Open thread: Still hot edition

Photo Sharing and Video Hosting at Photobucket

What's on your alleged mind . . . besides the heat?




Comments

A new Puppet blog

Called The Wild West. As if our friends in the mountain don't have enough BS rolling downhill to put up with.

If you don't read the Puppet blogs every now and then, you should. The only one that has any traffic beyond paid staffers is the Mecklenburg edition, which mostly attracts government haters who are riled up in the anti-transit/anti-planning direction. I sure hope these people aren't a cross-section of Charlotte's population, because if they are, we should give the damn city to South Carolina and put ourselves out of their misery.

Vote Swapping

Well, vote swapping has been on my mind. Ballot reform generally, in fact. Take those one at a time:

1. Vote swapping. I had this little piece in the News and Observer. Not clear there will be a viable "third" candidate this prez election. But, any interest in a vote-swapping site if there were?

My impression is that at least some people at BNC would like to see a little more pressure on the Dems from the left. A viable Green party candidate, for example, might help that. But folks are afraid of wasting their vote on a sure losing bid, and (worse) helping a Repub win overall. (Nader got 97,000 votes in FL in 2000, Bush "won" by 573, Nader cost Gore the election?)

WIll NC be in play in 2008? Would you "swap" your vote, if not?

2. IRV: Instant run-off voting. If NC, or even some of our cities, used instant run-off voting, there would be Green candidates, and Libertarian candidates, and maybe some others, and people could vote their hearts without fear of causing the victory of candidates they hate.

And some good research shows that lowering entry barriers to "third" parties makes the existing state-sponsored parties more responsive. The Dems in NC would clearly move to the left on several issues if there were a Green on the ballot and we had and IRV system. It would not change Dem control of GA, I think. But it would move Dem policy left, a smidgeon or maybe three smidgeons.

"It is to secure our rights that we resort to government at all." --Thomas Jefferson to Francois D'Ivernois, 1795.

Michael C. Munger

I would not use vote swapping

simply because I am a cynical soul, and I do not trust that I would not be taken for a fool on that one. Instant runoff voting, on the other hand, sounds like a very good idea to me, because it would open up the field to candidates who are not part of "the big 2". Would it push the Democratic Party to the left? I have my doubts, but I'm not a professor of political science - just a Dem who would love to see a more open process.


Be the change you wish to see in the world. --Gandhi

IRV facts and fiction

lcloud:

Unfortunately, your ideas about IRV are part of the common IRV mythology. IRV results in two-party duopoly, just like our system now. For instance, Australia's house of representatives (which has used IRV since 1918) has just 1 third party, out of 564 seats - so it is 0.18% third party. The U.S. is 0.09% third party, if you look at all state and federal seats combined. A negligible difference, even though Aussie third parties are respectable in the AU senate, because it uses proportional representation. Australian political analysts at Australianpolitics.com say IRV "promotes a two-party system to the detriment of minor parties and independents."

Ireland has used IRV to elect their relatively unimportant (mostly ceremonial) Presidential post since it began in 1938. The Fianna Fail party has won it every term save for one flukey exception, for a near-monopoly - in spite of the fact that the Irish Dail (legislature) has more than two vital parties. IRV has produced similar results in Malta and Fiji.

Here's some mathematical explanation for why this happens. By contrast, most of the 27 countries that use a genuine runoff (like France had some months back, between Royal and Sarkozy) have escaped duopoly.

IRV can also exhibit horrendous pathologies. Social utility efficiency calculations by Princeton math Ph.D. Warren D. Smith, have shown IRV to be essentially the second worst of the various voting methods that have historically been seriously proposed for use in public elections (the very worst is, you guessed it, our present system of plurality voting).

The solution is a simpler and hugely better method called Range Voting. It's the very system used to rate these blog posts (the 1-5 stars at the top). In fact, the simplest form of Range Voting is Approval Voting, which is identical to our present system, except that you change the "vote for one" rule to "vote for one or more" (per race). This simple change requires no ballot redesign, nor any substantial election equipment upgrades. It behaves much better than IRV, especially when voters act strategically. It also has the nice property that it is always safe to vote for your favorite candidate (unlike with IRV, where it is strategically best to top-rank your favorite of the apparent front-runners).

Please read our open letter to the Californians for Electoral Reform, and the IRV community at large.

We believe a change to the voting method is, hands down, the most important issue in the world.

Clay Shentrup
San Francisco, CA
clay@electopia.org
415.240.1973

Well, that's my fault. STV

Well, that's my fault. STV systems are a bad example, as my good friend Gary Cox has shown in his book MAKING VOTES COUNT.

So, it's fair enough to call me on that. Icloud was expressing the sentiment that opening the system to competition and different ideas was important. MY example was the problem.

I'm not convinced that IRV/other STV systems are WORSE than FPP, but Clay is quite right that it's not clear they are better in terms of selecting a single outcome, and could be worse under some plausible circumstances.

That makes the problem hard.....

The range voting alternative seems fanciful, though it is interesting. I'm a professional political scientist, and so I think schemes like that are great, of course. But we can't even get the state-sponsored parties to allow third party access in the CURRENT system. Why would they choose something that opens up political power to the actual will of the people. I'm skeptical, not on normative grounds but on practical grounds.

"It is to secure our rights that we resort to government at all." --Thomas Jefferson to Francois D'Ivernois, 1795.

Michael C. Munger

It would seem to me that . . .

some kind of legal challenge should be mounted to make third-party participation easier. Have such efforts ever been attempted . . . that is, pursuing legal rather than legislative remedies?

The state-sponsored parties are indeed an entrenched duopoly that subverts freedom of choice.

Third parties must change the voting method

Unfortunately, such tactics are effectively irrelevant. As long as you have a voting method that leads voters with a strategic incentive to vote for their favorite of the front-runners, regardless of who their sincere favorite is, you get duopoly - so third parties are out of luck.

But IRV has the same problem. So savvy reformers have to support Range Voting, or else. It's a harsh reality.

The upside is, Range Voting is simpler than IRV, and increases the average voter's satisfaction by a great deal, and decreases the importance of cash (which IRV really doesn't do, as it's supporters tend to claim), and reduces spoiled ballots (whereas IRV increases them), and is more resistant to the problem of strategic voting.

One of the biggest points we try to get across to third parties is that they need to focus their efforts on changing the voting method, above all else. Ballot access and other issues play a role as well, but no where near the kind of role that the voting method plays.

Clay

Why are range voting and approval voting so ignored?

Clay -- No one is using range voting and approval voting in the world for any governmental election. No legislature in any city, county or state is taking it seriously. Why do you think that is?

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Argumentum ad populum?

Rob,

I'm honestly more concerned with why governments should or shouldn't use one particular voting method over another, than with speculating on why they historically have or haven't.

As I have explained in some detail, Range Voting is simpler than Instant Runoff Voting, and has numerous benefits. So perhaps you could tell us why you think governments haven't adopted Range Voting?

Do you think it could have something to do with the fact that it has only recently started to be advocated (by the Center for Range Voting) whereas FairVote decided long ago to pursue IRV, and beat us to the punch?

Also, do you think that, in light of the evidence of Range Voting's simplicity and superiority to IRV, FairVote might consider changing course and supporting Range Voting instead? Its simplified form, Approval Voting, would certainly be a cheaper and simpler system to adopt than IRV.

Rob Richie is a dedicated

Rob Richie is a dedicated political activist, single-minded and highly partisan, that is, he has his goals and he will say whatever it takes to convince people to support them.

Sorry, this is long, but the issues are complex and if we want more than punches in a prizefight, I suggest exploring them in some depth.

It took those of us in the election community some time to figure out why Richie promotes IRV over other methods widely known to be not only better, but simpler and cheaper to implement. Turns out he wrote the reason out quite some time back.

Richie and others with some money to spend for activism, having decided that what the U.S. and progressive causes needed was Proportional Representation, a strategy was mapped out to bring this about. The first step, it was determined, was to promote IRV. Why? Well, IRV is the single-winner version of STV, a multiwinner method used for proportional representation. STV is much more complicated, but the process, applied when only one office is being filled, is IRV. STV is quite a respectable method, though not necessarily the best. But single-winner is different. In STV, a centrist candidate who would lose in IRV will win a seat, and so certain problems with the elimination process are noot.

Election method theorists who consider IRV a good method are rare as hen's teeth. However, Richie's goal is not IRV. It is Proportional Representation. The theory was that if they could get the U.S. accustomed to the voting method, then it would be easier to promote IRV, because the much more complex counting methods needed for STV would already be in place.

Everything Richie writes about IRV and other methods must be understood in that context. It is goal-oriented political debate, calculated to make his immediate purpose look good, and alternatives look bad.

Now, as to the question. Richie will write one thing in one place and something else in another. As a skilled activist, he will rarely lie; he carefully crafts what he says so that it is literally true. And often quite misleading. Here he doesn't even make a statement, he asks a question, assuming that his audience will fill in an answer. EVen though he *should* know that the answer someone will fill in is either false or misleading. By thus allowing the *reader* to supply the false information or analysis, he bypasses normal critical process. He is highly skilled at this.

Is Approval Voting in use in "governmental elections" anywhere? Well, that depends on what you mean by an "election?" Richie will want you to *not* use the most general meaning of the word, he will be quite happy if you take it as meaning secret ballot polls of the public deciding among candidates for office.

But "election methods" can be and are used to make other kinds of choices. And for other kinds of controversial public choices, Approval Voting is in current use commonly in the United States.

The election of an officer is a special kind of Ballot Question. A Ballot Question is a motion or initiative placed before the public, for the public to vote Yes or No on. In the full deliberative process of peer assemblies, all questions are Yes/No questions. "Election methods" are shortcuts used to avoid the repeated voting that would be used in a direct deliberative democracy. Since secret ballot is relatively cumbersome, election methods are most frequently used where secret ballot is the practice. But election methods bring paradoxes and can easily result in decisions that are not what the public, overall, would choose, if standard group decision process were used..

An election of an officer can be viewed as a series of conflicting ballot questions. For any question, if the Yes votes outnumber the No votes, and there is no conflict, then the question passes, otherwise it fails and nothing is done. Multiple conflicting ballot questions, then, create the problem that more than one could pass. What to do?

Plurality, FPTP, answers this by requiring voters to vote for only one. If you vote for more than one, your ballot will be discarded as an error. Why is it an error? Well, because you are not allowed to vote for more than one. That you might *intend* to vote for more than one was not something even on the minds of our founders, those who set up our traditional procedures. Rather, their idea was that the most popular candidate should win, even if not by a majority. Consider the ancient context: the most popular leader could raise the largest army, and could thus, one by one, defeat other leaders; thus it made sense, right at the outset, to simply award leadership to that one, to avoid the destruction of factional conflict.

But with ballot questions, the matter had to be faced. If we vote Yes/No on each one, and they are all considered independently by the voter, which one should prevail? Suffice it to say that the answer is standard: the one with the most Yes votes.

