPTP is very far from fine. James Gilmour - No virus found in this message. Checked by AVG - www.avg.com Version: 2012.0.2178 / Virus Database: 2425/5044 - Release Date: 06/04/12 Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] My summary of the recent discussion
On 4.6.2012, at 13.49, James Gilmour wrote: I think Plurality can be claimed to be the ideal method for the single-member districts of a two-party system, but then one should maybe also think that third parties should not be allowed to run, and we should stick to the same two parties forever. I don't get it. of course, if there are only two candidates, there is no problem with Plurality (because it's also a Majority). so how is Plurality so flawed if we accept that a two-party system is fine and dandy? if not Third parties, for Independents? what is the scenario with two parties where FPTP is so flawed? I think you already said it. If you want a system that allows also third parties and independents take part in the election, then Plurality is flawed. Only if you think that third parties and independents should nor run, and there should be only two parties, then Plurality is fine. These contributions to this discussion take an extremely narrow view of representation of voters as it is clear this discussion is not about a single-winner election (state governor, city mayor) but about electing representative assemblies like state legislatures and city councils. There can be major problems of representation if such representative assemblies are elected by FPTP from single-member districts even when there are only two parties. Even when the electorates of the single-member districts are as near equal as possible and even when the turnouts are near equal, FPTP in single-member districts can deliver highly unrepresentative results if the support for the two parties is concentrated in particular districts - as it is in most electorates. Thus party A that wins 51 of the 100 seats in the assembly may win those seats by small margins (say 550 votes to 450 votes) but party B that looses the election with 49 of the 100 seats may have won its seats by overwhelming margins (say 700 votes to 300 votes). Thus the 51 A to 49 B result is highly unrepresentative of those who actually voted: 42,750 for party A but 57,250 for party B. These numbers are deliberately exaggerated to show the point, but here in the UK we see this effect in every UK General Election since 1945, where the win small, loose big effect of FPTP has consistently benefited the Labour Party at the expense of the Conservative Party. And where such vote concentration exists - at is does everywhere - the result can be greatly influenced by where the boundaries of the single-member districts are drawn. Move the boundary, change the result. These are fundamental defects of FPTP in single-member districts that must be addressed if you want your elected assemblies to be properly representative of those who vote. Both of these defects has greater effect if the electorates are not so equal or if the turnouts vary with party support (as they certainly do in the UK). So even when there are only two parties, FPTP is very far from fine. James Gilmour Yes, that makes sense. FPTP in single-member districts is not fair in the sense that it gives only approximate results (due to the inaccuracy of counting the seats in single-member districts) and is vulnerable to gerrymandering. A system that counts the proportions at national level (typically a multi-party system) would be more accurate. Also gerrymandering can be avoided this way. The discussed proposal of changing the Plurality/FPTP method of the single-member districts to some other single-winner method has still many of the discussed problems (inaccuracy, gerrymandering). One could try to construct also two-party systems or single-member district based systems that would avoid those problems. But maybe proportional representation is a more likely ideal end result. Practical reforms may however start with whatever achievable steps. Juho Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] My summary of the recent discussion
what is the scenario with two parties where FPTP is so flawed? Only if you think that third parties and independents should nor run, and there should be only two parties, then Plurality is fine. On 4.6.2012, at 13.49, James Gilmour wrote: These contributions to this discussion take an extremely narrow view of representation of voters as it is clear this discussion is not about a single-winner election (state governor, city mayor) but about electing representative assemblies like state legislatures and city councils. There can be major problems of representation if such representative assemblies are elected by FPTP from single-member districts even when there are only two parties. Even when the electorates of the single-member districts are as near equal as possible and even when the turnouts are near equal, FPTP in single-member districts can deliver highly unrepresentative results if the support for the two parties is concentrated in particular districts - as it is in most electorates. Thus party A that wins 51 of the 100 seats in the assembly may win those seats by small margins (say 550 votes to 450 votes) but party B that looses the election with 49 of the 100 seats may have won its seats by overwhelming margins (say 700 votes to 300 votes). Thus the 51 A to 49 B result is highly unrepresentative of those who actually voted: 42,750 for party A but 57,250 for party B. These numbers are deliberately exaggerated to show the point, but here in the UK we see this effect in every UK General