Worth fightin' for?
What pulls me into politics at all is a sort of intellectual challenge - to look at something that you can take sides on and break it a part and see what it is made of. That's the way of the intellect.
Now with that admission, you may run away, at least if you at all imagined I had any noble motive towards furthering human spirit. No, it's just a game.
Oh, you didn't imagine I had noble motive? Well, I imagined that at least, but I accept that probably there is no nobility of the intellect. It is a cold thing, feed by its own power, and willing to devalue anything.
I titled this entry "Worth fightin' for?" recognizing whatever concerns or positions I might take in a political debate, I'm more interested in expression than follow through. That's again the intellect's limits - when the debate is won, it is forgotten, and whatever concerns or injustices it recognized can be fixed.
Well, just thinking about how easily I can drop a subject. On Friday I wrote a 1000 word opinion piece on my concerns of IRV in Minneapolis, and now for two days I've thought nothing of it. It is a sort of annoying fly I can swat at, but really I don't even care. Do I really want to put effort into defending a point of view?
Even as I ask, I see I'd open the door again and swat at it. I was annoyed last effort because my first focus was distracted by a more important one, and I didn't want to focus. Now I wonder if I ought to refocus on the new idea and attack full power there, while previously only a half hearted effort.
Specifically, I'm quite sure Minneapolis ought NOT to eliminate the primary round, even if IRV (and STV) can logically handle any number of candidates.
I can see why they want to eliminate it - because if it was partisan elections, there'd be no "general primary" to help, so why pretend it needs one?
I have my original issue of a top-two runoff as being most fair (to the plurality strength of candidates) in a single seat election, but I see a primary is much more valuable. Minneapolis already has a top-two primary, so majority rule is assured. So the issue is how to get more choices to the general election, and I see that a primary system that USUALLY has two candidates, sometimes three, and occassionally four would make for a much more orderly IRV process anyway.
So here I am, once again, imagining myself heroic, defending the primary as a "Right of passage" for candidates to be worthy of the general election.
It's a stronger issue than my little top-two mess, and seemingly neglected. I imagine I could write a kick-ass defense of the primary, and it would be great fun to imagine expanding the rules of the primary beyond top-two. I've already started on my Friday attempt, but I avoided the multiwinner elections.
Okay, let me try my worst here.
Supposition: Letting any candidate with a beating heart into a general election is foolhearty.
Why? Now I'm in trouble. We're all for "more choices", right?!
Well, said the mousy little voice of reason, I'm a rather busy mouse, springing mousetraps for my supper to feed my little ones every day, and I don't have time to research 52 candidates to make an informed choice.
Okay, you convinced me, but hold on, what sort of hoops do we want? Partisan elections for independent candidates require X signatures to gain access to the ballot. I expect city elections have no requirement. The primary itself is the proving grounds.
Right, but what conditions? For instance, like signatures, you could say a fixed number of votes ought to be considered, like 1000 votes for instance. In a city of say 200,000 voters(?), expecting to get 1000 people to vote for you in a primary ought to be reasonable, even if voter turn out is 10%.
A more general requirement would be like 5% of the primary vote. That's not asking for very much, and is low enough to guarantee at least two will pass, not mathematically certain, but the reality that highly known candidates will tend to grab the most votes.
Secondly there's a question on how people should vote in a primary. I lean towards a plurality result - everyone has one vote. IRV supporters might suggest IRV, but I'd say no. Heck some will say an "approval vote" is good, and might be, except it allows too many games to be played. No voting in a wide field is hard enough. Pick a candidate and get to 5%. There'll still be game playing - supporters of strong candidate have a freedom to "vote down" if they like, and with a low enough bar, it's a safe bet.
Okay, think of a nonpartisan primary as "parties of one". You have to pick sides. Rank elimination is fine when there's a majority requirement, but we're talking 5%. And if two candidates get 4%, they can cry spoiler if they like, but come on, why didn't they work together if they have a lovefest?
Here's an example election with many candidates:
http://en.wikipedia.org/wiki/French_presidential_election,_2002
7 candidates over 5%, and three over 15%. They had a top-two runoff, and the second round was miserable, with nearly 100% of votes going against the runner up.
Supporters will say IRV could have cut through this disaster, and I admit it could have, even with limited rankings.