And this is Approval Voting. Richie will argue, elsewhere, that it is not democratic (because the "first preference of a majority" may not be the Approval winner), he will blather about "Core Support," a criterion that he made up simply to make IRV good, he will raise misleading issue and objection over and over. He's a tenacious debater.

In discussing election methods, Richie and the way he works would not be relevant. But this is politics. People should know that he is highly skilled at crafting and presenting arguments designed to manipulate the way you think. In none of this would I suggest that you believe anything just because I say so. If the matter concerns you, research it. Ask the experts. You can, you know. There are two mailing lists, open to the public, where real experts read and write; there is the Range Voting list, which is really generally concerned about election issues and which is *not* a pure advocacy list, people are actively welcomed who disagree strongly with Range Voting as being a practical method; and there is the Election Methods list, which is much more general.

We in the election community have attempted to negotiate with Richie, but Richie has rebuffed all efforts to find common ground, even though, theoretically, we should agree on Proportional Representation. We would rather not oppose IRV so strongly, it is *probably* an improvement over standard methods here, though it is also probably not the first reform to implement. That honor would go to simply starting to Count All the Votes. Stop discarding overvotes, it usually only takes striking a coupleof lines in the election code, and the effect is obvious, in spite of all the smoke that Richie will raise: the spoiler effect is gone. If Florida had this simple and more just rule -- really a *lack* of a rule -- in place in 2000, we'd have elected a different President, and without any additional expense or fuss at all.

*Then* we can argue and debate about the next reform. Should we use ratings methods or ranked methods (if we are going to use ranked methods, why not insist that a candidate who would beat all other candidates, matched up two at a time against each other, win the election, if such can be determined from the rankings? IRV fails this, and for no good reason. It is just a product of the peculiar way that IRV drops candidates.

We do *not* have to make this decision now, and I'd really rather not debate the ratings vs. ranked issue at this time. Clay Shentrup, who brought Range Voting here, is a true Range Voting activist, but the general Range Voting community has decided to support Approval Voting as the immediate, simple, cheap, and effective first step. Just start Counting All the Votes!

Ah, yes, the original issue: Approval Voting is also standard, if invisible, whenever mutliple issues are decided in person, by voice vote or show of hands, whether the matter is governmental or not. It only becomes possible to discard overvotes if written ballots are used. I've seen explicit Approval Voting used for multiple choice questions in organizations, where the organization wanted to find the most broadly acceptable solution to a problem. I've seen it turn dedicated partisan fervor into *unanimous* acceptance of the result of an approval poll. Would this happen in public elections? Probably not, it takes full deliberative process and people who value consensus to pull off this trick. And, while that is possible, it is also quite another story.

And, as with many comments by Richie, there are layers of deceptive implication. Other than what I've described, I'm not aware of any usage of Approval Voting for governmental officer elections in the U.S. Unless you consider that the U.N. is located in the U.S.! I understand that the Secretary General of the United Nations is elected by Approval Voting. But that is not a "public election."

But something very much like Approval, in certain ways, was used in the U.S. i nthe past. Bucklin Voting is a hybrid method: it's like IRV and Approval combined. The ballot is a ranked ballot; if the first place votes don't determine a majority winner (precisely what IRV considers), the second place votes are added. Thus it *becomes* an Approval election. No candidate is dropped, but ranks continue to be collapsed until a majority winner emerges. I'm not arguing for Bucklin Voting, but it is certainly a respectable method. It was discarded in the U.S. for two reasons, and there is a paper on the FairVote web site that gives reasons. I'm pretty sure one of those reasons is misleading, it was not the real reason.

The *true* reason is that some state courts ruled Bucklin Voting unconstitutional, based on an alleged violation of the one-person-one-vote rule. That was, simply, an error in law, no experts today consider that Approval Voting violates one-person, one-vote, and the obvious counterexample is the matter of conflicting Ballot Questions.

And the misleading one is that allegedly it was discarded because most voters were not ranking candidates beyond first place. But this practice, called bullet voting in Approval and truncation in IRV, is not harmful to the public, it is only harmful to the voter who makes that choice, perhaps. If we look at Florida 2000, we can see that only a very small number of voters using an option to vote for more than once would have turned the election in a more fair direction. Approval helps even if most voters don't use it! And, unlike Bucklin, which uses a ranked ballot like IRV, expensive to implement and expensive to count -- unless you use those voting computers that we all love so much ---, Approval doesn't cost anything.

*Then*, if people really want to be able to express who their favorite is, and I think they will, further reform can be undertaken one small step at a time.

And why do I think Bucklin was dropped? Well, it ran contrary to the interests of those who held power at the time to allow third parties to gather strength without spoiling elections. It's just like Proportional Representation, which was dropped because it was working, and, horrors!, blacks and socialists started getting elected. Can't have that, you know!

bottom line: He is being paid to promote it

You clearly know Rob Richie well.

honestly, the "Fair Vote" tactics and strategy are amazing.

They use the same old talking points, and no matter what really happens, they keep using the same old feel good chants.

Instant Runoff Voting can take days and weeks to come up with a winner and it certainly requires either complex sorting of the ballots (for the 2nd or 3rd round) or more complex software (which our voting machines don't have, and which there is none federally certified to meet our law's standards).

Fair Vote was even smart enough to hire the Chairman of the Green Party to help promote the idea.

Ask how IRV does in San Francisco. But ask someone who isn't selling IRV.

San Francisco

Ask how IRV does in San Francisco. But ask someone who isn't selling IRV.

Funny you should mention that. I've lived in San Francisco for most of the past four years, and I don't support IRV. :)

Why are range voting and approval voting so ignored?

Rob Richie raises an excellent question, but he does not want us to actually think about the answer. As is typical for him, he asks a question, and the question is for the audience, and he knows what easy assumptions they will make. The implication of his question is that Range and Approval are poor methods because nobody is -- allegedly -- using them. Is this true?

No one is using range voting and approval voting in the world for any governmental election. No legislature in any city, county or state is taking it seriously. Why do you think that is?

First of all, Approval is in use in the United States, in "governmental elections" of a particular kind, where the "candidates" are not people, but Ballot Questions. If two ballot questions -- or more -- are on the ballot, and they conflict, i.e., it is "single winner," then, if more than one question passes, i.e., gets a majority, then the one with the most Yes votes wins. This is approval voting, pure and simple. It's already in use, and I've never seen it questioned.

But why not for public elections? It's really a good question if we take it as more than a rhetorical jab. There are two reasons that I know. First of all, the very simplest election reform is Approval. We get it using exactly the same ballot as we have now, but we count the votes just as we would count the votes in those Ballot Questions. We simply Count All the Votes. Is this politically feasible? Well, to my knowledge, it's not been tried. Perhaps it has not been tried because it's considered unfeasible. Wait a minute, that's circular! Catch 22!

That we discarded overvotes in the first place is also based on a circular argument, at least in the only place where I have ever seen a justification for it, Robert's Rules of Order. Ballots with overvotes are discarded and not counted because they are errors, so we cannot know the "real intention" of the voter. But, wait a minute! They are errors because they will be discarded! When we set up the election system in this country, it was assumed that people would simply vote for their favorite, and it took some time for the party system to develop. I think that someone would want to vote for more than one candidate simply did not occur to them.

But it did occur to people later. Bucklin Voting was used in this country for some time, and FairVote has some excellent, if misleadingly presented, information on it. Bucklin is a kind of IRV, where the voters submit a ranked ballot, perhaps with a second rank candidate on it. If no majority appears from the first rank votes, then the second rank votes are added in. Thus the first round is a simple vote-for-one election, just as in IRV, and, also as with IRV, if there is no majority winner in the first round, second rank votes are considered. However, instead of pulling in second rank votes by dropping the loser(s), and reassigning their votes to the second choices on those ballots -- which requires a complicated counting process -- the second round votes turn the election into a limited Approval election. Effectively, if you want to, you can vote for two.

Bucklin was used for quite some time in a number of states, but it was ultimately found unconstitutional in some states -- for reasons too complex to go into here, and which I seriously doubt would be applied to simply counting all the votes -- and was dropped in others, allegedly on the grounds that only 13% of voters were using the additional ranks.

However, my own strong suspicion is that it was dropped because it became inconvenient for those in power. Such a voting system opens the door for third parties to build vote strength without spoiling elections. In a strong two-party system, 13% is high! I expect that most voters will continue to vote as they have for a long time, for one, if we start to allow this form of alternative vote. However, supporters of third party candidates will now have an opportunity to express their support for their favorite *and* to still participate in the *real* election. In the FairVote discussion of Bucklin, it is alleged that voters did not use it because they were afraid that their vote in second place would help elect that second place candidate over the first, and, of course, that's true. But it's not a bad outcome, and only a seriously partisan voter would seriously dislike that result.

So, yes, in 2000, if a Green voter in Florida had voted for Gore *and* Nader, he would be effectively abstaining from the Gore/Nader pairwise election. It's not exactly that his vote helps Gore defeat Nader, rather he didn't *help* Nader defeat Gore. Now, how many Green voters do you think would have been worried about this possibility?

If your favorite is one of two frontrunners, why would you do anything more than vote for one only?

Well, there is a reason. If you want to push your party in the direction of a third party, you may want to give some voting power to that third party. And, again, doing so, if enough people agreed with you, might elect that third party candidate.

So don't do it if you'd be upset by that outcome!

So why is Approval not being used in the U.S.? Partly because the simplicity of it has not been widely realized (and the inventor and many supporters of Approval are partly responsible for this, by calling it 'approval' and connecting the votes with 'approval' of the candidates, when, in fact, they are just votes, actions, and it's possible that you would sensibly vote for a candidate -- in addition to your favorite -- when you do *not* 'approve' of this candidate, whom you consider the least of two evils).

Approval is not being proposed as the absolute best election system, though there are some very knowledgeable people who think it is. Rather, it is being proposed, in particular by Range advocates, because it *is* a Range method, the very simplest one, it is cheap and simple and should generate no opposition from election officials, it actually makes their job easier, not harder, it involves no additional public expense at all beyond figuring out new ballot instructions, and it actually does a much better job dealing with the spoiler effect than IRV.

And the other reason it is not used, I strongly suspect, is that Bucklin, which is partly Approval, was dropped because it was working, just like Proportional Representation was dropped because it was working, and, horrors!, we can't have *them* being represented and having power in the political process!

Just one thing I want remembered from all this: Counting All the Votes is simple, easy to understand, cost-free, and it ends an old injustice. If people still want IRV, well, in my view, it is up to the people. But the obvious first-step reform is not IRV, it is simply starting to count every vote cast. You can be sure that if someone is against this, they are ignorant or have some other agenda.

For FairVote, where most IRV people get their information and analysis, the other agenda is paving the way for STV, a common method that is used for Proportional Representation. They are worried that if we get Approval, the pressure will be off, they *want* us to suffer from the spoiler effect until we give them what they want, which is the increased voting complexity of IRV, from which STV is a small step.

