A “perfect” voting system

http://www.drmaciver.com/2011/04/a-perfect-voting-system/

by cavedave • 15 years ago

158 146 15 years ago

TamDenholm• 15 years ago

I find CGP Grey does an excellent job of explaining things very simply in his youtube videos. Personally, i'd like to see these run on British TV.

http://blog.cgpgrey.com/the-politics-in-the-animal-kingdom/

yannickmahe• 15 years ago

I was going to show that link as well. His explanation of instant runoff convinced me.

DennisP2• 15 years ago

Instant runoff has some mathematical problems. Ranking your favorite candidate highly can actually work against you. http://www.rangevoting.org/IrvPathologySurvey.html

Empirically, every country with instant runoff still has two-party dominance.

A better system is range voting: http://www.rangevoting.org/

For a really great book on different voting systems and their advantages and disadvantages, see Gaming the Vote by William Poundstone.

MetalAndWheels• 15 years ago

Thanks for that. I'm the creator of those videos and I'm really glad to know that the AV one actually changed minds.

DRMacIver16• 15 years ago

Thanks. I've only just got around to watching these, and they're really very good.

StavrosK2• 15 years ago

For some reason, I like this. However, I think that a voting system that 95% of the population can't even understand won't be very popular overall...

DRMacIver16• 15 years ago

I don't think it's that hard to understand. Certainly as a voter it's very easy to understand what it is you should do - easier than under FPTP where you are forced to think tactically. The consequences are slightly hard to think through if you're not good at probability, but it's easy to boil those down into a set of pithy statements that will tell you most of what you need to know.

greendestiny2• 15 years ago

It's not hard to understand how it works. It is harder to understand how it's fair.

Of all the voting systems that will never see the light of day, I think I like this one most of all. However I would say despite its similarity to the status quo it would completely change the nature of government when parties couldn't count on who would get elected despite the safeness of their constituency.

limmeau5• 15 years ago

It is rather simple, if you compare it to e.g. the D'Hondt method used in Germany.

arethuza8• 15 years ago

Discussing this with colleagues someone suggested a splendid addition - give the person whose vote was used in a constituency a nice bonus (say £100,000) to encourage people to come out and vote.

[Edit - think how interesting election nights would become rather than the usual few hours of predictable tedium that I always feel compelled to watch.]

cavedave• 15 years ago

Or they get all the deposits of those who lost their deposits from not getting enough votes?

jan_g• 15 years ago

You generally do not want to know who voted for which candidate. E.g. voting should be anonymous.

arethuza8• 15 years ago

I'm pretty sure voting isn't 100% anonymous in the UK - it isn't public but don't the ballot papers have a unique serial number on them?

As for collecting the cash prize - we could make people select if they want it anonymously or not (like the lottery).

gyom2• 15 years ago

Bitcoins ! :P

billpg• 15 years ago

Yes! Cancel the referendum and replace it with one that has this option as well!

limmeau5• 15 years ago

This needs random numbers whose randomness can be demonstrated.

You could ask every candidate to commit to his random number, publish a hash of it, and afterwards compute a hash of all candidates' random numbers to get the final random number. As long as most candidates distrust each other, this should work.

gyom2• 15 years ago

And then you need to explain how hash functions work to a bunch of people who dropped out of high school to work on the farm so that they don't feel cheated by the new electoral system.

In the long run, though, it shouldn't be so hard to get people to accept these kind of solutions.

arethuza8• 15 years ago

Think of the prestige this would bring to CS... :-)

limmeau5• 15 years ago

We could use sealed envelopes made of heavy paper instead of the first hash function usage, and then have a guy in a lab coat point at a colourful picture of a fractal to demonstrate the randomness of the second hashing step.

radu_floricica• 15 years ago

There is less feedback this way. Somebody could find a way to game the system, and all the statistics in the world couldn't prove he cheated.

Plus, randomness is a lot more random and a lot less uniform then people usually give it credit for.

bluedanieru• 15 years ago

You'll still know how many people voted for each candidate in this system, but instead of that determining who wins, you pick a random vote from the pool.

eli• 15 years ago

I think it fails the "easily understandable by the average voter" test. And that's kinda a big one if you actually want change.

d4nt• 15 years ago

I love it. We Brits would never go for it though. I can see the Daily Mail headlines now: "Fascists Given Seats in 'Election Lottery'".

arethuza8• 15 years ago

Surely the Daily Mail would be delighted with such a result?

estel• 15 years ago

No, their particular brand of hateful racism is only acceptable if they can portray as despicable another group even more hatefully racist than them.

tobylane• 15 years ago

I'm voting for AV because I want a chance to vote for PR. Proportional Representation is the proper/good/final solution.

