Finnish Court Accepts E-Voting Result With 2% Lost

Nailor writes "The Helsinki Administrative court accepted the municipal voting result in an election in which 2% of votes cast were not counted at all. We discussed this situation at the time. The court noted that the e-voting machinery has a feature, that should be considered as an issue. However, it also noted that 'a little over two percent failure rate can not be considered as such as a proof that the voting official would have acted erroneously.' Does this mean 98% of votes is enough to figure out how the other 2% voted? Electronic Frontier Finland has a press release about the court decision (Google translation; Finnish original)."

Comments

Failed to Finnish

[dov_0]

If it is a first past the post system this is a problem. If it is majority rules, then as long as there is enough of a majority.

2% failure rate is a bit much though?

Re:Failed to Finnish

[Novus]

Finnish municipal elections are always by the D'Hondt method, so the result can be strongly affected by a few additional votes. In fact, if the missing votes were all for one candidate, that candidate would have received the most votes.

Re:Failed to Finnish

[Jurily]

Finnish municipal elections are always by the D'Hondt method, so the result can be strongly affected by a few additional votes.

Doesn't really matter. If you let them vote, count all the fucking votes. It's that simple.

I have my own problems with any voting system skewing the results in favor of the two candidates most likely to win ("Don't vote on the little guy, your vote will be lost!"), but this is ridiculous.

Did they offer any reimbursement for the people whose vote they didn't count? I'd be pissed off if they did that to me. I'd also start screaming around about someone cheating, and likely sue as well.

Re:Failed to Finnish

[jaria]

> Did they offer any reimbursement for the people whose vote they didn't count?

I am considering paying 2% less taxes this year. Clearly, 2% is within the allowable government tolerances.

Re:Failed to Finnish

[Anonymous Coward]

But it's not that simple. This e-voting debate has always been tainted by a complete lack of scientific logic, which your post typifies.

No measurement is perfectly accurate. Counting votes by hand is not a perfectly accurate method. We expect computers to be perfect but the measuring system isn't perfect so the results aren't either.

It doesn't matter. Our voting system has always had a margin of error, and always will. Thankfully one court in the world understands this and is bold enough to say 2% is an acceptable error.

If the election is not close, it clearly has no effect on the end result. It can only matter if the election is close.

And here's the thing - if the election is close enough, then no system will accurately measure the result. We have a situation, well understood in the world of science, where the noise is louder than the signal. The result of any binary discriminator in this situation is effectively random.

Closely-called elections have always been randomly decided, and they will always be randomly decided.

Let's just accept this and get on with the real debate which is, what is an acceptable margin of error? Do you agree with 2% or not? What would you like the error to be? How much are you willing to spend to reduce the error?

You'll note that in this properly cast debate, anyone saying that only 0.000000% is acceptable counts as an extremist who won't be listened to.

Re:Failed to Finnish

[Jurily]

If the election is not close, it clearly has no effect on the end result. It can only matter if the election is close.

It's the principle that matters. You're basically saying they should randomly not let people in the booth saying "You're the error margin, good bye."
How would you feel?

You'll note that in this properly cast debate, anyone saying that only 0.000000% is acceptable counts as an extremist who won't be listened to.

Make a law saying anyone who didn't get their vote counted doesn't have to pay taxes until the next election. I'm sure they'll get it sorted out in no time.

Re:Failed to Finnish

[Ornedan]

2% is bloody well not an acceptable error rate when manual counting error rate is 0.5% on average. As such, the court saying 2% is acceptable is utter bullshit.

Re:Failed to Finnish

[powerspike]

In Australian, we have what they call preferences, if you don't put them in, the person you voted for gets to cast them for you. Alot of the smaller parties are setup to lose, so they can throw 3-4% of the vote to one of the two big parties. You can either out a 1 in a box, or put 1-50 on the vote form...

They Take advantage of the system either way.

Re:Failed to Finnish

[Anonymous Coward]

> Let's just accept this and get on with the real debate which is, what is an acceptable margin of error? Do you agree with 2% or not?

