Ask Slashdot: How Would You Use Computers To Make Elections Better?

shanen writes: Regarding politics, is there anything that Americans agree on? If so, it's probably something negative like "The system is broken," or "The leading candidates are terrible," or even "Your state is a shithole." With all our fancy technology, what's going wrong? Our computers are creating problems, not solutions. For example, gerrymandering relies on fancy computers to rig the maps. Negative campaigning increasingly relies on computers to target the attacks on specific voters. Even international attacks exploit the internet to intrude into elections around the world. Here are three of my suggested solutions, though I can't imagine any of today's politicians would ever support anything along these lines:

(1) Guest voting: If you hate your district, you could vote in a neighboring district. The more they gerrymander, the less predictable the election results.
(2) Results-based weighting: The winning candidates get more voting power in the legislature, reflecting how many people actually voted for them. If you win a boring and uncontested election where few people vote, then part of your vote in the legislature would be transferred to the winners who also had more real votes.
(3) Negative voting: A voter could use an electronic ballot to make it explicit that the vote is negative, not positive. The candidate with the most positive or fewest negative votes still wins, but if the election has too many negative votes, then that "winner" would be penalized, perhaps with a half term rather than a full term.

What wild and crazy ideas do you have for using computers to make elections better, not worse?

Comments

Simple

[oldgraybeard]

I would not use computers! Paper ballot feed in to (Yes computer based) non connected totaling systems. Just my 2 cents ;)

Re:Simple

[Motard]

Right. Computers are good at counting shit. Let's just let them do that.

Elections are about people. Until skynet fixes this, this is going to have to do.

Re:Simple

[0100010001010011]

Paper ballot with scanners.

Rather than a mono-culture require that no 2 bordering counties may use the same brand scanner.

After the early election results are in share ballots with 2 neighboring counties to use on their machines.

If the machines report different numbers you hand count them.

Bonus companies for how accurate they are. Those that are more than 1% off get no payment for the machines and the company is not allowed to make machines for the next election cycle.

Re:Simple

[networkBoy]

select victor from rand(candidates);

Re:Simple

[dryeo]

Paper is pretty secure here, where most anyone and especially members of all parties, can watch the whole process. Equally important is that the process is so simple that the average person can easily understand the process.
With your idea, no matter how secure it actually is, the average person will not understand it and it will appear to be a black box. Trust in the system is as important, if not more so, as having a trustworthy system.

Re:Simple

[doom]

Paper is pretty secure here, where most anyone and especially members of all parties, can watch the whole process.

Yes, that's pretty much it. There are scams you can use with paper ballots, but they're harder to get it to scale [1]. Arguably, a hybrid system (as is common these days) of paper ballots counted electronically could be better than a pure paper system-- then you can use computer techniques to look for problems, and paper hand counts to check afterwards.

Equally important is that the process is so simple that the average person can easily understand the process.

Yeah, exactly. You might have your spooky crypto-magic uncrackable system deployed perfectly, but it's too complicated for citizen's to understand, you could get demagogues whipping up distrust for the system. Counting the vote accurately doesn't help if no one believes you.

[1] One year, the coast guard found the lids of ballot boxes floating around in the San Francisco bay...

Re:Simple

[Chas]

I simply don't have enough mod points for this post.

Sorry, but computers simply are NOT a trustworthy medium for something like this.

I LIKE computers. And I trust them to work as they're told.

I simply don't trust the assholes who're doing the telling. Nor that someone couldn't subvert them and become the assholes who're doing the telling...

Re:Simple

[ShakaUVM]

Agreed. All digital voting is the biggest threat to democracy ever seen.

Re:Simple

[clovis]

SELECT candidate AS victor FROM candidates ORDER BY bribe_amount DESC LIMIT 1;

I registered to vote as
Clovis'); DROP TABLE candidates;

Re: Simple

[dskoll]

Really? You ever try to rig a paper ballot election? You'll find it almost impossible to commit massive fraud, something that's trivial with computer voting.

Re:Simple

[bidule]

The paper system is far to easy to control and rig.

Paper trace is the hardest thing to fake.