I'm slightly confused about voting multiple people for a larger area. Is this PR, or something else good?

It will always be a boring night, even more so if there are coalitions formed. More election night comedy, and transparent coalition agreements.

DRMacIver16• 15 years ago

> I'm slightly confused about voting multiple people for a larger area. Is this PR, or something else good?

It may or may not be PR. It depends on how you elect those multiple people: They may be elected under a PR system, or they may be elected under something non-proportional.

I'm for PR in general. I'm not sure how I'd feel about a PR-per-constituency approach.

perlgeek4• 15 years ago

Random numbers in election systems are an absolute no-go.

Suppose that a court rules that one district must be re-counted, and based on the result there will be a new, random candidate.

This creates incentives for going after re-countings if your political opponent has won the lottery, not because the result will be much different, but simply because you get a second chance in the lottery.

It boils down to the thought that repeatability is import for fairness. Somebody claims that your DNA is identical to the one collected from the crime scene? You can make a second test and verify the first one. You think you have been treated unfair in court? You can repeated the trial at a higher court. All the doping tests in sports are verified by repetition.

The list is nearly endless. Even if a random experiment does treat you fair, you can't known.

The second point is that randomness only creates the appearance of fairness on the large scale, single individual still get treated unfair.

dabeeeenster2• 15 years ago

How can you have a recount of a single vote?

JoachimSchipper9• 15 years ago

Imagine that some civil servants hates, for instance, the BNP. If the proposed system is implemented carelessly, (s)he may elect to "lose" the randomly-chosen vote that would otherwise put a BNP politician in power, and hope that the "next" vote will be for a party more to his/her liking.

"Recounts" are nonsense, but I can imagine that an audit might help against the above scenario.

JoachimSchipper9• 15 years ago

As stated elsewhere in the comments, you can make a number of people (e.g. anyone who asks[1]) commit to some number[2] (e.g. by giving out SHA1(their_number)). Then calculate a random number from that (conceptually, addition modulo number of voters; actual implementation left as an exercise to the reader).

If the vote with that number was actually cast, that is the vote that is actually used. If not, try the next one until one actually was cast. Pretty repeatable, and recounting is easier.

[1] Note that "all candidates" may not be sufficient, since it allows the candidates to collectively choose the outcome by picking their numbers such that one of their own/friends' ballots is chosen. This could be used to elect a popular Labour politician in a remote district where only the major two parties are running, in exchange for electing two unknown Tories.

[2] If you're not into crypto, put tickets containing every possible ballot number in a big barrel. Have a trusted/blindfolded guy pick tickets until he finds one that corresponds to a vote that was actually cast. There are lots of solutions, of varying degrees of cleverness.

[EDITs: added two footnotes]

perlgeek4• 15 years ago

This does account for the repeatability, but the point still remains that it treats individuals unfair, and that the unfairness only averages out over many districts.

How would you feel if you got 80% of the votes in your district, but the lottery (even if repeatable) picked your opponent? How would those feel who gave you your vote? I'd certainly felt cheated.

lsd5you• 15 years ago

I'm not sure that this is enough as it is stated here and in the original comment. There may be a technical solution that allows it to be done in one go, but an alternative would be to do the count, have a window in which the count can be challenged and then do the draw in a pseudo random deterministic way with seeds coming from a number of sources (e.g. different candidates).

arethuza8• 15 years ago

Those of us who live in safe seats where we don't support the locally dominant party are arguably already being treated unfairly. It wouldn't be so bad if there weren't quite so many safe seats - but given that 382 out of 650 seats fall into this category I think something should change.

http://en.wikipedia.org/wiki/Safe_seat#United_Kingdom

DRMacIver16• 15 years ago

It's worth noting that this solution utterly blows the safe seats problem out of the water. This is probably my favourite feature of it.

yummyfajitas3• 15 years ago

For a perfect example of this, see the presidential election in 2000. The vote totals in Florida were within the margin of error, making the election a coin flip with the pretense of counting [1]. Politicians behaved exactly as you describe.

[1] At the time I was in college, and had just learned stats. I advocated just flipping a coin in full view on TV and moving on. Everyone thought I was crazy.