Limited rankings actually represents the primary limitation of IRV in practice. Basically a fair limitation is to say you shouldn't have more candidates than one more than the rankings. So a 3-ranking ballot ought not to have more than 4 candidates.
Although it can equally be argued that even in a 4-seat election, the top 1-3 preferences will likely hold the majority the vote, so it's not easy to defend limited rounds merely on ranks available.
Is it REALLY a necessary issue? It could be a primary is not needed for technical limitations, but entirely on arguments of limitations of voter attention.
I basically accept AT LEAST twice as many candidates can exist in an STV election as winners, although that makes single-winner IRV trivial! Okay allow up to 2*seats+1?
A second issue is over what sort of primary should exist for multiseat elections? I mean for proportional representation like STV? A majority choice isn't acceptable. I'd lean towards an "Equal&Even Cumulative vote" since it is semi-PR. It allows you to split a single vote as widely as desired. Surplus transfers ought not to be valuable in a primary because we'd like more voters to choose the set, while surpluses come from voters who already got a winner in.
The simplest semi-PR method is one-vote! I only reject it because it discounts voters who wish to compromise, and creates false strategy (which candidate needs my vote more?)
So I basically come out for regular CV, N-votes for N-seats, distributable. I like E&E better because it is simpler to vote.
Anyway, whatever "single vote" method is used in the primary, what standard should pick the winners? I'd accept a top-2N argument, and all above a 1/5th quota. A low threshold is acceptable in exchange for no surplus transfer. (10% for 1 winner, 6.7% for 2 winners, 5% for 3 winners, 4% for 4 winners, etc.)
It seems "complicated" to try to defend any given rule. The only defense at all is the "old rule" was top-2 or top-2N for multiseat elections. So I'd follow those PLUS a vote threshold to include more that that.
I'm too much in love with E&E-CV as being a simple as "plurality-at-large", an approval vote. It could be limited marks by winners, probably reasonable. Certainly never perfectly safe to vote more widely, but I would claim WORSE compromises occur otherwise. In a one-seat election primary, if I like A and B, and A looks strong, I'll vote for B, and if many like me "mis-estimate" A, then they may lose both. The value of splitting a vote on both is it gives the best results if all supporters did that - best at least if there's a chance both can make the threshold.
Well, overall I'm stuck, unable to promote CV without a fight, not in any form, EXCEPT single-vote. Defending single-vote means unacceptable strategy issues.
The obvious solution is to apply STV fully in a rank primary, and that seems impossible, since the whole point is to make a threshold for starting a rank vote. And what quota do you set in a STV primary? You could say quota/2. So that's 25% for a single winner election, and 16% for a 2 winner election, 12.5% for a 3 winner election.
In practice, even with surplus transfers, the quotas won't be easily met with many "bullet votes". That's harmless (surplus transfers surrendered basically.)
Okay look at my favorite month data, have 2 winners, and pretend 5 for primary. Setting a 16.67% quota, it comes up with Aug, Jul, May, Sep, Apr, and loser Oct at the bottom.
Then in a 5-candidate general, with 2-winners (33.3% quota), September and May come out ahead 32.86% with July a razor close third at 31.79%. But that election had well ranked ballots. (Only 2.1 votes lost outside these 3)
Anyway, I can see a clear value of a primary on the grounds that voters will be more likely to rank them.
So my simple suggestion is take 2*seats+1 candidates from the primary. That's PERFECT for keeping a single winner election down to a two round system! (Two birds with one stone!)
Yes, keep a single-vote primary, and take the top 2*seat+1 candidates, plus further thresholds for including more. I'd almost just say 2*seats, unless a third is "worthy", but I can go either way.
On defining worthiness, I always accept a fixed percent threshold (for a given number of seats). What that threshold is, is debatable, anywhere from 1/(seats+1) to 1/(2*seats+2) basically.
The purpose of the threshold is to help candidates recognize if they are viable or not. It is easier to know if you may be close to 10% than to know if you're in the top-5. Maybe not, but a fixed threshold holds power for candidates to shoot for, a fair line to accept failure. Better to barely miss 10% than barely miss 4th placing.
Still the system does reek of strategy and failure. Only a low threshold evens the field enough to accept "wasted" votes on top. An E&E-CV is so far superior in result, I'm sad I can't propose it. Strategy is muted by willing compromise.
