Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Government Audited versus Public Source Electronic Voting

By greenrd in Op-Ed
Sun Jul 21, 2002 at 12:03:45 PM EST
Tags: Politics (all tags)
Politics

In Palm Beach, Florida - one of the focuses of controversy in the 2000 presidential election - once again, the spectre of electoral incompetence, or worse, rears its ugly head. AP reported last week: "... a Boca Raton man's claims that he lost a City Council election in March because the new machines malfunctioned. Former Mayor Emil Danciu's suit seeks to have the results overturned and a new election held.

"The voting machines that replaced butterfly ballots and hanging chads are checked by an 'Enron-style of auditing' and don't provide voters any assurance that their votes are being cast, an expert testified Tuesday. Rebecca Mercuri, a CS professor at Bryn Mawr College in Pennsylvania, said questions remain about the $14 million machines Palm Beach County purchased to improve its voting system, because they are designed to audit themselves."


Supervisor of Elections Theresa LePore says such a review would void the machines' warranty and that they've been reviewed twice by labs appointed by the federal government and also by a state worker.

However, is this sufficient? It's a well-known truism that pre-release testing of software cannot be guaranteed to catch all bugs or even security problems (intended or not) - so by the same token, would it not be prudent to allow independent testing of any machines or components that have been alleged to be faulty, at any time? We don't accept Microsoft's extensive beta-test program as "proof" that the latest Windows will have zero bugs - why should we accept a limited government lab test, or a "mock election", as "proof" that an electronic voting system works flawlessly? After all, if some party had rigged the system, they would probably do it by rigging only machines that were not going to be sent off for testing.

Two reasons are given for denying an independent audit: it would void the warranty of machines that cost $3,500, and some of the information is a trade secret of the company that produced the voting machines.

Both reasons appear to be completely bogus. The first is a sleight-of-hand. Electronic voting systems are presumed to be good value-for-money: the cost of independent auditing should have been factored in as an expected cost - and taxpayers are unlikely to complain about spending a few thousands to check that a voting system is fair, when there are far, far worse taxpayer bondoogles to be worried about, as a recent k5 article highlighted, courtesy of Ralph Nader.

The second is highly suspicious, as I'll explain below. What legitimate reason could the company have for classifying its records trade secrets? (Whatever the reasons, it's worth noting that this is not an isolated case of "trade secrets" stymying public access to public interest information - it's an endemic problem in the borgeouning "privatisation industry" - running rings around Freedom of Information rights in both the US and the UK.)

In any case, the AP article, perhaps unwittingly, presents an overly narrow debate: should there be independent auditing or not? This leaves aside the question of whether this system should ever have been adopted in the first place. It seems clear to me that the following minimum standards should be required of any electronic voting system used for public elections:

  1. An independent paper trail, recording individual votes: either generated by the voting machines at the time of voting, or paper ballots which are used as input to the machines and kept for auditing purposes. This is fundamental, because it is not feasible to dismantle every piece of hardware before/after every election and check that it has not been tampered with. With pencil and paper voting systems, or other simple systems, this is not so much an issue because the "paper trail" is precisely what is used for tallying - but with an electronic system no indendent paper trail implies no evidence if someone tampers with a machine and then resets it. (An additional feature: Public key cryptography could even be used to provide anonymous traceability for individuals to check that their own particular vote(s) had been recorded correctly.)
  2. This should go without saying - but essential paper records of elections should not be destroyed or dumped.
  3. No trade secrets over any aspect of the process. Elections must be a totally transparent, independently auditable process, in case of accusations of fraud which need to be investigated through the judicial system. Companies can still use patents to protect any truly non-obvious "innovations" they make (leaving aside the issue of the USPTO granting dubious patents for now) - and if their systems are devised in an obvious way, they have no legitimate reason to classify them as "trade secrets" anyway.
  4. By implication, all source code to all software used - including operating systems - should be available for unrestricted public review at no charge. (This does not imply that it should necessarily be fully open source - i.e. freely available to copy and modify - which would be nice, but is not essential for transparency.) There is no good reason that I can see that the source code should be hidden - except of course if the system has vulnerabilities, and in this case the vulnerabilities should of course be fixed rather than hiding the source code, which is a mere temporary band-aid - and a very ineffective one at that.
  5. If a serious vulnerability is found after an election, the election should simply be re-held after an extensive code audit, just to be on the safe side - this is again much preferable to a closed-source system, which gives the company owning the code a huge incentive not to be open about failures which could have jeopordised the impartiality of the system. (There should not be a "landslide cut-off point" beyond which it is not necessary to hold a reelection, because that would be vulnerable to political manipulation to shift it too low.)
In theory, minimal conditions such as these should be deducible from consitutional rights to a fair election process and basic common sense. However, it does not seem like the place of the courts to create such detailed rules as "interpretations" of constitutional law, even if they are in theory a consequence of constitutional rights. Perhaps the US government should pass a federal law laying down minimum standards for vote tallying systems, and electronic ones in particular - which would still allow individual counties etc. to choose their own election systems subject to these standards. After all, government procurement is commonly subject to detailed rules and regulations - why should this not be the case for something that is so vital to a healthy democracy: a transparent and auditable election process?

One last thought: if we are going to keep a paper trail of every single vote anyway - and this would need to be consulted every time there are "credible allegations" of fraud after an election - why not simply count all votes manually in the first place? It might be more expensive, but you'd have to successfully bribe quite a few people without getting caught to significantly influence a manual count - providing time-outs don't enter into the equation as they did in the 2000 presidential election.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
Best voting system?
o Closed source, electronic 3%
o "Public source", electronic 24%
o An electronic system that meets all the minimum standards given to the left 5%
o Pencil and paper / pen and paper 35%
o Electronic voting, manual counting 3%
o Other 4%
o Bah, elections never change anything anyway, so who cares? 5%
o This poll is fixed - I demand a recount!! 17%

Votes: 108
Results | Other Polls

Related Links
o controvers y
o AP reported
o a recent k5 article highlighted
o anonymous traceability
o dumped
o Also by greenrd


Display: Sort:
Government Audited versus Public Source Electronic Voting | 135 comments (132 topical, 3 editorial, 0 hidden)
When are you Americans going to learn? (4.62 / 8) (#1)
by sticky on Sat Jul 20, 2002 at 08:06:02 PM EST

Do it the Canadian way. All you need is a some pencils, ballots, cardboard boxes, and a bunch of people hired off of the street to run an election. Granted, there is a little more to it but that is the basic gist of it. The system is accountable since the ballot counter signs off on the ballots they count. Partisan scrutineers from any party that wants to send them can oversee the counting process but have absolutely no say in how they are counted; they can contest the questionable ballot(s) with the electoral body (which, incidentally, is non-partisan).


Don't eat the shrimp.---God
Yupper. (5.00 / 4) (#8)
by haflinger on Sat Jul 20, 2002 at 08:26:11 PM EST

And, if you really want computers, you can set them up the way Toronto did in the last municipal election. (I'm not there now, but I was there then. I actually got to vote for a councillor who I liked and who won. I was amazed. He's still doing a good job too. But I digress.)

That is, you've got paper ballots, which are fed through a scanner which scans for which circle has been blacked out.

Did people from the future send George Carlin back in time to save rusty and K5? - leviramsey
[ Parent ]

Absolutely (5.00 / 6) (#28)
by ajf on Sun Jul 21, 2002 at 05:31:44 AM EST

I have not seen a compelling argument for these mechanical systems America uses. Even if they are cheaper than hiring enough people to count pencil-and-paper votes manually, the resulting legal action surely more than outweighs any benefit. And it is completely idiotic to accept a contract where such vital machinery cannot be independently tested. It seems to me that governments these days are accepting ludicrous conditions imposed in their contracts with private enterprise.

"I have no idea if it is true or not, but given what you read on the Web, it seems to be a valid concern." -jjayson
[ Parent ]
"Trade Secret" (3.60 / 10) (#2)
by momocrome on Sat Jul 20, 2002 at 08:12:42 PM EST

I have a feeling the trade secret is a cell-phone glued to the motherboard. Dial-a-vote functionality for whoever orchestrated the chain of kickbacks that got these multi-million-dollar machines purchased in the first place.

The people of that community need to rip open one of these boxes pronto, warranty be damned.

"Give a wide berth to all that foam and spray." - - Lucian, The Way to Write History

paper voting (5.00 / 5) (#3)
by linca on Sat Jul 20, 2002 at 08:15:40 PM EST

Why is there any need for voting machines? Paper voting works perfectly throughout Europe, and is reliable, fast and cheap. The counting is done by citizens, usually voluntarily, working for the democracy. It is over in a couple of hours except when recounts are needed. Recounting doesn't change the actual votes. And human counting is amazingly reliable.

Why bother with technology when there is absolutely no need for it?

Amazingly reliable? (4.00 / 3) (#4)
by greenrd on Sat Jul 20, 2002 at 08:20:43 PM EST

And human counting is amazingly reliable.

Perhaps reliable enough - but amazingly reliable? Compared to what?

Typically you get a different answer every recount.


"Capitalism is the absurd belief that the worst of men, for the worst of reasons, will somehow work for the benefit of us all." -- John Maynard Keynes
[ Parent ]

to Florida Machines... (4.33 / 3) (#9)
by linca on Sat Jul 20, 2002 at 08:30:40 PM EST

Even elections over 70000 votes with a margin of a few dozen votes are not contested. I doubt you get error rates over one in a thousand.

[ Parent ]
Tit for tat (3.00 / 1) (#30)
by Nick Ives on Sun Jul 21, 2002 at 06:57:22 AM EST

Maybe the reason why handcounted elections are not contested in a margin of a few dozen votes is because otherwise you'd end up with endless tit for tat handcount recounts as both sides took advantage of the "different answer every time"...

--
Nick
morning! time for shower...

[ Parent ]

That happens (5.00 / 2) (#48)
by Simon Kinahan on Sun Jul 21, 2002 at 03:26:17 PM EST

In the UK, the returning officer generally tells the candidates to get lost if they ask for recounts without clear cause. The system seems to work pretty well: some seats come down to a few tens of votes difference, and yet the counting comes out consistent.

Simon

If you disagree, post, don't moderate
[ Parent ]
Tribulations of a Returning Officer (5.00 / 4) (#40)
by thebrix on Sun Jul 21, 2002 at 01:25:01 PM EST

My maternal grandmother was a Returning Officer (the official, usually nominated by local solicitors, who declares the final constituency result) in five United Kingdom General Elections (1945 to 1959) and, in one case, returned a majority of 88 out of a good 50,000 votes cast.

The difficulty with paper ballots is interpreting poorly-made marks. What she did was set aside the absolutely uncontestable ballots; then she got every candidate to agree on most of the rest; then she was forced to make a ruling on those that remained (to one party or the other, or striking out the vote). As I recall there were about eighty such ballot papers which fell into the third category, so her decisions swung the result.

No voting system is perfect, but the whole process right from the start of the count was in full view (with the disputed ballot papers being laid out on a table and scrutinised there) and everyone agreed it was fair, even although the final result was not to everyone's satisfaction.

Things appear not to have changed much since then, as this excellent BBC News Online page describes.

Plug in passing: for students of UK politics David Boothroyd's site is a phenomenal resource.

[ Parent ]

democracy by the numbers (none / 0) (#45)
by eLuddite on Sun Jul 21, 2002 at 03:14:37 PM EST

Human counting becomes amazingly reliable if only a random sample of the population is selected and allowed to vote. What is more, statistics tells us the probability of an election error in this random sample will be less than counting everyone's vote. All that is true, but uninteresting. What is interesting is the belief that voting is some kind of inalienable right rather than a method for choosing representation.

---
God hates human rights.
[ Parent ]

"You're a voter!" (none / 0) (#114)
by ZorbaTHut on Mon Jul 22, 2002 at 11:02:36 PM EST

I gotta admit, I really like this idea. It also has the bonus that people might get more interested in voting - "well, my vote won't matter" starts being a little harder to accept when you know full well that only one person in, say, a thousand is chosen to vote. So maybe we'd get more turnout.