MartinCron• 15 years ago

In the last Washington State Governor's race, we had a similar condition, where the margin of victory was well within the margin of error. It seems that in such case, the only way to determine the actual will of the people is to do a re-vote.

Unfortunately, whatever candidate was ahead a any point in time would say "it's time to just move on".

I'm afraid that too many people don't get the concept of "margin of error" and will insist on getting a count that 100% accurate (instead of merely 99% or something).

nkurz2• 15 years ago

While appealingly efficient, I don't think this is sound practice. "In the margin of error" does not mean that it's a 50/50 tossup. But you're usually pretty accurate, so what am I missing?

mike-cardwell• 15 years ago

Just make it so recounts can only be done before the final random pick.

vrotaru• 15 years ago

Not that I endorse anything but a two-step procedure can enforce a kind of repeatibility.

  * Step 1. A thousand votes are selected randomly. And counted and the counts are made public
  * Step 2. The winner vote is selected at random in public procedure.

Now step 1 can be repeated at a contestant request, but if the counts are roughly the same, step 2 doesn't happen. Actually if the counts are too different, a revote may be in order

joe_the_user• 15 years ago

Obviously there couldn't be recounts but since there wouldn't be "outlier" elections (where someone would win by an unusual five votes), that wouldn't matter as much. (Of course, every election could be an outlier so this might not make it pleasing)

The result clearly wouldn't be fair in our normal sense of fairness and that would indeed be a huge roadblock.

The only appeal would come if a given system looks both unfair, unprincipled and corrupt. Then "just unfair" might be better.

BoppreH• 15 years ago

Electronic voting?

evgenit• 15 years ago

Here's a way to avoid it: Change the process. Do counting, re-counting, appeals, etc. first, until a final decision as to the correct count and valid votes is made (all appeals exhausted or no longer possible due to time elapsed). We have final decisions on all these matters in current systems. This way, you will "know" that the election was fair, since this depends on the ballots, not the candidate.

Then do the random step. The drawback is that there may be a month or two between casting votes and having the results, but I don't see why this is a big deal.

iwwr2• 15 years ago

The problem is not the allocation of power, but the centralization of power. Once you have a lot of power in the hands of the central government, it will be captured and misused by cliques and special interests.

The way to 'fix' democracy is to reduce its scope, to make it as localized as possible. In this way, political decisions tend to be made closer to those affected.

The idea that the various branches of government can block and check each other's power has proven quite misguided; they will always cooperate to increase their collective power.

guelo2• 15 years ago

I'm guessing you're american even though this story Is about a british voting system. So you want local military, local FAA, local social security, local CDC, local EPA, local interstate highways, local FDA, local FDIC, local NASA, local weather satellites... I could go on but it seems like every little city providing all these services might be a tad wasteful.

iwwr2• 15 years ago

Why would that be wasteful? Note that having local authority does not preclude having larger organizations in place.

Changing the voting system does not change the dynamics of political power.

DennisP2• 15 years ago

Or maybe he's from Switzerland, which actually does have a fairly decentralized system: http://www.bhutan-switzerland.org/pdf/Kaelin_Switzerland.pdf

lhnn2• 15 years ago

Did he list any of those as being what he wants localized? Did he say every "little city" should do it?

In principle, he's absolutely right. A weaker military, less intrusive FDA/EPA, no social security would be nice.

knowtheory2• 15 years ago

You are familiar with the Articles of Confederation, and their abandonment?

The Federalist-Anti-Federalist debate has been going on since the Revolutionary war. It is neither a settled debate, nor is it without precedent.

The branches of government do block and check each other's powers, and to assert otherwise is just false. The fact that the Federal government doesn't do what you want isn't a sign that the US system of government is broken.

The fact that the government can't pass policies is a sign that US politics is broken (although i will grant that each branch of government has been complicit in this individually). But that's hardly collusion.

judofyr• 15 years ago

A simple way to generate random numbers:

Let's assume we have N positions that we want to elect MPs for. Each ballot contains an ID and a secret number between 1 and N (both numbers are physically on the ballot, but you can only see it by scratching an area). Next you define a hash function that takes a vote and returns an integer. For each 1 to N, you simply take all the ballots with the secret number and sum the hash of the votes. You can then use that value you choose choose a ballot ID.

The good thing about this solution is that you can easily re-count the votes and you end up with the same answer. And, you can place all the information in public so other parties can verify it.