Election since 1945, where the win small, loose big effect of FPTP has consistently benefited the Labour Party at the expense of the Conservative Party. And where such vote concentration exists - at is does everywhere - the result can be greatly influenced by where the boundaries of the single-member districts are drawn. Move the boundary, change the result. These are fundamental defects of FPTP in single-member districts that must be addressed if you want your elected assemblies to be properly representative of those who vote. Both of these defects has greater effect if the electorates are not so equal or if the turnouts vary with party support (as they certainly do in the UK). So even when there are only two parties, FPTP is very far from fine. Juho Sent: Monday, June 04, 2012 4:06 PM Yes, that makes sense. FPTP in single-member districts is not fair in the sense that it gives only approximate results (due to the inaccuracy of counting the seats in single-member districts) and is vulnerable to gerrymandering. It is not a question of not fair (which can be a highly subjective assessment), it is simply that the result is not properly representative. And the distortion is not due to inaccuracy - the defect is inherent in the system as it is based on single-member districts. And it is a defect, given the purpose of the election - to elect a representative assembly.. Such a system is vulnerable to gerrymandering, i.e. to the DELIBERATE manipulation of the district boundaries. But the real point is that these boundary effects occur even when there is no gerrymandering, i.e. no deliberate manipulation. A system that counts the proportions at national level (typically a multi-party system) would be more accurate. Also gerrymandering can be avoided this way. Yes, the votes could be summed at national level and the seats allocated at national level. But you do not need to go to national level to achieve proper representation. Where the electors also want some guarantee of local representation, a satisfactory compromise can be achieved with a much more modest 'district magnitude' than one national district. The discussed proposal of changing the Plurality/FPTP method of the single-member districts to some other single-winner method has still many of the discussed problems (inaccuracy, gerrymandering). One could try to construct also two-party systems or single-member district based systems that would avoid those problems. But maybe proportional representation is a more likely ideal end result. Practical reforms may however start with whatever achievable steps. All single-member district voting systems will have similar defects. But remember my comments were made in direct response to the statements quoted at the top: (more or less) If there are only two parties, FPTP is fine.I think the problem with what may be regarded achievable steps is that many contributors to this list start in the wrong place. Elections are for electors - so where the objective is to elect a 'representative assembly' (state legislature, city council), the first requirement should be that the voting system delivers an assembly that it is properly representative - all else is secondary. James - No virus found in this message. Checked by AVG -
Re: [EM] My summary of the recent discussion
On 4.6.2012, at 19.18, James Gilmour wrote: A system that counts the proportions at national level (typically a multi-party system) would be more accurate. Also gerrymandering can be avoided this way. Yes, the votes could be summed at national level and the seats allocated at national level. But you do not need to go to national level to achieve proper representation. Where the electors also want some guarantee of local representation, a satisfactory compromise can be achieved with a much more modest 'district magnitude' than one national district. In Finland there was a reform proposal that counted the proportions at national level, but the seats were still allocated in the existing districts. (Current government doesn't want to drive that proposal forward.) One can do this trick also with quite small districts. In Finland the size of the smallest districts is 6, but even smaller districts could work. Both targets can thus be met simultaneously, accurate proportionality and local(ish) representation. All systems will however have some rounding errors. In this proposal the seats of the parties are allocated to the districts so that the total sum of seats per district and seats per party are exactly correct, which means that some of the last seats have to be forced to go right, and this may violate the personal interests of some candidates (some other party may get the seat with fewer votes), but in a rather random / unpredictable / unbiased way that people are likely to accept. Political proportionality in the districts is also not as accurate as at the national level, but I guess the national level proportionality is the one that counts. In theory one could use this system also with single-member disticts, but the forcing operations would already be quite violent. If current single-member district countries want to keep the idea of very local representation, one approach could be to use only slightly larger districts tahn today (maybe 3, 4), calculate proportionality at national level, and then allocate the seats to the districts using some similar algorithm as in the Finnish reform proposal. Just an idea, to keep as much of the familiar and maybe liked feaures of the existing system. Juho Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Another reason why Greens won't vote Dem, due to previous count results.