I just don't think I can advocate a single-vote system for multiseat election that disallows a split vote!
I think I may yet have to abandon my fixed threshold. The problem with that is that it encourages "strategic downvoting" for supporters of candidates far above the threshold.
No, I think a pure ranking is the best we can do in a one vote system, and depend on statistics to average out the strategy. Fine, without split votes, I abandon fixed thresholds.
Top-(2N+1) primary has merit. It is simple, gives choice, and high standards.
On my month election? 2 winners? Top 5? Sep=19%-May=16%-Aug=12%-Jul=10%-Apr=9%. That's 65% of the vote. Looks good!
I'm convinced A top-(2*N+1) primary in a one-vote system is a good compromise.
Even if we had a split vote CV system, it's not so bad to just take a fixed set.
Can I can ANY primary data for Minneapolis? .
http://www.ourcampaigns.com/RaceDetail.html?RaceID=154964
2005 - [Rybak 44.48%, LcLaughlin 35.34%], Hakeem 13.84%, Kscielski 1.76%, Johnson 1.27%, Franson 0.81%
There was ONE candidate below third potentially worthy.
And % data from 1993: http://mapnp.geeks.org/pipermail/mpls/2001-July/004474.html
Belton=28%, Derus=20%, Cramer=16%, Rapson=16%, Jefferson=7%, Wodele=7%, Other=6%
That's a wild case for a virtue tie for third. Keeping both could be justified.
Just looking at the potential of IRV, ANY of the top SIX candidates could theoretically rise in a slow runoff via transfer votes to win.
With names, in fact, I found a link above, with an analysis, reasonable or not:
>Eighteen candidates ran for Mayor in 1993, and the primary results
>looked like this:
>Belton 28% Cramer 16% Derus 20%
> Rapson 16%
> Wodele 7%
> Jefferson 7%
> Belton and Derus, the most left-liberal and most right-conservative,
>went on to the general election. Problem was, the moderate 47% of the
>electorate split their moderate votes four ways, and no one of the four
>"moderate" candidates got into the finals.
Good enough for me. Can I draft a quick opinion piece here?
*******************
I congratulate Minneapolis voters for their progressive support of Instant Runoff Voting for their city elections. The breath of support was amazing and Minneapolis will be able to lead the state in the demonstration of how elections can be done better.
However I must offer one point of concern, one which can easily be addressed before implementation of new elections in 2009. I believe the call to remove the nonpartisan primary as a mistake. As far as I can tell there were three reasons for the step: (1) To save money (2) Low voter turnout (3) To increase choices in the general election.
I must strongly disagree with the first two reasons - politics is a participation sport, and requires more than just the time of voting, but learning about the candidates. In short, I expect the minority of voters who go to the primary are exactly the voters who will have spent the time to make informed choices among the candidates. The financial cost of the primary is a small affair compared to the costs involved to voters in their time and energy in making an informed choice in the primary.
Now on the third point I fully accept, but with IRV we can do the primary better - allowing more candidates through. How many? My conservative position would be keep the top-three candidates for single seat elections, and in general twice the seats plus one for multiseat elections.
Is that too narrow? Perhaps, but I'm sure having no primary is much too wide. I am not doubting that IRV and rank preference ballots can't handle more candidates. My concern is first to candidates and second to voters. I believe candidates ought to feel free to run in the primary, but ought to see inclusion in the general election as something that must be earned. The currency of political capital is votes, and the primary ought to be seen as the date of destiny for candidates to measure their base of support.
Secondly it is a mark of respect to voters to give them a legimate list of candidates to choose between. You can't expect to give voters a phone book and make them pick names. Sure voters ought to be informed, but we can make it a little easier on them, and make sure there are standards for candidates.
Sure, popular names like "Sharon Anderson" might still get votes to pass the primary, but if the set of candidates is small enough by the general election, they'll be properly handled through the media for who they are beyond just a nice name.
Despite romantic ideas of democracy, voting is a hard process. It will continue to be a process filled with half-knowledge and intuition despite all our hopes for fact and reason to lead us.
I hope that Minneapolis reconciders the elimination of the primary. I encourage all thoughtful voters to also support this. Choice is good, but knowledge is better, and keeping the a primary round is too important to let go without a fight.