That's not a bad goal, in itself, but I happen to think that deceiving people, in order to get what you think is best, is part of what is wrong with contemporary politics, which has become about winning, not about finding and agreeing on the truth.

Exactly Right.

Icloud was expressing the sentiment that opening the system to competition and different ideas was important.

I appreciate that you got that - I don't really understand the finer points of the IRV, STV, FPP, etc. Too many initials for me! :)

I am a Democrat, and probably will be until I die. But I would like to see a greater range of ideas represented within the party, and I don't see that happening until a greater range of ideas are represented on a national, or statewide stage. It makes sense in my head, at least. More people talking, more ideas out in the open - it can't be bad. It has to be good. That's just how it works.


Be the change you wish to see in the world. --Gandhi

So you need to eliminate favorite betrayal

If you are a Democrat, you should consider the strategy of using Range Voting for the Democratic nomination process, on straw polls and the caucuses. This would better represent the will of the people, because it allows voters to fearlessly support their favorite candidates.

Plurality, IRV, and most other systems do not.

Defending IRV

Michael,

Instant runoff voting is not even close to plurality voting in the values it creates for democracy -- it deals with nearly all of the "spoiler" concerns people have, and when it doesn't, it's far too opaque for a voter to try to outsmart the system. IRV is a winner-take-all system -- indeed majority voting is one of its values -- so not designed to represent minority opinion in office, but it forces the major parties to be on their toes. If they nominate weak candidates or take voters for granted they can lose -- witness Mary Robinson's win in the Irish presidential race in 1990, Ken Livingstone's win in the London mayoral race in 2000 and Bob Kiss' win in the Burlington mayoral race in 2006. All the alleged "spoiler" dynamics Clay talks about apply equally to runoffs, and he freely admits runoffs can shake things up.

The reality is we have an entrenched two-party system that's unlikely to go away within winner-take-all elections, so the question is how one opens up political choice and expands debate within it while separately building the important case for proportional voting. IRV clearly can be a win-win in this situation, which explains why both the likes of Howard Dean and John McCain alongside the Libertarians' Blll Redpath and Greens like David Cobb embrace it. And IRV is getting some serious consideration and use in North Carolina. Gary Cox and theoreticians aside, IRV has a lot of history to point to that's a clot more nuanced and important than Clay suggests.

Range voting and approval voting are likely to break down in the real world of how people vote (if they ever get tried anywhere -- no governmental election in the world uses them). It's transparent what you should do -- vote for your candidate and no one else. If some voters naively use the system sincerely and others don't, the naive ones lose. A system with approval voting -type logic called Bucklin voting resulted in exactly that dynamic when tried in important elections like Alabama gubernatorial primaries a century ago, where nearly 90% of voters were bullet voting for one person. For more, see:
http://www.fairvote.org/global/?page=2077 - Bucklin voting
http://www.fairvote.org/rangevoting -- some thoughts on range and approval voting

Rob Richie
FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Some corrections

This post is typical of Rob Richie's serious misunderstandings of the finer details of voting methodologies (on which we've tried to correct him many times before).

Richie here implies that IRV either solves the spoiler dilemma or, when it doesn't, makes strategy too difficult for a voter to figure out. For example, in this scenario, a group of Nader supporters get Bush if they sincerely top-rank Nader. But if they strategically top-rank Gore, their second choice, then Gore wins, and they get their second instead of their third choice. The point Richie tries to make here is that voters wouldn't have any good way to know that this would happen, and so if a voter truly supported Nader, he'd just vote for Nader. But Richie fails to understand basic statistics that we have explained in great detail. Essentially, all a voter has to know in order to game the system is which is more likely:

1. That his favorite candidate (e.g. Nader in this scenario) is going to be elected, but if we betray him by making this strategic dishonest voting move (insincerely top-ranking Gore), we'll foolishly prevent that, or...
2. Our making this move will successfully cause our second choice (e.g. Gore here) to win.

What computer models, and real world ballot data both show, is that #2 is significantly more likely, meaning that a voter's best strategic option with IRV is to essentially always insincerely top-rank his favorite apparent front-runner, not his favorite overall candidate - whether he knows it will be good for him in that particular election or not. History seems to have empirically confirmed our theory, because IRV leads to two-party duopoly, even though it shouldn't have, if voters were generally being honest. It is interesting to note that plurality voting with a runoff seems to psychologically "trick" voters into increased honesty, which may account for why most of the 27 countries that use that system have escaped the duopoly widely found in plurality and IRV countries.

I stress that we have explained this at length to Rob Richie, on numerous occasions. Not only does he refuse to accept this relatively trivial math, but he suggests that voters in a Range Voting or Approval Voting election would tend to strategically "bullet vote" - that is, vote for just one candidate (e.g. give one candidate a "10", and everyone else a 0). but it is easy to show that this is not a good strategy. Consider taking any ballot where the voter had used this vote-for-one tactic, and then asking him to additionally vote for all the candidates he likes better than that one candidate he voted for. That could only help the voter - it simply could not hurt him. In justifying his contention, Richie has argued that a voter wouldn't want to vote for a second candidate, because it could hurt his more preferred candidate (that is, you have given your favorite a 10, but by giving your second favorite a 9 instead of a 0, you might help him beat your favorite). But that argument doesn't make sense when we consider, for example, a Nader voter who would have strategically voted for Gore in our current system. With Range/Approval Voting, he may start by voting for Gore, but then it does not hurt him to additionally vote for Nader - Range Voting makes it ALWAYS safe to maximally support your sincere favorite candidate (and anyone you like more than the candidate you've had voted for if yo could only vote for one). Note: Richie is effectively arguing that this voter wouldn't want to additionally vote for his more preferred Nader (after having already voted for Gore), because that might cause Gore to beat Nader. Simply put, Richie's logic just doesn't make sense here.

Richie also contends that "majority voting" is one of IRV's values, even though IRV can elect X when a huge majority prefers Y over X. Example. Range Voting can do this as well, but tends to do it in circumstances where it can be sensibly defended, whereas IRV does it rather randomly.

Richie contends that IRV keeps major parties "on their toes", but does not give any evidence that this applies any more to IRV than to our current vote-for-one method. Weak candidates can lose with any voting method. And it is ironic that Richie brings up Mary Robinson, who won the IRV-elected Irish presidential race in 1990. She was with the Labour Party, and hers was the only deviation from monopoly rule of that office by the Fianna Fáil party. That's right, aside from that single exception, IRV has produced monopoly rule in the Irish presidential post since 1938 - and that's in spite of the fact that Ireland has more than two competitive parties in their legislature, and the presidential post is mostly ceremonial (so there's not even much incentive to game the system).

Let's look at Richie's other examples. Bob Kiss was elected mayor of Burlington, Vermont in their first use of IRV, in 2006. Ken Livingstone was elected as mayor of London as an independent in the first election after it adopted a form of IRV. It takes voters a few election cycles to learn to game an election system, but history shows it will inevitably happen. Consider that it took the U.S. some time to settle into two-party domination even with our vote-for-one system. And as for Ken Livingstone, note that he was actually a popular member of the Labour Party, only switching to an independent when he failed to get the Labour nomination for the initial 2000 mayoral race. In the 2004 race, he ran, and won, as the official Labour candidate. That's not exactly a blow to duopoly. It is similar to the Joe Lieberman vs. Ned Lamont situation in the U.S. not so long ago, in which Lamont won the Democratic nomination, but lost to Lieberman, who ran as an independent. Does Rob Richie think that this is evidence that plurality voting helps fight duopoly? If not, then why is he trying to use Livingstone as an argument that IRV does?

Richie is right that the spoiler dynamics of IRV can apply to genuine runoffs (although the methods are mathematically different with more than 3 candidates). However, we believe this is essentially due to voter psychology. Essentially the two-step processes of genuine runoffs "tricks" voters into being more honest in the first round, because they intuitively feel that they can correct this mistake in the runoff. As anyone who has studied elections knows, voters can do some very irrational things (e.g. about 1/3 of voters are registered as independents, even though that simply makes them less powerful, because they then cannot vote in the primaries). And either way, we do not support genuine runoffs any more than we support IRV. We (the Center for Range Voting folks) see them both as essentially unsupportable. We are merely pointing out how IRV's properties cause it to result in duopoly, even though other single-winner methods can change that.

The irony continues as Richie mentions this, but then says that our duopoly is unlikely to go away with single-winner elections. My very point was that this is demonstrably false, since most countries that use a genuine runoff have escaped duopoly (somehow Richie missed this point, and thought I was praising runoff systems). There is no reason that Range Voting, for instance, should result in duopoly. This is because Range Voting allows voters to safely vote for their favorite candidates, regardless of whether they run in a major party or not (and regardless of whether they seem "electable"). Again I point out that IRV does not have this important quality.

Many with the Center for Range Voting - myself included - agree that there's a place for proportional representation. Indeed we have this insightful page on the issue. But we believe there is a strong case to be made that we cannot expand proportional representation without first breaking up duopoly in single-winner elections, which will then allow us a chance to change legal impediments to its adoption. So we think that Range Voting is a much better stepping stone to proportional representation than IRV is. And Reweighted Range Voting and Asset Voting are simpler than STV, and are superior in several objective ways (e.g. are monotonic - so raising your preference for a candidate cannot hurt him). I understand that Richie believes RRV will be susceptible to (un)strategic exaggeration, but this ignores the obvious fact that doing so will cause a voter to waste power in later rounds of voting. For instance, if you are electing a council of 3, and you exaggerate your honest 7 into a 10 in the first round, then your power in the later rounds is diminished. Therefore that exaggeration hurts, and is not strategic. I thought this was obvious, but it seems that Richie does not spend a great deal of time analyzing much of the work he criticizes. That is unfortunate, because he is a man with a great deal of public relations expertise and political experience. His skills could be put to much better use if he'd be a better researcher/scientist.

Richie and I both cite examples of both supporters and detractors of Instant Runoff Voting. The difference between them is that the detractors have analyzed the issue in far greater detail. For example, the Libertarian Reform Caucus has called IRV a "bullet in the foot" in an article in which they analyze its serious flaws, as well as the duopoly it has historically produced. They advocate and use Range Voting for internal processes. Michael Badnarik, a former Libertarian Presidential candidate, has publicly spoken of his preference of Approval Voting (simplest form of Range Voting) over IRV. In fact he recently had as a guest on his Austin radio show, Rob LeGrand, of Citizens for Approval Voting, who spoke about the intricacies of the two methods. And as I noted above, Australians, having used IRV in their house of representatives since 1918, take it as a point of fact that IRV produces duopoly (and I am sure their third parties would prefer Range Voting, if they were to thoroughly research the two methods).