Someone is trying to screw a nail because he doesn't like hammers.

Re:Simple

[ClickOnThis]

Agreed. All digital voting is the biggest threat to democracy ever seen.

This.

It's not the people who vote that count, it's the people who count the votes.

Apocryphal, and attributed to many sources (including Joseph Stalin) but in an anti-democracy dystopia, ain't it true?

IMHO, don't let computers have full control over votes. Always, always have a hard-copy ballot. Let computers tally them, but leave it open for humans to examine what happened.

Re: Simple

[ShanghaiBill]

The reason the USA bundles elections is because more people get out and vote

Serious question: Is there any evidence that higher voter turnout is correlated with "better" government?

Higher turnout may just dilute the vote of people that took time and effort to understand the issues.

Voting is mandatory, and thus very high, in Greece, Argentina, Turkey. I don't think many people would consider any of these to be "well governed".

Compulsory voting

Re:Simple

[swillden]

Arguably, a hybrid system (as is common these days) of paper ballots counted electronically could be better than a pure paper system -- then you can use computer techniques to look for problems, and paper hand counts to check afterwards.

Even better than that. You can apply techniques from cryptography to enable mathematically-provable election results, providing any voter with a receipt they can use to verify that their ballot was counted correctly -- but which they cannot use to prove to any third party who they voted for, and enabling anyone to download the set of files representing the final tally and verify the totals themselves, and trace every ballot back to those printed and verified pre-election.

In that system, computers are used both before and after the balloting. Before to generate all of the codes printed on all of the ballots and to enable integrity-checking of the printed ballots, and after to electronically scan all of the ballots and perform various post-election verification processes. And of course the paper ballots remain if there's a need to check them as well, though honestly the mathematical guarantees are tighter than any recount.

This system exists, has gone through several iterations of improvements to make it more and more practical and cost-effective, and has been tested in real-world elections. It's the brainchild of Ronald Rivest (the "R" in "RSA") and David Chaum (inventor of the first practical digital cash system, which long predated Bitcoin), among other top academic cryptographers and it's called Scantegrity II

Re:Simple

[mridoni]

Paper is pretty secure here, where most anyone and especially members of all parties, can watch the whole process.

Yes, that's pretty much it. There are scams you can use with paper ballots, but they're harder to get it to scale

That's important, but there's more: with paper ballots, literally anyone, on a small scale, even without any formal education at all, can understand the principles involved and monitor the process, before or afterwards. With electronic voting, you need people with experience (and very possibly degrees) in cryptography and security. Not only this severely restricts the number of people who are able to assess if the process is rigged, but also it makes the process "less democratic", given that the greater part of the population, in practice, is hindered from exercising their right to check that the election process was really fair.

Re:Simple

[currently_awake]

Vote buying works much better if the people you pay for votes can prove they did. How much would you need to offer the 60% of Americans who don't vote to sell out? You only need 5% to throw an election.

Take the average of the desires of the voters

[jader3rd]

Instead of voting for a candidate, have the electorate vote on a number of issues (combination of recent past issues and issues on the docket). Then take the average, and the candidate of the political party that is closest to the average, wins. Parties can do whatever they want to determine candidates.

Down side is that for those who already feel like voting is like busy homework, this will add to the load.

Re:Take the average of the desires of the voters

[Kjella]

Down side is that for those who already feel like voting is like busy homework, this will add to the load.

If they think picking red or blue every few years is too much effort, maybe they're better off staying home. And honestly that could actually bring people out to vote primarily on a particular issue they care about rather than vote for Hillary vs Trump. Because I can see how neither would be particularly appealing....

Re:Take the average of the desires of the voters

[omnichad]

Its ok to promise 100% employment and low taxes but if you don't deliver you forfeit your salary, go to jail, banned from holding a government position

This is not what you vote for. You vote for people who know how to make informed decisions, and you trust them when they get into office because they have access to way more information and resources than you do. If that doesn't work, the problem is your candidate. Campaign promises should not exist outside of general statements.

Re:Take the average of the desires of the voters

[shanen]

Interesting idea, but I don't see how the accountability would work. The winners of the elections might fail to live up to the promises they had made.

