Slashdot Mirror
ICANN At-Large Results
troyboy writes: "The ICANN election results are up. Go here for the complete results. The North American election went down to two candidates before there was a majority winner: Karl Auerbach." Only for the North American election was the instant run-off system needed; for all other regions there was an immediate majority. Viewing the election results in practice is fascinating and I can't help but wonder how much the U.S. elections would improve if we used a similar system.
A quote from Auerbach's candidate webpage:
"My candidacy is one that is founded on the belief that the Internet should not be controlled and dominated by those who aspire to nothing higher than mass marketing. It is my position that individual people ought to have a major voice in the governance of this revolution we call the Internet."
See also Auerbach's platform.
After a double-homicide in the quiet suburb of Oak Park, Illinois both residents of the Harry S. Adams House designed by Frank Lloyd Wright, were found dead. Police are now suspecting both ICANN candidates Emerson Tiller and Barbara Simons as the murderers. Motive has not been established yet, but these ICANN candidates are still at large and presumably heading towards a trade-show in Indianapolis.
What is interesting about Europe is that we elected a radical democratic with many many more votes before the first commercial representative (from German Telekom, nearly 6000 vs. 990). Andy Müller-Maghun may indeed have a chance to influence ICANNs work in a positive, read: non-restrictive, for free flow of information, way. Let's see if "they" (read: commercial intrests) give him a chance and if he has the power to meet theese high expectations.
To repeat:
* The majority of the public prefers Hsub to Notnilc (55% to 45%), in BOTH cases.
* Whether or not Torep runs has no effect on the actual desires of the voters with regard to the question of whether Hsub is preferable to Notnilc. They ALWAYS prefer Hsub to Notnilc.
* As stated, in BOTH cases Hsub is preferred to Notnilc; however, in only ONE case does Hsub win.
Let's think of it another way. Let's pretend 1% of the US are Nazis, and 99% are Jeffersonians. Let's say someone gets ahold of Thomas Jefferson's DNA, and someone else gets ahold of Hitler's DNA. Jefferson is cloned 5000 times, Hitler is cloned only once. In the next election, all 5000 Jeffersons run against Hitler. Who wins? Let's say that right after Hitler wins, he is challenged to a 1-on-1 election by Jefferson #536. Who wins that?
__
Despite the trolls, Instant Runoff Voting (IRV, aka Preferential Voting) is generally considered far superior to plurality winner-take-all. In the US, there is no constitutional problem with IRV (in fact, it has been endorsed in a consenting opinion by the Supreme Court, and by figures as diverse as Ralph Nader and Rush Limbaugh).
Due to a recent supreme court decision which invalidated blanket primary systems, the state of WA is revamping its primary system. Instant runoffs, due to their ability to collapse multiple virtual runoffs into a single round, are a cheaper replacement for primaries. It is very possible that the many groups interested in better democracy in WA (the Grange, the League of Women Voters, and the minor parties) will use the citizen interest in this issue to run IRV as a state initiative. If you're interested in this issue, contact me via email (it's not even obscured; leave the cookie in, please, even though I can handle it with or without).
ObOnTopic:
Anyone interested in voting systems should know about Arrow's Theorem, which states that there are no "perfect" ; voting systems. The only way to have a group of people preferentially rank a group of options so
1) new options will fit neatly into the ranking without mixing things up;
2) if everybody agrees on a ranking that ranking is chosen;
and 3) new voters who prefer A to B never cause B to win over A
is to have a dictatorship (ignore all voters except one). My personal choice of "ideal" system is to elect executives via borda selection among the condorcet-winning group. And then a house selected by proportional representation and a senate by approval voting. Hey, a boy can dream.
Preferential Voting: easy as 1-2-3
I was pleasantly surprised that the results basically mirrored the candidate ordering from the 'Slashdot slate' that appeared on Thursday. I wonder what percentage of the at-large members are slashdot readers. A fairly good amount from the looks of it. The total number of votes cast was also surprisingly low - anyone know how many at-large members there are?