Given that 1% or even 0.5% is the limit where a party can ask for recompensation for their expenses (speaking about Germany), 2% is definitely far, far too much, and obviously anything above 0.1% is too much.
But that is not the real issue: the real issue is that nobody knows or can proof why which votes were lost, and the electronic voting systems make it completely impossible to find out even if 60% of votes were lost. You can then also not determinine if there was any system or regularity in which the votes were lost. This means that just speaking statistically, the margin for error must be far lower for electronic systems in order to allow the same accuracy as traditional voting.
If someone had had the possibility to let the green party loose 2% at will it is quite likely it would never have grown - a party that now regularly gets around 10% of votes.
If you say 2% loss is acceptable with a voting system that makes reliable tracking impossible (since different to pen and paper, electrons do not leave any traces we can track these days), that means you essentially o.k. any arbitrary vote manipulation as long as it does not exceed 2%.
That is not much if all you have is a party with 60% and one with 40% but if you have e.g. 1 with 6%, 1 with 5%, 4 with 3% maybe 3 more with about 1% that is basically giving the permission to mess up the system as you like (particularly if above/below 5% decides whether you get seats in parliament or not).

Re:Failed to Finnish

[Thiez]

> We expect computers to be perfect but the measuring system isn't perfect so the results aren't either.

The measuring system IS perfect. You either vote for a candidate or you don't. Adding integers is something computers can do very well without errors.

> You'll note that in this properly cast debate, anyone saying that only 0.000000% is acceptable counts as an extremist who won't be listened to.

A 0.000000% error margin is perfectly acceptable when you are adding some positive integers when the sum of all integers is guaranteed to be between 0 and 6 million (finnish population 2008 estimate 5.3 millios).

Do you call a teacher an extremist he when says 1 + 1 != 2.000000001 (assuming both 1s are integers)?

Re:Failed to Finnish

[Jurily]

Given that 1% or even 0.5% is the limit where a party can ask for recompensation for their expenses (speaking about Germany)

1% in Hungary.

But that is not the real issue: the real issue is that nobody knows or can proof why which votes were lost, and the electronic voting systems make it completely impossible to find out even if 60% of votes were lost.

No, it's not impossibe. We're talking about computers here, they can be audited.

Also, I'd like to see, how

votes[canditate]+=1;

has an error margin of 2%.

Re:Failed to Finnish

[Anonymous Coward]

> ...reimbursement ... pissed off ... screaming ... cheating ... sue...

Calm down. The votes were not counted because the voters themselves screwed up. They didn't press the confirm button after entering the candidate's number code. 98% of people did manage to follow the instructions. The person to sue, and the one that you should yell at, is in the bathroom mirror.

The voting machines will be made more idiot proof in the future, but there will always be people who will manage to circumvent the most ingenious idiot proofing. With pencil and paper voting there are always votes for candidate #1 (numbering starts at 2 in all elections; there is no candidate 1. The list of candidates is *right* *there* in front of the voter in the voting booth; no #1 anywhere in sight.)

Re:Failed to Finnish

[Jurily]

Re:Usability Glitch? (Score:5, Insightful)
by Antique Geekmeister (740220) on 2008-10-29 8:47 (#25552091)
The card should have been locked into the machine until the voter said 'OK' or cleared the screen, and locked it in with an alert and a deactivation warning if the person left the booth without doing either. Anyone can get confused about simple directions for an entirely new system. How many of us have tried to walk away from an ATM with our card still in it because we were distracted?

Re:Failed to Finnish

[Anonymous Coward]

> No, it's not impossibe. We're talking about computers here, they can be audited.

At least after the fact, you can only audit the non-volatile parts (and an audit during the election would probably be hard to align with the desire to keep vote secret).
To my knowledge, current voting machines unfortunately use von-Neumann architecture, not Harvard architecture.
Thus all executed code is in volatile storage, any manipulation on it can not be proven with today's technology (the same probably applies to anything but hard-wired ROM, where it should depend mostly on how much effort each side invests).
The closest you can get is showing that it is quite unlikely that someone would be able to manipulate it - but it is likely that it will always be possible to manipulate, and most likely impossible to show if such a manipulation happened or not even when you have the possibly manipulated piece of hardware in your hands.
This _is_ different to ordinary paper, and IMHO it is a significant difference.
I think it is significant, since what do you do when someone claims that there was manipulation? Do you accept that claim just when they prove it is possible at all? Expect a lot of re- and rerevoting.
Don't you accept it unless they can prove such a manipulation actually happened? That seems to be the stance of the German parliament, but isn't that a rather high bar when it is "impossible" to prove?
I realize that none of my arguments so far are really a argument against voting machines in general, but my opinion is that currently voting machines are so far away from the ideal that they can be called "non-auditable". Even if they weren't there is still the issue that even the most basic level of auditing is possible for less than 1% of the population - in comparison I think given enough time and some instructions there would be more people that could detect even an exceptionally well done manipulation to an ordinary paper vote, particularly if they are allowed to attend the voting and counting itself (which these days is the norm I think).