I say again, the academic argument does not meet the real-world. My vote is not going to be influenced by these arguments, and since I'm the only voter in my district likely to read them, they are not likely to match real-world voter experience. NOBODY's expectation is really a sum over anything. That's just a an analytical tool to try to attach a number to how we form our votes. On EM, it's been shown by at least three people, in at least two ways, that the expectation-maximizing strategy of Approval is to approve the above-expectation candidates. It has been shown in my district that nobody except me reads the EM list. So most vote without ever having encountered a phrase such as the expectation-maximizing strategy. Americans in general aren't interested in any message that requires more than an eighth-grade education. No offense, but until the part of the EM community that is advocating something dis-entangles their advocacy from the study of EMs in general, advocates for one EM or another are, as we say in the American South, pissing into the wind. -Original Message- From: election-methods-boun...@lists.electorama.com [mailto:election-methods-boun...@lists.electorama.com] On Behalf Of Michael Ossipoff Sent: Thursday, May 31, 2012 9:39 PM To: election-meth...@electorama.com Subject: [EM] Another reason why Greens won't vote Dem, due to previous count results. As I was saying in a recent previous post about this, Approval's count results will tell Green-preferrers whether or not they need Dem to protect against Repub. And I gave a reason why that is: Preferrers of the middle of 3 parties have no reason to approve either extreme. I told of a reason why that is. Now I'd like to tell of another: On EM, it's been shown by at least three people, in at least two ways, that the expectation-maximizing strategy of Approval is to approve the above-expectation candidates. It's obvious why that's so: Your expectation is the sum, over all of the candidates, of the product of a candidate's win-probability and hir utility. It's obvious that when you increase the win-probability of a candidate who is better than your expectation (you do that when you approve hir), that will raise your expectation. Well, suppose you're a Democrat-preferrer (if there really are any). Say it's Green, Dem, Repub. If it's certain that some particular candidate will win, then your expectation is the utility of that candidate. Otherwise your expectation is somewhere between the utility of the Green and the Dem, or somewhere between the utility of the Repub and the Dem. Say it's somewhere between the Green and the Dem. As I said above, your best expectation-maximizing strategy is to approve (only) all of the above-expectation candidates. By assumption, the Green is farther from you than is the point representing the utility equal to your expectation. So you don't approve the Green. What would it take to make your expectation worse than the Green? No, even if it were almost certain that the Repub would win, that wouldn't do it, because Dem and Repub are so close that you couldn't squeeze an amoeba between them. The expected utility for you would have to be a candidate farther away from you than the Green and the Repub. And that would be impossible with just 3 candidates. So then, what if there were more candidates? Maybe there's a candidate who, as seen by you (a Dem-preferrer) is beyond the Green, in the same direction. Maybe the it looks as if someone that far away from you, in that direction, will win. Well, if that's so, then we can forget all about the Repub as a threat, can't we. In that case, the Green preferrers certainly have no reason to approve Dem. So, if there are any Democrat-preferrers, they aren't going to vote for the Green, except under conditions that would make the Rep is so unwinnable that the Dems aren't needed as a compromise. And with the Dems not approving in either direction, the count totals of Greens and Repubs will be a good estimate of their preferrers' numbers. And when the Green shows as bigger than the Repub (as s/he immediately will) it will be obvious that Green can beat Repub, and that Green preferrers don't need to approve Dem. As I was saying, in fact, I suggest that, in the 1st Approval election, the Repubs count-total will be so low the suggestion of Dem as a necessary compromise would be quite out of the question. Aside from all this, remember that, when non-Republocrats are seen as viable, there will be statisticians and (honest) poll-takers who are very interested in finding out about the relative numbers of preferrers of the various parties. Mike Ossipoff Election-Methods mailing list - see http://electorama.com/em for list info Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] My summary of the recent discussion
On 5.6.2012, at 1.52, James Gilmour wrote: On 4.6.2012, at 19.18, James Gilmour wrote: A system that counts the proportions at national level (typically a multi-party system) would be more accurate. Also gerrymandering can be avoided this way. Yes, the votes could be summed at national level and the seats allocated at national level. But you do not need to go to national level to achieve proper representation. Where the electors also want some guarantee of local representation, a satisfactory compromise can be achieved with a much more modest 'district magnitude' than one national district. Juho Sent: Monday, June 04, 2012 7:48 PM In Finland there was a reform proposal that counted the proportions at national level, but the seats were still allocated in the existing districts. (Current government doesn't want to drive that proposal forward.) One can do this trick also with quite small districts. In Finland the size of the smallest districts is 6, but even smaller districts could work. Both targets can thus be met simultaneously, accurate proportionality and local(ish) representation. All systems will however have some rounding errors. In this proposal the seats of the parties are allocated to the districts so that the total sum of seats per district and seats per party are exactly correct, which means that some of the last seats have to be forced to go right, and this may violate the personal interests of some candidates (some other party may get the seat with fewer votes), but in a rather random / unpredictable / unbiased way that people are likely to accept. Political proportionality in the districts is also not as accurate as at the national level, but I guess the national level proportionality is the one that counts. In theory one could use this system also with single-member districts, but the forcing operations would already be quite violent. If current single-member district countries want to keep the idea of very local representation, one approach could be to use only slightly larger districts than today (maybe 3, 4), calculate proportionality at national level, and then allocate the seats to the districts using some similar algorithm as in the Finnish reform proposal. Just an idea, to keep as much of the familiar and maybe liked features of the existing system. Iceland currently uses a system that sounds very like the Finish proposal. Votes are tallied at national level and in six constituencies, each of which has nine constituency seats in parliament. Nine additional equalization seats are distributed to constituencies and allocated to political parties so that the parliamentary representation of each party and each constituency will reflect as closely as possible the total votes received. This is done by solving a pair of simultaneous equations! It does have the effect you describe, forcing out some constituency winners and replacing them with equalisation candidates. This seems to be accepted because the constitution demands that every vote shall have equal value. In the Finnish proposal there are no equalization seats. There are only the regional seats that have been allocated to the districts in proportion to their population. But of course, you don't need to do it this way, nor does the proportionality have to be just party PR. With STV-PR in multi-member districts the voters have the power to choose the winning candidates on whatever PR basis matters to those voters. I do appreciate that STV is totally unacceptable to quite a number of the more vociferous members of this list, but STV-PR does address effectively many of the issues that arise in electing properly representative assemblies. party PR - can achieve accurate proportionality between parties - does not (usually) support accurate proportionality between different sections of a party - allows high number of candidates and seats per district - some methods allow votes to parties only, some allow votes to individual candidates STV-PR - can achieve accurate proportionality between parties - supports proportionality between different sections of a party - allows only smallish number of candidates and seats per district - allows voters to cast mixed votes that list candidates from more than one party - works also in (non-political) elections with no party structure Juho James - No virus found in this message. Checked by AVG - www.avg.com Version: 2012.0.2178 / Virus Database: 2425/5044 - Release Date: 06/04/12 Election-Methods mailing list - see http://electorama.com/em for list info
Re: [EM] Gerrymandering solutions.