As for Howard Dean and John McCain, I doubt that you will find any deep analysis by them of the relative strengths and weaknesses of various voting methods. Like most IRV advocates, they probably have taken a cursory glance at IRV, identified that it helps with the Perot/Nader "problem", and spoken in favor of it. I'll bet they wouldn't know what "monotonicity" or "independence of irrelevant alternatives" meant, to save their lives. A great number of myths and misconceptions about IRV persist in the public sphere, and are there largely due to the work of people like Rob Richie himself. Even the League of Women Voters is partly to blame. See some discussion of this. Such myths largely account for why so many ostensibly well-intentioned people make the grievous mistake of supporting IRV. And in my own experience talking to David Cobb personally over the telephone, such myths seem to perfectly explain why he too supports IRV. He has simply been misled.

Richie flails on his closing paragraph, claiming that Range Voting will break down in the real world. But he provides no evidence for this - just his own intuition and assertions. Our organization is headed by a Princeton math Ph.D. and a credentialed engineer, and we've actually analyzed a good deal of empirical data on this issue (for instance, I've done exit polling using Range Voting, in Beaumont, Texas). Our findings simply do not support Richie's claims. We have even performed election simulations using a wide range of election models, and giving IRV the benefit of the doubt by assuming that it will encourage more honest voting (which may well be completely backwards). Even then, Range Voting comes out robustly superior to IRV.

Richie's argument that no government election systems currently use Range Voting is also particularly ironic. Would he apply that same argument to the first IRV advocates, if we could theoretically send him back in time? If not, then I think it only fair of him to show the same respect to Range Voting, and judge it on its merits. If we argue against new ideas because they're new, then we do not advance.

It's transparent what you should do [using Range Voting] -- vote for your candidate and no one else.

This is simply false, and the fact that Richie continues to make this claim after it has been explained to him otherwise (numerous times), is truly astounding. We have even looked at computer-based statistical trials to compare various strategies, for the economists out there.

If some [Range Voting] voters naively use the system sincerely and others don't, the naive ones lose.

The same is true of Instant Runoff Voting, only the problem is much more severe with IRV.

A system with approval voting -type logic called Bucklin voting resulted in exactly that dynamic when tried in important elections like Alabama gubernatorial primaries a century ago, where nearly 90% of voters were bullet voting for one person.

This statement again illustrates Richie's use of deceptive arguments. Bucklin voting is radically different from Range Voting, and is in fact a rank-order voting method like IRV - not a cardinal voting method, like Range Voting.

http://www.fairvote.org/rangevoting -- some thoughts on range and approval voting

This page was pre-emptively responded to in my links above, but again I cite a critique by a Princeton math Ph.D. and voting method expert, Warren D. Smith:
http://rangevoting.org/RichieRV.html

I would hope those with an objective eye for honesty and scientific rigor will appreciate our commitment to the facts here. Range Voting is substantially superior to IRV, and light is increasingly being shed on the myths and misconceptions associated with IRV.

I invite those interested in continuing the discussion to drop by our Yahoo discussion group.

Clay Shentrup
415.240.1970
clay@electopia.org
San Francisco, CA

Clay's "corrections"

Hmm... Rather than go through point by laborious point, two things:

1. Clay has this whole thing about how IRV creates a spoiler dynamic that will impact voters, but has presented no direct evidence that it plays out in real elections-- I'd be interested in a single article about the thousands of IRV elections that have taken place for national legislative and executive offices to see how voters changed their preferences due to such strategic calculations. But note: the alleged spoiler impact would be even more serious in runoffs (because the field is truncated to two after one round), but Clay doesn't have as many problems with runoffs. I don't get it.

2. Clay, the fact that Bucklin voting was an ordinal system (ranking candidates first an two) has NOTHING to do with how it operates like approval voting in psychology. All the rankings did was give voters MORE options than they have with approval voting, as their second choice wouldn't immediately count against their first choices. But with Bucklin, all second choices were added to all first choices when no candidate won a majority. So basically voters could "approve" of two candidates, with only their first choice counted in the first round. The fact that nearly 90% of voters in a hotly contested gubernatorial primary race would not rank another candidate second ties directly to people widely figuring out the problem of having your second choice count against your first choice. The 13% that did rank a candidate second almost certainly were making a strategic mistake due to being naive about the real-world consequences of their system. And the winner fell far short of having votes from a majority of voters.
By not getting this point, then, Clay doesn't see why with approval voting very few people would "approve" of more than one candidate. It's so transparent what to do to help your favorite candidate -- bullet voting. Note that some exit poll survey of a relatively few voters who had never thought about the system before is nothing like how a system would be used in the real world, where all the political players would have engaged in it and informed people how they wanted them to vote.
Approval voting has rarely been used for anything important, but where it has, like the IEEE (the Institute of Electrical and Electronics Engineers, Inc.), this is exactly what happened. It turns into plurality voting, which explains they the IEEE got rid of it after a few elections. Meanwhile, numerous major private NGO's (and of course many nations) use IRV and keep using it.l

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Selfish altruism???

Clay has this whole thing about how IRV creates a spoiler dynamic that will impact voters, but has presented no direct evidence that it plays out in real elections

Actually I did. Remember, IRV has historically led to two-party duopoly in all four countries where it has seen widespread long-term use, which is exactly what we would expect if people were strategically (or perhaps naively) exaggerating their rankings. And I could just as well ask you, where is your evidence that IRV doesn't result in strategic voting, especially when strategic voting with IRV is so much simpler to describe and follow than with Range Voting?

It is interesting that you demand evidence to support our claims that IRV users will do what is strategically wise for them to do, yet you claim that Range/Approval Voting would result in a bullet voting (vote for just one) behavior, which actually is strategically bad for the voter, and worse even for society as a whole. That is, you don't even believe that the voter will help himself at the expense of society as a whole (what normally happens with strategic voting), but you actually believe he'll hurt society and himself.

The ultimate point is that, even if we generously assume that Range Voting will result in vastly higher numbers of strategic voters than IRV, Range Voting still does a better job of satisfying voter preferences - meaning you, Joe Voter, are statistically more likely to be more satisfied with Range Voting elections than with IRV elections. This effect is especially pronounced the more strategic the voters are in general (as in, the more voters there are who would try to game either system).

I agree that there's something to be said for the comparison you make between Approval Voting and Bucklin voting, at least insofar as bullet voting behavior is concerned. However, I believe your analysis is mistaken. Say we go with your figures, and assume that only 10% of voters will vote for more than one of the candidates, in a typical 3-candidate spoiler situation let us say the Democrat would beat the Republican by 5% (a realistic margin). But then let us say a Green joins the race, and garners 15% of the votes. We will make a fairly realistic approximation (for the sake of simplicity) that these votes are taken from the Democrat. Now, if even 67% of the Greens additionally vote for the Democrat (to maximize the odds of not getting the Republican, and having a result they can live with), then the Dem still wins, and that's only 10% of the electorate voting for more than one candidate. The point is, it's fine if most Range/Approval voters just vote for one candidate. It's primarily the voters who could traditionally create a spoiler problem whom we want to utilize the option to vote for more than one candidate.

And now consider what happens if a third party (we'll just continue with the Greens, from this example) continues to grow in size, until the voters' ranked preferences might look something like this:

% of voters - their preferences
28% Green > Dem > GOP
21% Dem > Green > GOP
6% Dem > GOP > Green
45% GOP > Dem > Green

This scenario illustrates that with IRV, a third party that grows to be competitive with the most similar major party, is now a liability for its supporters. If they vote sincerely for Green, they get Republican - whereas if they "lie" by insincerely top-ranking Dem, they get Democrat. Since this scenario is generally more probable than the odds that Green will beat GOP, Green supporters know (or come to realize in time) that it's statistically/economically unwise of them to cast a sincere vote for Green. And so, they don't. And so, duopoly reigns.

But with Range Voting, Green supporters can always cast a sincere vote for Green. If they get to the point where they think Green has a realistic chance of beating GOP, they can then decide not to be on the safe side by supporting Dem.

This point bears repeating. With Range Voting, a voter can ALWAYS show his support for his sincere favorite, whether the candidate appears "electable" or not. With IRV this is simply not the case.

but Clay doesn't have as many problems with runoffs. I don't get it.

I guess what I said before bears repeating:

we do not support genuine runoffs any more than we support IRV. We (the Center for Range Voting folks) see them both as essentially unsupportable. We are merely pointing out how IRV's properties cause it to result in duopoly, even though other single-winner methods can change that.. (somehow Richie missed this point, and thought I was praising runoff systems).

By not getting this point, then, Clay doesn't see why with approval voting very few people would "approve" of more than one candidate. It's so transparent what to do to help your favorite candidate -- bullet voting.

That's a great argument, if you believe a voter's motivation is to altruistically help his favorite candidate, rather than to help himself. But I just don't think this is very realistic.. For instance, we know that a lot of Nader supporters strategically voted for Gore back in 2000. How many? Well, that might take a little digging. But in any case, you are in sparse company if you believe voters will choose to selflessly help their favorite candidates. Most people accept the commonly held belief that voters tend to do what is best for themselves, which is why they often won't vote for a candidate that they feel can't win. For instance, a lot of Democrats will probably vote for Obama or Hillary in the Dem primaries, even if a "second tier" candidate like Kucinich or Gravel is their favorite - simply because they don't want to waste their vote.

Note that some exit poll survey of a relatively few voters who had never thought about the system before is nothing like how a system would be used in the real world, where all the political players would have engaged in it and informed people how they wanted them to vote.

Yes, we're well aware of that. But we have rigorously analyzed various possible strategies, from a game theory perspective, and it seems that no matter what model we employ, if it is anywhere even approaching realism, Range Voting outperforms IRV.

One thing I admit though, is that we generally looked at voters as either being sincere, or selfishly strategic. You propose a fundamentally different model, in which voters are are selfish enough to want to help just their favorite candidate, but selfless enough to want to waste lots of their ballot power by helping no one else (like their second or third choices). I can't see any reasonable argument that voters would behave in such a way, and I've really never encountered this theory in all of my encounters with election reform. I think we can answer this question scientifically by trying to find out what fraction of voters will betray their favorite candidate to get a better result. Looking at the 2000 election would be a good place to start.

Do you have some evidence to support your counter-intuitive perspective on this?

Approval voting has rarely been used for anything important, but where it has, like the IEEE (the Institute of Electrical and Electronics Engineers, Inc.), this is exactly what happened. It turns into plurality voting, which explains they the IEEE got rid of it after a few elections.

We actually have an alternate explanation for why they reverted. And note that lots of other large organizations full of math/engineering/science geniuses still use Approval Voting to this day - but few use IRV.

Meanwhile, numerous major private NGO's (and of course many nations) use IRV and keep using it.

There are some that do, but if you ask them why, they will likely point to the same kinds of typical IRV myths that you yourself have perpetuated, and which we address here.

Maybe you can explain for our audience why that Science magazine article back in 2001 falsely claimed "A voter's best strategy is to sincerely rank the candidates."

Last post in this exchange..