On the other hand, it's got the potential flaw that the population might not be selected randomly. I can just see someone figuring out how to weight the machines towards, say, the midwest :P

Not that I have any better ideas, I just like pointing out the flaws in everyone else's. Look, I'm a typical internet forum reader! *runs*

*ahem* as I said, I like the idea ;)

[ Parent ]

Different answer every recount? (none / 0) (#103)
by lunatic on Mon Jul 22, 2002 at 02:53:47 PM EST

Typically you get a different answer every recount.

And on what evidence do you base this statement? I have served as a scrutineer in three federal, two provincial and two municipal elections in Canada. In each of these elections, I was watching a half-dozen or so polls. I have seen maybe three polls change on recount. This includes more complicated ballots, like municipal councillor races (pick 6 of the 10 candidates).

Each of the polls were counted at least twice, in full view of the scrutineers (representatives of candidates) and returning officers. Very few ballots were disputed, as the rules for what constitutes a properly marked ballot are laid out clearly in the statutes.

Moreover, the ballots are counted in the same room where they were cast. At close of voting, the doors are locked, the seals broken, and the ballots counted. The RO's are responsible for balancing ballots out with ballots in. Candidates observe the count, and upon completion of the count, the ballots are placed in envelopes (candidate X_1..X_n, rejected) which are sealed and initialled by the scrutineers and RO's.

[ Parent ]

Florida got a different number every recount(NT) (none / 0) (#132)
by acronos on Wed Jul 24, 2002 at 03:27:05 AM EST



[ Parent ]
this is all a result of... (4.33 / 3) (#5)
by momocrome on Sat Jul 20, 2002 at 08:21:05 PM EST

These electronic voting initiatives springing up in the US are the result of the controversy surrounding our last Presidential Election.

There were numerous improperly scored ballots that became hotly contested. The election was so close that a handful of votes in Florida became crucial in determining the Presidency. The electronic system is meant to circumvent this problem.

"Give a wide berth to all that foam and spray." - - Lucian, The Way to Write History
[ Parent ]

er, (4.50 / 2) (#7)
by momocrome on Sat Jul 20, 2002 at 08:23:27 PM EST

the electronic voting systems are meant to circumvent the improperly scored (or marked, punched, whatever) ballots, not the hadnful of votes swinging ther election bit.

feh.

"Give a wide berth to all that foam and spray." - - Lucian, The Way to Write History
[ Parent ]

But... (4.66 / 3) (#18)
by J'raxis on Sat Jul 20, 2002 at 11:54:17 PM EST

The electronics are not needed. Paper ballots can be designed to not fall prey to the inaccuracies that Florida’s particular paper ballots had. And Florida’s hole-punch things seemed to have had quite a few design flaws. Massachusetts ballots, for example: these ballots have a thick black arrow pointing to each candidates name. Each arrow has a gap in it. Voters fill in the gap in the arrow next to the candidate they want to vote for (Special thick markers are handed out with the ballots, so you don’t have to worry about people using dried-up pens or light pencils, or whatnot).

— The Raxis

[ J’raxis·Com | Liberty in your lifetime ]
[ Parent ]

no argument there (4.00 / 2) (#20)
by momocrome on Sun Jul 21, 2002 at 12:15:44 AM EST

just trying to describe the motivation for these gimcrack stopgaps. :)

"Give a wide berth to all that foam and spray." - - Lucian, The Way to Write History
[ Parent ]
A lack of gravity? (none / 0) (#96)
by ellF on Mon Jul 22, 2002 at 01:02:49 PM EST

Massachusetts ballots, for example: these ballots have a thick black arrow pointing to each candidates name. Each arrow has a gap in it. Voters fill in the gap in the arrow next to the candidate they want to vote for[...]

Interestingly, that wasn't the case when I voted - we had a punch-card setup in the town where I voted, Amherst. The fact that even in a single state such a variety of voting machines can be found is a manifestation of a larger problem: we have not taken our elections seriously at any point in the recent past, and probably longer.

It truly frightens me that in the wake of the 2000 election any state, let alone Florida, would be so careless as to use a system that is not publicly controlled and watched.

The problem, I think, lies not in what voting machine is being used, but rather in the underlying attitude taken by both state officials and individual citizens toward the electoral process. No bandaging measures - such as Palm Beach's recent switch to electronic voting booths - can address this fundamental weakness. In my opinion, it will not be until elections are treated with some degree of gravity by everyone involved that the process can improve.

[ Parent ]
you cannot use paper voting on the knowledge tree (1.00 / 2) (#79)
by johwsun on Mon Jul 22, 2002 at 06:02:01 AM EST

Well, you may can, but it is too difficult to implement. thats why we need electronic voting, to make things move faster.

[ Parent ]
Individual post-voting auditing dangerous (5.00 / 7) (#10)
by winthrop on Sat Jul 20, 2002 at 08:34:27 PM EST

If we are to have a secret ballot, it is very dangerous to allow individuals to audit their own ballots after the election is over. If it is possible for an individual to confirm who they voted for, it is possible for them to prove to somebody else who they voted for. This allows people to intimidate or bribe other people into voting for a particular candidate. While I doubt it would be possible to have a really thorough operation doing this without getting caught, it is definitely possible in certain situations with power imbalances: some marriages or employer-employee relationships, for example. Indeed, in those situations, it might even be that the possibility of such a post-vote audit would be enough to scare people into voting against their wishes, which makes this even scarier than the existing vulnerability absentee and mail-in ballots have to intimidation and bribery.

After one votes not necessarily well afterward... (4.50 / 2) (#17)
by J'raxis on Sat Jul 20, 2002 at 11:47:33 PM EST

I suppose post-vote auditing, if we’re using electronic equipment, could be something as simple as a You have voted for $FOO. Is this correct? dialog on a screen. As long as you cannot take something with you out of the ballot box, this eliminates the ability to prove how one voted.

— The Raxis

[ J’raxis·Com | Liberty in your lifetime ]
[ Parent ]

expanding that idea (4.60 / 5) (#26)
by izogi on Sun Jul 21, 2002 at 04:15:57 AM EST

I think that's the way to go, but if it's just on the screen then there's no way that the voter can be absolutely sure that what's shown on the screen is what's being recorded. There could still be a bug in the program.

A slightly more reliable way to do it would be to have the machine print out a paper receipt stating who the vote was for. The voter confirms that it's correct on the paper, and then it's dropped into a box, without any personally identifiable information attached to the vote. This probably happens automatically, although there would need to be a system to ensure that unconfirmed votes don't get dropped in or are clearly marked as being voided.

Once voting has finished, the system can provide a fast and immediate tally of the votes from memory. This would probably be correct, pending any software bugs. The concrete record is in the paper receipts that were confirmed by the voter though. These get counted manually afterwards for confirmation of the number.


- izogi


[ Parent ]
Till roll type thing... (4.50 / 2) (#47)
by gordonjcp on Sun Jul 21, 2002 at 03:25:03 PM EST

If you printed it out on paper, with nothing that identified the voter as such, you could wind the paper back onto another roll.  When you vote, it shows the bit of printout pertinent to you on the piece of paper behind a perspex window.

That way if the electronic vote recording screwed up, there would be  a hard copy that you could verify.

Give a man a fish, and he'll eat for a day. Teach a man to fish, and he'll bore you rigid with fishing stories for the rest of your life.


[ Parent ]
Position dependent (5.00 / 3) (#64)
by wnight on Sun Jul 21, 2002 at 09:01:12 PM EST

The problem with this is the sequence of voters is recorded. You could start with a known voter who votes for some obscure party (a write-in ballot for Mickey Mouse, maybe) and then with a list of the voters who used a particular booth after that, the vote-counters would have access to who votes for who.

The list of voters is easy to determine because the fact that someone voted isn't now protected, you could eaily loiter (especially with a team) around the voting booths recording this information.

It needs to be cut from the spool.

[ Parent ]

You're right. Never thought of that. (4.00 / 1) (#78)
by gordonjcp on Mon Jul 22, 2002 at 05:09:34 AM EST

Well, you would still need some way of making sure the printed vote wasn't taken out of the booth.

Perhaps cut it, and let it drop into a viewing window?  That way the voter would see the ticket cut loose from the till-roll, so no chance of identifying them from the sequence.  A wire ejector or perhaps a puff of compressed air would then dump the ticket into the box before the person left.

It's not hard to see how you could fit that with *two* bins, with clear chutes.  One would be for correct votes, one would be for dumped votes.  The voter would have final call when they saw the ticket in the viewing slot, and could then say "Yes, go ahead and vote" and the ticket would go to the "keep" pile, or "No, I want another go" and the ticket would go to the "dump" pile.  In both cases you'd see it go into the appropriately labelled bin.

Give a man a fish, and he'll eat for a day. Teach a man to fish, and he'll bore you rigid with fishing stories for the rest of your life.


[ Parent ]
Another idea (5.00 / 1) (#99)
by CaptainSuperBoy on Mon Jul 22, 2002 at 01:48:05 PM EST

Use a laser to engrave a wooden ball with the name of the candidate. The wooden ball can't be faked due to the grain of the wood. Then Tom Cruise will come out and wave his arms around for a while.

--
jimmysquid.com - I take pictures.
[ Parent ]
doesn't solve the problem (4.66 / 3) (#71)
by X3nocide on Sun Jul 21, 2002 at 11:17:43 PM EST

its quite simple to put up something like "you've voted for $f00. confim" and then call a vote(Repulican).

pwnguin.net
[ Parent ]
Yes, but it's repudiable (5.00 / 2) (#49)
by marxmarv on Sun Jul 21, 2002 at 03:28:30 PM EST

If we are to have a secret ballot, it is very dangerous to allow individuals to audit their own ballots after the election is over.
If you don't want to audit the vote, or if the audit poses dangers for you, you lose the card or put it in a drop safe for some Leage of Women Voters worker to account for by proxy, breaking you from the chain of custody. If you trust the LWV more than you trust your own circle, that may be the thing to do. Verifying a signed hash of your ballot instead of the raw ballot retains confidentiality, but prevents the public verification of the totals. An unsigned hash of your ballot is vulnerable to dictionary attacks if one knows the structure of the ballot, and thus doesn't increase confidentiality much over verifying the raw ballot.

Is the state more corrupt than its people? If not, maybe we deserve the government we get.

-jhp

[ Parent ]

I think voting should be kept simple (5.00 / 3) (#80)
by izogi on Mon Jul 22, 2002 at 06:16:18 AM EST

I think you have to be careful with this. As soon as you make it possible for someone to verify who they voted for after they've left the polling booth, the problem re-emerges of making sure that someone doesn't come and point a gun to your head demanding you hand over the means to determine who you voted for... Even if it meant you going to the league of women voters and telling them you wanted to "check your vote". I don't know if this is exactly how you meant it to work.

From another angle, what's to prevent the future-elected government from passing a law that requires the League of Women Voters to place their private key into a government operated key escrow service so that past votes can be reviewed?

Another possibility is that 20 years from now, technology might have improved so that the cryptography of today can be easily cracked.. maybe it exists now and we just don't know about it. If they're kept on record anywhere in any form, future governments in a dramatically changed world could look through the votes of past elections to identify potential friends and people who should be executed immediately to be safe.

Probably the most objectionable problem about a signed hash is that it's not easy for everyone to understand how it works. The best thing about a paper-and-pen vote is that it's extremely simple, and everyone can relate to and understand it. You indicate who you want to vote for on paper, drop it in a box, and someone counts them. The whole process is very understandable by nearly everyone. As soon as computers start getting involved, 99% of the people don't understand or comprehend 99% of the process. Instead, they're required to trust something they don't understand.

I guess there will always be limits with anonymous voting booths. You have to trust someone somewhere along the line, including people who count the votes. There's probably a way to formally prove that you can't have total and secure anonymity as well as verification, but I'd be interested to actually see the proof.