Also still subject to abuse from the appeal to single-issue voters. That's actually how the GOP is getting a lot of votes these days. Just tell them you agree with THE issue.

Ranked voting

[Anonymous Coward]

Computers would make this easier but are not required.

https://en.wikipedia.org/wiki/Ranked_voting
http://www.fairvote.org/rcv

Re:Ranked voting

[swillden]

Computers would make this easier but are not required.

https://en.wikipedia.org/wiki/Ranked_voting http://www.fairvote.org/rcv

I used to be big fan of ranked voting, especially with Condorcet evaluation with Schwarz Sequential Dropping. Then I tried to explain it to a few people and changed my mind. Instant-runoff is a little simpler, but still pretty complicated -- and actually a bit tricky to execute correctly since it's inherently multi-pass (Condorcet is simpler to execute). Simplicity matters because what's just as important as having a fair election, is having a fair election that voters can understand and trust.

I think the best scheme overall is approval voting. The mathematical properties of approval voting are almost as good as the best ranked voting schemes. It's a little more vulnerable to strategic voting (which is when voters might have reason to vote other than their true preferences, as is the norm in plurality-rules schemes), but really not very much. In theory it also doesn't capture quite as much nuance of voter intent since it doesn't allow one to express a preference between two acceptable candidates. But it does allow voters to express another important element of intent which ranked ballots don't allow: acceptability. And it's brain-dead simple to understand.

If you don't know how it works, here you go: An approval voting ballot has all of the candidates listed. You mark all of those that are acceptable to you. The candidate with the most marks wins.

Such a system eliminates the strong two-party bias that plurality-rules systems have (Duverger's Law, that bias is called). In very few cases does it ever make sense to vote other than your true preferences. And it encourages parties to field broadly-acceptable candidates.

Tallying is a single-pass process and counts can be provided by sub-regions for totalling (unlike IRV, where the runoff phases require reinterpretation of the ballots at each runoff). If it's desired, you can even specify a minimum win threshold -- if no candidate gets, say, 50% approval then no one wins and you re-run the election with a new slate of candidates. There's an obvious risk of never getting a winner here, so such a system should probably progressively lower the required approval level to be sure that someone eventually wins, but the flip side is that such a system would mean that the 2016 US presidential election would never have put either Hillary Clinton or Donald Trump on the ballot; both (all) parties would be looking for someone with broader appeal.

However, approval voting can be done with or without computers, so it's not really relevant here. IRV can also be done without computers, though it's kind of tedious without them.

Re:Ranked voting

[Richard Kirk]

Approval voting gets my vote.

I have considered systems where you could vote for and against people. My local elections usually some up with about 8 candidates, where the best are faceless, and the worst are the nasty sort of nationalist. I can't pick one Iike, but there are certainly ones I hate. People with strong opinions for one person would vote for them and blackball everyone else. So, being able to vote for and against any candidate almost reduces to approval voting.

Approval voting also is easy to understand, and gives a sensible result in one vote. If there is a simpler system without the problems of the single vote for, I can't see it.

Democracy is how you remove a government

[FeelGood314]

The electorate are always going to be emotional, easily fooled and right stupid. Look at the average person and then realize half the people are dumber than him. The best you can hope for is to have a way to remove an incompetent government after 4 or 5 years. The best democracies are ones where people successfully remove a bad party and don't have it return unless it seriously changes it's way. These are countries where new parties can be created and eventually form the government. Poor democracies are ones where the government doesn't change, switches back and forth between two parties with no hope of a third party forming, or countries with endless coalitions where the same people stay in government forever.

Long Live The Republic

[CarterMeyers]

Go back to a republic - where only land owners get to vote. Then use computers to adjust the weight of each vote based on how much land you own.

Abolish gerrymandering by using computers

[klingens]

Use an algorithm to create congressional districts with census data so each district has approx. the same amount of possible voters and the smallest circumference. No more rigging to create safe districts for either party with ridiculous borders.