Actually Barbara Simons quoted Karl when she was discussing the issue of TLDs and agreed with him. She was my second choice. Lessig was third.
It's not enough to bash in heads, you've got to bash in minds. - Captain Hammer
For those who think that the US elections should change to this method, it doesn't really matter very much because we operate under a two party system for the most part (with some notable exceptions) and different voting methods only take effect with more than two candidates
It would matter quite a bit because people wouldn't fear throwing their vote away on a third party candidate. That would actually give us a shot at electing a non-republicrat for a change rather than voting for a third-party candidate, thus helping to ensure that the candidate that you want the least gets into office. Democrats fear this effect and many third-party supporters vote democrat rather than green or reform or libertarian, etc, just to try to make sure a republican doesn't get in because the democratic votes got siphoned off by the third party.
It's not enough to bash in heads, you've got to bash in minds. - Captain Hammer
In order of greatest eligible, to least:
Asia/Australia/Pacifica: 38,246 eligible, 17,745 voted, or 46%
Europe: 23,442 eligible, 11,309 voted, 48.2%
North America: 10,632 eligible, 3,449 voted, 32.4%
Latin America/Carribean: 3,548 eligible, 1,402 voted, 39.5%
Africa: 315 eligible, 130 voted, 41%
So, in no case could any of this be said to be representative of a majority.
And, once again, the Seat of Democracy® can't even get a third of eligible voters to vote. Are we bored with democracy, or what?
t_t_b
--
I think not; therefore I ain't®
I'm on PJ's "enemies" list! Are you?
Well now.
The US has a reasonably well-deregulated telecommunications and internet business. However, the second-biggest Internet market, in Europe, is still dominated by former state monopolies, lack of local flat rate phone fees, badly-developed net infrastructures, and general exploitation of customers.
And into this environment, Andy Mueller-Maguhn was elected 'Region 2 Director for Europe'. Yes, Andy Mueller-Maguhn. The former head of the German Chaos Computer Club, a team regularly posting information about cracking phone cards, cable TV decoders, and computer security information. They're the German '2600' equivalent, reviled and detested by state institutions, telecomms companies, etc. They're the rebels of the underground.
And now their former head is the Director of the European division of ICANN. Imagine Eric Corley (Emmanuel Goldstein) being the head of the US ICANN. Now imagine what important aspect this brings to ICANN in Europe.
Or, to quote Homer Simpson: Woohooo!
Alex T-B
St Andrews
I'm not sure what that's an example of. At the page linked, they have a situation where a slight plurality prefers milk, and a majority prefers one of two alcoholic beverages, with their vote split between the two. Under the instant run-off, one of the alcoholic beverages wins rather than the milk. So? That's a not a paradox, that's the system working properly!
A clearer example: let's say that there are two liberal candidates and one conservative candidate in an election. And let's say that the population is 60% liberal and 40% conservative. The "correct" result with a 60/40 population is that they should elect a liberal candidate to represent them, right? Well, with the U.S.'s current system, they might get a conservative: vote ends up 30%/30%/40%, and the conservative wins. Under the instant run-off, they'll get one of the liberal candidates, the "correct" choice - in other words they aren't penalized for having more than one candidate, and they can express a preference between the two without fear.
--
Michael Sims-michael at slashdot.org
Quoth Michael:
Yeah, good thinking there, Michael! The U.S. elections would be much better if they were conducted like ICANN elections. I'd love it if the majority of members of Congress were chosen by an unrepresentative group rather than by public election, and the only voice we got was one token member from each region of the country. It'd be even better if there were completely unreasonable barriers to participation in the process, and election "rules" that change at a moment's notice based on whether or not said unrepresentative group feels that an election might go its way or not. That sounds GREAT!
In all seriousness, it's good to see that at least some people at ICANN will have a clue now. I just hope that they move towards greater transparency and participation so that we don't see a repeat of this year's election fiasco ever again.
Read my blog.