Clay,

- Anyone who doesn't vote sincerely with IRV is likely to hurt themselves, because elections have enough unpredictability for you to "outguess" the system. So I absolutely stand by the 2001 letter in Science.

- You're quite wrong that it's hard to know how to vote strategically in range voting. You give the maximum number of points to your favorite and no points to anyone else. Doing otherwise will hurt your favorite's chances. You can be absolutely sure of that.

Onto real work, and this will be my last post in this exchange. But Clay, with all your work, where are you making concrete headway? Do you have anyone adopting your proposals for real elections? If not, why do you think that is? Might it just be possible that you're wrong?

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Errors about Range Voting -

Errors about Range Voting

- You're quite wrong that it's hard to know how to vote strategically in range voting. You give the maximum number of points to your favorite and no points to anyone else. Doing otherwise will hurt your favorite's chances. You can be absolutely sure of that.

Once again, Richie mixes half-truths with outright false statements, he's been pretty careless here.

Voting "strategically" in Range is a misnomer, because, unlike ranked methods, "strategic" merely means "applying the full weight of your vote in an effective manner," something which most of us prefer that voters actually do, whereas in ranked methods it means reversing preferences.

So how to vote effectively in Range? Richie has not stated the general rule, he has only given what someone who actually prefers a frontrunner and dislikes everyone else would do. In that situation, he's quite correct. The maximally effective vote would be as he describes, *unless there are more than two frontrunners.* If there are more than two, it gets complicated, but it's still not truly hard.

So, here, I will just note some counterexamples to what Richie claimed. You are a supporter of Nader in 2000. Sure, you will max rate Nader. But you will also, if you care about actually influencing the outcome with a full vote, max rate Gore.

You are a supporter of Gore in 2000. But you like Nader and you want to move the Democratic Party in his direction. So you max rate Gore, and you rate Nader as you like. You are completely free, unless you are worried that Nader might actually win, and then you have a more difficult choice.

Let's not make it complicated, at first. If there is a two party system, if third party candidates don't have a chance to actually win, effective voting in Range is completely easy, and the general solution is not as Richie claims. You really have two choices, both of them reasonably effective.

First, so-called approval-style voting. You max rate candidates you want to support and you min rate all others. Contrary to claims of some critics of Range, there is nothing wrong with this.

Secondly, so-called "effective sincere voting," which is that you consider, first, the frontrunners, and you max rate your favorite and min rate the other. Then you max rate anyone you prefer to the frontrunner, and min rate any candidate whom you like less than the other frontrunner.

And then there may be some other candidates. You can do what you like with them, it will do no harm, unless you are wrong about identifying the frontrunners, in which case, if you have voted sincerely with them (i.e., used intermediate ratings to express your actual preference strengths) you will have performed a public service, by honestly expressing true preference strengths in a system which is designed to maximize overall satisfaction for society..

Yes, in such a situation your vote could cause your favorite to lose (remember, this is only if a third party candidate has a chance to win), but only to one whom you actually considered good. It won't cause your favorite to lose to the worst fruntrunner or anyone worse than that, because you are already voting your maximum strength in those pairwise elections.

Richie bails, his usual

Richie bails, his usual practice

Onto real work, and this will be my last post in this exchange.

We have seen this over and over from Richie. He pops into a forum, lobs some grenades, and then leaves. He uses, over and over, the same bankrupt arguments, but in each forum, they are new to the participants, and he's put a lot of work into making them effective. And if someone does the work to dismantle them and see what is actually being said and the real implications, it becomes long and tedious. Lots of people won't read it, and thus are left with exactly the impression he wants people to have. He is *highly* skilled at this, it's easy to understand why he is the one being paid to do election reform.

From my point of view, though, he represents much of what is wrong with the U.S. political system. It's not designed to find consensus solutions. It's a contest, where the contestants try to make everything black and white, develop the most effective sound bites. We could solve this problem, and there are initiatives under way. They will be successful, I predict. But for now, we have Richie and U.S. Presidential 2008 to worry about. That may not be a close election. But what if it is?

But Clay, with all your work, where are you making concrete headway?

Shentrup is a dedicated volunteer. He does not represent the Center for Range Voting, and his views are highly polarized. But he's also working for a good cause. Usually there is something behind what he claims, and he is sincere. He's not trying to deceive you, he believes what he's saying, and he's thought about it. Some of us think he hasn't thought enough, but this really reaches to some rather fine points. He's right. Range Voting is the best method on the table. But it's also more complicated than Approval (Approval is really a Range method, but the simplest, there are only two possible votes: Yes or No for each candidate, and No can be assumed from the absence of Yes).

Shentrup has done quite a lot of effective work; one of our goals is to foster the use of Range and Approval voting in non-governmental organizations. Then, when people are familiar with it, and we know that the horror stories promoted by Range critics actually don't occur in real life -- that's likely -- its chances for public adoption are much better. Range Voting is already in wide use for multiple choices, in business and in other organizations. Not usually for elections, as such. Shentrup has been effective in promoting this, so far.

Richie has routinely dismissed Shentrup and other Range advocates as politically naive. That may be true. But I'm not sure I'd want to be as "politically sophisticated" as Richie. I'm not willing to lie or to conveniently forget the truth that has been explained to me over and over, merely to make an effective argument. Frankly, I'd rather lose the argument than deceive to win it, something that might disqualify me from participation in U.S. politics.... But that's not my goal. My goal is public education. I trust democracy. I trust that an informed public will make good decisions. Richie, most certainly, does not, he has been fairly explicit about that at various times.

Do you have anyone adopting your proposals for real elections? If not, why do you think that is? Might it just be possible that you're wrong?

Notice the qualifier "real." It's a typical Richieism. He means governmental elections, even though there are other elections which can sometimes move more power and money than a governmental election. He's got a very narrow focus.

Shentrup is not always right. But he's quite honest, and he is fighting for what he sees as the truth. I'd much rather walk in his shoes than Richie's, for Richie is paid to promote a political cause, not the truth, and he has nothing but contempt for those who are true experts on election methods, such as other supporters of the Center for Range Voting. He thinks them idle dreamers, but, it's more accurate that he does not understand them, he does not take the time.

As I understand that Richie is an activist, and that constant debate would indeed be unproductive for him, I'd suggest that he designate someone to act as his proxy in debates about election methods, someone who could present his arguments and stick around to sift what is true about them from what is not. And who, might it be possible?, would tell him if it turns out that he's been backing the wrong cause. Would Richie listen to his choice for proxy?

Well, if he'd listen to anyone, I suppose it would be that person. Meanwhile, we'd all have a chance to get to the bottom of all this debate. Is this just a matter of seeing who tosses the biggest grenades? Or does the truth matter?

Sticks and stones...

Sigh. Lots of loaded statements here. But FairVote isn't against full debate. We just don't want to have it over and over and over again.

We're not alone in our decision to support IRV over plurality voting, runoffs and range voting. Several charter commissions have recommended IRV; none that I know of have recommended approval voting. Several state League of Women Voters have endorsed IRV after substantial, thorough studies. none have recommended approval voting as far as I know. Several city councils have voted to put IRV on the ballot; none have done approval voting. We have dozens of members who are highly respected political scientists who share our basic critique and reform analysis.

Among Lomax's various personal attacks, let me say that I don't do the work I do for the money. Whatever money FairVote has largely came out of winning more support for the work we do, from sincere individuals and foundations. But we started with nothing but a vision of improving American elections and people coming together for that common good.

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

...but facts will never hurt me.

It is illustrative that we (three credentialed mathematicians who study voting methods, and others) have shown you to be severely factually incorrect (for instance, about strategy with IRV and with Range Voting), and your best rebuttal is to call our statements "loaded".

You say you don't want to have this debate over and over again, yet you choose to repeatedly ignore devastating refutations of many of your arguments (by experts) and to make the same misleading and false arguments to new audiences, in forum after forum. I have never seen any indication that you have even the slightest interest in adjusting to new evidence. This is because your organization started long ago as Citizens for Proportional Representation ("CPR!") back in 1992, and made the decision to support Single Transferable Vote (STV), a proportional representation voting system that becomes IRV in single-winner races. For you, IRV is nothing more than a "stepping stone" to STV (although, we contend, not as good a stepping stone as Range Voting is to even better PR methods than STV). So the self-selected board chose IRV and never seriously considered anything else. Thus, you don't particularly care about IRV's problems, and your attempts to defend it are, more or less, purely tactical. You claim that a voter is strategically wise to vote honestly with IRV, even though that's easily disproved. You claim that a voter is strategically wise to "bullet vote" for just one candidate in Range Voting, even though that is also patently false. But it doesn't matter to you whether these claims are true or false. As long as the majority of people just see the pretty packaging that you wrap IRV in, and decide not to actually look inside the box, your agenda is advanced - and that's apparently all that matters to you. Other hard-working reform activists, such as Joyce McCloy, founder of the NC Coalition for Verified Voting, consider you to essentially be a snake oil salesman (my paraphrasing).

Another especially illustrative example of your typical misleading rhetoric is your mention of various groups who use and/or endorse IRV. These groups have usually adopted IRV on the basis of the very myths and misconceptions that you have so cunningly propagated. The League of Women Voters even made some flagrant errors in their "substantial, thorough studies" of IRV, which are pointed out by a Princeton math Ph.D. here. And yet you tout this popularity among non-experts as a substitute for actual evidence: "IRV is better, because the people who were misled into adopting it mistakenly think it's better."

But if you want to use an appeal to popularity that is actually meaningful, look no further than the large math/science professional organizations who use Approval Voting rather than IRV:

* Mathematical Association of America (MAA), with about 32,000 members;
* American Mathematical Society (AMS), with about 30,000 members;
* Institute for Operations Research and Management Sciences (INFORMS), with about 12,000 members;
* American Statistical Association (ASA), with about 15,000

These kinds of organizations are relatively immune to your misinformation campaign, and so they use Approval Voting rather than IRV - because it is objectively simpler and better. Your response to these is typically to point out a single example of an organization that reverted from AV to plurality voting - likely for political, not technical, reasons.

When you speak of nameless "highly respected political scientists" who allegedly agree with you, I wonder why you do not share with us the evidence that convinced them of IRV's quality. All indications I've seen support the view that there is a general scientific consensus that Range/Approval Voting is superior to IRV. In this discussion alone, 3 politically astute Ph.D.'s (2 in math, 1 in economics) have pretty thoroughly shot down your arguments. We have actually seen you argue with established mathematical/statistical facts.

I really don't know what more there is to say here.

Among Lomax's various personal attacks, let me say that I don't do the work I do for the money...But we started with nothing but a vision of improving American elections and people coming together for that common good.

I believe you truly did start out with good intentions. But you unfortunately chose to pursue the wrong reform, well over a decade ago. Any switch at this point would be an admission that you were wrong all that time, and that you wasted a great deal of time and energy. And so you now stubbornly refuse to admit you were wrong, to the great detriment of our democracy.