- izogi


[ Parent ]
Good questions and points (none / 0) (#109)
by marxmarv on Mon Jul 22, 2002 at 04:41:40 PM EST

Even if it meant you going to the league of women voters and telling them you wanted to "check your vote". I don't know if this is exactly how you meant it to work.
No, as in you leave your proof of voting with the LWV and they then cart your ballots off to some location. Essentially you transfer your right to audit the system (and all the information tying your vote to you) to them. Your pseudonymous key never leaves the polling place, in that case.
From another angle, what's to prevent the future-elected government from passing a law that requires the League of Women Voters to place their private key into a government operated key escrow service so that past votes can be reviewed?
The LWV doesn't need a private key. Anyone can read out the vote information on the card, which is the pseudonymous key, the raw vote data, and the signatures. As to what happens twenty years from now, that's an operational problem, and the cards can time-bomb themselves after some period of time, say a few weeks after the election, or on command, thus obliterating any connection of your own identity to the vote cast.
Probably the most objectionable problem about a signed hash is that it's not easy for everyone to understand how it works. The best thing about a paper-and-pen vote is that it's extremely simple, and everyone can relate to and understand it.
This is very much true. Though from the user standpoint it can all be made very simple, iff home computers had smart card readers. From the user standpoint, verifying their vote could mean nothing more than firing up their browsers, keying four or five letters into a web form, and searching down the list for the rest of their "vote serial number". They can even hold the receipt up to the screen to compare. It's not a requirement that everyone check their vote after the results are tabulated, but even if only a small number of people bring discrepancies in their own votes to the attention of their social circle and others, that there is a fly in the soup won't be unknown.

As for signed hashes, only the geeks need to understand them. As far as the user is concerned, "This set of numbers is your vote, scrambled to protect your privacy. If this set of numbers (on your voter receipt|on your smart card) does not match the set on this web page, contact your local Election Commissioner immediately at [...]"

I guess there will always be limits with anonymous voting booths. You have to trust someone somewhere along the line, including people who count the votes.
"The people who cast the votes decide nothing. The people who count the votes decide everything." -Josef Stalin

You only have to trust that the system as deployed is honest and not subvertible. Through long experience I do not trust humans to accurately relay my desires up the chain of command, but I would trust this system.

There's probably a way to formally prove that you can't have total and secure anonymity as well as verification, but I'd be interested to actually see the proof.
You can always destroy the media. Remember that the only link between the identity of the voter and their vote is in the card. The voting machine must be trusted to not tie your identity to your vote, yes, and vote fraud may cause anonymity breaches if the owners of miscounted ballots step forward.

It's unfortunate that no matter what, you can't design a secret ballot completely impervious to fraud, but you can still get pretty close.

-jhp

[ Parent ]

Low tech verses High tech voting (none / 0) (#135)
by izogi on Thu Jul 25, 2002 at 04:21:24 AM EST

Thanks for the response. There are a few areas where I'm not sure I agree with you to do with simplicity of voting, but I won't bother with a point-by-point thing here. It's mostly summed up in the following:

You only have to trust that the system as deployed is honest and not subvertible. Through long experience I do not trust humans to accurately relay my desires up the chain of command, but I would trust this system.

In that case, you're still trusting the programmers who developed the system. Or in the case that it's open sourced every line of code has been audited to your satisfaction, you're trusting that the hardware developers have designed hardware that correctly executes the software. You're trusting that the deployers of the system have deployed the technology correctly, and haven't made an error or put in place a false polling booth, for example. Either way, you're trusting people somewhere.

It's a two sided problem. By using technology you've dramatically reduced the number of people capable of abusing it. You've also dramatically reduced the number of people capable of auditing it as well as understanding other people who audit it. Everyone relies on a minority. The advantage to low-tech voting is that although it may be more open to potential abuse, it's also easier for everyone to keep track of what the system implementors are doing, and understand the issues.

I'm personally more in favour of low tech voting because IMHO, it's important for as many people as possible to be able to understand the system and audit and trust people based on their own judgement. With high-tech voting, 99% of people will believe whatever the media tells them about an election being rigged, without any further information or the ability to make an informed decision. Unfortunately the modern western media will nearly always go with the most sensational viewpoint of any story.

eg. If a vote counting person is accused of making a mistake that might have resulted in them possibly mis-counting every 40th vote due to mis-understanding the layout of a voting form, everyone understands that and can make an intelligent judgement on it. On the other hand a if a programmer is accused of not catching a bug that caused the system to possibly mis-count every 40th vote, the general public would almost automatically assume a worst case scenario. Nobody understands it so somebody's probably trying to decieve them anyway.

I'm not trying to rule out technology in the voting process. It can definitely speed things up, and when implemented correctly it's probably more accurate. But I also think there should be a simple system behind it that everyone can understand, used as verification of the correct result.


- izogi


[ Parent ]
Well (none / 0) (#131)
by acronos on Wed Jul 24, 2002 at 03:20:29 AM EST

From another angle, what's to prevent the future-elected government from passing a law that requires the League of Women Voters to place their private key into a government operated key escrow service so that past votes can be reviewed?

If you can stop the problem now, why couldn't you stop it then. The same thing that allows you to stop the government from using a hash will stop the government from abusing that hash if it existed.

[ Parent ]

Chicago Style Voting (4.40 / 5) (#11)
by Sherman Peabody on Sat Jul 20, 2002 at 08:58:43 PM EST

As a point of interest, the city of Chicago uses the 'butterfly ballots' of the type that caused the fiasco in Florida. In the last election, they kept the machines, but had another machine to check that all the votes were accurately counted. I think this was a cheap way to ensure your ballot was correct because the machine let you know if you had a problem with your ballot, including undervoting (not voting for an office).

The main problem was that since Chicagoans also need to vote for judges and many other offices, almost everyone undervotes. The undervote warning is then ignored when it could be telling you your vote for President didn't get counted.

Idiotic and completely misses the point. (4.25 / 12) (#12)
by kitten on Sat Jul 20, 2002 at 09:52:34 PM EST

Can someone please explain to me why the hell individual districts should be able to use any kind of ballot they wish for a national election?

Butterfly ballots, punchcards, paper/pencil, optical scan, etc etc etc.. ladies and gentlemen, the last time I checked, this was the 21st century, the dawn of the information age, and we're still conducting our affairs like it was 1776.

All districts should be using the exact same ballot and voting system when it comes to national issues. There is not a single reason that I can conceive of that the preceeding sentence is false.

People sit around and bicker endlessly about what ballot is most effective and with what format and run "mock elections" and the hue and cry of hanging chads, dimpled chads, pregnant chads, illegible handwriting, broken machines, blah blah blah. Any fifth grader who has ever taken a standardized test in the US public school system could come up with the answer..

In my school we called them "Scan Tron", which was the propietary name of a machine that looked at what bubble you filled in with your #2 pencil. These machines are cheap, fast, exceedingly reliable (in 12 years of schooling, I can only recall three times where I had to complain to the teacher that the machine screwed up).. so why can't we use these?

See, here's how it would work:
WHO DO YOU WANT TO BE PRESIDENT?
A. ELVIS PRESLEY
B. CAPTAIN KIRK
C. NEAL STEPHENSON
D. AXL ROSE
Say the voter wants Captain Kirk as president, he takes his pencil, fills in "B" on his form, and drops it in the slot.
No muss, no fuss. No more bickering, no more confusion. And in a worst-case scenario, where the votes must be tallied by hand for whatever reason, it's easy to see what the voter wanted: Either he filled in the bubble or he didn't, and a human can easily tell if the voter was an idiot and only partially filled it in - no more trying to visually decode archaic punchcards from machines invented in the 1950s.

I honestly don't see what the problem here is.
mirrorshades radio - darkwave, synthpop, industrial, futurepop.
no national elections (5.00 / 7) (#13)
by Delirium on Sat Jul 20, 2002 at 10:07:48 PM EST

Officially, anyway. The only official national elections are where the representatives of the electoral college get together and vote for president. The elections you and I vote in, before that, are entirely up to the state to administer. In other words, it's up to each state to figure out on its own to whom its electoral votes should go.

[ Parent ]
I was gonna say that (3.00 / 1) (#14)
by Anonymous 242 on Sat Jul 20, 2002 at 10:28:11 PM EST

You beat me to it.

[ Parent ]
Yeah, which is also idiotic.. (4.25 / 4) (#25)
by kitten on Sun Jul 21, 2002 at 04:04:04 AM EST

Can someone please explain to me why we're still using the antiquated electoral college system? On some distant level, I can comprehend why it might have been useful two hundred years ago. Today, we can move information around at the speed of thought. I see absolutely no reason why we shouldn't conduct national elections by direct vote.

This would have a variety of benefits, not the least of which would eliminate situations like the 2000 election, in which Bush won the electoral votes, but Gore won the popular vote. Even if we all agree that the Florida debacle was clearly in Bush's favor, those votes from Palm Springs don't alter the numbers - more people voted Gore than Bush, yet Bush wins. Logic here, anyone?

(And I'm not just being a dick. I really do want to know if there's a reason we're still doing this.)

Furthermore, it seems to me that the electoral college more or less makes some people's voices unheard, and useless. Consider an area which leans very heavily Republican, say. A Democrat in this area might as well not even bother, right? If it's virtually forgone that all the electoral votes will be going Republican, why should the Democrat even go to the polls? His vote won't count for anything.

But if it was done via a one-to-one, direct vote, then it doesn't matter who he lives around and what their political leanings are. A vote for a candidate will be a vote for a candidate, regardless of his neighbors.

mirrorshades radio - darkwave, synthpop, industrial, futurepop.
[ Parent ]
Read this... 'How to Rig an Election' (5.00 / 2) (#29)
by mr strange on Sun Jul 21, 2002 at 06:48:18 AM EST

'How to Rig an Election' (from the Economist magazine). I think you might find it interesting.

intrigued by your idea that fascism is feminine - livus
[ Parent ]
Interesting article (4.00 / 1) (#69)
by Ian Clelland on Sun Jul 21, 2002 at 11:09:07 PM EST

And not surprising, of course, that people are willing and able to manipulate the electoral boundaries to entrench their own power.

But is it really as bad as the article makes it out to be?

The article mentions that there were fewer closely contested elections in 1998/2000 than in a long time. Shouldn't this imply that more people are being served by their favoured candidate than before? In a close vote, after all, closer to half of the constituents will be unhappy with the results.

Similarly, in the example of the Florida district containing so much of the beach areas -- doesn't it make sense to have the interests of such a (presumably) uniform demographic served by single representative with solid support? If the intention of a representative democracy is to have someone in government to represent your concerns, wouldn't every citizen rather be part of the majority in their district? If Florida's 22nd district was to be split up and amalgamated with others, the individual pieces might find themselves in the minority of every district of which they were a part, and without any proper representation as a group.

It seems that drawing electoral boundaries to homogenize the voting patterns of the enclosed constituents should actually work to make more people happy with the election results -- or is this a situation where increasing the local satisfaction actually skews the global results?

[ Parent ]

In some cases, it's worse. (none / 0) (#70)
by haflinger on Sun Jul 21, 2002 at 11:12:35 PM EST

Check out what's been happening in Maryland the last few months.

The Governor's proposed redistricting plan was ruled unconstitutionally undemocratic by the State Supreme Court. They're going bananas right now.

Did people from the future send George Carlin back in time to save rusty and K5? - leviramsey
[ Parent ]

K5 to the rescue (4.00 / 1) (#31)
by Irobot on Sun Jul 21, 2002 at 08:42:12 AM EST

Did you happen to see this article on K5 from awhile back? I remember it was a fairly well-balanced piece...

Irobot

The one important thing I have learned over the years is the difference between taking one's work seriously and taking one's self seriously. The first is imperative and the second is disastrous. -- Margot Fonteyn
[ Parent ]

Whoops (3.00 / 1) (#32)
by Irobot on Sun Jul 21, 2002 at 08:50:23 AM EST

I'm not sure you'll have access - I just looked at the voting record and it only had a score of 66. I'm not sure what that means as far as stories go...

Irobot

The one important thing I have learned over the years is the difference between taking one's work seriously and taking one's self seriously. The first is imperative and the second is disastrous. -- Margot Fonteyn
[ Parent ]

Why not BOTH? (4.50 / 2) (#35)
by treat on Sun Jul 21, 2002 at 10:54:56 AM EST

Can someone please explain to me why we're still using the antiquated electoral college system?

Those that argue in favor of the electoral college do make compelling arguments. But then it occurred to me - wouldn't it be best if the winner had to win both the popular vote AND the electoral college? Winning both is the usual case anyway of course, and surely if they do not win both, they do not have a clear mandate.