Perhaps more significant than the fact that Karl won is that Harris Miller got the lowest votes in the North American at-large contest--Miller was the only one of those candidates who supported strong intellectual property, commercial, interests in ICANN. The other candidates, including Karl, all expressed more or less the same ideas--that ICANN should restrict itself to technical rules, not policy that is intended to enforce legal matters such as trademarks in domain names.
But all of the candidates did remark that commercial interests still dominate the entire ICANN board. So we are now counting on Karl to to mobilize us users against those commercial interests when they seek to use ICANN wrongly. (Karl does note he works for Cisco, so we ought to hold him to his effort to be independent of even his boss.)
For that reason Slashdot ought to allow Karl some space here to solicit feedback from users and let us know what is happening in ICANN. Good idea, Wellspring!
Perhaps the most interesting thing about the run-off is that it didn't matter. Contestants were eliminated in the same order as the votes in the first round. It could have been a plurality vote and it wouldn't have mattered.
I like how North America Got A Cryptic Chart
chart
while everyone else got gimpy ones.
I feel sorry for the Popov guy, sorry he didn't get that many votes cause i sure love his cheap vodka.
Be happy. The country that invented the internet could also ruin it. Be glad you did not get some kind of MicroSoft stooge that would fragment the largest part of the net.
Friends don't help friends install M$ junk.
Just fine 'em if they don't vote like we do in Australia!
Voting is a civic duty, just like driving responsibly, parking sensibly, spitting carefully - and if you don't do these, you get fined.
Have a system (again like Australia) where legitimate excuses are clearly stated and easily verified, it costs you very little to administer, and the effect is that basically it's not worth your time not to vote, so you do...
But for the rest of us, here's the other winners and their profile pages:
Masanobu Katoh - Asia/Australia/Pacific
Nii Quaynor - Africa
Andy Mueller-Maguhn - Europe
Ivan Moura Campos - Latin America and Caribbean
Boffoonery - downloadable Comedy Benefit for Bletchley Park
Shit, she seems a bit cool. And good at what she does
Ivan was/is one of the most influent government officials in Brazil, specially in Computer/Internet/Telecom related issues. He was also a important player in defining the structure of the Internet in Brazil. He is indeed good at what he does, and was by a large margin the best canditate for the Latin America zone.
Be sure to read Curtis Ganz' testimony before you make up your mind that direct elections are more desirable. If you feel that minority opinions are too easily ignored now, or that the "tyranny of the majority" is something genuinely worth worrying about (as did the Founders when they wrote the Bill of Rights), direct winner-take-all elections aren't the way to go. They sound good, because simple, but in this case they are (to quote Mencken), the solution that's "simple, neat, and wrong."
See article bottom of page, America Ready for Third Ketchup Brand . How about that?
Friends don't help friends install M$ junk.
and presumably abulafia is a troll.
Boffoonery - downloadable Comedy Benefit for Bletchley Park
You do realize that by using Hitler in an argument, you automatically kill the thread and lose, right? That's USENET law!
Anyway, that aside, it doesn't seem like you actually read my post, or maybe just the top part of it. In your first example, the primary election is missing. That primary would enable us to bring it down to two candidates. Occasionally, you do get a case where a third candidate appears. Then the party closer to their view creates a compromise/consensus position which lures away their supporters while minimizing the loss of moderates. Or if the new faction represents a middle ground between the two parties, they compete for the centrist vote while trying to hold their radical wings.
So let's reverse the names again on your candidates. Perot, Bush, Clinton. What happened to the reform party? Well Perot stood really for three things:
1) generic moderation (the parties are too different, and can't work together enough to get things done.)
2) balanced budget
3) protectionism (ie no free trade)
Each party moderated; just look at the debate last night to see how much closer the two parties are. To the point that we have Ralph Nader and Jon Katz claiming that the parties are too similar now. The budget got balanced (kudos to programers and the Web for giving us the economy to make it possible.) And protectionism? Well, it just doesn't have much public support anymore. People realize that trade helps the poor in other countries while helping us here in the US. The few protectionists who are left either support Pat Buchanan or Ralph Nader.