Clay Shentrup
clay@electopia.org
415.240.1973

How IRV would affect U.S.

How IRV would affect U.S. elections is largely a matter of speculation, there is little experience on which to base it. However, a less biased analysis would not be as rosy as the picture Richie presents. Let's look at the individual claims:

Instant runoff voting is not even close to plurality voting in the values it creates for democracy -- it deals with nearly all of the "spoiler" concerns people have,.

Comparisons with plurality are misleading. Few think that IRV is worse than plurality, except for the possibility that certain kinds of results, to which IRV is seriously vulnerable, would sour the U.S. public on election reform in general. Those effects, though, would not be terribly common, and the U.S. already tolerates truly lousy results from Plurality. So I don't buy the "good is the enemy of the best argument."

Does IRV deal with "nearly all of the 'spoiler' concerns"? Well, there are two concerns that I know of, IRV deals with one of them, what I call the "first order spoiler effect." Here, a candidate diverts a small percentage of the vote, causing the overall preferred frontrunner to lose. This is the spoiler effect that we are most aware of, it's an immediate and present problem, and it probably elected Clinton in 1992 (Ross Perot) and Bush in 2000 (Nader). However, all election methods on the table, and most notably the extremely simple and cheap Count All the Votes proposals, solve this problem

Then there is the center squeeze effect, which is known to afflict top-two runoff (think France). IRV is quite similar, and center squeeze *will* affect IRV when a third party approaches parity. If you support third parties, and you'd like to see them succeed, at least occasionally, IRV is not for you. Now, Richie knows about this objection, so he tries to head it off:

and when it doesn't, it's far too opaque for a voter to try to outsmart the system

Richie knows that critics of IRV will note that it fails the Favorite Betrayal Criterion, that voters can improve their outcome by voting the opposite of a true preference. However, the alleged complexity is beside the point. FBC is a concern because *without* a strategy, with sincere votes, the election chooses what is, quite obviously, the wrong winner. Center Squeeze means that a candidate who would win, paired off against the IRV winner, even by a supermajority, loses. The supporters of this candidate can avoid the problem by coordinating their votes. And, in fact, voters do this -- political parties in Australia distribute "how to vote Liberal" or the like -- and it *improves* the outcome.

Center Squeeze, though, is probably responsible for the fact that all countries using IRV have maintained strong two-party systems. Center Squeeze makes it far too hazardous for a third party to approach parity. If one gets near, but fails to win, that's it. That party is history, for its supporters have been very badly burned, they got a terrible outcome. As they say, if you are going to shoot the king, don't miss.

Essentially, IRV accomplishes some good things, but there are better and less hazardous methods of accomplishing those things. Should we be content to imitate the less-desirable methods of other countries, merely because they are ahead of us in some respect, or should we take the opportunity to truly study our options?

For immediate reform, there is a very simple solution that solves the immediate problem and starts to allow third parties to build support without spoiling elections. It's not a perfect method, it too can be improved. But it is cheap, -- actually free -- it requires no changes to election equipment, in fact it makes counting *simpler*. Ballots look the same, though ballot instructions might change a little. And suddenly third party supporters and independents have a chance to show in public elections their true support, without spoiling the elections.

And, yes, Proportional Representation. While STV is quite good (Mr Shentrup may not agree, but most of us do), there are better methods still, that accomplish the goal of broad representation even more accurately. Certainly we should not be content with solving the first-order spoiler effect. Approval continues to function well as a third party approaches parity, but the question becomes much more complicated....

but easy to fix, using what is really quite traditional. Hold a runoff when the majority has not clearly chosen one of the candidates. It would be unusual. But far more democratic than "instant" runoffs. (There are problems with runoffs, for sure, but a runoff triggered by multiple candidates getting a majority would not have a bad outcome no matter what happens in the runoff. There are more sophisticated considerations needed to look at this in detail.)

Blah blah blah

Y'all realize most of us have long since stopped reading what you're saying. I'm not saying stop....I'm just saying...whatever.

I'm still voting straight ticket D. The hell with this. :-D


Be the change you wish to see in the world. --Gandhi

Yeah, me too,

I try to keep up with voting issues and I was happy to do things to help get the NC Verifiable Paper Ballot law passed but this thread has dissolved into something boring.

Have you called to support H. Res 333 Impeach Cheney Today? call 202-224-3121 & ask for your Congress member by name

Boring?

Well, everyone has different areas of political interest, but people like myself and Abd find election methods to be a fascinating subject. And I whole-heartedly believe that the issues is far greater in importance than issues like election finance, or global warming. That is because election methods determine our ability to choose whatever path it is that we, as a society, want to take. If we want one thing, but get another - because of a bad voting method - then we aren't able to respond to the issues that matter to us.

Here are some links that make my case, that Range Voting could make an enormous difference to several significant world problems:
http://rangevoting.org/WorldProblems.html
http://rangevoting.org/LivesSaved.html

But whatever your political passion, best of luck.

Clay

My political passion

But whatever your political passion, best of luck.

My political passion is waking people UP! You know - stop talking about it - and DO SOMETHING. If you think Range Voting is better - tell me how I can make it happen. I can read the statistics for myself, and I will. Who do I have to talk to? Are there any actions being taken on this in the US? Where? Is a letter to the editor called for?

Is there a movement for this?


Be the change you wish to see in the world. --Gandhi

Activism

If you check out our discussion forum and our web site, there are some action items. We're a very young organization, so at this point we're trying to focus on education. I'm working to get various organizations to adopt Range Voting or even Approval Voting. We do op-eds and that sort of thing. I'd like to start getting some literature (pamphlets, etc.) made, but of course that takes time and money, and we've got to grow our base in order to have funding.

Let us know if you have any helpful ideas.

Your website

provides a lot of very useful information, and I actually learned quite a bit from it. I do have a suggestion for you. If you are trying to reach a larger audience, use the "simplified" version of your site as the entrance. The language in it is a bit more accessible to people who don't live and breathe political statistics, but who might be interested in what you've got to say.

For what it's worth, your simplified version checks out to a reading level of about 10.5, which means someone reading on a 10th grade level should have no trouble with it. Unfortunately, most documents meant for the general public need to be written at 8th grade or lower level. Many newspapers are written at 6th grade or lower level! If your stuff is not meant for the general public, then it's cool.

When you get to the point of actually having literature (even if it's just Adobe PDF files that can be printed for distribution from your site), you should have it checked for readability. It's simple to do: in most word processors, it's an option on the spelling and grammar check. To simplify the document, remove unnecessary words, use shorter words, and use shorter sentences.

Visually, your website is a bit overwhelming - if I hadn't gone specifically looking for the information after reading what you've written here, I would have gone back to Google to see if there was a more professionally written site. Sorry. You need some more white space on the page to give old eyes like mine a chance to rest before I'm hit with the next concept. Inserting a space would help.

Please don't take this a negative comment. I've bookmarked the site - it's a great source for information, and seeing your passion for this topic, I assume that it's regularly updated. It will be a good way for me to keep up with what's happening.

Thank you for what you're doing.


Be the change you wish to see in the world. --Gandhi

Thanks for the suggestions.

Thanks for the suggestions about the web site, rangevoting.org. I don't control the site, though some material I've written is posted there. Many of us have suggested that the main page be made simpler, but doing that while at the same time conveying the necessary information is not necessarily easy.

And this brings me to what readers can actually do, beyond merely discussing this stuff.

(1) If we spent one-tenth the time working on the system, as we do on the content, i.e., in understanding the purpose of democracy and improving its expression, as distinct from trying to control the outcome of specific elections, we would find it much easier to get better, more broadly acceptable outcomes. I'm not suggesting not working on elections and the details of decisions! But if that is all we do, we will never resolve the recurring mess we find ourselves in. The mess is not the fault of the Republicans! Beating them in 2008 will not make us safe. So: pay some attention to the system. It was put together over two hundred years ago by some pretty smart people, but they did not anticipate everything, and aspects of what they did were heavily colored by their times. With caution, we can do better, not in the name of the Democratic Party or the Green Party or the Libertarian Party, or the Republican Party, but in the name of democracy and freedom.

(2) The dangers in the system are like an iceberg, we see pieces sticking up and think they are the problem, so we try to shave them off. The system itself is not, in fact, a "problem," it is a solution to a very old problem, and the solution has many interconnected facets. Trying to improve it by looking only at one facet is likely to lead to burnout and frustration. We can and should make small improvements, as practical, but there are also global improvements possible that completely bypass the system. We should be aware that better solutions may come from places we don't expect. As progressives, for example, we tend to distrust the practices of business; but business was forced to solve certain problems long ago, or fail. The way that business addressed the problem of scale in democracy, long ago, is actually far more democratic than what we see in political systems, where the business solution is generally *prohibited* by law. Why?

I've never seen the question debated until recently. If we look, there are lots of these little contradictions, elephants in our living room. One path out of our mess is to start looking around, not thrashing around. Thrash around, we break things. Jumping off, half-cocked, to reform things, without taking the time and trouble to understand the existing system and the possible changes, is pretty much guaranteed to make things worse. The twentieth century was one sad story after another of fantastic ideals and ideas, very attractive, that led to terrible consequences. So *think.* And, yes, debate. Debate is good if we learn from it, if all we are doing is trying to convince others that we are right, it is worse than useless, it is part of the problem.

(3) And for something very practical, the RangeVoting.org web site could use some editing. If there is a page there, including the home page, that seems too hard to understand, confusing, unnecessarily complicated, whatever, then rewrite it, and send it to the mailing list rangevoting@yahoogroups.com, where it will certainly be considered, and we all might learn something. And for those of you who are opposed to Range Voting, you can help too. Tell us why you are opposed, and stick around for the discussion. Again, somebody will learn something! But do be aware that there are some very smart and informed people reading the Range list and sometimes writing there, and some of them can be blunt. I haven't said it will be easy! (And others will be very welcoming, even of critics.)

It is not at all necessary for everyone to understand election methods, but if nobody is discussing and debating their merits, we are quite likely to continue to make the same mistakes, over and over and over again.

As to those who comment derisively, claiming that nobody is reading all this, well, we just proved them wrong, didn't we? I'm not writing for the people who don't read, but for those who do, and I'm content if there is only one who reads. Let those who read inform those who do not, everyone has their role to play.

There's a lot going on for IRV in North Carolina

If you like the means to accommodate more voter choice, majority voting, less expensive campaigns and higher turnout in decisive elections, check out www.fairvote.org./irv and be in touch with Elena Everett at elena@fairvote.org- - she's working full-time on instant runoff votingin North Carolina, focused in particular on helping Cary and Hendersonville in their transition to IRV elections this fall and helping other communities consider it.