About gerrymanmdering; PR would be a solution to gerrymandering, but certainly not the only one: 1. Proxy Direct Democracy wouldn't have a gerrymandering problem either. If Proxy DD can be made count-fraud-secure, then it would make PR obsolete. 2. Whatever can be accomplished by PR can be accomplished by an at-large single winner election, because every single winner method can output a ranking of candidates instead of just one winner: Elect the winner. Then delete the winner from the ballots and count them again. That will elect the rank 2 winner. Then eliminate the rank 2 winner too, and count the ballots again. Each time, delete every previous winner before counting to determine the next winner. So you can elect N winners at large in a state, or nationally, for a body such as Congress (or its separate houses, if you want to keep them) or a state legislature. Of course, with Approval, it only requires one count, and you elect the N candidates with the most approvals. 3. But districting needn't have a gerrymandering problem, even if single-member districts are kept. Who said that districts have to be arbitrary and freehand-drawn?? Where did we get that silly assumption? Draw the district lines by some simple rule that doesn't leave any human discretion or choice. It would be completely automated, but it would be so simple that it would be very easy for anyone to check. For example: You could divide the country (or state) into N1 latitudinal bands such that each has the same population/average longitudinal width. Then divide each latitudinal band into longitudinal sections, in such a way as to give each section the same population, and so that there are the right number of such sections overall. But of course you wouldn't have to use latitude and longitude if you don't want to. On a map, on any projection, that you choose, use a rectangular grid of lines, drawn similarly to the way described above. If you use a gnomonic projection, then all of your district lines will be straight lines on the Earth (great circles). If you use a cylindrical projection, then it will be as described in the previous paragraph. But it could be any projection you like. I'd suggest that gnomonic and cylindrical (using parallels and meridians as described in the previous paragraph) would be the main two choices. Districts divided by parallels and meridians, or by straight lines (great circles). The point is that it could be done by a simple rule that would have no human input, no human choice. What if it divides a county or a city? So what? No problem. The rule could be that houses would be all counted on whichever side of a line most of the house's area lies. It could be automated of course, but the result could easily be checked by anyone. To change the subject a little, I'd like to bring up another geographical government suggestion, while I'm at it: Partition. It doesn't make any sense for people to have to live under a government that they don't like, with people whom they don't agree with or don't like. So why not just divide the country up into separate countries, according to what kind of government people like? It's ridiculous to make everyone share the same county, when they want different kinds of country. It would be like a PR election, except that it would be for square miles instead of for seats. Though, like districting, the partitioning of the country could be (1) by an automatic rule, with those same rectangles (I like that), or (2) it could also be done by national negotiation in a PR negotiating body, or maybe by a proxy DD negotiation. I like the quick simplicity of (1). But (2) could _maybe_ be done in such a way as to ensure that each new partition-country has, to the best extent possible by negotiation, equally good land, by whatever standards its people want to bargain for. Of course an overall census could be taken periodically, and the process repeated, to take into account people who have voted with their feet. But those adjustments wouldn't be necessary, because the initial partition would let everyone live in the govt they like best. But, though I don't watch tv, I used to watch it along with the family I was part of, so what about a family like the one in All in the Family? Should Archie Bunker's daughter have to remain in his country? Likewise the family in A family Affair (if I've got the show-name right). It is not your fault what country you're born in. So there would be a strong case for letting people continue to choose what country they want to live in, even after partition. Of course then it would be necessary to repeat the initial partition process to adjust the national borders to the new populations. Maybe migration should only affect borders when it's by people who were born in the country that they're in, as opposed to people who chose that country at partition time. But migration must be distinguished from fecundity, for this purpose. A country shouldn't be able to