Hm... Any randomness problems in this solution?

perlgeek4• 15 years ago

Have fun scratching off the tissue from millions of votes.

Even if it's statistically sound, it's not practical, because it doesn't scale.

rudiger• 15 years ago

There may be issues with the above solution, but scratching off tickets isn't one of them. Do you know how inefficient the process of counting votes can be now? You might think that it is impractical, or that it "doesn't scale", but they manage.

yesbabyyes9• 15 years ago

This is a refreshing and interesting suggestion for a voting system, and it seems like David and others here are far more knowledgeable about voting systems than I am.

I have been thinking about a voting system for a while and I really need feedback from people who know this stuff - I have a feeling there's large flaws with it. Here's the general idea:

· Each voter can delegate his/her vote to any other voter, in any number of levels. I.e., if Alice delegates her vote to Bob, and Bob delegates his vote to Carol, Carol carries three votes.

· Delegation of votes can happen at any time, so Alice can delegate her vote to Bob until further notice.

· Each voter can change their delegate at any time.

· Each voter can vote for a motion. If the voter has previously delegated his/her vote, the delegate will not carry that voters voting power in this motion. So, if Alice uses her vote in motion 1, Bob only carries one vote and Carlos 2, but only in that motion.

· Each voter can delegate his/her vote to different voters in different issues; Bob may delegate his vote to Carlos in financial issues, and to Alice in justice issues.

My hope is that this would scale to e.g. a nation state, as an alternative direct democracy. Today in Sweden, we delegate our votes to 349 MPs every four years. When introducing my system, that would be the initial state. After that, voters would start delegating their vote to other people that they trust, while using their vote in issues that they know or feel strongly about.

This is probably an old idea, and as I stated earlier, I'm guessing there are major flaws with it, so fire away!

JoachimSchipper9• 15 years ago

I don't think your categories work very well (even absent politicking, is a bill to raise teachers' salaries an educational, financial or social bill?). The rest of your scheme is complicated but seems to work as designed.

That said, a big advantage of the currently popular party-based systems is that it ensures at least a minimum of coherent policy-making - a party that votes for both more spending and tax cuts can at least be called on it. I didn't follow it closely, but I've seen suggestions that California's experiments with direct democracy have at least partly caused financial troubles.

yesbabyyes9• 15 years ago

Thanks Joachim! The categories are meant as examples only, but I think that when applied to a nation state the various departments might be a good start.

It might seem complicated but I don't think it is, rather I think I am bad at explaining it. Then again, in the words of Swedish poet Esaias Tegnér, "that which is dimly said is dimly thought", so I will try to clear it up!

The main problem, I think, is how to count the votes, i.e. when Alice has delegated her vote to Bob, and Bob votes, and then Alice votes, I think there might be race conditions where people could game the system. So, it might be that you would have to have "voting windows", where people who carry the most voting power has to vote first, for instance.

vaporstun• 15 years ago

One interesting point about this concept is that one of the limiting factors was likely that it was too complex to keep track of who had whose vote. Traditionally, with paper and bureaucracy, it would have been a near impossible task to calculate that on the day of the vote, Carol has been delegated Bob's vote and Alice's and that neither have rescinded such delegation prior to the vote. The paper trail made it infeasible.

However, now that we have computers to do most of the figuring, things can happen in near real-time thereby removing this constraint and making this (more) possible. Certainly there are many other side issues (authentication of Alice delegating her vote to Bob, did she do so under duress?, etc.)

Very interesting idea though and, as you said, probably an old idea, but one that may have been scrapped due to these technical issues. Now that we can overcome them, seems worth revisiting.

yesbabyyes9• 15 years ago

Thank you vaporstun, that's exactly what I think as well. Delegating of votes is not a new idea, it's used in lots of societies - I got the idea from my housing cooperative. But it quickly gets out of hand after just one level of delegation, and if people are allowed to delegate their vote and then vote, well such a system has to be electronic.

As I said, I'm sure there are ways for ambitious politicians to game the system so please try!

limmeau5• 15 years ago

This sounds like the Liquid Democracy system, which the Pirate Party of Germany uses to decide on party-internal policy questions. (Other organizations may have had it before them, but they're the first I heard it of)

The major point of dispute there is that this system cannot be simultaneously anonymous (nobody knows whether I voted for or against nuclear weapons for public schools) and independently verifiable (anyone can check that everybody's delegations have been computed correctly).

yesbabyyes9• 15 years ago

Thanks limmeau, indeed it does sound exactly like Liquid Democracy. I have thought about the problem you raise and it is indeed a drawback.

aamar• 15 years ago

Right, and liquid democracy is really a specific application of proxy voting: http://en.wikipedia.org/wiki/Proxy_voting

A startup working with this idea: http://www.telegraph.co.uk/technology/social-media/8394883/J...