Ralph Nader probably won't do that well in this election. But if he does, the democrats will swing back to the radical left- endorsing radical environmental policies, more government, higher taxes and isolationist trade policies. So your example is actually a great one which, when its real world equivalent is looked at, has worked perfectly.
I had to think for a minute before I responded, because you do have a point. However, the idea isn't just a matter of electing compromise viewpoints. Picking something which eventually will be the consensus is different from picking a candidate who has been pushing for that consensus in an election. This is a subtle difference, and I wonder if I am phrasing it correctly now.
Differences in opinion can exist on (at least) three levels:
Priority: This issue is important, while that issue can wait. Example: the Farm Bill will be coming up for another vote in 2002. Losing sleep over that? Probably about as much as a farmer did over the Clipper Chip.
Extremity: In other words, we may both agree that taxes are needed, but I may want rate X%, while you want rate Y%. The correct compromise might not be in the middle! Some approaches are each at a local minimum, and the compromise is worse than either one.
Approach: Republicans tend to want to solve problems by allowing free markets to produce the solutions. Democrats want to solve things by the passage of laws or use of public funds. So, when faced with a broken education system, Democrats want to strengthen the US Dept. of Education (a federal office which regulates school systems) and spend more money on public school. They feel that the government system isn't properly designed, or isn't sufficiently supported. Republicans think that the system itself is the problem-- it has no market forces to back it up and therefore doesn't self-optimize. Republicans prefer giving parents choice (depending on the person, either between public schools, or with vouchers, among all public and private schools). There isn't a compromise between these outlooks-- the solutions are mutually exclusive, and result from their differing model for just what the problem is. Each believes the that other will only make the problem even worse.
Priority, Intensity, Approach: Synthesizing these views is not so much a statistical process-- it is a social and intellectual process. If we had a fascist with 40% of the vote, and a Republican and a Democrat with 30% each (remaining are undecided), then the Republican and Democrat must compromise (possibly with the fascist, more likely with one another). But the differing outlooks mean that there is a problem with regard to Approach. Priorities and Extremeties are easier to average out-- I get some priorities, and you get some, too; we split the difference on extremeties.
But what about Approach? The trick there is to frame the issues to one's supporters in a way which makes the similarities stronger than the difference. To keep your fans' eyes on the ball, so to speak (after all, the more important difference is between either party and the fascist). This is a social trick which a mere mathematical ranking doesn't cover. More importantly, rating the Republican a '#1', the Democrat a '#2', and the Fascist '#3' forces the voter to express something they don't feel: that either candidate is infinitely preferable to the fascist!
Anyway, the point is that political leadership is needed to bring large groups to endorse compromise positions. Preferential voting systems attempt to automate the process of consensus and compromise, rather than making it something which everyone is a part of.
I'm not sure if I phrased all of this correctly, because as I said, you did make me think about it. But I hope you understand where I'm getting at with this.
I think this ought to be interesting. Auerbach has been very critical of the ICANN since it's inception. He has some good ideas about how to make the domain system more democratic (like having one or more TLD where trademark holders do not get first pick). It will be good to have someone on the board who represents something besides the corporate interest.
Not only that, but Karl was one of only three member-nominated candidates on the North American ballot. The other four candidates were nominated by the existing memebers of the ICANN board. I'm glad that people decided to thumb their noses at the board's attempt to stack the election with people friendly to their agendas. It ought to send them at least a small message (not that they will care).
Speaking of the U.S. elections (which the article mentioned, please don't mod me down for being off-topic), here is a recent article at policy.com that discusses getting rid of the Electoral College and replacing it with a direct popular vote. It discusses historical reasons for the creation of the current system, and provides some interesting links at the bottom.
Free Hans!