Now with that admission, you may run away, at least if you at all imagined I had any noble motive towards furthering human spirit. No, it's just a game.
Oh, you didn't imagine I had noble motive? Well, I imagined that at least, but I accept that probably there is no nobility of the intellect. It is a cold thing, feed by its own power, and willing to devalue anything.
I titled this entry "Worth fightin' for?" recognizing whatever concerns or positions I might take in a political debate, I'm more interested in expression than follow through. That's again the intellect's limits - when the debate is won, it is forgotten, and whatever concerns or injustices it recognized can be fixed.
Well, just thinking about how easily I can drop a subject. On Friday I wrote a 1000 word opinion piece on my concerns of IRV in Minneapolis, and now for two days I've thought nothing of it. It is a sort of annoying fly I can swat at, but really I don't even care. Do I really want to put effort into defending a point of view?
Even as I ask, I see I'd open the door again and swat at it. I was annoyed last effort because my first focus was distracted by a more important one, and I didn't want to focus. Now I wonder if I ought to refocus on the new idea and attack full power there, while previously only a half hearted effort.
Specifically, I'm quite sure Minneapolis ought NOT to eliminate the primary round, even if IRV (and STV) can logically handle any number of candidates.
I can see why they want to eliminate it - because if it was partisan elections, there'd be no "general primary" to help, so why pretend it needs one?
I have my original issue of a top-two runoff as being most fair (to the plurality strength of candidates) in a single seat election, but I see a primary is much more valuable. Minneapolis already has a top-two primary, so majority rule is assured. So the issue is how to get more choices to the general election, and I see that a primary system that USUALLY has two candidates, sometimes three, and occassionally four would make for a much more orderly IRV process anyway.
So here I am, once again, imagining myself heroic, defending the primary as a "Right of passage" for candidates to be worthy of the general election.
It's a stronger issue than my little top-two mess, and seemingly neglected. I imagine I could write a kick-ass defense of the primary, and it would be great fun to imagine expanding the rules of the primary beyond top-two. I've already started on my Friday attempt, but I avoided the multiwinner elections.
Okay, let me try my worst here.
Supposition: Letting any candidate with a beating heart into a general election is foolhearty.
Why? Now I'm in trouble. We're all for "more choices", right?!
Well, said the mousy little voice of reason, I'm a rather busy mouse, springing mousetraps for my supper to feed my little ones every day, and I don't have time to research 52 candidates to make an informed choice.
Okay, you convinced me, but hold on, what sort of hoops do we want? Partisan elections for independent candidates require X signatures to gain access to the ballot. I expect city elections have no requirement. The primary itself is the proving grounds.
Right, but what conditions? For instance, like signatures, you could say a fixed number of votes ought to be considered, like 1000 votes for instance. In a city of say 200,000 voters(?), expecting to get 1000 people to vote for you in a primary ought to be reasonable, even if voter turn out is 10%.
A more general requirement would be like 5% of the primary vote. That's not asking for very much, and is low enough to guarantee at least two will pass, not mathematically certain, but the reality that highly known candidates will tend to grab the most votes.
Secondly there's a question on how people should vote in a primary. I lean towards a plurality result - everyone has one vote. IRV supporters might suggest IRV, but I'd say no. Heck some will say an "approval vote" is good, and might be, except it allows too many games to be played. No voting in a wide field is hard enough. Pick a candidate and get to 5%. There'll still be game playing - supporters of strong candidate have a freedom to "vote down" if they like, and with a low enough bar, it's a safe bet.
Okay, think of a nonpartisan primary as "parties of one". You have to pick sides. Rank elimination is fine when there's a majority requirement, but we're talking 5%. And if two candidates get 4%, they can cry spoiler if they like, but come on, why didn't they work together if they have a lovefest?
Here's an example election with many candidates:
http://en.wikipedia.org/wiki/French_presidential_election,_2002
7 candidates over 5%, and three over 15%. They had a top-two runoff, and the second round was miserable, with nearly 100% of votes going against the runner up.
Supporters will say IRV could have cut through this disaster, and I admit it could have, even with limited rankings.
Limited rankings actually represents the primary limitation of IRV in practice. Basically a fair limitation is to say you shouldn't have more candidates than one more than the rankings. So a 3-ranking ballot ought not to have more than 4 candidates.