Re:Abolish gerrymandering by using computers

[Baron_Yam]

The problem is that while you can use mapping data to break districts in somewhat logical ways, we also tend to want to divide voting by 'neighborhoods' when possible. This is both a problem and a solution, since neighborhoods are pretty much by definition clusters of similar people who probably vote similarly.

Nobody wants to see their vote ignored because they're a small neighborhood adjoining a larger but politically different one, and nobody wants to see someone else's vote given more effective weight because the line was bumped to keep their neighborhood distinct.

Re:Abolish gerrymandering by using computers

[jonwil]

Do it like they do in Australia.
Changing the electoral boundaries is done by an independent body (the Australian Electoral Commission).
It is done on a regular basis and is designed to ensure that each state has a suitable number of representatives based on its population and that each electoral district has a similar number of electors.

Generally the AEC will try and keep related areas in the same district (e.g. a specific suburb will all be in one district for the most part) and they also take submissions from people into account when drawing up the boundaries.

Since its done independently and people get to have a proper say, no-one can argue its unfair to them (e.g. they cant say "hey we dont think its fair that our poor black neighborhood is in a district of mostly rich white people" because they had the opportunity to complain/object to that decision and because no-one can argue that the decision was in any way political.

Re:Abolish gerrymandering by using computers

[Wrath0fb0b]

Not Australian, forgive my ignorance, but as far as I can see from Wikipedia, the AEC is answerable to parliament and run by a minister of the government. I don't get how it's reasonable to call it an 'independent body', when presumably the government (with support of parliament) can have it do whatever they please.

It might be run non-partisanly as a matter of tradition, I suppose.

Re:Abolish gerrymandering by using computers

[billyswong]

Are there any objective definition of 'neighborhoods'? If yes, then we can group houses/buildings/estates in census data so they can stay in the same district even with computer district genearation. If there is no objective definition, then your desire to "divide voting by 'neighborhoods'" is just a desire to gerrymandering.

I would use computers...

[hey!]

To look up all the excellent work done by mathematicians, economists, political scientists and cryptographers on (a) how to conduct votes and (b) how to use votes to select candidates, before I bang together my own half-baked proportional representation scheme.

The maybe I'd write an R routine to detect gerrymandered states (actually quite easy if you've taken the first step above) and then hack into politicians' social media accounts so I could blackmail them into outlawing partisan gerrymandering.

here is what an expert says about your idea

[oldbox]

So you want to reform democracy?

Https://medium.com/civic-tech-thoughts-from-joshdata/so-you-want-to-reform-democracy-7f3b1ef10597

Voting Can Be Improved But Not With Computers

[crunchygranola]

There are is a very active field of "voting theory" research about how voting systems can be improved, but little of it has to do with computers per se, though they can make implementation of the post-vote processing more convenient. That is to say, it is not the "computer" that is improving anything. Various forms of preferential voting have a lot to recommend them, along with variations like "instant run-off".

In general it is a good idea to identify actual problems (e.g. widely unpopular candidates winning due multiple candidates splitting the vote; spoilers being run to siphon of votes from specific candidates, etc.) and propose actual fixes that are subjected to formal analyses, large scale simulations and such to validate that they are improvements.

The suggestions of the OP mostly sound like "let's just try something different" rather than carefully considered improvement proposals.

Re:Voting Can Be Improved But Not With Computers

[swillden]

You should look into Scantegrity II. It applies computers before and after the balloting to actually increase the integrity of the vote -- irrespective of what voting system is used.

Manufactured difficulty

[grasshoppa]

The only thing that makes computer based ballots exceedingly difficult to implement is politics, not the technology. It's actually quite simple.

A user is assigned a random physical token on registration at the voting place, which "START TRANSACTION"s the process. They take this to the machine, which "unlocks" the voting capability. They make their selections, which the computer stores AND prints out a human AND computer readable receipt ( so no bar codes ), which the user then takes back up to the registration desk for secure storage in case of a recount. They also turn the token back in at this time too, and the agent runs it over the "out" scanner, which effectively issues a "COMMIT" to the vote.

What's the token for? A few different purposes.