Region's Total votes 130
% of World Total 00.4%
Votes for ICANN Nominated Candidates 100 ( 88%)
Latin America
Region's Total votes: 1,402
% of World Total: 04.1%
Votes for ICANN Nominated Candidates 1,166 (83%)
North America*
Region's Total votes: 3,449
% of World Total: 10.1%
Votes for ICANN Nominated Candidates 1,114 (32%)
* - N.A. counts 1st choice votes only. Note also that if one counts the three candidates who expressed the most doubts about ICANN (Lessig, Simons, Auerbach), they got about 75% of first round choices.
Europe
Region's Total votes: 11,309
% of World Total 33.2%
Votes for ICANN Nominated Candidates 3,066 (27%)
Asia
Region's Total votes: 17,745
% of World Total 52.1%
Votes for ICANN Nominated Candidates 16,996 (95%)
Comments:
Participation rates did not correlate well with what I'd guess numbers of hosts or estimated numbers of users are, except arguably at the low end.
Five regions fell into three groups:
I have a blog.
As one might say, that happened, but you can't prove it happens every time nor whether it will happen again. There is therefore no justification for the equivalence.
:)
What you should observe, however, is that in every stage but the 3rd, Barbara Simons attracted more 2nd, 3rd, etc place votes than Karl Auerbach. He still won on the strength of his no.1 ballot. One could draw the conjecture that Simons had broader but weaker support than Auerbach.
This is obviously not a true run-off system in that a recast with the additional information (who won, and by what margin). Consider political conventions in most nations (no, not the US). It's truly rare for a candidate to win on the first ballot, and significant horse-trading does take place between ballots. This horse-trading and positioning is a valuable feature of the run-off election in politics. I'm not sure what difference it would make for the ICANN Member-at-large, though.
--
--
There is no premature anti-fascism. -Ernest Hemingway
This is a little roundabout, but it gets where its going. So please bear with me.
Labels like 'liberal' and 'conservative' are deceptive. I have seen many who share the same brand name argue viciously over their beliefs, while (and the VP debate last week is a good example of this) people of different political traditions can have similar beliefs. So what is the practical result? Your example was this:
In this example, let's remove the labels. There are three candidates. One is supported by 40% of the people, and the other two by 30%. Who should be elected? If none of the three can reconcile their differences to create a majority voting block, the 40%!
Is removing the labels ok? It feels wrong, because we remember the ideologies in the first example, and want to somehow account for the two liberal candidates wearing the same political label. But from an outside viewpoint, it doesn't matter what the particulars of their policies are. The point of the election is to generate consensus whereever possible, minority rights wherever threatened, and majority rule whenever it is needed. If the two 30% candidates can't agree to a common viewpoints, then it doesn't matter what their label is-- they are different political groups. A minor difference is only minor in the eyes of the beholder.
Ultimately, after the election, whoever didn't vote for the winner will have to live with the results of the election. A working majority will have to be formed in the Congress/Knesset/Diet/Parliament. Otherwise, the government would collapse. Once he wins office, the leader must forge a working majority. That means compromising and working with his opponents-- and where that fails, building a majority.
In multiparty systems, the voter sees plenty of labels and badges, but ultimately, once they win, these parties are able to go into a back room and decide what is critical and where they can compromise. In other words, you are rewarded for the fanaticism of your supporters (luring them away from other parties) and then quietly leave the consensus-building to be done in secrecy.
In a two party system, here's what happens in your example: 40% guy wins. But in the next election, the two 30% candidates decide they have more in common with each other, and ONE runs, fielding a selection of issues which represents the best of both (and, often, compromises between the two). Then the 40% candidate loses, until HE can find a enough voters to compromise with that he can build a faction of more than 50%. In this case, leaders compromise BEFORE the election, and their supporters can each judge the quality of their compromise before voting. That's why the two major parties have primaries! Each major party is made up of a thousand minor factions, and the faction which can forge the best representation of the party as a whole gets to be the party's candidate. Each party races for 50% in the general election. And if one faction decides to try to railroad all the others, it will get outrun by its competitors.