ZachPruckowski3• 15 years ago

This seems really vulnerable to vote buying, unless the person receiving the vote delegation isn't informed that they're voting for two. Otherwise, your boss will say "delegate me your vote, or else". If he doesn't straight-up fire you for refusing[1], he can write bad performance reviews for two straight quarters and then dump you. Or imagine if a church threw caution into the wind (sacrificing their tax-exempt status) and said "delegate your vote to your preacher/pastor/etc., or you're going to Hell". Unless you can lie about delegating the vote, then you may as well not have it.

In the above cases under the present system, I can just say "sure boss, I'll vote for Obama/Romney/whoever" and then walk into the booth and vote for whomever I like.

[1] - This is arguably not illegal in some "right to work" states in the US.

yesbabyyes9• 15 years ago

Hi Zach, yes, that's true. I definitely think it's a good feature to let people know their voting power though. It might be solvable by only letting them know in segments: 1-5, 6-10, and so on, or something like that.

logicchop4• 15 years ago

This isn't a different voting system; it's only a special case of the current system. Keep in mind, in your system it's got to be possible to NOT delegate your vote; it has to be possible for everyone to just vote on their own without delegation. In that case, what happens? You haven't really specified the voting system, just an additional feature that a voting system might take on.

yesbabyyes9• 15 years ago

That's true - if everyone votes on their own, it's direct democracy.

merloen• 15 years ago

You need to make it expensive (in time or money) to run for office. Otherwise everybody would simply vote for themselves. Or people would vote for the special interest candidate of their pet special interest, and mainstream parties would vanish.

DRMacIver16• 15 years ago

The proposal someone provided on my blog was that you should require X signatures for a nomination. e.g. if you have a constituency with 10,000 people and you want 10ish candidates require 1000 signatures for nomination.

known• 15 years ago

I think no voting system is perfect till people can/will vote as per their conscience.

thelema314• 15 years ago

The entire point of this system is to produce a voting system that does not discourage people voting per their conscience. All deterministic systems so far have flaws that encourage people to vote "strategically" to improve the chances of their candidate being elected or other flaws.

atacrawl2• 15 years ago

Until voting is made mandatory, a "perfect" voting system is a non-starter, because no system could therefore truly reflect the will of the people.

jpadkins• 15 years ago

you can't reflect the will of the people with representative democracy. People have different value judgement, therefor it is impossible to create laws that are 100% reflective of the people.

Democratic representatives do not represent you like a lawyer or financial advisor. They have to act in your and your interest alone, or they breach ethics and possibly get sued. 1:1 vs. Many:1 relationship.

atacrawl2• 15 years ago

I'm referring to the result of an election, not the actions of those elected.

logicchop4• 15 years ago

Even if this is a necessary condition for a perfect voting system, it doesn't change the underlying problem. Even if everyone voted, you wouldn't consider it a system that "truly reflected the will of the people" unless that voting system satisfied a few basic principles, like you can vote for anyone you want, there's no dictator, only information on the ballots is taken into account, etc. But those principles are Arrow's, and Arrow shows that there can't be a voting system that satisfies them. So there can't be a perfect system at all. What's the point then in making everyone vote?

colanderman• 15 years ago

Random voting is stupid for one reason: there is little incentive for any individual to vote for a competent candidate. You have the same chance of choosing the next president whether you vote for Obama or Joe the Plumber. Yes, reasonably-minded citizens will vote for one of a handful of competent candidates, but a vast number of people will simply vote for themselves.

The equivalent in a plurality system would be if all write-in votes were counted together, and then a write-in chosen at random should the write-ins outnumber the votes for any other single candidate.

I guess the "solution" would be to require a minimum # of signatures for a candidate to be on the ballot (and votable). But that's just starting to sound like approval voting, so you might as well just go all the way with it!

JoachimSchipper9• 15 years ago

I hadn't realized this, and your comment is too late to gain many upvotes, but yes - very good point.