Although it can equally be argued that even in a 4-seat election, the top 1-3 preferences will likely hold the majority the vote, so it's not easy to defend limited rounds merely on ranks available.
Is it REALLY a necessary issue? It could be a primary is not needed for technical limitations, but entirely on arguments of limitations of voter attention.
I basically accept AT LEAST twice as many candidates can exist in an STV election as winners, although that makes single-winner IRV trivial! Okay allow up to 2*seats+1?
A second issue is over what sort of primary should exist for multiseat elections? I mean for proportional representation like STV? A majority choice isn't acceptable. I'd lean towards an "Equal&Even Cumulative vote" since it is semi-PR. It allows you to split a single vote as widely as desired. Surplus transfers ought not to be valuable in a primary because we'd like more voters to choose the set, while surpluses come from voters who already got a winner in.
The simplest semi-PR method is one-vote! I only reject it because it discounts voters who wish to compromise, and creates false strategy (which candidate needs my vote more?)
So I basically come out for regular CV, N-votes for N-seats, distributable. I like E&E better because it is simpler to vote.
Anyway, whatever "single vote" method is used in the primary, what standard should pick the winners? I'd accept a top-2N argument, and all above a 1/5th quota. A low threshold is acceptable in exchange for no surplus transfer. (10% for 1 winner, 6.7% for 2 winners, 5% for 3 winners, 4% for 4 winners, etc.)
It seems "complicated" to try to defend any given rule. The only defense at all is the "old rule" was top-2 or top-2N for multiseat elections. So I'd follow those PLUS a vote threshold to include more that that.
I'm too much in love with E&E-CV as being a simple as "plurality-at-large", an approval vote. It could be limited marks by winners, probably reasonable. Certainly never perfectly safe to vote more widely, but I would claim WORSE compromises occur otherwise. In a one-seat election primary, if I like A and B, and A looks strong, I'll vote for B, and if many like me "mis-estimate" A, then they may lose both. The value of splitting a vote on both is it gives the best results if all supporters did that - best at least if there's a chance both can make the threshold.
Well, overall I'm stuck, unable to promote CV without a fight, not in any form, EXCEPT single-vote. Defending single-vote means unacceptable strategy issues.
The obvious solution is to apply STV fully in a rank primary, and that seems impossible, since the whole point is to make a threshold for starting a rank vote. And what quota do you set in a STV primary? You could say quota/2. So that's 25% for a single winner election, and 16% for a 2 winner election, 12.5% for a 3 winner election.
In practice, even with surplus transfers, the quotas won't be easily met with many "bullet votes". That's harmless (surplus transfers surrendered basically.)
Okay look at my favorite month data, have 2 winners, and pretend 5 for primary. Setting a 16.67% quota, it comes up with Aug, Jul, May, Sep, Apr, and loser Oct at the bottom.
Then in a 5-candidate general, with 2-winners (33.3% quota), September and May come out ahead 32.86% with July a razor close third at 31.79%. But that election had well ranked ballots. (Only 2.1 votes lost outside these 3)
Anyway, I can see a clear value of a primary on the grounds that voters will be more likely to rank them.
So my simple suggestion is take 2*seats+1 candidates from the primary. That's PERFECT for keeping a single winner election down to a two round system! (Two birds with one stone!)
Yes, keep a single-vote primary, and take the top 2*seat+1 candidates, plus further thresholds for including more. I'd almost just say 2*seats, unless a third is "worthy", but I can go either way.
On defining worthiness, I always accept a fixed percent threshold (for a given number of seats). What that threshold is, is debatable, anywhere from 1/(seats+1) to 1/(2*seats+2) basically.
The purpose of the threshold is to help candidates recognize if they are viable or not. It is easier to know if you may be close to 10% than to know if you're in the top-5. Maybe not, but a fixed threshold holds power for candidates to shoot for, a fair line to accept failure. Better to barely miss 10% than barely miss 4th placing.
Still the system does reek of strategy and failure. Only a low threshold evens the field enough to accept "wasted" votes on top. An E&E-CV is so far superior in result, I'm sad I can't propose it. Strategy is muted by willing compromise.
I just don't think I can advocate a single-vote system for multiseat election that disallows a split vote!