1) If the paper ballot is lost before being returned to desk, the user's vote can be eliminated ( ROLLBACK ) and they can vote again. Paper ballots are important.
2) Ensure that only those who should be voting are voting, as verified by the registration desk. Voting machines are not randomly open to anyone who walks up to it. In fact, you can take this a step further and make the secured tokens the only way to unlock the database on disk, cryptographically.

See? Easy. The real question should by; why haven't we done this yet?

Re:Use to stop illegal voters

[doom]

Not that it matters to you, but no one can find any evidence of these hordes of fraudulent voters swaying any recent elections. You've gotta go back decades before you find even half-way plausible stories to that effect.

As a doorstop

[Opportunist]

Or maybe a paperweight.

You MUST NOT use computers for voting purposes. Even if every bit used is open source, even if you open source hardware, software and everything can be audited and everyone can verify that the hash of the binary is the same that a binary you compiled from the source is the same, even if you do EVERYTHING to make sure that anyone is able to audit it, it's a VERY VERY BAD idea.

Not because it can be manipulated. But because you cannot silence the ones claiming it's still fraud and that all the computer savvy people cooperate to overthrow democracy and humanity altogether. Because they can't audit it, because you need to know how computers work and how to audit computers to actually perform one.

Paper and pencil have one key advantage: EVERYONE can audit it. It takes the ability to see and the ability to count. Even reading is entirely optional because all the ballot slips are identical and you can simply go and count the ones with the cross at the same position. Every party can send whoever they want to supervise the election, no special education or skill needed.

It's less about actual election fraud. It's more that nobody can sensibly claim there had been one. We live in a time of fake news and creative reporting. Is it that far fetched that any party who loses an election would start rumors about rigged voting machines that could of course be audited, but only be a select few (aka "the elite")?

With paper and ballot, it's trivial to debunk anything like this. They could have sent literally ANYONE to supervise the election process. They could send ANYONE to recount the ballots. Any claim to fraud would instantly fizzle.

Re:As a doorstop

[Baron_Yam]

I would object to complete exclusion of computers: computer tallying of paper-and-pencil ballots greatly speeds up the initial counts. So long as the paper ballots exist and can be counted should a problem be expected, leaving out the computers is simply an exercise in delay and increasing the average cost of human involvement in the process.

Or a Condorcet methd.

[robbak]

Instant run off is reasonable, and has the benefit of being easily doable with people and papers, and so is more auditable. Yes, this is a benefit of the IRO method.

But it does lead to some unreasonable decisions when a poll produces 3 close frontrunners, where the voters who voted for candidate #3 essentially decide the vote by their lower preferences. If you are going to put a computer in charge, then there are better ways, ways that find the Condorcet winner - the winner who, based on everyone's preferences, would have won a one-on-one election against any other candidate. (The methods vary when there is no Condorcet winner, when the preferences are circular - a situation that you might think rare, but there are times where the extreme left and right find they have more in common with each other than with the more boring center.)

These methods, however, are complex, and require a computer to work out - although that computer result can often be confirmed in a hand count.

Hackproof way to use computers in elections:

[rnturn]

Stack computers on top of one another until you have a surface tall enough to fill out a paper ballot.

Re: Lower the voting age to 15

[c6gunner]

Sorry, did you just suggest that minorities are all parasites?

You are one seriously racist mofo.

Computers have no place in elections to begin with

[LeftCoastThinker]

Keep computers out of elections. They are far too easy to tamper with in ways most elections officials are incapable of detecting.

Paper ballots and any other form are also imperfect, but at least they physically exist and if properly guarded and supervised by election observers, they are reasonably difficult to rig, assuming you get rid of absentee ballots for everyone who is able bodied and require a signed doctors excuse on a prescription pad for the disabled.

derp

[nyet]

My top two:

Condorcet voting, perimeter/area ratio limits for districting.

In Austraila, the paper ballot is extremely secure

[aberglas]

All votes counted in front of scrutineers, every time. And then independently tallied.

The USA's love of machinery combined with the complete lack of independent oversight looks like it has been deliberately designed so that it can be rigged.