Couldn't you solve this by requiring any randomly-elected candidate to have at least, say, 1% of the popular vote, though? (This means that you don't entirely eliminate strategic voting, unless people can specify a fallback option; but it reduces it by a lot.)

lyudmil4• 15 years ago

This is a cool thought, but I think it's a bad idea to use it in actual elections. The author failed to really address the "crazy minority party" concern. His argument was that, while his system would give such people a chance in getting elected, it's only fair that it did since they indeed had a constituency and their chance of winning is proportional to their constituency size. He adds, "The fact that I consider someone to be utterly despicable doesn’t give me the right to disenfranchise them."

I'm not sure the point is valid. I certainly believe in freedom of speech and I accept my duty to protect the right of even the most offensive people to be heard (in fact, they should receive a greater degree of protection). However, by supporting a voting system that gives them no chance of being elected before they convince enough people for their ideas to be considered mainstream I'm not disenfranchising them since I'm not denying them the right to vote. I'm just marginalizing them, which is the proper response to all crazy claims and ideas. I offer two hypotheticals in support for those who still feel I'm infringing on someone's liberties by writing what I did:

(1) Suppose we live in a country that's a nuclear power. Suppose also there is a fascist party completely bent on nuclear war with Israel (for example) in our country and that they have a constituency that amounts to 1% of the electorate. Suppose nuclear war means total destruction on a global scale. I'll argue that giving those people a chance at forming a government is immoral and, in fact, disenfranchises everyone outside of your country since they have to control over who you elect, yet they suffer the horrible consequences.

(2) Suppose my party ran on a platform of dismantling democratic institutions (including elections of any sort) and vowed to distribute the majority of resources to my constituents, basically re-instituting feudalism. Random selection would give us the opportunity and the power to dismantle the democracy it was meant to be the engine of. It would also technically make it illegal for other countries to intervene, since our "feudal" government would be democratically elected.

innes• 15 years ago

"Suppose ... there is a fascist party ... and that they have a constituency that amounts to 1% of the electorate"

Given a legislature of 600 representatives, a party with 1% support would have a vanishingly small chance of winning a majority.

lyudmil4• 15 years ago

Even if you consider running consistently over a long period of time?

Also, my hypotheticals were meant to be such that you wouldn't take even vanishingly small chances. If you are willing to, then I failed in my argument.

DRMacIver16• 15 years ago

It's worth noting that the consequences of random election are not "the crazy party gets into power". It's "the crazy party gets a couple of seats". The chances of someone with even 5% of the election are astronomically low (http://twitter.com/johnmyleswhite/status/62863260566827008 - the probability of someone with a 5% minority getting the majority over the course of 50 elections).

The likely scenario here is that your crazy party gets 1, maybe 2, seats at the worst and promptly demonstrate themselves to be incompetent tools, spoiling their chances of reelection.

lyudmil4• 15 years ago

I don't think that's what the author was describing. I think it was a case of everyone gets a vote, then we pick someones vote at random.

IsaacL• 15 years ago

This is an issue with most (all?) PR systems. The author's argument was not that crazy minority's should get into power, but they should at least be represented in the political system (Parliament or whatever). They should at least have a chance to air their views on the national level.

I think this has two side-benefits: one is that such crazy minorities are often shown as incompetent when given such 'legitimate' representation; the other is that it removes their PR boost of being "downtrodden outcasts".

lyudmil4• 15 years ago

With regard to your first point, I don't think that's what the author is describing:

"Once all of the votes have been cast, you pick one voter at random and use their choice"

It's not a matter of representation, which would not only be OK with me, but necessary if we claim freedom of speech. The author is describing a random dictatorial mandate weighed based on constituency size. Maybe I'm completely misunderstanding this...

aranazo• 15 years ago

Using random factors in election is a really old practice going back to Athenian Democracy at least.. It's known as Sortition.

DRMacIver16• 15 years ago

Both of which terms are explicitly used in the article. :)

However this is not a sortition. It's an application of random-dictator to the problem of electing a per-constituency representative. This turns into a sortition only in the degenerate case where everyone runs and votes for themself.

TravisLS• 15 years ago

There's an interesting side effect of this system which I think means that it would have lessened impact on the policies implemented by the elected officials.

As in his example, pretend the Kitten party appeals to 10% of the electorate distributed evenly across districts. So in roughly 10% of districts, the Kitten party would earn a seat, but would only represent 10% of their constituency.

In this case, the incentives of the elected Kitten party representative shift. Instead of keeping to their Kitten party ideals, they would seek to maximize their following within their district by shifting to a median political position. This would ensure them the maximum likelihood of re-election in the following campaign.