Re:Failed to Finnish

[grgon]

I am considering paying 2% less taxes this year. Clearly, 2% is within the allowable government tolerances.

This is actually a really good point. Another example: if during one day in one country all the transactions in one bank contained 2% errors, that is money going to the wrong recipent, wrong amounts etc. it would be totally unacceptable. maybe if we involved a bank and put a euro coin/note in each envelope....

Re:Failed to Finnish

[canonymous]

Dunno why parent is being modded Funny! 98% taxation for a 98% change of being represented...

Re:Failed to Finnish

[Gordonjcp]

Doesn't really matter. If you let them vote, count all the fucking votes. It's that simple.

If you read the article, you'll see that what the problem really is, is that 2% of users were too stupid to register their vote properly.

I would suspect that here in the UK you'd see approximately 25% of votes discounted, with 10% of users just standing in front of the touchscreen going "When's Strictly Come Dancing on?", 10% of users trying to roll it up and smoke it, and the remaining 5% trying to challenge it to a knife fight.

Re:Failed to Finnish

[tepples]

Also, I'd like to see, how

votes[canditate]+=1;

has an error margin of 2%.

You didn't show how you locked votes in case multiple voting machines try to access it concurrently. A lot of languages that use this sort of syntax don't specify concurrency semantics for +=. Nor did you show how your program handles a hardware or operating system failure between when the user presses "submit vote" and when the line you quoted gets executed, or between when the line you quoted gets executed and when votes is committed to nonvolatile storage. Nor did you show the code that obfuscates the variable canditate so that it doesn't match candidate 2% of the time.

Re:Failed to Finnish

[gzipped_tar]

Also, I'd like to see, how votes[canditate]+=1; has an error margin of 2%.

Software design principle: All input is invalid.

Even simple input as a touchscreen click should be validated. And there's a long way to go between a click and "votes[candidate] += 1;" and it can go Wrong.

Re:Failed to Finnish

[Anonymous Coward]

You have just done the best hidden 'first post' post I've ever seen! Well done!

Re:Failed to Finnish

[Anne Thwacks]

Return to PDP11's. The x++; operation is impemented in hardware in a PDP11, and as I hope you all know, C is the assembly language of the PDP11, which was designed as a "hardware Fortran machine", Even "for (i = 0; i

Do the voting machines use a really sad processor? How sad can processors get in the 21st century?

I know compilers can get pretty bad, but this bad?

Re:Failed to Finnish

[RiotingPacifist]

If the election is not close, it clearly has no effect on the end result. It can only matter if the election is close.

And here's the thing - if the election is close enough, then no system will accurately measure the result. We have a situation, well understood in the world of science, where the noise is louder than the signal. The result of any binary discriminator in this situation is effectively random.

While you are right that all elections ave a margin of error and that 2% is better than 3%, in a proportional representation system (as the link in GGP suggests) 2% can make a difference (changing the lead by up to 8 seats)

Re:Failed to Finnish

[mppareto]

In fact, if the missing votes were all for one candidate, that candidate would have received the most votes.

Not really. Assuming that the screw-ups are random (hence do not derive from some attribute such as jurisdiction, sex etc...), then with 98% of the votes, you should have the same proportion of votes as you would otherwise (at least with a very high probability), and hence the same seating arrangements - even using the Finnish method.

The problem, therefore, is to determine what exactly is causing these errors and whether or not they can be considered random.

Re:Failed to Finnish

[Sebilrazen]

Unfortunately you were probably going to have an informative example, but /. stripped out everything after the "<" on that line. (Assumption based on standard "for (i=0; i<number; i++)")

Re:Failed to Finnish

[Beardo the Bearded]

Doesn't really matter. If you let them vote, count all the fucking votes. It's that simple.

That's not generally how it works, at least not in Canada.

They count the votes until there's a statistical likelyhood that one candidate has an overwhelming majority. If Candidate A has 75% of the popular vote with 25% of the votes counted, they usually don't count them all. (They do count from every polling station, however.)

Only if there's no clear winner do they actually count all the votes.

Re:Failed to Finnish

[ultranova]

No measurement is perfectly accurate. Counting votes by hand is not a perfectly accurate method. We expect computers to be perfect but the measuring system isn't perfect so the results aren't either.

Bullshit. We aren't talking about measuring physical system at the level where interference from the Uncertainty Principle or other sources would be a factor. We are talking about communicating an integer from a set of a few hundred at most from one human being to another.