[ Parent ]

what happens? (5.00 / 1) (#36)
by rhyax on Sun Jul 21, 2002 at 12:28:18 PM EST

what would happen if they didn't win both in your system?  (i.e. they won one and not the other, like the last presidential election)

[ Parent ]
Obviously (5.00 / 2) (#68)
by carbon on Sun Jul 21, 2002 at 10:53:57 PM EST

We just have Rusty serve as president for 4 years, with an overclocked Zaurus as vice president, and next election we do it all over again.

Well, that's how I'd do it.


Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Original intention (5.00 / 1) (#72)
by X3nocide on Sun Jul 21, 2002 at 11:35:07 PM EST

Was to hash it out in Congress. The idea was that the states would vote for all their favorite sons and then Congress would chose between the top 2 or 3 if I recall. Of course, they also intended the runner up to be the Vice President. Thankfully the Constitution itself is not holy writ, but rather a modifyable document of good intentions.

pwnguin.net
[ Parent ]
the problem with that (4.00 / 1) (#51)
by Delirium on Sun Jul 21, 2002 at 04:25:15 PM EST

Is that it would essentially mean that the major urban areas (New York, Chicago, Los Angeles, San Francisco, Atlanta, Houston) would control the United States, due to their overwhelming population. Thus only things of interest to urban residents would ever be considered.

[ Parent ]
No, no, no! (4.00 / 2) (#56)
by kitten on Sun Jul 21, 2002 at 05:37:27 PM EST

Is that it would essentially mean that the major urban areas (New York, Chicago, Los Angeles, San Francisco, Atlanta, Houston) would control the United States, due to their overwhelming population.

No, no, a thousand times no. That's my whole point, you see. The way it stands now, population centers are the key. Texas, California, etc. And a minority vote in one of those areas doesn't count for anything.

By using a direct, one-to-one voting system, a Democrat vote in New York counts just as much as a Democrat vote out in the middle of nowhere. One person, one vote. John lives in Atlanta, and Joe lives in a place where his closest neighbor his fourteen miles away - and both of their votes count precisely the same: One.


mirrorshades radio - darkwave, synthpop, industrial, futurepop.
[ Parent ]
Not quite (4.00 / 1) (#58)
by Anonymous 242 on Sun Jul 21, 2002 at 06:14:05 PM EST

<blockquote><em>The way it stands now, population centers are the key. Texas, California, etc. And a minority vote in one of those areas doesn't count for anything.</em></blockquote>

The same effect would be greatly exagerated in a direct election. The twenty largest cities in the US hold close to one fifth of the entire US population. In the current system, some effort must be made to reach rural areas because one must receive a majority of electoral votes and the electoral votes are spread out amongst all the states proportional to population. In a direct vote, the large cities would dominate and in most cases predominantly rural states would have absolutely <em>no</em> say in electing the president.

<p>As something of a tangent, it is also a fallacy that one's vote is wasted if one's candidate does not win or if one's vote is not significant in the election of who one's vote is cast for. While the primary purpose of an individual's vote is <em>usually</em> to attempt to put a candidate into office, there are also many other purposes for voting.


[ Parent ]

Hmmm... (4.00 / 1) (#87)
by Rhinobird on Mon Jul 22, 2002 at 08:40:33 AM EST

Is there any precident for the splitting into parts of a state? To become a state you have to have a minimum population. I wonder if there might be cause for making a maximum population. This might help spread out some of the electoral votes. Of course it would probably be best to leave the ultimate descision up to the states in question (Califonia, New York, etc). But I wonder if the issue could be brought up to the states on the behalf of the rest of the republic..
"If Mr. Edison had thought more about what he was doing, he wouldn't sweat as much." --Nikola Tesla
[ Parent ]
that's not the point (4.00 / 2) (#59)
by Delirium on Sun Jul 21, 2002 at 06:14:50 PM EST

Nearly 80% of the U.S.'s population lives in urban areas. Thus politicians will only cater to issues of interest to urban voters -- anything that would be only (or primarily) of interest to rural voters will never be discussed at all, because it's not worth the effort to get a measly 20% of the vote. So federal money, projects, etc. will be essentially exclusively focused on urban areas.

[ Parent ]
Why is this bad? (4.00 / 1) (#61)
by ShadowNode on Sun Jul 21, 2002 at 06:48:44 PM EST

Shouldn't a democracy be more concerned with where the majority lives? Should other minority interests be catered to in this way as well?

[ Parent ]
minority interests should hae some voice (4.00 / 2) (#62)
by Delirium on Sun Jul 21, 2002 at 07:41:35 PM EST

In fact, one of the founding principles of the U.S. is that the majority should not have absolute power. This is why the Supreme Court can overturn laws that were passed by a majority vote and supported by a majority of the population, for example.

But yes, it should be more concerned with where the majority lives, just not to an extreme. And that's how it currently works -- populous states get more votes than less-populous states. But the less populous states get more representation than they would in a direct-election scheme. For example, the 12 least populous states in the U.S. (counting D.C. as a "state", since it has votes) have a combined 44 of 538 electoral votes -- 8.2%. Not a ton, but enough to at least have some voice. In a direct election, they would combine for 4.1% of the vote -- half as much. At that level, a presidential candidate could very easily completely ignore the interests of even the 15-20 least populous states.

So while people now complain that California, Texas, New York, Florida, and a few other popular states are the only ones candidates concentrate on, you'd have the same problem in a direct election system, only it'd be far worse.

[ Parent ]

*some* voice, yes (4.00 / 1) (#84)
by ShadowNode on Mon Jul 22, 2002 at 07:05:35 AM EST

But why is geographical location so special? Should there be an electoral college for religious believes as well? Don't minority religions deserve a voice as much sparsely populated areas?

[ Parent ]
you're kidding right? (3.50 / 2) (#85)
by Rhinobird on Mon Jul 22, 2002 at 08:35:00 AM EST

Well for one we don't live in heaven, we live on earth. Second that would clearly set up the government to respect the establishment of a religion. Damnit, doesn't anybody read the flipping constitution anymore? amendment one: Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the government for a redress of grievances. If you want to set up your religion like the US government then go ahead. Then you could get your own electoral college for you religion. Or you could just become Catholic...only there they use cardinals to elect the Pope, but it's the same principal.
"If Mr. Edison had thought more about what he was doing, he wouldn't sweat as much." --Nikola Tesla
[ Parent ]
minor quibble (1.50 / 2) (#86)
by Rhinobird on Mon Jul 22, 2002 at 08:36:35 AM EST

US is NOT a democracy. It's a representative repulic. They really ought to do a better job in high school civics classes.
"If Mr. Edison had thought more about what he was doing, he wouldn't sweat as much." --Nikola Tesla
[ Parent ]
How about a balance between the two (4.00 / 1) (#73)
by archmedes5 on Mon Jul 22, 2002 at 12:30:08 AM EST

After tallying the popular vote you take the winner of votes within each senatorial district and add 1/100th of the popular vote to the total.

An example

Candidate one gets 750,000 votes

Candidate two gets 1,000,000 votes

Candidate one won 60 of the 100 senatorial districts

Candidate two won 40 of the 100 senatorial districts

Total is 1,750,000 votes so

Candidate one gets 750,000+(1,750,000/100)*60 which is 1,800,000

Candidate two gets 1,000,000+(1,750,000/100)*40 which is 1,700,000

This may not seem quite fair at face value, but you have to consider that pure popular vote puts the power into the hands of the urban centers.  Low population states like alaska or montana, or rural states like Iowa or nebraska don't have enough people to push the popular vote in the proper direction far enough.  By including the votes for each state and giving them enough weight to influence the outcome of the vote, you force those running for presidency to consider the smaller states the same as the larger states.

[ Parent ]

Your vote never counts (4.00 / 1) (#74)
by dipierro on Mon Jul 22, 2002 at 01:03:57 AM EST

Furthermore, it seems to me that the electoral college more or less makes some people's voices unheard, and useless. Consider an area which leans very heavily Republican, say. A Democrat in this area might as well not even bother, right? If it's virtually forgone that all the electoral votes will be going Republican, why should the Democrat even go to the polls? His vote won't count for anything.

Actually, with the electoral college system your vote is more likely to count. What are the chances that the popular choice for the entire country will come down to a single vote? Virtually nothing. What are the chances that the result for your state will come down to a single vote? Well, virtually nothing, as well, but a much bigger virtually nothing.

In any case, you're certainly right that the system is antiquated. Ever since the 16th Amendment the federal government has had far too much power to leave the choice of the president up to the states to decide.



[ Parent ]
Depends on what you mean by "won" (4.00 / 1) (#91)
by The Private Fedora on Mon Jul 22, 2002 at 10:27:39 AM EST

"...Bush won the electoral votes, but Gore won the popular vote."

These two statements do not mean the same thing. Bush "won" the electoral vote because a majority of the electoral votes were cast for him. Assuming the same definition of "won," Gore did not win the popular vote because he failed to have a majority of the popular votes cast for him. Gore only won 48.38% of the popular vote. This gave Gore a plurality, but not a majority. Bush won 47.87% of the popular vote. This leaves 3.75% of the population voting for someone other than Bush or Gore. So 51.62% (Bush + other) of the people voted for "not Gore" and 52.13% (Gore + other) of the people voted for "not Bush."

If one were to abolish the electoral college, only replacing the electoral vote with the popular vote, neither major candidate would have a majority, and the election would be decided by the House of Representatives.

All statistics from fec.gov

-------
"Is life so dear, or peace so sweet, as to be purchased at the price of chains and slavery?"
Patrick Henry, The War Inevitable, March 23, 1775
[ Parent ]

I many other countries.. (none / 0) (#98)
by Weezul on Mon Jul 22, 2002 at 01:28:57 PM EST

..this kind of problem is solved by making deals with that 3%.  I don't think this solution would work very well for the president (you should just call a national revote instead of sending it to congress), but it would work for any possition where you have multiple people win.

Example: We could modify the house to not represent individual states and instead represent regions of 15 million people.  There would be 15 representitives, 5 ellected every 2 years, with 6 year terms.  You would simply vote for your prefered person from a list of 10-20 candidated.  The 5 candidates with the highest precentage would win and get seats.  Candidates with lower precentages would have the option to give their precentage to one of the 5 winners.  Once it was all said and done, the 5 winners would not have one vote per representitive, instaead they would vote their final precentage.

This one dose a reasonably nice job of getting someone to represent everyone without having too many representititves.  By the arrow theorem, no voting system is perfect, but I think you could show that the classic voting system problems only occure here for very small parties.  I think there are "marketing related problems" that could occure for larger parties (suppose the two major parties allways win 4 or 5 seats and the smaller parties give their support to the least objectionable large party guy, it might be better to use strategic voting to win 3 seats and have a higher chance to recieving more 3rd party precentage), but one dose not really expect to remove marketing related problems.

"Fascism should more appropriately be called Corporatism because it is a merger of state and corporate power." - Benito Mussolini
[ Parent ]

Good ideas... (none / 0) (#100)
by The Private Fedora on Mon Jul 22, 2002 at 02:08:12 PM EST

I would rather see deals made or a national re-vote than a president chosen by Congress. Currently, the House of Representatives decides the outcome of any election where no one candidate has a majority of electoral votes. This allows a president to be chosen much sooner than in the case of a national re-vote. The amount of time needed to conduct a re-vote would have been seen as too great to the writers of the US Constitution. This was during a time when the vote count would not be complete until three or four months after the election.

-------
"Is life so dear, or peace so sweet, as to be purchased at the price of chains and slavery?"
Patrick Henry, The War Inevitable, March 23, 1775
[ Parent ]
speaking from being in school... (4.00 / 1) (#15)
by yodason on Sat Jul 20, 2002 at 10:47:47 PM EST

scantrons screw up ALOT more then that =p

[ Parent ]
odds of scantron screwup (none / 0) (#16)
by yodason on Sat Jul 20, 2002 at 10:49:17 PM EST

ah damn, hate when I rember something second after I post, but the error rate is maybe 1 in 100, 1 in 75 if your messy.