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

More IRV myths

Rob,

When are you going to learn that when you lie and mislead, we will call you on it? Sigh. Here we go again.

more voter choice

The best generalized strategy in IRV is to top-rank your favorite of the two front-runners, regardless of who your sincere favorite is. So it's no surprise that IRV has historically resulted in entrenched two-party duopoly. In the Irish presidential post, which has used IRV since 1938, there has been a virtual monopoly with the Fianna Fail party. So history says that with IRV, voters generally have 2 (or even just 1) choices.

majority voting

Ludicrous. IRV can elect candidate X, even though Y was preferred to X by a huge majority. For instance:

#voters - their vote
10 G > C > P > M
3 C > G > P > M
5 C > P > M > G
6 M > P > C > G
4 P > M > C > G

M wins the election with IRV, even though P is preferred to M by 22 of the 28 voters. Thus you can imagine my frustration when Rob Richie talks about how Range Voting "breaks down" in the Olympics when judges are dishonest. IRV can horrendously break down when voters/judges are 100% honest.

less expensive campaigns

1) Some estimates say IRV would increase costs. Are you saying it's okay to increase costs to the taxpayers in order to save the candidates money?

2) What is your evidence that IRV even does save candidates money? This is probably based upon your seriously mistaken belief that IRV makes voters free to vote for a candidate without worrying about whether the candidate is electable - which I should add that Range Voting actually does. But since IRV incentivizes voters to top-rank their favorite of the clear front-runners, having lots of money to appear like a front-runner (and thus, electable) is still crucial.

higher turnout in decisive elections

I'm not even going to argue whether this is correct, or substantial. The point I would make is simply that Range Voting performs better than IRV, even if less voters show up (although we believe there's solid evidence that it would increase turnout far more than IRV).

But again, I fear that many of our points are too esoteric. Rob Richie will continue to fool people with clever one-liners, e.g. "IRV is as easy as 1-2-3." That's what he's good at.

Find me some excellent candidates to vote for

that's what i care about. Teach people civics so that Americans know why they should vote. Wrench the power from the hands of a few media giants who control our minds. Restore the rule of law and democracy so that voting even matters.

It's that old chicken or egg problem.

Have you called to support H. Res 333 Impeach Cheney Today? call 202-224-3121 & ask for your Congress member by name

If you look over our site, I

If you look over our site, I think we make a pretty compelling case (although perhaps initially unintuitive) that adopting Range Voting would do substantially more to address these issues than pretty much any other single issue (or lots of them combined). One thing it does is dramatically decrease the importance of cash, which gives more honest politicians a better chance with the well-funded corporate lap dogs. It also plausibly would have a substantially greater positive effect than eliminating vote fraud, for instance.

I know it's hard to believe, but look through the pages I've mentioned.

Range Voting literally doubles the benefit of having democracy.

what Raleigh City Council said

its a damn shame this didn't get reported in the local media, so I had to transcribe the meeting from the video posted at the Raleigh City Govt website:

What Raleigh NC City Council Members said about IRV

City Council member Dr. James P. West opposed IRV: “I indicated that I have some concerns about this especially in disenfranchises certain segment of voters…especially those of lower socio economic level…”

Council member Thomas Crowder: “Just like blackjack in Las Vegas, we are going to see a lot of game-men's-ship trying figure out the odds on putting people into office…”

Council member Tommy Craven: “To me this is something that would certainly serve the convenience of the board of elections… but it's certainly not in the best interest of the voting public.”

Poor objections

These objections to IRV are pretty flimsy. Thomas Crowder seems to be making the point that IRV will make elections less predictable. But on the contrary, IRV stops the weak-third-candidate type of spoiler effect, meaning it probably makes elections more predictable. But let's just hypothetically assume he's right. What's wrong with that? For instance, Range Voting would cause there to be a greater number of competitive match-ups than just the one between the two major party candidates. That's less predictable, but it's also good - it means voters effectively have more choices.

And how is IRV (actually, NC will be using "instant" top-two runoff, not IRV) going to disenfranchise poor/minority/etc. voters? I suppose the ballot might be a little more complicated, but as I recall, they are allowed to rank just one, if they like. Different cities implement slightly different variations of rules, so I'm not sure how NC is doing it. But it's hard for me to buy the "oh gosh, ranking some choices is too hard for an adult" argument.

Although I can't help but point out, Range Voting and especially Approval Voting are simpler and less prone to error.

Theory and practice

I'm not convinced that IRV/other STV systems are WORSE than FPP, but Clay is quite right that it's not clear they are better in terms of selecting a single outcome, and could be worse under some plausible circumstances.

Well, that's a bit complicated. In direct comparison to plurality (First Past the Post) voting, IRV is a moderate improvement, in an objective social utility sense. But in order to get the full picture, we have to consider practical facts. IRV increases spoiled ballots, typically by a factor of 7. It also seems to catalyze the implementation of fraud-conducive (electronic!) voting machines, according to leading election integrity expert Rebecca Mercuri, a computer science Ph.D. who has been called upon to speak on election fraud and related issues in various court cases. And some estimates say it can actually increase costs, even when taking the elimination of (often relatively infrequent) runoff elections into account.

I think the biggest argument against IRV is a problem that many engineers are well-aware of - there's no impediment to a good solution like a "good enough" mediocre solution. Once IRV is implemented, reformers often feel the leak has been patched, and so they move on. Eventually they might see that they still have many of the same problems they had before, including two-party duopoly. But at that point it's too late. If they try to correct their mistake, everyone else will say, "Hey you progressive fruit cakes - we gave you what you wanted, and you're still not satisfied?" And there just won't be the momentum or the stomach to undergo another change to the voting method. This helps to explain why reformers such as myself have worked to encourage reformers not only to support Range and Approval Voting, but to not support IRV. We feel there's a pretty strong case to be made that it's a net negative, when all of these facts are taken into account.

This also seems to be the general consensus among credentialed math and political experts. Consider the work of Columbia's Steven Brams, Dartmouth's Robert Norman, and Harvard/Unity08's David King. Experts who deeply study the game theory behind election systems (and yes, also factor in the important real world considerations, as well as empirical evidence) seem to overwhelming prefer systems like Approval Voting to IRV. The public sphere is just a little behind them, largely because of misleading propaganda by groups like FairVote.org. This is unfortunate, because they also do some very good work, and seem to have good intentions.

Inaccuracies

"But in order to get the full picture, we have to consider practical facts. IRV increases spoiled ballots, typically by a factor of 7. "

Burlington VT in its first IRV election had 99.9% valid ballots, with five candidate going for an open mayoral race -- with the very highest rate of valid ballots in the lowest income ward. San Francisco in its first citywide election with IRV had 99.6% valid ballots. That's a bit lower than the validity rate in some of the non-IRV races, but much higher than most of our elections around the nation. Using hand-counted paper ballots, Ireland typically has about 99.5% valid ballots.

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Sure

From a brief scan p;roblems with your page include:

- Australia has mandatory voting and mandatory ranking to have a valid ballot. People will tell you in Australia those are two huge factors for its rate -- a lot of people are protesting being the polls.

- Ireland in its presidential races with IRV consistently has lower rates of spoiled ballots than in the U.S. using plurality voting. You suggest ballots that don't count in the final round are "spoiled," but voters have the right to abstain from a ranking.

- Most ridiculously you suggest undervotes are "spoiled." So let's take November 2004, when a whole lot of people are out to vote for president and don't go all the way through the ballot. That's typical, and indeed several of the city races had less dropoff than higher ballot races in the city like for state legislature.

I'm not sure how many people are following what may well be a tiresome debate, so I'm signing off. But I will note that you didn't engage with how Bucklin voting indeed is like approval voting, and the fact that nearly nine in ten voters didn't want to have a second choice count against a first choice is very instructive.

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Rob Richie
Executive Director

FairVote
6930 Carroll Avenue, Suite 610
Takoma Park, MD 20912
www.fairvote.org

Is Bucklin like Approval?

Richie raises some fairly minor points here, even though he said he'd not be writing more. He's right it can get tiresome. He has put together a series of arguments, like an attorney for a cause, which he is. These arguments are dense with incorporated assumptions and misleading implications, which can cause discussion, if it tries to answer the points he raises, to be come *quite* tedious. A few here will read all of this. Many will not. I'd ask the few to bring back and dwell on what is important from all this. That way, the few represent the many. This kind of process can work to find consensus, *without* everyone spending all their time reading long arguments.

I'm not going into the vote spoilage issue, only to note that many ballots are currently spoiled due to overvoting, and that would continue with IRV, and it is almost certain that the rate would increase. But, to me, it's not a major issue. If that were all that was wrong with IRV, it would probably be foolish to oppose it!

However, something more important:

I'm signing off. But I will note that you didn't engage with how Bucklin voting indeed is like approval voting, and the fact that nearly nine in ten voters didn't want to have a second choice count against a first choice is very instructive.

The story of Bucklin is indeed interesting. Bucklin, as Richie asserts, is "like" Approval Voting. But it is also like IRV. It uses a different procedure for transferring votes. Essentially, instead of running a series of "instant runoffs," dropping the last place candidate each time, it runs a series of Approval elections, adding additional approvals each time. In Bucklin, while your second place vote can indeed hurt your first place choice -- depending on how others vote, i.e, if your first place vote does not win a majority, then the second place votes come in and, of course, your second place votes go to some one else. Bucklin is quite like Approval, but staged so that you *do* have a chance to express your first place vote and have your second place vote not hurt the first place one. But if that candidate isn't going to win -- i.e., would not win in Plurality -- then you are approving additional candidates. If you care to. If you don't you won't

Who would care? Well, it is indeed the same as approval. Ten percent of voters caring to approve an additional candidate in public elections in a two-party system, that is quite healthy, actually. The spoiler effect is avoided. I strongly suspect that Bucklin was *successful,* and was ejected because of opposition from entrenched party interests. It was too helpful to third parties, so a concerted political and legal battle was begun to remove it, and that succeeded.

In any case, we are not proposing Approval as the absolute best election method. Rather it is a very quick, cheap, and quite effective fix for the immediate problem, which it solves quite as well as IRV, at much lower cost than IRV (and than Bucklin, which *did* solve the problem, before, and would do so again.)

None of these methods, by themselves, is going to eliminate the party duopoly. But Approval would do just as good a job as IRV, at far lower investment. Then, later, we can worry about refinements. Bucklin is also simple, and should perhaps be considered again. It has some actual history, though I suspect, again, that the history has been presented to us selectively by people with an axe to grind. We need more and deeper research, something that is definitely not going to come from "FairVote." FairVote is only going to present to us what it thinks will lead us to the conclusions they want. The Center for Voting and Democracy is certainly not for democracy, real democracy, now. They don't believe in it, and we can know that because they don't use it.

Problems?

Australia has mandatory voting and mandatory ranking to have a valid ballot. People will tell you in Australia those are two huge factors for its rate -- a lot of people are protesting being the polls.

Actually, I agree. Perhaps we should compare Australia's IRV spoilage rates with rates from when they used plurality, instead of comparing them to countries that use plurality voting but do not have mandatory voting.