It doesn't matter. Our voting system has always had a margin of error, and always will. Thankfully one court in the world understands this and is bold enough to say 2% is an acceptable error.

Losing 1 vote out of 50 is not acceptable, especially since hand-counting the votes in Finland typically takes just a few hours, and the results are ready by the election day evening anyway.

Then again, I suppose I shouldn't except anything better from Finnish courts. I mean, in the past they've come up with such brilliant gems as premature ejaculation being a mitigating factor in a rape case, or convicting a man from stealing... from a landfill. Or maybe being batshit insane is just a feature of courts everywhere.

Re:Failed to Finnish

[ultranova]

But that is not the real issue: the real issue is that nobody knows or can proof why which votes were lost, and the electronic voting systems make it completely impossible to find out even if 60% of votes were lost.

Even worse: for all we know, 60% of votes really were lost and replaced with computer-generated ones. The more cynical side of me suggests that this is the real reason for pushing for electronic voting systems, for as Comrade Stalin said: "Those who cast the votes decide nothing. Those who count the votes decide everything."

Re:Failed to Finnish

[Anonymous Coward]

What you are not getting here, is that the question is not wether 2% is an acceptable rate of error, but that wether e-voting is secure or not in the big picture.

Secure means much more than the rate of error. Actually, 2% is acceptable rate for me, but wether e-voting is secure is a whole another issue.

In this case, the e-voting software is closed source. Three university doctors were allowed to peer-review the code before the elections, but they said that this was only formal, because they weren't given copies of the source code, but rather had to inspect the code in one single audition session that was limited in time. Also, they had to sign a non-disclosure agreement before the session!

How do we know, if the method the computer uses to caluclate the votes is correct? Also, how do we know, that the machine in the voting venue is running this particular software, or a modified version?

Very, very hard questions to answer. That's why I think counting methods that are done by humans (counting votes isn't such hard anyway) are much more reliable.

Re:Failed to Finnish

[thegnu]

Another example:
if during one day in one country all the transactions in one bank contained 2% errors, that is money going to the wrong recipent, wrong amounts etc. it would be totally unacceptable.

maybe if we involved a bank and put a euro coin/note in each envelope....

I dunno about completely unacceptable. If anything, I'd make money.

But my father pointed out that the reason why banking machines are more accurate is that there's more money in banking than in voting. Same with gambling machines. Voting is run by a bunch of retired volunteers with little money, and it doesn't earn anyone any money.

So voting machines are cheap pieces of shit with no QA. Not that that doesn't bother me.

Re:Failed to Finnish

[thegnu]

It's the principle that matters. You're basically saying they should randomly not let people in the booth saying "You're the error margin, good bye."
How would you feel?

But then they wouldn't be the error margin. And that's not how it works. The thing about the error margin is that if you've got a 2% error margin, i'm betting there's well over 90% probability that those lost votes will be a representative sample.

Make a law saying anyone who didn't get their vote counted doesn't have to pay taxes until the next election. I'm sure they'll get it sorted out in no time.

What you're suggesting is to increase taxes to pay for better voting machines. Get involved if you care. Don't if you don't.

Re:Failed to Finnish

[BOFHelsinki]

Actually it was modded "Finny" but Slashdot got it only 98% correct, which we find perfectly acceptable.

"Finland -- the democracy where 49% decide what the other 49% can do."

Re:Failed to Finnish

[Meski]

Why make voting idiot proof? Make it more complicated, and get a better result.

Re:Failed to Finnish

[N1AK]

Although anything done on a large enough scale will have an error margin, it does not make 2% an acceptable error for an election in a democracy. I don't like pulling figures out of the air, but if I had to then I doubt even a 0.5% error rate is good enough.

Also, in addition to the basic principle that peoples votes should be counted in an election the larger the error rate the easier to hide vote tampering on a scale that may affect the result. Especially in a system that isn't first past the post (which I believe was the case here). If you consider that in many elections in Europe that use proportional representation there could be a dozen or more parties fighting for votes your error rate is likely to be higher than the voting figure for some of these parties.

i'm betting there's well over 90% probability that those lost votes will be a representative sample.

And I'm willing to bet that the risk that a 2% error margin might lead to an incorrect result (especially in a prop rep system) isn't low enough.

Re:Failed to Finnish

[fgouget]

But that is not the real issue: the real issue is that nobody knows or can proof why which votes were lost, and the electronic voting systems make it completely impossible to find out even if 60% of votes were lost.

No, it's not impossible. We're talking about computers here, they can be audited.