I think I may yet have to abandon my fixed threshold. The problem with that is that it encourages "strategic downvoting" for supporters of candidates far above the threshold.
No, I think a pure ranking is the best we can do in a one vote system, and depend on statistics to average out the strategy. Fine, without split votes, I abandon fixed thresholds.
Top-(2N+1) primary has merit. It is simple, gives choice, and high standards.
On my month election? 2 winners? Top 5? Sep=19%-May=16%-Aug=12%-Jul=10%-Apr=9%. That's 65% of the vote. Looks good!
I'm convinced A top-(2*N+1) primary in a one-vote system is a good compromise.
Even if we had a split vote CV system, it's not so bad to just take a fixed set.
Can I can ANY primary data for Minneapolis? .
http://www.ourcampaigns.com/RaceDetail.html?RaceID=154964
2005 - [Rybak 44.48%, LcLaughlin 35.34%], Hakeem 13.84%, Kscielski 1.76%, Johnson 1.27%, Franson 0.81%
There was ONE candidate below third potentially worthy.
And % data from 1993: http://mapnp.geeks.org/pipermail/mpls/2001-July/004474.html
Belton=28%, Derus=20%, Cramer=16%, Rapson=16%, Jefferson=7%, Wodele=7%, Other=6%
That's a wild case for a virtue tie for third. Keeping both could be justified.
Just looking at the potential of IRV, ANY of the top SIX candidates could theoretically rise in a slow runoff via transfer votes to win.
With names, in fact, I found a link above, with an analysis, reasonable or not:
>Eighteen candidates ran for Mayor in 1993, and the primary results
>looked like this:
>Belton 28% Cramer 16% Derus 20%
> Rapson 16%
> Wodele 7%
> Jefferson 7%
> Belton and Derus, the most left-liberal and most right-conservative,
>went on to the general election. Problem was, the moderate 47% of the
>electorate split their moderate votes four ways, and no one of the four
>"moderate" candidates got into the finals.
Good enough for me. Can I draft a quick opinion piece here?
*******************
I congratulate Minneapolis voters for their progressive support of Instant Runoff Voting for their city elections. The breath of support was amazing and Minneapolis will be able to lead the state in the demonstration of how elections can be done better.
However I must offer one point of concern, one which can easily be addressed before implementation of new elections in 2009. I believe the call to remove the nonpartisan primary as a mistake. As far as I can tell there were three reasons for the step: (1) To save money (2) Low voter turnout (3) To increase choices in the general election.
I must strongly disagree with the first two reasons - politics is a participation sport, and requires more than just the time of voting, but learning about the candidates. In short, I expect the minority of voters who go to the primary are exactly the voters who will have spent the time to make informed choices among the candidates. The financial cost of the primary is a small affair compared to the costs involved to voters in their time and energy in making an informed choice in the primary.
Now on the third point I fully accept, but with IRV we can do the primary better - allowing more candidates through. How many? My conservative position would be keep the top-three candidates for single seat elections, and in general twice the seats plus one for multiseat elections.
Is that too narrow? Perhaps, but I'm sure having no primary is much too wide. I am not doubting that IRV and rank preference ballots can't handle more candidates. My concern is first to candidates and second to voters. I believe candidates ought to feel free to run in the primary, but ought to see inclusion in the general election as something that must be earned. The currency of political capital is votes, and the primary ought to be seen as the date of destiny for candidates to measure their base of support.
Secondly it is a mark of respect to voters to give them a legimate list of candidates to choose between. You can't expect to give voters a phone book and make them pick names. Sure voters ought to be informed, but we can make it a little easier on them, and make sure there are standards for candidates.
Sure, popular names like "Sharon Anderson" might still get votes to pass the primary, but if the set of candidates is small enough by the general election, they'll be properly handled through the media for who they are beyond just a nice name.
Despite romantic ideas of democracy, voting is a hard process. It will continue to be a process filled with half-knowledge and intuition despite all our hopes for fact and reason to lead us.
I hope that Minneapolis reconciders the elimination of the primary. I encourage all thoughtful voters to also support this. Choice is good, but knowledge is better, and keeping the a primary round is too important to let go without a fight.
posted by Ares Olympus at 1:56 PM
0 Comments:
Post a Comment
<< Home