Re:In Austraila, the paper ballot is extremely sec

[aberglas]

I might add that it is very quick. Takes about an hour to count all votes in a booth completely by hand. No scanning involved.

Bennett Haselton

[drafalski]

Has anyone asked Bennett Haselton? I'll wager he has an idea or three...

Duh...

[Ferretman]

...you DON'T.

Hard, primitive, paper ballots is the only way to go. They can be scanned for a fast count and if there's a need for a recount you can do that by hand.....and the parties of the candidates post monies the day before the election so that this can be paid for. If the loser decides not to bother with a recount, the parties get their monies back.

Ferret

Re:Absention gets the seat

[Mal-2]

That's a good way to keep newcomers and single-issue candidates out. Only the popular would dare to run, and we'd be left with even more of a celebrity shitshow than we have now.

That would be good, not bad

[raymorris]

Over half of Americans don't know who the vice president is. That's how interested many of us are in policy and the political process. A supermajority can't distinguish the Republican platform from the Democrat platform when it is handed to them with the party name redacted.

I don't have my car fixed by someone who doesn't know what an "engine" is, I don't have dental cavities filled by someone who can't point to my bicuspids, and I don't want national policy decided by people who don't recognize the name "Mike Pence", nor know how many senators there are.

> I think it's clear that if you want representative democracy to work and be considered legitimate, you need fewer barriers to voting, even if people like you think a DMV visit is reasonable.

And that's the reason the founders created a republic, not a democracy. The federal budget isn't American Idol. If you're not interested enough in participating in society to either have a driver's license or swing by and pick up a (free) ID, maybe you're not the person who should be deciding federal law and other national policy, based on "I heard he was born in Africa"or "because she's a woman". Maybe the decisions of national policy SHOULD be made by people who have enough interest to do more than "text your vote to 1-800-bumper-sticker".

Meet Arrow's Impossibility Theorem

[Solandri]

Before asking what is the best voting system, someone was smart enough to ask if it's even possible to determine what the best voting system is.

The answer is no. If you start by requiring some common sense criteria for what an election should accomplish, it turns out no voting system can meet those requirements. If you rank those criteria in a specific order, it's possible to mathematically calculate which voting system generates "unfair" results more often for that particular ranking of criteria (which is to say, a "worse" system can still sometimes produce a "fairer" result by your predefined standards). But if you change the ranking of the criteria, a different voting system ends up being best and producing a different winner. A clever documentary back in the 1990s even came up with an example where a dozen candidates participated in a mock election whose votes were tallied using a dozen different voting systems, and each voting system produced a different winner (i.e. each candidate won under one voting system).

Comment removed

[account_deleted]

Comment removed based on user account deletion

Fixed election dates

[currently_awake]

Fixed election dates are Evil. The length of an election is much longer with fixed election dates because everyone knows when it's happening, where snap elections you have to wait for the election to be called before you spend money on campaigning (don't want to spend all your money and not have an election). Longer elections favour the rich (cost), and nobody likes year long elections.

Digital solutions are easy without a secret ballot

[Paul Fernhout]

The USA did OK without a secret ballot for 100 years: https://en.wikipedia.org/wiki/...

Sure, there may be voter intimidation and vote buying and so on without a secret ballot. But will the consequences really be worse than widespread electronic election fraud?

And the fact is, you can find out who many people probably voted for by looking at campaign donation records anyway.
http://classic.fec.gov/finance...

We expect elected representatives to generally vote in a recorded way and to defend their votes. Why do we think that can work but doing the same for individuals won't?

Otherwise, use paper ballots -- ideally counted by a group of humans from different political affiliations like is done in many other countries.

Some bigger issues than technology for the USA:

We could return to the original constitutional number of Representatives so that each vote for one counts 10X more -- which might reduce the role of money in such elections.
https://economix.blogs.nytimes...

And maybe go back to having Senators appointed by State Legislatures.
https://www.senate.gov/artandh...

And also consider a Parliamentary system where Congress selects a Prime Minister instead of a direct election of the President (given what a money-driven circus such elections have become):
https://www.minnpost.com/eric-...