So theoretically the actual policies coming out of each district would still represent the median view of the electorate in that district, and the Kitten party would remain disenfranchised.

DRMacIver16• 15 years ago

Hmm. I'm not sure I believe that. I'm not sure I don't believe that either mind you. Consider the following questions:

* If they're willing to do anything to stay in power, why were they running for a minority party in the first place?

* How likely are the voters to use their single vote to vote for someone who is still running as kitten party but is acting like a watered down version of the people they actually want to vote for?

shasta• 15 years ago

I think you might improve this scheme by doing it twice. First, a random sampling of votes cast are used to create a candidate pool. Next, one lucky voter's vote is selected to pick a candidate from the pool. The upside to this is that it helps get rid of self interest / crazy voting. If the scheme David MacIver proposed were used for determining president, I expect 25% of people would vote for themselves. (Well, he proposed starting from a small pool, so I guess I'm wondering about what would happen if you used this scheme to produce the pool in the first place)

protomyth2• 15 years ago

The NBA draft works this way (most votes = lowest finish previous year). It has not done a good job and I think this would create riots in the streets.

kgermino• 15 years ago

I'm sorry but how is this similar? I'm probably missing something but I don't see the connection.

protomyth2• 15 years ago

If I understand this scheme, we put the "votes" in a box and pick a random winner from them. The NBA draft puts its "votes" (# based on finish order of bottom) and picks a random winner from them.

lmkg• 15 years ago

Some background on the problem this is trying to solve:

There are a handful of mathematical theorems that basically say "there is no perfect voting system." More rigorously, they enumerate a list of desirable properties that you would like voting systems to have, and show that NO voting system can have all of these properties, because they are actually contradictory.

Some of these properties are things that are "obvious" ones that any practical voting system should satisfy: voting matters, there is no dictator, candidates can win, determinism. Two of them are non-trivial: absence of tactical voting, and irrelevance of independent alternatives (IIA). Tactical voting basically means you have a game-theoretic incentive to vote in a way that differs from your true desires. IIA basically means that if Gore would win a 2-man election against Bush, adding Nader to the race should not make Bush the winner. These properties are related, but not the same.

All of these properties are desirable, and a "perfect" voting system would have all of them, but you can't have all of them. You have to pick at least one to give up. Tactical voting in particular is extremely hard to remove, even if you give up IIA, while systems that satisfy IIA are somewhat complicated. The article suggests that it is very easy to satisfy all of the other criteria, if we let go of one of the "obvious" criteria: determinism.

So it's basically a question of picking your poison. You can have determinism, or you can get rid of tactical voting, but you can't do both. The author is contending that determinism is the best property to give up.

Further reading: http://en.wikipedia.org/wiki/Arrows_theorem is one of the mentioned theorems, dealing with ranked-preference voting systems (vote-for-one-candidate is a special case). There are a few others for different classes of voting systems.

DRMacIver16• 15 years ago

As a side note, this does not directly address the problem proposed by Arrow's theorem, although it could easily be made to do so, but the post isn't so interesting in that context. It's more about how it interacts with a constituency system.

If you want a randomised solution to Arrow style preferential voting, the following is the best I currently know of:

1. If you have one candidate, stop. That candidate is elected.

2. If you have more than one candidate, pick a candidate at random. Call that candidate Pivot. Every candidate who the majority thinks is worse than Pivot drops out of the race. If there is no one the majority thinks are worse than Pivot, Pivot drops out.

3. Goto 1.

Edit: Err. Correction. That's an algorithm for selecting a single winner from Arrow style preferential voting. It's based on an only slightly more complicated one for giving you the full range though.

logicchop4• 15 years ago

The condition you identify as "IIA" isn't Arrow's condition. IIA in Arrow's theorem is best understood as the principle that only information on the ballot is relevant to the vote. The condition you cite is a much stronger counterfactual property: that if x beats y, then x beats y even if we introduce z. But this is not Arrow's condition (see Bordes and Tideman on this).

leot2• 15 years ago

A particularly problematic aspect of this system is that, in close elections, it provides little incentive to the elected representative to do a particularly good job once they get there. If the pol won with 51% of the vote, then perhaps if they work their arse off they could get to 55%, and if they do jack they might only get to 45%.

rmc• 15 years ago

In Ireland, we use PR (single transferable vote), however we do have local politicians. PR does not mean no local constituency stuff. If anything there is too much locality in ireland.