[ Parent ]
also remember (4.50 / 2) (#37)
by rhyax on Sun Jul 21, 2002 at 12:40:40 PM EST

while thinking about it, i would think scantron/optical-scan (OS) would be the best too, it's simple, easy to read etc. i don't think that many people who are familiar with the system mess it up nearly 25% of the time. but when you're voting in a big election you have old people to think about, people that have never seen a os entry-system. also people from other countries.

it's not entirely obvious to people using it for the first time how the system works, clearly you could show examples on the page (like scantron does) but people are fairly dumb. in austin, tx we used os entries and a friend of mine volunteered to count ballots. i asked her, "should you really have to count very many, since they were scanned?" she just laughed, apparently people are even dumber than i thought, the forms we used were not multi-use, meaning the candidates names were on the same sheet to be sent through the reader, to make sure there was no confusion about who they voted for many people circled the candidate they wanted, often writing additional information suck as "this other candidate X sucks!" not to mention all the extra marks which don't let the scanners work... personally i think if you can't figure out the instructions we don't need your vote, but others may not agree :)

[ Parent ]

More accurate (4.83 / 6) (#19)
by J'raxis on Sun Jul 21, 2002 at 12:05:16 AM EST

A more-accurate system might involve allowing the voter to use a machine to mark the ballot. He presses a button and the machine makes a clear, bold, black, unambiguous, indelible mark on the paper. This will eliminate idiotic mistakes like putting little tick marks in the circle, or not pressing hard enough, or scribbling all over it, or whatever.

This would unfortunately be similar to the punch card system, however the mark would be in ink, not a hole in the paper, and it would not rely on the pressure the user applied, so you would not end up with “hanging” or “pregnant” chads. Just press the right button, the machine stamps a big black on the paper.

— The Raxis

[ J’raxis·Com | Liberty in your lifetime ]
[ Parent ]

Use a pen. (4.00 / 1) (#21)
by delmoi on Sun Jul 21, 2002 at 12:21:18 AM EST

In the last Iowa elections we used felt tip pens. The reason to use a pencil is so that you can erase an answer on a STD test if you change your mind.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
in two words? (3.50 / 4) (#22)
by /dev/trash on Sun Jul 21, 2002 at 02:42:51 AM EST

State's Rights.

---
Updated 02/20/2004
New Site
[ Parent ]
Difficulty of compromising the systems (4.16 / 6) (#23)
by fencepost on Sun Jul 21, 2002 at 03:05:24 AM EST

One advantage of having a variety of disparate systems is that it's much more difficult to compromise the entire system if there is no single system to compromise.  

This hasn't been a huge issue in the past in part because of all the disparate systems and in part because compromising all those paper-based systems reliably would require significant financial and time outlay.  Having a single national standard system would make it simpler, but not vastly so - producing counterfeit ballots pre-punched with a somewhat random distribution that matched the results you wanted would be only slightly simpler with a standard ballot format.

With electronic ballots it becomes a much more serious issue:
Consider if you will a national standard closed-source electronic voting turnkey system produced by a large and politically influential software company.  
Consider that system communicating encrypted results electronically back to a central tallying system over the most convenient and reliable network available - one designed to keep working after a nuclear war, perhaps.  
Consider someone finding a way to crack those voting machines.
Consider CodeRed done right, done low-key, done to subtly shift perhaps only a couple percent of the vote.

I'd much rather have a heterogeneous environment of voting systems.
--
"nothing really says "don't hire me, I'm an idiot" quite as well as misspelling "pom-pom" on your resume." -- former Grinnellian
[ Parent ]

Exactly.. (4.66 / 3) (#24)
by kitten on Sun Jul 21, 2002 at 03:50:52 AM EST

..which is why I don't suggest using an electronic system across the board. Just a clear, unambiguous, definite black mark on a piece of paper. A piece of paper with choices clearly marked A, B, C... and candidate's names clearly corresponding to one of each. Simple.

Each district then tallies the results, and sends them in using whatever means they've been using. No amount of electronic tomfoolery is going to be able to play havoc with such a system.


mirrorshades radio - darkwave, synthpop, industrial, futurepop.
[ Parent ]
More difficult to compromise? (none / 0) (#97)
by lunatic on Mon Jul 22, 2002 at 01:28:22 PM EST

One advantage of having a variety of disparate systems is that it's much more difficult to compromise the entire system if there is no single system to compromise.

This assumes that the system is a parallel system and can tolerate failures in individual components (the precincts), so long as there are at least some components are still functioning. I don't believe this is an adequate model of the electoral system.

If some of the components fail. The entire system should be considered to be in a failure state.

Failure states in components can be detected much more easily if the components are in wider use.

[ Parent ]

the problem is (none / 0) (#107)
by aphrael on Mon Jul 22, 2002 at 03:33:49 PM EST

that for financial reasons (and because turnout is higher), most states combine national and state elections into the same election, and most localities combine their local elections into the national-state elections. Which (a) makes a national standard for national elections a de facto standard for state and local elections, and (b) causes problems because different local political districts actually have elections for different things.

[ Parent ]
I agree, except for one critical point (4.33 / 6) (#27)
by zvpunry on Sun Jul 21, 2002 at 05:06:22 AM EST

You write:
If a serious vulnerability is found after an election, the election should simply be re-held after an extensive code audit, just to be on the safe side

I'm actually convinced this will be accepted by the Republicans and Democrats. It's too politically advantageous not to.

Imagine a technically savvy Democratic consultant finds a potential security compromise. He tells his superiors, and it works its way up, until the decision is made: we say nothing, unless we lose the election.

So he loses, and they announce the problem. First, he's given a new election, and several more weeks of campaigning to make a difference. Then, the mere allegation that there was a potential security compromise, and that certain Republican consultants might have had motive to rig the election (nevermind the means to do it), might swing the election in the Democrat's favor.

It almost happened two Novembers ago with the more traditional ballots. This would be a timebomb, because all software has glitches. Then there's the conspiratorial angle, where each party will attempt to legislate new "features" into the voting system, each with their own known security holes, just in case they need to use it in a future election. This can be alleviated with full open disclosure to all citizens (not only party consultants), but not completely eliminated.

No, the onus should be on the loser to clearly and accurately demonstrate that there was an election compromise, before an election is reheld. There are election irregularities, well, regularly. Imagine if all it took was that some single citizen somewhere had the motive and means to berate his wife into voting for his candidate, in order to order an election reheld. That's not too far afield from what you're suggesting.

america's new political system (none / 0) (#115)
by ZorbaTHut on Mon Jul 22, 2002 at 11:17:00 PM EST

"We'll pay you a thousand bucks to vote for our opponent, and if he wins, say he bribed you."

"Okay."

[ Parent ]

Missing option poll (4.25 / 4) (#33)
by Betcour on Sun Jul 21, 2002 at 09:21:19 AM EST

How about "paper" ? Why would you need a pencil ? Over here you just pick the paper with the name of the candidate you want on it and put it in the enveloppe. There's no doubt what the vote is, no problem with electronic and nobody contesting the results.

vote for many people (5.00 / 2) (#38)
by rhyax on Sun Jul 21, 2002 at 12:43:24 PM EST

the elections for the us are usually for a lot of positions from national down to city level.

[ Parent ]
Easy (4.00 / 2) (#39)
by Betcour on Sun Jul 21, 2002 at 12:59:44 PM EST

Then use different enveloppe, one per election (or if applicable vote for groups and not just individuals)

[ Parent ]
California. (4.75 / 4) (#46)
by haflinger on Sun Jul 21, 2002 at 03:21:04 PM EST

There can be in some cases hundreds of items to fill out on a California ballot. That is, one for each referendum, plus a large slate of municipal, state and federal legislators, executive branch, and judicial branch. All being elected.

Did people from the future send George Carlin back in time to save rusty and K5? - leviramsey
[ Parent ]
Then prioritize (4.00 / 1) (#82)
by QuickFox on Mon Jul 22, 2002 at 06:52:34 AM EST

Then use the practically failsafe preprinted slips for the most important elections, such as electing the President. Use marks on paper for the hundreds of referendums.

Give a man a fish and he eats for one day. Teach him how to fish, and though he'll eat for a lifetime, he'll call you a miser for not giving him your fi
[ Parent ]
(OT) Direct democracy? (4.00 / 1) (#83)
by QuickFox on Mon Jul 22, 2002 at 06:53:45 AM EST

Choosing and voting for hundreds of different offices and issues sounds like a huge amount of work for every voter. Then only people who have lots of free time can explore and make informed choices on every issue.

This sounds like direct democracy rather than representative democracy. The great problem with direct democracy is precisely this: Only people who have lots of free time get a say.

Give a man a fish and he eats for one day. Teach him how to fish, and though he'll eat for a lifetime, he'll call you a miser for not giving him your fi
[ Parent ]

The Biggest Human Error (3.12 / 8) (#41)
by holycola on Sun Jul 21, 2002 at 02:41:05 PM EST

I'm surprised at the amount of attention given to the mechanics of voting in the 2000 presidential election, compared to the lack of attention given the obscene conflict of interest in having Katherine Harris decide who would get the college votes for Florida. The woman ran the repulican campaign in Florida, and then is left to decide who wins the state? Who fucking cares how the votes are counted when this kind of shit covers the steering wheel?

-----
This is not a sig.
Katherine Harris played it by the book. (3.87 / 8) (#44)
by Skywise on Sun Jul 21, 2002 at 02:56:11 PM EST

Personal opinions of her political ambitions aside, she did nothing illegal.

Florida state law was, and is, that there was a 48 hour window in which to contest the election results.

Gore contested TWO counties, not the entire election, TWO counties.

ONE WEEK LATER, Harris signed off on the votes of the other counties approving them as THE count.  At no time during that week were any other (official) contestations made about the election results.

Could she have NOT signed off on the results?  Sure, if she were Democract she probably would have taken that course of action.  And that would've been questionably illegal too.

[ Parent ]

The sky is falling, the sky is falling! (3.75 / 4) (#42)
by Skywise on Sun Jul 21, 2002 at 02:43:31 PM EST

No voting system is tamper-proof or idiot proof, and you're going to have to assume that people are going to tamper with the system for their own gains.  (That is why we have checks and balances in the US Government, right?)

Puched paper voting was supposedly "fool-proof" because you had a paper trail and it was "easily" distinguishable what constituted a vote because the hole would be punched out.  And we see how easily that turned out.

Scan-Trons/pencil/pen marking will suffer the same consequences.  Does half a mark count as a vote if the machine didn't read it?  What if there were 1 whole mark and 1 half mark for the same office?  Invalid vote, or "obvious" assumption that the whole mark was the correct vote?

(Meanwhile, in California, whole boxes of paper ballots were... misplaced in a county office... and didn't make it to the counting centers...)

So now we go to machines which will do the tabulation for us...  And now the argument is that we don't *know* if the votes were counted or not because there's no paper audit trail!?  (Like the paper ballot system above that's so "fool-proof"?)

Machines will stop people from overvoting (voting for more than one candidate) but they can't stop people from undervoting because it's legal not to vote for a particular office.

As for a paper receipt of voting, that could lead to other abuses where people could be forced to show their voting papers of how they voted.  It's a SECRET ballot so that there can be no backlash.

So, maybe the machines should display a summary of your actions at the end with a pop-up window(r) that says "Are you Sure?  (Y/N)"  The problem with THAT is that some elections will have tens if not close to a hundred things to vote for?  So how will any person be able to keep track of all the choices, especially in light of the fact that ALL machines allow you to review and change your vote until you commit?

I'm not saying that we shouldn't try to make the voting system more vote proof/idiot proof.  AND the machines should be open to audit.  (The reasons and particulars here are hazy, but I suspect that any auditing will probably allow the capability of manipulating the results, and the voting machine company will no longer guarantee their vote worthiness.  I've worked in elections with the old mechanical voting machines and all they store is a tabulation of the votes.  No particulars.  And you can only erase all the votes to 0, or lock the current tabulation.  So where will auditing get you?)

But to file a voting complaint because people aren't sure how they voted on the machines is asinine.  Because that's basically an intelligence test of the people and that's illegal.

If the people can't be trusted to be intelligent enough to vote and need leaders to protect them from themselves, well, we might as well just go communistic now...


As they say in Minority Report.. (3.66 / 3) (#43)
by strlen on Sun Jul 21, 2002 at 02:46:53 PM EST

The system is only human. It _must_ be flawed (and no I'm not giving out any spoilers at this time). This is why I dislike electronic voting systems. You need to maintain the machines, hence you'll need a special creature known as a systems administrator. Commonly known to consume twinkies, and able to write Perl code in their sleep, they too can hold a political opinion. Then there is the issue of politically motivated crackers. If a cracker can deface a pro-Israeli site, what's to stop them from removing a couple of ballots that are for the pro-Israeli candidate? Hence, here's a suggestion. Many countries in Europe still use it. As Betcour put it, paper and pencil. The two are manually counted on site, and results are telephoned in to the central comitee. Simple, fair (you don't need to be a rocket scientist to put an X next to your candidate on the ballot), and harder to forge without causing tons of stir up.