Both two party and three+ party systems have this feature of compromise and consensus-building. The difference is that a two party system has this process before the election, and it is conducted in public. In a parliamentary system, it is conducted after the election, and out of the prying eyes of the public.
For something small like ICANN, there is too much overhead for a two party system to be worth it (though the trademark issue would seem to be creating two factions anyway). The ranking system is appropriate for the size and influence of the organization. But when it comes to a whole nation, with a military and police, you need a fairly extensive process.
Two party systems are simple-- if you aren't too involved and only vote at the general election. But if you aren't happy with the nominees from your party (or the party that you are closest to), then just vote in the primaries for the guy who better represents your views. It takes some research, but you are rewarded by a party which is more like you. If you still aren't happy (tough customer!), then volunteer for the candidate who best represents your views. Rough estimates: A campaign for the House takes about 25 people. A campaign to be Senator or Governor (depending on your state's size and population, of course) takes about 25-100 people. One volunteer, even at these heights, makes a huge difference-- especially if you have organizational, technical or writing skills.
Or better yet, help a county commissioner/freeholder or state senator/representative. You are bound to be one of maybe three or four people on the campaign! If you work hard, you are almost certain to win-- and if there isn't any candidates you like, run yourself! It isn't that hard, and it is tons of fun.
I am very happy that Karl won. One thing we need to remember, though, is that he is only one voice in the board. We all need to be ready to help him if he is going to get stuff done on the board.
Rob, Jeff: what are the odds of getting an interview with him? He was basically /.'s pick!
I read through Karl's policy statement on his website and a couple things made me wonder if he understands the full set of tradeoffs involved.
He did a fine job of explaining why the technical stability of the name system is not a significant problem that should inhibit change, but I kept wondering if by "stability", the ICANN traditionalists were concerned with non-technical stability issues that were at stake, such as stability of brand name and trademark recognition online. Isn't that the heart of the issue? Is anyone really arguing about name system uptime?
I think the ability of big companies to squash any use of any particular brandname or trademark they come up with should be limited, but I don't think that companies and their trademarks, whether "McDonalds" or "Slashdot," should be without protection from looter/moocher types who attempt to register recognized names under other global TLDs, whether ICANN adds 20 more or 20,000 more. I know that aspects of this problem are dealt with in current ICANN policies, albeit not fully satisfactorily, but in my limited examination of Karl's writings, I haven't seen Karl articulate that he either recognizes this issue or has a balanced solution in mind. Corrections or pointers welcome.
--LP
(This is not an attack on Karl. This is an attempt to point out an issue of concern and ask for informational pointers or responses.)
Not to mention the fact that the VAST majority of Americans do not have access to an Internet capable computer. These may be commonplace in our (geek) houses; however, we must remember that computers are still a luxury, not a right.
Do you realize that wars have been started against the will of majorities because anti-war candidates have "split the vote"? In such an instance, would you consider this as having "worked perfectly" if in the next election the vote-splitting was remedied by convergence of the vote-splitters??
So what? It is apparent (at least not unreasonable) that the majority of Americans prefer McCain to both Bush and to Gore; the fact that, in spite of this, McCain could not become a contender for the presidency, is an outrage and an affront to democracy.Honestly, I can't even see why you would consider the above listed effect good. Sure, it means that third parties can influence the major parties by drawing away votes. Who cares? That influence is almost always contrary to the will of the public -- it certainly would be in the Nader example you list. It means better representing the views of Democrats at the expense of representing the views of Republicans; what we want is a comprimise between the two groups. If Naders started replacing Gores in the Democratic Party, that would have the potential of even more seriously dissatisfying all Republicans, while appealing more only to a group of Democrats much smaller in comparison. Plurality in general polarizes elections, guaranteeing that half of the populus will be seriously dissatisfied, rather than Condorcet which tends to produce a populus that is all somewhat dissatisfied, but only in those areas where the individual's views are unpopular.
__