logicchop4• 15 years ago

Not to be dismissive, but this is a pretty unhelpful approach to the voting problem. The idea of picking a random person is motivated by the thought that the more support a candidate has the greater chance they have of being selected. But then why not just select the person with the highest probability of being selected? The difficulty is that even this "harmless" process is infected. The real upshot of Arrow's theorem is that you can't aggregate preferences: there is no "measure theory" underlying the concept of group will.

bobwebb• 15 years ago

The thing that interests me about this article is that ~200 years ago, compared to the simplicity of FPTP, most other voting systems would be quite difficult to implement, because all bureaucracy/vote counting would have to have been carried out by hand.

I wonder what voting systems would have looked like, had electronic systems been around at the time of (public) democracy's inception.

omouse3• 15 years ago

The BNP have a percentage of the voters, they should get a percentage of the say. The fact that I consider someone to be utterly despicable doesn’t give me the right to disenfranchise them.

I completely disagree with that. Anyone that would deny you the right to do something should not be allowed to have that right. They wouldn't give immigrants and non-British citizens a right to vote, so why should they be allowed to have a say at all?

Democracy and freedom need to be defended. Allowing people who actively subvert it to win an election would be absurd.

lucasjung2• 15 years ago

Put another way: marginilization of fringe positions could be considered a desireable quality of a voting system, as long as sufficient free speech protections are in place.

In a sense, elections act as a sort of filter between ideas and actions. Freedom of speech can be used to disseminate ideas, and some truly awful ideas will disseminate rather widely. However, they won't usually gain enough adherents to get enough votes to make it past the "filter." That's why there's little harm in letting people espouse "dangerous" ideas with their speech[1]: most people will recognize bad ideas for what they are, and the proponents of the worst ideas generally will not be elected to positions where they can influence policy directly.

This "filter" is far from perfect: some bad ideas will become laws nonetheless, and some good ideas will get blocked. More of the former than the latter, though: if an idea is good enough, it will eventually garner sufficient support to become a significant factor in elections.

There have been cases in history where entire societies did adopt bad ideas (e.g. fascism). However, I think that when an entire society takes such a dramatically wrong turn, even the most perfect voting system will not save it.

[1]It's important to draw a line between espousing ideas and encouraging actions.

omouse3• 15 years ago

You're revising history a bit; when fascism was rising there were a lot of people that were fighting against it such as the socialist and the communists and the anarchists. Unfortunately, other societies were trying to contain those socialist forces and this reduced their power to counter the fascists. You can see this happen in Spain and Italy as well (there were a large number of collectives in Spain and Italy has had anarchists and communists for a long time, in the 1970s they were quite the active movements).

The "filter" shouldn't be just at the voting booths. We should be reminded before and after election day why the ideas of those fringe positions is a bad idea.

Espousing ideas in such a way as to hide the true nature of them is something that political parties are getting good at. If there are dangerous ideas, they're hidden behind nice language or buried in the back of the political party's plans. So it's harder for people to recognize those bad ideas.

I think we would do well to learn from the past and be more active in denouncing ideas such as those of the BNP and other fringe parties.

jacques_chester3• 15 years ago

The most reliable way to defend against fringe politics from becoming overly influential is to make voting mandatory, as Australia does. See Hotelling's Law for why.

http://en.wikipedia.org/wiki/Hotellings_law

omouse3• 15 years ago

What do you consider fringe politics? And why would forcing people to vote keep them from becoming overly influential?

FrojoS2• 15 years ago

Great, but I think he is missing one more big requirement. Your vote must be secret. That is, there must be no way for YOU to prove others how you voted. If not, bribery and enforcement become possible - and will happen.

To make things even more complicated, it would also be good if you can be sure and CHECK at any point that you vote was counted the way you intended. I don't know if there is a solution to this, but I'm imagining something like a hash code, that you see the moment you vote, together with feedback. That is, you can actually see, that your vote got counted the moment you vote and then later you can use the hash code to see, that the data base was not corrupted. Of course you should not be able to prove how you voted.

ajslater• 15 years ago

http://news.ycombinator.com/item?id=2495775

Compares One Vote, Instant Runoff Voting, Acceptance Voting, Ranked Choice Voting, Condcorcet Method and various methods of Rated Voting via simulation testing for maximum voter happiness with the result.

A “perfect” voting system

Related Stories