--
[T]he strongest man in the world is he who stands most alone. - Henrik Ibsen.
pencil vs mechanical (none / 0) (#127)
by spasm on Tue Jul 23, 2002 at 05:30:08 PM EST

Still use pencil on paper in Australia too, and that's for preferential voting where (in the upper house) the number of candidates sometimes exceeds 100. And you have to write the *numbers* 1 to (n candidates) on the ballot paper - much less room for reliable mechanical counting.

I used to work for the federal electoral commission - we'd have the poll workers count votes by hand on election evening (with reps from all parties present to "scrutinize" the process and make sure no ballots got added or removed; our staff would then produce a two-party prefered result for that polling station and call it in by phone; if a third party go a large enough number of primary votes in the district to invalidate the two-party preferred method the electoral commission would count the whole lot properly (and would do so to get 'exact' results weeks after the election anyway). The whole thing would be done by about 10:30pm after close of polls at 6pm.

Now I live in San Francisco - by happy coincidence the last Australian federal election (~12 million compulsory voters) took place the weekend before the San Francisco election (~500,000 *registered* voters, dunno how many bothered to show). The Australian result was in by 10pm, even with preferential voting and pencilled numbers on paper and counting by hand; the San Francisco ballot took two weeks to count with first-past-the-post voting and mechanical counters. Oh, and then they had to have a run-off election because no candidate got 50% of the vote. But that's another rant : )

Pencil and paper is reliable and, given an even vaguely well organised system, can be extremely quick. Not to mention cheap, even with the marginal extra number of staff you have to hire for a few hours after close of polls to help with the count.

[ Parent ]
Electoral incompetence in India (4.60 / 5) (#50)
by Brenner on Sun Jul 21, 2002 at 04:09:04 PM EST

You thought you had it bad here? Let me give you an example of electoral icompetence at its worst: the presidential elections in India. The president is elected by the elected members of parliament and members of state assembly. There are 4785 votes possible and there were 2 candidates. So to elect our new president, all you had to do was mark one of the 2 boxes (yes! just 2 choices). Apparently 174 of them didnt know how to vote and their votes were counted as invalid!!

A Radical Hypothesis (3.00 / 3) (#60)
by elysion on Sun Jul 21, 2002 at 06:31:13 PM EST

What if any participation in a political process whatsoever interfered your higher brain functions? Occasional exposure to politics, such as voting each year, might have only a temporary effect, while prolonged exposure, such as holding office, might cause permanent damage.

Hell, it explains [insert your least favorite politician's name here].

[ Parent ]

government incompotence (2.50 / 2) (#52)
by mlong on Sun Jul 21, 2002 at 04:50:43 PM EST

If they have just bought electronic machines that printed out a backup receipt, this wouldn't even be an issue. If there was a question they could just manually tally the votes. But Florida seems to be too stupid to do the obvious.

correction (2.33 / 3) (#53)
by mlong on Sun Jul 21, 2002 at 04:51:47 PM EST

I meant had not have

If they have just bought electronic machines that printed out a backup receipt, this wouldn't even be an issue. If there was a question they could just manually tally the votes. But Florida seems to be too stupid to do the obvious.

Tracked ballots (4.33 / 3) (#54)
by Chris Andreasen on Sun Jul 21, 2002 at 05:06:23 PM EST

How 'bout something like this:
Every ballot has a preforation along the top with a unique random number printed just above and just below. The voter selects his/her candidates, tears off the top to keep, and cast the rest of the ballot. After the results are tallied, a complete list of the ballots is made publically available showing which number voted for which candidate. The voters check the number they tore off against the list and see if it matches. If it doesn't match, the voter calls up the town hall and says "Hey, you screwed up. Give me a new ballot."
This should eliminate any problems due to questionably filled out ballots, and allows for independent tallying of the results as well. The only problem I see is that there's a potential for loss of anonymity: The ballot counters take everyone that voted the candidate they don't like and screw up the votes. When the voter come and ask for new ballots, the counters know who voted for the repugnant candidate since those were the only ballots that were screwed up. Of course, if such a sizeable percentage of ballots were erroneously counted in favor of the wrong candidate, it would be rather obvious and there would probably be a huge public outcry.
Any thoughts?
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

Then . . . (4.50 / 4) (#63)
by acceleriter on Sun Jul 21, 2002 at 09:00:10 PM EST

having bought your vote, the boss of the local political machine collects your receipt, verifies via the random number on the ballot stub you gave him and list of ballots made available that you voted for whom you agreed to, and your legs don't get broken.

[ Parent ]
good idea, but go further (5.00 / 1) (#65)
by zenofchai on Sun Jul 21, 2002 at 09:03:12 PM EST

instead of just taking the number tab, why not just have an entire carbon copy of the ballot to take home, and then go ahead and do the posting of what numbers voted for what. then you have the carbon with your number, and actual PROOF that you voted for candidate XYZ. the way you have it, any joe schmoe could claim they voted for XYZ, but not really have anything. sure, the system could "reproduce" the "original" ballot, but would that really satisfy many people?
--
as gold which he cannot spend will make no man rich, so knowledge which he cannot apply will make no man wise.
[ Parent ]
Not a secret ballot (5.00 / 2) (#77)
by thebrix on Mon Jul 22, 2002 at 04:45:15 AM EST

In principle, this is nearly there with the United Kingdom system (which has been in place since 1872); you get a numbered polling card through the post telling you where your polling station is, then your ballot paper is numbered accordingly before you make the cross and drop it into the box. So, if something goes horribly wrong, voters and ballots can be matched up ... or not.

There are two little local difficulties, though:

- the system has not had to be used in a Parliamentary election since 1911 so is probably redundant;

- it means the ballot is not secret.

The last point is fatal because the European Convention on Human Rights and parallel treaties mandate secret ballots; now that the Convention is incorporated in UK law the numbers have to go.

(To the relief of many; there were persistent rumours that, if there was a rather outré candidate standing, such as Communist or BNP, the names of those voting for them went to MI5. Never proved, but a good conspiracy theory).

[ Parent ]

It is secret (none / 0) (#112)
by jwb on Mon Jul 22, 2002 at 07:11:06 PM EST

This doesn't compromise the secrecy of the ballot. The user is issued a random ballot. They simply keep the strip from the top. There is no way to take the ballot and find the citizen who voted it.

[ Parent ]
Unfortunately, it isn't (none / 0) (#119)
by thebrix on Tue Jul 23, 2002 at 05:01:11 AM EST

This page explains the process a bit more clearly.

Postal votes also have a serial number on all the pieces of paper: you put your voting form into an envelope and seal it, then a witness signs a form to confirm that you are who you say you are, then the first envelope and the form goes inside a postal envelope.

[ Parent ]

Potential Problems (4.50 / 2) (#94)
by wnight on Mon Jul 22, 2002 at 11:49:09 AM EST

The problem with a receipt that lets you "anonymously" check the status of your vote is that someone could demand that you vote in a certain way, and return a receipt that proves it.

This could be fixed by allowing for fake receipts, if you request them. Perhaps there're two receipt on each ballot, and an arrow pointing to one (on the ballot) saying "fake", you take either the real one if you don't fear reprisals, or the fake one if you do.

Or, we allow people to spoil a ballot by marking a big "FAKE" checkbox, the receipt reports the vote on this ballot, except for the 'Fake' status. So you enter a real ballot and then (possibly) a fake one and collect the receipt you care about. (This in some ways is easier than having two receipts per ballot). Perhaps when you vote you're given two ballots, one pre-marked FAKE, you can vote the fake if you wish, or toss it in the garbage can inside the booth.

When votes are counted, the 'Fake' votes are ignored in the final count, yet they are registered in the computer that allows people to check their votes. If a recount is needed, perhaps because millions of people have a receipt for the loser, they open up the vaults and look at the (now sequentially stored) vote cards, and you can see which ones are fake. (But because this only comes out afterwards, and when the whole country/city demands it, hopefully it'd be safer.


[ Parent ]

Hmmm... (none / 0) (#102)
by JahToasted on Mon Jul 22, 2002 at 02:50:57 PM EST

interesting system, but would my 76 year old grandfather understand it?
______
"I wanna have my kicks before the whole shithouse goes up in flames" -- Jim Morrison
[ Parent ]
The obviousness factor (none / 0) (#117)
by wnight on Mon Jul 22, 2002 at 11:48:47 PM EST

Well, while I think the revised system is fairly obvious (read the other posts under mine), I don't think it's really important.

Few people understand cryptography, yet there are enough independent experts who verify the systems we use, that we feel safe trusting them.

Even if your grandma didn't understand the intricacies she could simply take someone's word for the increased security aspects. Unless she wanted a receipt, or a fake receipt, she just votes like she normally does. The only catch would be to not put your real vote on the one with the big 'FAKE' printed across it.

She'd need to understand the fake receipts if she was being coerced, but they'd have to explain the issue of receipts to her first, at which point the fake receipt is a fairly obvious concept.


[ Parent ]

Fake cards (none / 0) (#110)
by Chris Andreasen on Mon Jul 22, 2002 at 04:43:09 PM EST

The potential for bullying the vote out of someone was definitely a huge oversight on my part. I'm not sure that fake cards are that viable a solution, though. Suppose that the group tabulating the votes just decides to say that the candidate they don't favor had enough fake votes for him/her that the favored candidate is put into the lead. They can make up as many fake cards as they want, since the people who legitimately cast fake ballots can't step forward without fear of reprisal by [insert local bad guy who bullies votes out of people].
Hmm... So we'd need a system whereby A) only the person who cast the ballot can verify it, B) the information needed can't be transferred to another person, and C) the vote remains anonymous. Frankly, I'm stumped as to how one would implement such a system.
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]
Ok, try #2 (none / 0) (#116)
by wnight on Mon Jul 22, 2002 at 11:44:04 PM EST

Have obviously fake ballots, with the word 'Fake' stamped on them in an obvious fashion. People are issued one fake, and one real ballot. If you don't care about reprisals, garbage the fake.

This way, the "fakeness" of a ballot is easily determined when looking at the cast ballots. Independent reviewers can come in and examine the results.

Mark all the receipts with a large random number so that you can reference the ballot they came from, when needed. Make the checking system automated so that you type in (on a phone, or the web) the receipt number and it returns the results. No need to show someone your receipt to check it, and it can be done anonymously from a pay-phone, perhaps for free at an 800 number.

The only real drawback I see is that the receipt doesn't contain your vote, just a link to the ballot. The means that if the enemy changes votes for A to B, you will know you voted for A but you don't have any proof.

This last bit could be solved by using public-key encryption and having the voting computer write 'This Vote is for A on Ballot #x', signing it, and writing the signature to the ballot and receipt. That way it's still not obvious looking at it which candidate it's for, but you can link it to a physical ballot and also have the cryptographic security to prove the ballot you find later is the one matching the receipt. This isn't really easy to understand, so it's not something everyone would do, but perhaps enough trusted verification centers could exist (universities, etc) where curious people could go to check their ballot (the signature would have been stored online when created) and register complaints semi-anonymously (the best you can really do) if they detect a problem.

[ Parent ]

Taking the fake ballots another step (none / 0) (#122)
by Chris Andreasen on Tue Jul 23, 2002 at 11:35:48 AM EST

What if there was a minimum number of fake ballots that had to be included in the final tally? Here's my idea: The ballots are clearly make real and fake, but small percentage of the final vote (eg: 1%) is required to be fake. Some of them would be the fake votes cast by the voters themselves, the rest would be added by the vote counters after tallying the submitted non-fake ballots. The percentage of new fake ballots cast in each candidate's name would be equal to the percentage of real votes they recieved (eg: if a candidate received 75% of the real votes, 75% of the fake votes would have to be for him as well).