Ireland in its presidential races with IRV consistently has lower rates of spoiled ballots than in the U.S. using plurality voting.

But more than in Irish plurality voting.

You suggest ballots that don't count in the final round are "spoiled," but voters have the right to abstain from a ranking. Most ridiculously you suggest undervotes are "spoiled."

Whether we call them "spoiled" or not, the page clearly distinguishes these types of ballot from those which are not even usable in the first round. So the page is accurate.

Consider the case with San Francisco, which you did not seem to take issue with:

So from this we see that the overvote error rates in San Francisco ranged from 3 to 11 times higher with IRV than with plurality voting, typically 7 times higher. (If double-ranking a non-top candidate in IRV were also considered – we haven't – then IRV overvote error rates would have been even higher.)

Here we are clearly talking about ballots that you would claim are truly spoiled (not even usable in the first round).

But I will note that you didn't engage with how Bucklin voting indeed is like approval voting, and the fact that nearly nine in ten voters didn't want to have a second choice count against a first choice is very instructive.

Actually, I did respond, and one of my essential points was:


..it's fine if most Range/Approval voters just vote for one candidate. It's primarily the voters who could traditionally create a spoiler problem whom we want to utilize the option to vote for more than one candidate.

This is one of those fine points that you consistently gloss over, because it doesn't fit with your lob-and-run strategy.

Warren D. Smith (Princeton math Ph.D) weighs in

>Australia has mandatory voting and mandatory ranking to have a valid
ballot. People will tell you in Australia those are two huge factors
for its rate -- a lot of people are protesting being the polls.

--"People" will tell me? Odd, the Australians I ask are not among
those people. I wonder
who they might be.

Here are the facts.

Some Australian districts abolished mandatory ranking (still mandatory
voting though) and ALL
Australian substates exhibit higher ballot spoilage rate than USA
plurality voting.
Nor are the states without mandatory ranking exhibiting smaller spoilage.
CONCLUSION #1: So we know mandatory ranking is not it.

I cannot compare Australian IRV vs plurality races (can I?)
cross-time is so much time it would be a ridiculous comparison. You
might be able to do it cross-race if there are any Australian races
using FPTP and if you regarded those races as comparable?

(Also all Australian states allow "party line" voting where you just
following the ranking suggested by some party without having to input
it yourself.)

Now according to John W. Dean (remember him from the Nixon admin?)
"Some thirty-three countries - all democracies - have mandatory voting laws."
Meanwhile Wikipedia says 32 countries:
http://en.wikipedia.org/wiki/Compulsory_voting .
and names these as enforcing it:
* Argentina
* Australia (compulsory enrolment and voting for both state* and
national elections)
* Belgium
* Brazil (non-compulsory for 16 & 17 year olds and those over 70)
* Chile (enrollment voluntary)
* Cyprus
* Ecuador
* Fiji
* Greece
* Liechtenstein
* Luxembourg
* Mexico
* Nauru
* Peru
* Singapore
* Switzerland (Schaffhausen)
* Uruguay

Trying to find some numbers:
Argentina (80-85% turnouts) had 2.3% and 7.9% spoiled ballots
Uruguay (85-95% turnout) had 3.7% and 2.6% spoiled ballots
these figures pre- and post-liberalization respectively (data from a
2004 paper by
Sebastien Dube).

In a Peru election:
"blank ballots=1.47%, spoiled ballots=3.29%"

The http://rangevoting.org/SPRates.html page already gave spoilage
rates of 2.16%
for Mexico 2006.

http://www.ecsiep.org/index.php?option=com_content&task=view&id=579&Itemid=92
says:
The programmes director at the Fiji Citizen's Constitutional Forum,
Jone Dakuvula, says the complex system has meant too many voters
haven't been able to state their intentions. "Fiji has the highest
invalid ballot papers in the world. In the last election it was 11 per
cent of the ballot papers were invalid. In 1999, it was over nine per
cent." Note Fiji used IRV.

CONCLUSION#2: based on these figures, we conclude it isn't mandatory
voting that is
causing the discrepancy, it is IRV. Comparing the IRV
mandatory-voting countries
with the non-IRV mandatory voting countries, the IRV ones have the
higher spoilage rates in EVERY case except for post-lib Argentina (but
the avg of the
two Argentina figures is 5.5% which is comparable to Australia).

In all I've cited rates from a total of 6 mandatory voting countries
finding the two IRV countries
top 1 and tie for 2&3 between Australia & Argentina. The probability
that 6 random numbers
would by chance happen to have the two IRV ones top and either 2nd or 3rd top,
is about 7%, yielding 93% confidence that IRV causes higher spoilage
rates (all 6
were mandatory voting countries) and this is regardless of
mandatory-ness for ranking.

>Most ridiculously you suggest undervotes are "spoiled".

--utterly false charge. Read the web page before you make the criticism,
not after, please. We discussed the distinction between undervotes
and other kinds
of spoilage quite correctly.

>I will note that you didn't engage with how Bucklin voting indeed is
like approval voting, and the fact that nearly nine in ten voters
didn't want to have a second choice count against a first choice is
very instructive.

--if this attack was intended to imply that approval and/or range
voters will not vote for "second choices" (or that only 10% will) then
you are exactly wrong as has been shown by numerous
real world studies of these voting systems, e.g. the French studies
http://rangevoting.org/FrenchStudy.html
http://rangevoting.org/OrsayTable.html
in which the vast majority (about 80% or 90%) did NOT vote plurality-style.

Indeed, as far as I know, the rates of plurality-style voting with IRV
(in Australian areas that
allow this as an IRV vote - this is called "plumping") actually are
HIGHER than the rates
of plurality-style range & approval voting in these two French
studies. So again,
you are exactly wrong, IRV actually makes plurality-style voting more,
not less, likely
than range and approval voting based on this data.

(Also, it would be nice if your Bucklin claims were backed up by actual data.
So far, I have not seen any such data from you.)

- Warren D. Smith (posted by Clay)

Instant Runoff Voting, Black Flight and San Francisco

The article says that San Francisco, the most well known IRV jurisdiction in the United States - has a "white power structure" and that African Americans are fleeing San Francisco in droves. The Mayor has ordered a "study" to find out how to retain the black population.

San Francisco hopes to reverse black flight
By John Ritter, USA TODAY
August 26, 2007

SAN FRANCISCO — Wayne Cooksey joined the flight of African-Americans from this city last year to escape soaring rents and buy a home. Michael Higgenbotham left six years ago for a safer neighborhood and better schools for his three children. Adell Adams retired and wanted to downsize but knew her home's equity wouldn't go far in a market where decent condos start at $500,000.

Aubrey Lewis was among the first to go, to nearby Oakland in 1977. "We left because of the housing situation," says Lewis, 77. "And that was early. It hasn't gotten much better."

African-Americans are abandoning this famously progressive city at a rate that has alarmed San Francisco officials, who vow to stop the exodus and develop a strategy to win blacks back to the city. In June, Mayor Gavin Newsom appointed a task force to study how to reverse decades of policies — and neglect — that black leaders say have fueled the flight.

Black flight can alter a city's character. "It's important for a city's future that it be a diverse place, and San Francisco is drifting toward being an upper-middle-class city," says Ed Blakely, director of Katrina recovery for New Orleans.

According to Census estimates, the number of blacks here shrank from 13.4% of the population in 1970 to just 6.5% in 2005 — the biggest percentage decline in any major American city.

San Francisco began holding IRV style elections in 2004.

Has IRV helped the community?

WHO is IRV helping?

How IRV has worked in San Francisco

I am not ready to support changing our system, but right now we have to face the very real threat that "FairVote" aka Rob Richie is posing to our elections here in North Carolina. Their goal is to spread this voting scheme to 10 counties in 2008, after "proving" it to be a success in Cary and Hendersonville.

Fair Vote is selling IRV as:

a way to help third parties - that appeals to third parties
a way to "avoid costly runoffs" - that appeals to some govt officials
a way to "avoid the spoiler effect" - that appeals to the fears of democrats

None of these claims are panning out.

How do I know this?

Because we can examine how IRV has worked out in San Francisco.
San Francisco has used IRV in municipal elections since 2004. They spent considerable money on upgrading their voting machines (about $1 million) and about $800,000 on voter education.

The current scandal plagued Mayor was elected through IRV, and beat the Green Candidate. A real election with real debate might have provided a win for the Green Candidate.

Now scandal plagued Mayor Gavin - the guy who boinked his campaign manager's wife? and who has a publicly exposed and serious drinking problem? and a city that appears to be going down-hill? has no serious contenders in the upcoming mayoral race.

see this page on IRV at NCVoter and you will also find links to show that IRV doesn't prevent the spoiler effect.

IRV doesnt help third parties, it strengthens 2 party monopoly

Look at what IRV is doing for San Francisco! The scandal plagued Mayor of San Francisco is a shoe-in:

SAN FRANCISCO
"Newsom, a sure bet for 2nd term, faces eclectic mix of challengers"
Saturday, August 11, 2007

...Newsom's challengers include a performer whose nickname is Chicken John; a blogger, Josh Wolf, who spent seven months in prison for refusing to give authorities video outtakes of a protest; and Grasshopper Kaplan, a homeless taxi driver who sleeps in his cab.

The biggest name among the hopefuls is former Supervisor Tony Hall, known for his conservative politics and disdain for the mayor, which in part stems from the fact that the mayor appointed him to run Treasure Island and then had Hall fired 15 months later...But while Newsom has his critics on the Board of Supervisors and inside City Hall circles, his approval ratings remain high among voters, even in the wake of a sex scandal and his admission of a drinking problem.

...In addition to Hall, the chicken, wolf and grasshopper, a nudist activist and the owner of the Power Exchange sex club also filed papers to run against Newsom, bringing the total number of challengers to 13.

Newsom's chief political strategist, Eric Jaye, said Friday that the lack of a viable threat in the race means the mayor, who already has raised more than $1.6 million in contributions, will not necessarily mount a campaign against anyone, but rather for the message he is trying to send....

Update, the one candidate considered to have any decent chance of winning, Tony Hall - has dropped out of the race even after qualifying for matching funds.

Tony Hall quits race for S.F. mayor, calling Newsom too entrenched Chronicle Staff Writer Friday, August 31, 2007
Tony Hall - the best-known candidate challenging San Francisco Mayor Gavin Newsom for reelection - is pulling out of the race, citing a lack of opposition in the city to a second term for the mayor.

Hall, a former San Francisco supervisor, made that announcement on ABC 7 Thursday, two days after he qualified to receive $50,000 in public funds for the race. Hall said he could not generate support among big-name donors.

That just leaves the Chicken Johns, and San Francisco with its same problems.

Its certainly hard to justify the nearly $800,000 spent per year on voter education and 600 public IRV info meetings a year so that people can re-elect a scandal plagued incumbent member of the 2 party monopoly.

www.ncvoter.net

Pages