Voting Patterns
Voting PatternsNote that nowadays there's Electorama for you: http://wiki.electorama.com/wiki/Main_Page
Herein is described some patterns we democratic citizens use for voting. They are part of a larger set of ConsensusPatterns? which in turn is part of a larger set of SocietyPatterns?. But, right now, let's focus on patterns of voting.
Anything to do with voting is fair game. From the voting system, to how to choose a vote, to how the results are interpreted, to whatever.
It would be nice to avoid cultural assumptions. AmericanCulturalAssumption(s), or otherwise. In order to do this, lots of people must contribute! Lots and lots! This means you.
The stupid thing about democracy is the assumption that the majority knows better what is good for me than I know.
Not quite. The stupid thing about democracy is the assumption that you know what is good for you.
When the whole world becomes democratic, the government will be a Chinese-Indian coalition. I doubt they know better what is good for me than I do.
Did I miss some point? Who said (or defined), that voting would be democratic? I think it is democratic only under the precondition, that there already is consensus to use and accept voting for decision finding.
Democracy does not work that way. The majority makes their decisions. No matter what you like or agree with.
Which is why there are no existing democracies, under any sort of 'pure' definition.
Choosing your vote
Was this an IsaacAsimov story? A computer chooses randomly a person. Then, that person was analyzed by the computer and then the computer was able to say who was elected. How would this VotingPatterns be named? -- JeanMarcHeneman
Yes, as I recall, it was titled, "Franchise". -- RussellGold
So, what about a FranchiseVote? VotingPatterns. Let's have that modification: the person is still chosen randomly but the person votes directly. No need for analysis. -- JeanMarcHeneman (time for refactoring?)
Why bother? If you are going to pick voter randomly, just pick the winners randomly. The premise in "Franchise" was that the computers already could predict who would win if the elections were held normally, so the actual voting was superfluous. The selected person just supplied some missing data so that the computers could make their predictions. The franchise voter did NOT get to pick the winners. -- RussellGold
You are right (I often forget simplicity... but the franchise voter may be taken as a scapegoat by others in the population so the computer may less be blamed). Let's refactor once you've read this... By just a comment on this interesting way of voting. And kill all other comments. -- JeanMarcHeneman
This sounds a bit like my thought experiment with a voting system: You hold the election as usual, but instead of counting up all the votes, you just pick one of the votes from a big (really big!) hat. The effect of this is that the most preferred candidate is likely to win, but nothing is sure. Also, you don't have to do a BlockingVote? (that is, you would really vote for A, but the race is between B and G, and you certainly don't want B to win, so you vote for G.) Just my $2E-2. -- JohannesBrodwall
Just came across this discussion... back in 1996, there was an article published in The Perl Journal on the topic of voting systems. One of the points the article made was that research had shown that elections are always unfair :o) . To be precise, the article listed just 4 principles which you would expect from a fair voting system, and referred to a piece of research which proved that the goals were inconsistent. So, you must always compromise. The article is online here: http://www.itknowledge.com/tpj/issues/vol1_3/tpj0103-0002.html - unfortunately, you need to be a subscriber to read it. I was back in '96 but I'm not now - otherwise I'd refer to the original research. Anyone got a copy to hand? -- BrianEwins
If that particular article is hard to come by, it may help to know that the "no fair elections" thing is officially called "Arrow's Theorem"; it was proved by Kenneth Arrow, who IIRC won the Nobel Prize in economics for his work on voting systems.
See http://condorcet.org/rp/arrow.shtml for a no-subscription-necessary overview of ArrowsTheorem.
Check out a review of voting systems at http://www.scientificamerican.com/askexpert/math/math2.html. -- MikeHo
Surprising that the authors ended up favouring a) Borda and b) Approval voting, both of which suffer from similar problems related to tactical voting. Because both involve assigning a score to a person's vote for each candidate, and totalling these scores, they cause votes cast for second and third choice candidates to count against the first choice.
In contrast, the SingleTransferableVote never does this, and DeCondorcet does it only rarely. (The first author judged the fairness of systems by CondorcetsCriterion?, which is why the choice of Borda was surprising.) Academic opponents of SingleTransferableVote usually point to the MonotonicityCriterion.
SingleTransferableVote just seems bizarre, and DeCondorcet is thoroughly opaque to most voters. ApprovalVoting is still the best, simply because it is better than InstantRunoff? and the current system for satisfying the conditions, and is understandable to a layman. Condorcet would be rejected by a populace as voodoo - it's just too complex. -- MartinZarate
See also YouCantForceVoting, TheTroubleWithConsensus, ScalableSyncracy, PatternsForEffectiveMeetings
EditText of this page
(last edited September 16, 2014)
or FindPage with title or text search