The number of people who vote would be publically available, verifiable by checking the sign-in log at the place in which the ballot was cast, so the vote counters can't just stuff the ballots with fake ballots as they want. If the vote counters decide to fill out all the fake ballots to further their own political gains, as I mentioned in a previous post, the number of votes they can add is too negligible to make a big difference. If the final tally comes really close to a tie, and those few votes do count, independent auditors could recount the ballots under state supervision and after signing a non-disclosure agreement stating that they won't reveal any of the numbers on the fake ballots under penalty of law.

Since the average citizens have no way of determining which ones were real and which ones weren't when they see the final results, there's no way to prove that someone did actually vote for a particular candidate or not, so the local bully can't forcefully coerce someone to vote in a certain manner. If the number of people who actually submitted these fake votes at the voting booth exceeds 1% (or whatever the percentage was decided upon), then it can be assumed that the election is being rigged and the police can pursue criminal investigations.

There's still a problem with this, though: suppose candidate A has a clear majority, but there are at least 1% of voting population that favors candidate B. The B supporters can all vote with the fake ballots and essentially grind the election to a halt. Any ideas on how to prevent this?
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]

Fake votes (none / 0) (#123)
by wnight on Tue Jul 23, 2002 at 02:02:27 PM EST

I don't see the requirement for a certain number of fake votes. How does casting extra fake votes distort the count?

I don't see any way to prevent all attempts at determining if a vote is fake. You could triple encrypt the receipt number and print that on the ballot, meaning that only with three secret keys (ideally held by three organizations) you couldn't reverse the ballot number to get the receipt number, or encrypt the receipt numbers you'd seen to get the ballot numbers, without the cooperation of all three parties. (Each vote could actually be encrypted with three session keys, each one is encrypted with each organization's key as well, so they can decrypt the session key and reveal it without revealing all votes they signed.)

But, if you don't have any trusted organizations, the system falls apart. (Myself, I'd feel safe if the ACLU had one key, the NRA the other, etc. Organizations I don't individually trust, but who would never cooperate.)

The receipt->result link is made, unencrypted, when you vote. Anyone can look up a receipt number to see what it points to, but only a wide-conspiracy can find the actual ballot cast and find out if that was a fake ballot or not.

But when counting comes, you simply take all ballots, discard the fakes, and count the results as normal. For further checks, you make sure the number of receipt->results in the computer matches the number of fake + real ballots.


[ Parent ]

Keeps the vote counters in check (none / 0) (#125)
by Chris Andreasen on Tue Jul 23, 2002 at 05:18:54 PM EST

I don't see the requirement for a certain number of fake votes. How does casting extra fake votes distort the count?
Quickly reviewing my post at the very top of the thread, after the final count is taken, the full list of which number voted for which candidate is published - anyone can see the full list and independently count it, and anyone can verify that their own vote was recorded properly. This way recounts shouldn't be necessary unless a large number of people were incorrectly recorded as voting for a different candidate. No one can be sure of which votes were fake, though. The vote counters could very well lie and say that 15% of the vote was fake, and they were all cast for candidate A, which makes candidate B the winner. There would be no way of knowing if they were telling the truth or not without knowing which votes were fake.

Simple example:
Ballot 1 is cast for candidate A
Ballot 2 is cast for candidate B
Ballot 3 is fake, but cast for candidate A
Ballot 4 is fake, but cast for candidate B
It should be a tie, but the vote counters are biased towards candidate A, so they say on the published list that only ballot 4 was fake. There's no way of verifying this without knowing that both 3 and 4 were fake. Everyone seems content that their votes were recorded properly, and 3 and 4 are content to know that they aren't being revealed as fakes, but the vote counters just put their own candidate in office.

By setting a cutoff point, the counters can't just make up any percentage they want. They announce before the vote that the final tally will be padded so that 1% of the votes are fake. When they publish the results, the public can see that the number of people who voted has indeed increased by 1%, and that 99% of people who casted their vote can verify that their vote was counted correctly. The other 1% either won't come forward because they have been threatened into voting for a particular candidate, or they don't exist. Since it's only 1% of the total voters, it shouldn't skew the final tally that much anyway.
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]

Quick correction (none / 0) (#126)
by Chris Andreasen on Tue Jul 23, 2002 at 05:22:34 PM EST

It should be a tie, but the vote counters are biased towards candidate A, so they say on the published list that only ballot 4 was fake.
Quick correction: they don't actually say publish that ballot 4 was fake, they just say one of the votes was fake instead of correctly saying that two of them were.
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]
Okay, I see that much (none / 0) (#128)
by wnight on Tue Jul 23, 2002 at 05:45:06 PM EST

That's okay then, but why mix the counted ballots with the results used to check receipts?

When someone votes with a fake ballot, you record the result and the ballot number. The person can then call in with the receipt to check the results.

The real count happens later, and you sort out all the ballots with the word 'FAKE' stamped across them (in a very obvious way).

The computer could remember which ballots are real and which are fake, so that it could give you an accurate count, but in case of questions, simply count the paper ballots.

I can't see a way for someone to lie in this scheme.

[ Parent ]

Receipts (none / 0) (#129)
by Chris Andreasen on Wed Jul 24, 2002 at 12:45:31 AM EST

Those who cast the votes decide nothing; those who count the votes decide everything
-Josef Stalin (I think)

After the vote is counted, the public needs some way of verifying that the count was done correctly. This is where the receipts come in. Anyone can verify that they were indeed counted, and counted for the correct candidate. They can count the final tally themselves, if they so choose, with a 1% margin of error factored in for the fake votes.
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]

Here's a problem with that. (none / 0) (#106)
by aphrael on Mon Jul 22, 2002 at 03:30:46 PM EST

In parts of California, you are required to sign in when you are handed a ballot, so that it's easier to prevent double-voting (eg., if you've signed in twice, they'll catch you if they audit the books) and fraudulent voting (you're only allowed to sign in if you're registered).

The order of those signatures is directly related to the number on the ballot. Eg., from the signature order, and the fact that someone has to know which ballots went to which polling place (in order to prevent someone from absconding with a set of ballots, voting them all, and returning them illegitimately), such a system would make it possible to determine how individuals voted.

[ Parent ]

Randomly selected (none / 0) (#111)
by Chris Andreasen on Mon Jul 22, 2002 at 04:51:59 PM EST

The idea is that the ballots are numbered and/or selected randomly. This might entail something such as, after signing in, the voter is told to "pick a ballot - any ballot" from the stack. When returning the ballot, it could be cast into a large bin, thus mixed in haphazardly with the other ballots. This could very well be a huge pain for the people who have to neatly sort and count the ballots afterwards, but it would certainly defeat an attack whereby evil ballot counters tried to verify a voter's identity by checking the sign-in log.
--------
Is public worship then, a sin,
That for devotions paid to Bacchus
The lictors dare to run us in,
and resolutely thump and whack us?

[ Parent ]
Democratic vote fraud vs. GOP vote fraud (2.83 / 6) (#55)
by nomoreh1b on Sun Jul 21, 2002 at 05:25:01 PM EST

I liked the thrust of this article on requirements for voting machines. Now, I don't think this is going to happen any time soon for one simple reasons:
both parties are dependent upon vote fraud.

Democratic vote fraud is generally low tech in nature-which means things like stuffing ballots, dead people voting, spoiling ballots for non-democratic candidates. GOP vote fraud is high tech in nature. I'd bet Republicans own most voting machine companies. You saw this in the last election. The democrats really wanted a hand count and the GOP a machine count in Florida.

IMHO the real way to make vote reform happen is to get Non Governmental Organizations to use cryptographically sound, open source voting mechanisms. At some point, folks somewhere are going to get sick of all the lies.

Huh? What? (4.50 / 2) (#67)
by HypoLuxa on Sun Jul 21, 2002 at 09:46:43 PM EST

GOP vote fraud is high tech in nature.

Care to give an example? No, the 2000 Presidential race doesn't count; the Republicans wanted to use the machine count since it was in their favor. From what I know, in the US, voter fraud has been the time tested means of intimidation, bribery, ballot box stuffing, false absentee ballots, etc., and has been used by everyone.

Who used "high tech in nature" voter fraud? How and when?

--
I'm guided by the beauty of our weapons.
- Leonard Cohen
[ Parent ]

Dems vs. GOP (4.00 / 1) (#75)
by nomoreh1b on Mon Jul 22, 2002 at 01:08:04 AM EST

Care to give an example? No, the 2000 Presidential race doesn't count; the Republicans wanted to use the machine count since it was in their favor.

The question is what was the real vote-and how much did the vote recorded differ from that vote. I personally think the GOP has a capability to use mechanisms in voting machines to their advantage. I wouldn't trust an election with voting machine just because some corporate types say the voting machine is honest.

From what I know, in the US, voter fraud has been the time tested means of intimidation, bribery, ballot box stuffing, false absentee ballots, etc., and has been used by everyone.

Think about it. Democrats tend to have lots of volunteers from special interests. The kind of folks that work as a clerk counting ballots by hand tend to be Democrats. Stuffing ballots the old fashioned way tends to favor Democrats--it utilizes those capabilitites/resources in which Democrats have an edge. The GOP historically has lots of money compared the the democrats. The value of the major makers of voting machine equipment is small compared to the leverage that control of these mechanisms give. That playing field tends to favor resources the GOP has.

If either side was in favor of truly honest elections, they'd be favoring proposals like the one the author of the original article proposed.

It is pretty clear historically that the Democrats used extensive vote fraud in the 1960 presidential election(especially in Daley's Chicago and LBJ's Texas). Do you really think the GOP would let that go without a response? That just doesn't strike me as credible.



[ Parent ]

Random speculation (4.50 / 2) (#89)
by HypoLuxa on Mon Jul 22, 2002 at 10:13:28 AM EST

The question is what was the real vote-and how much did the vote recorded differ from that vote.

No, the question was who used high tech voter fraud, when and how. I'm assuming your first assertion, that GOP voter fraud was "high-tech in nature" was just random speculation, since you haven't offered any examples of how it has occurred.

As far as the rest of your comment, I think you are seeing what you want to see. Democrats and Republicans both have armies of volunteers, and they both raise about the same amount of money to spend on elections. You mention the 1960 presidential race, which was probably the most dramatic use of voter fraud in the US in the 20th century as your example, and while that certainly was the high water mark of voter fraud, it certainly wasn't unique. I don't see how either party in the US has used an sort of unique or high-tech methods of voter fraud, and I doubt that changing to electronic voting machines will change that.

--
I'm guided by the beauty of our weapons.
- Leonard Cohen
[ Parent ]

I agree (5.00 / 1) (#93)
by porkchop_d_clown on Mon Jul 22, 2002 at 11:19:41 AM EST

You don't need to be a republican to mess with the machines. heck, hacking the machines philadelphia uses can be done by anyone with a propane torch - they have soft lead seals on them, the seals can be melted, the machine tallies "adjusted" and then the machine resealed. No problem.


--
The gift that lasts a lifetime: Give your child "mental blocks" this christmas!


[ Parent ]
Realistic Expectations (4.00 / 2) (#57)
by n8f8 on Sun Jul 21, 2002 at 05:58:56 PM EST

As history has shown us, some people are just unreasonable when it comes to the level of "proof" they want to prove somthing. Like it or not, there is no way of providing irrefutrable, tangible proof of somthing recorded on digital media. A reciept? Proves nothing. We'll have to go back to archaic punchcards and checkboxes and even then we'll have to rely on computers to do the tallying in big districts. Even if the voting device were "open" or "open source", you'll never convince the paranoid and technologically ignorant that the system is foolproof.

Sig: (This will get posted after your comments)
Right (4.00 / 1) (#88)
by peace on Mon Jul 22, 2002 at 09:32:52 AM EST

But you might be able to convice Greenrd, or me, for instaince. Arguing that you can not have a perfect system is no excuse for having an obviously broken one. By your logic we may as well drop market clay shards in pots. Come on! :)

A system should always attempt to improve. If it does not it will fall into disrepair.

Kind Regards

[ Parent ]

Which is exactly why it's a bad idea. (none / 0) (#105)
by aphrael on Mon Jul 22, 2002 at 03:27:53 PM EST

The legitimacy of the government --- eg., the degree to which the governed accept the right of the people doing the governing to do so --- is, in a democracy, measured by the faith the people have in (a) the preferability of democracy to other systems of government, and (b) the degree to which the mechanisms actually reflect the will of the people.

The problem with complicated voting systems that depend on technology that the majority of people don't understand is that it undermines both of these pillars. Regardless of what the technical types say, if people believe that a voting sytem can be corrupted easily, they won't have faith in it; if they don't have faith in it, it will undermine their belief in the legitimacy of the government *and* possibly in the preferability of democracy over other systems.

Now, granted, a system which is trusted by all but the overly paranaoid is fine. But the trouble is, nobody really knows if electronic systems *will* be trusted by everyone but the overly paranoid, or if they'll be widely distrusted --- and if they're widely distrusted, then even if they are accurate, they will undermine our political cohesiveness.

[ Parent ]

Rebecca Mercuri's papers on the subject (4.66 / 3) (#66)
by goonie on Sun Jul 21, 2002 at 09:20:59 PM EST

Rebecca Mercuri's electronic voting website makes a pretty compelling case, as far as I am concerned, that electronic balloting is a dumb idea.

Not quite as dumb as having elected partisan officials, and elected, partisan judges, making rulings on the elections of their buddies, but almost.

yeah.. (3.00 / 1) (#90)
by AnalogBoy on Mon Jul 22, 2002 at 10:20:59 AM EST

Thats almost as bad as having elected partisan morons in the whitehouse, and *un*elected partisan justices in the supreme court, making rulings on the elections of their buddies.  

Sorry to hear it happened in your country too, though.

=)

--
Save the environment, plant a Bush back in Texas.
Religous Tolerance (And click a banner while you're there)
[ Parent ]

You're right... (4.50 / 2) (#81)
by jeep on Mon Jul 22, 2002 at 06:22:29 AM EST

We can't trust proprietary solutions to be involved in votes of any importance. See the K5 article I wrote Why Electronic Voting Software should be Free Software.

Furthermore over the past 4 years of studying these issues I am becoming more and more skeptical of the value (let alone security) in letting a computerised voting system run our elections. The opportunities for fraud, disenfranchisement and general problems as well as the huge costs involved just aren't worth it. And this is from somone who has written an Internet Voting system. I was a believer.

Jason


--
The FREE e-democracy Project
Promoting Free Software in Government

Electronic voting is just as bad as chads (5.00 / 1) (#95)
by tcomeau on Mon Jul 22, 2002 at 12:13:02 PM EST

I recently completed a project for a software design class that included designing a vote counting system for punch card ballots. In the course of my research I found this:

Punchcard methods and systems using direct recording electroninc devices (DREs) had significantly higher average rates of spoiled, uncounted, and unmarked ballots than any of the other systems.
The Caltech/MIT Voting Technology Project

Many of the problems with DREs have nothing to do with the correctness of the count. Rather, it seems DREs are as likely as punchcard systems to produce "voter error", particularly in voters who are not comfortable with related technologies such as ATMs.

There does seem to be a common view that "it's just like an ATM, and everybody uses those." There are two statements there, and neither is true.

One major distinction between ATMs and voting systems is that ATM errors can be readily reviewed, since the transaction is personally identifiable. One major goal of voting systems it to assure that transactions are not personally identifiable.

Many people do not use ATMs, or use them only for the simplest transactions. My wife, who is well educated, holds a professional degree (J.D.) and is not terribly tech-phobic, uses ATMs only to get "fast cash" -- she pushes a total of seven buttons (including a PIN) to get money. Navigating a ballot of a dozen candidates and three or four issue questions is significantly more effort, more like making a deposit, moving money between accounts, and only then getting a specific amount of cash. She has no idea how to navigate any other part of the system, and has repeatedly indicated her willingness to sue if our local Election Commission moves to electronic ballots.

Paper ballots, particularly those configured to support optical counting, seem to be an ideal technology for voting.

[ Parent ]

Receipts? Journal tape, even? (5.00 / 2) (#92)
by greenshift on Mon Jul 22, 2002 at 10:44:48 AM EST

Name a "mission-critical" computer system doing a relatively simple task like counting votes that doesn't keep a paper record.  

Just like any register at any store, there should be some form of journal tape that can later be counted and cross-checked with the electronic results.  

Or how about printing a receipt upon completion of the voting "transaction?"  You vote, take your receipt, make sure it's as you intended to vote, and if so you drop it in a box before you leave.

Would that be so difficult?

What would be the point, exactly? (none / 0) (#108)
by marxmarv on Mon Jul 22, 2002 at 04:13:47 PM EST

If the votes still need to be tabulated from paper, where's the labor savings? If the receipt can't be tabulated by machine (and that crap thermal paper is not easily machine-feedable) then where's the labor savings? If humans and computers don't read the same thing off the receipt, then how is it more secure and what's the point?

Journal tape, on the other hand, is a fine idea and I'm all for it.

-jhp

[ Parent ]

Cross-checking and auditing... (none / 0) (#118)
by greenshift on Tue Jul 23, 2002 at 01:33:56 AM EST

Who cares about the labor savings?  We need votes that can be quickly and accurately counted.

The preliminary tallies would be taken instantly from the voting machines, so you'd have instant election results.  In most cases they'd still decide the actual winner even if they're off by 5%.

The point of printing out the result of each person's vote is to eliminate computer miscalculations due to faulty code or crap hardware.  

On one end you'd have your electronic voting machine:  a "lock-box" if you will, that no one can open to inspect.  Obviously, you can't trust its results.  But it is very fast and very cheap, and gives instantaneous results.  

Since you can't audit the results using just the computer, you must print out a paper record of the vote so that the voter can inspect to check for accuracy, otherwise how do you know if your vote was counted?  How does the election board know?

The inaccuracies so far have been, what, 3% or so on the computers?  That'll be within the margin of error to conclusively decide most contests.  Where the paper votes come in is to make sure that you have the same number of votes on paper as recorded in the computer.  Sure, it will take some time to count all the paper votes, but it's a necessary evil.  Perhaps an open source computer system could be used to aid in tallying the paper vote by cross-checking with the human tally.  If you print a barcode out on each receipt, you could just hook up a barcode reader, and scan each paper vote receipt in relatively little time.  Of course, you're also relying on the original machine's accuracy, which is suspect.

I conclude that the only way to have truly accurate voting is for there to be a gigantic cardboard "ballot" the size of a wall placed in the voting area.  The giant cardboard ballot will have all the candidates listed, like a regular, ballot, only 20 times the size.  It will be similar to a punch card ballot, where there is a perforated round hole corresponding to each candidate for each office.  To vote, you phisically punch out the hole of the candidates you which to vote for.  Then a picture is taken, filed, and counted.  Your voting is now, simple, easily audited, and most importantly, fun.

[ Parent ]

Why do so many people think people (none / 0) (#130)
by acronos on Wed Jul 24, 2002 at 03:13:03 AM EST

can count better than a computer. A computer isn't just a little more accurate - it is at least a factor of a million times more accurate. A computer could tally the entire Florida election over and over again and come up with exactly the same answer every time. People cannot.

[ Parent ]
Why so complicated!?! (none / 0) (#101)
by GreenCrackBaby on Mon Jul 22, 2002 at 02:37:56 PM EST

I'm amazed at the complexities of this whole voting system(s). Why does it have to be so difficult?

Have voter fill in circle next to their choice(s). Voter drops card into computer slot. Computer says, "You have voted for <blah> Is this correct?" Voter answer yes...computer keeps card. Voter answers no, computer marks card as rejected, voter fills in new card.

This leaves a complete audit trail, while ensuring people are voting for who they think they are voting for!

anonymity. (none / 0) (#104)
by aphrael on Mon Jul 22, 2002 at 03:22:44 PM EST

It's also considered important in our political system that votes be anonymous. In the scenario described, they aren't --- it's possible to tally a record of what order the cards were taken in, and of what order people came in to vote (because you have to record who voted to prevent double votes).

[ Parent ]
Open source vote authentication (3.00 / 1) (#113)
by Blarney on Mon Jul 22, 2002 at 08:54:15 PM EST

The obvious answer to these problems is to provide a convenient way for community members to monitor the accuracy of voting in their community. I suggest a method whereby information about each vote is broadcast.

Each polling place will have a large loudspeaker mounted on the roof, capable of being heard for several miles. Every time a vote is cast, the loudspeaker will announce the corresponding number and candidate, such as "Bush 475" or "Quayle 92". These announcements will also be transmitted on the FM band for convenient taping, with digital tonebursts following each announcement carrying the same information in machine-readable format.

Now everybody can be part of the vote count!

Vote Fraud Made Easy (none / 0) (#134)
by SEWilco on Wed Jul 24, 2002 at 10:11:33 AM EST

"I'll pay you $100 if you vote for Fred. The loudspeaker just has to announce that Fred's votes increased by one."

Oh, so you'll only announce very ten votes?
"I'll pay all 20 of you $100 each if you vote for Fred..."

[ Parent ]

Conference coming up (none / 0) (#120)
by jeep on Tue Jul 23, 2002 at 08:30:51 AM EST

If you're really interested in these kinds of issues there's a conference coming up this November in London, UK. It's got a lot of the key players at it including:

TheMinister responsible for elections (Nick Raynsford MP),
the e-Envoy (Andrew Pinder),
the Chairman of the Electoral Commission (Sam Younger) as well as the heads of SOLACE, Local Government Association and the Associatione of Electoral Administrators.

Quite a lot of interesting stuff going on - I went last year and found it very thought provoking. It's only a day but they still pack a lot in!

More info: http://www.lga.gov.uk/Event.asp?lSection=0&id=SXBDC5-A780FB42


--
The FREE e-democracy Project
Promoting Free Software in Government

Voting systems (none / 0) (#121)
by metreperhour on Tue Jul 23, 2002 at 09:00:25 AM EST

I am not sure if I should really ask to have my commentary really setup as a new article. Anyhow...
Personally, the machines that New York City use are pretty intresting.  I have not yet done the research on when/why they were built but my guess would be that the reason is that New York always has to have things BIGGER. That includes corruption. (tammany hall)  
The voting machines used in New York based on work from the orginal Edison Patent aparently called a Shoup 3.2 seem to work pretty well.
  I think that any system setup in the US should have these machines good features. Having a pure electronic system IMHO is not a good thing.  Electro-machanical systems have some advantages of being verifiable.

   I do have a proposal on a machine based on these things which should be:
   1) simple
   2) easy to reconfigure
   3) reliable

What would K5 readers recommend I do with my proposal?

From an MIT/Caltech announcement:
"States and municipalities using lever machines will have to replace them in the near future, and the two most common alternatives are punch cards and optical scanning devices. Ironically, many localities in Massachusetts have recently opted for lever machines over punch card ballots because of problems with punch cards registering preferences."

Bzzzzzt! Wrong! (none / 0) (#133)
by SEWilco on Wed Jul 24, 2002 at 10:06:18 AM EST

Electromechanical systems are as verifiable as the odometers on used cars.

The counting mechanisms can be shaved so counting errors are more likely for certain candidates, stickers can hide nonzero counter starting values, counters can be rigged to jump forward, magnets can make mechanisms change...

Oh, you want the box to be clear plastic so everyone can see it working? Nobody should be able to know how an individual voted, so the mechanism has to be hidden.

[ Parent ]

Further Reading (5.00 / 1) (#124)
by rhino1302 on Tue Jul 23, 2002 at 02:56:00 PM EST

Bruce Schneier's wonderfull book "Applied Cryptography" includes a section on secure elections complete with numerous references to different protocols.

He starts the section with the following requirements for an ideal protocol:

  1. Only Authorized voters can vote.
  2. No one can vote more than once.
  3. No one can determine for whom anyone else voted.
  4. No one can duplicate anyone else's vote. [This turns out to be the hardest requirement]
  5. No one can change anyone else's vote without being discovered.
  6. Evry voter can make sure that his vote has been taken into account in the final tabulation.

Additionally some voting schemes may have the following requirement:
  1. Everyone knows who voted and who didn't.

He ends the section with the following observation:

Voting protocols work, but they make it easier to buy and sell votes. The incentives become considerably stronger as the buyer can be sure that the seller votes as promised. Some protocols are designed to be reciept-free, so that it is impossible for a voter to prove to someone else that he voted in a certain way.

I think every self-respecting geek should have a copy of this book.



Government Audited versus Public Source Electronic Voting | 135 comments (132 topical, 3 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!