The code is a trade secret and thus cannot be audited. I'm not sure about Finland, but nobody, not even the state, can audit the voting computer's code in the USA or in France.

Re:Failed to Finnish

[fgouget]

No measurement is perfectly accurate. Counting votes by hand is not a perfectly accurate method.

You seem to imply that manually counting the ballots has an error rate of 2% or higher. I would really like you to provide a link to a study that would confirm that. Because I have participated in many ballot counts in France and I would take exception to that assertion. I don't see how you could even get a 0.1% error rate if the procedures are followed properly. But of course if other countries have broken elections (like voting on 60 different things at once) or broken ballot counting procedures (like moving the ballots to a central location before counting them), then all bets are off, no matter how you count the ballots.

Re:Failed to Finnish

[thegnu]

Well said.
Cheers,
Nathan

Re:Failed to Finnish

[Anonymous Coward]

Also, I'd like to see, how

votes[canditate]+=1;

has an error margin of 2%.

I usually vote for a candidate and not a canditate. Maybe that's the problem.

Re:Failed to Finnish

[bluefoxlucid]

That's a bad example. If you count 60% of the vote, and reach 51% of the vote for one candidate, you already know that at least 9% goes to the other candidate and at most 49% does. If you're going to accept 51% as "winning," you stop counting. If you lose 2% of the vote, a candidate must win with more than 52% (note: the math is a bit more complex); otherwise, you can't reliably determine the winner.

In a bank, every penny counts as 100% of a penny.

Re:Failed to Finnish

[shnull]

in that case, this is a definite fail

What's the margin of victory?

[AuMatar]

If one person had 70% and the other 30%, the 2% won't matter and they should accept the election while fixing the problem for the next. If it's 51/49 victory, then its an issue now.

Re:What's the margin of victory?

[Toonol]

I did a little digging, but couldn't find an article in English that mentioned what the actual gap was. That probably means that the gap was greater than 2%, and the various bloggers are afraid that publicizing the fact that the missing votes don't matter might blunt our outrage.

It's a problem, certainly, but it may not have any bearing on the outcome of this particular election.

Re:What's the margin of victory?

[Anonymous Coward]

but the elections are not done only to determine who is the winner. Elections also show how much support the smaller parties and candidates have. They also show whether a smaller party or candidate suceeded in getting more votes than another smaller party or candidate. Such results can help smaller parties or candidates build coalitions, disband, get more members, or change their politics.

Re:What's the margin of victory?

[Front Line Assembly]

According to the article a few votes could have changed the results (different persons elected).

Re:What's the margin of victory?

[Anonymous Coward]

This was municipal voting where difference to get elected or not was down to one vote. Now when you need something like 50 votes in small town to get elected and then something like 40 votes didn't get counted in that town so these 2% could have easily changed results in the towns that electronic voting was tested.

Re:What's the margin of victory?

[Card]

This was a municipal election, carried out in three small municipalities with just several thousand voters. The victory margin in those municipalities was just a few votes. This error clearly affects the outcome of the vote.

Effi's site had an english version, too:
An error margin of 2% in municipal elections ruled acceptable in Finland

Re:What's the margin of victory?

[jaria]

The margin of victory was *0* votes. There are people who got elected by roll of dice in the voting board, because they got the same number of votes as someone else.

Clearly, even one additional vote would have changed the situation.

Re:What's the margin of victory?

[BSAtHome]

It's a problem, certainly, but it may not have any bearing on the outcome of this particular election.

It doesn't matter whether the votes would have made a change or not. It is the basics of fairness that dictates that all votes should be counted. The margin of error should not play a role. If people are asked to vote, what is the problem with accepting the responsibility for counting them? The precedent set here says: we don't care about your vote since it doesn't matter. That is a very dangerous precedent because people will start losing interest and faith in the democratic system this way.

Re:What's the margin of victory?

[Anonymous Coward]

It's a problem, certainly, but it may not have any bearing on the outcome of this particular election.

It doesn't matter whether the votes would have made a change or not. It is the basics of fairness that dictates that all votes should be counted. The margin of error should not play a role. If people are asked to vote, what is the problem with accepting the responsibility for counting them? The precedent set here says: we don't care about your vote since it doesn't matter. That is a very dangerous precedent because people will start losing interest and faith in the democratic system this way.

All votes should be counted? Really? Did you look at any of the PDFs of the actual ballot scans from the Coleman/Franken election? Some of those were pretty indeterminate as to who they were voting for. Some of them should have just been ignored. But often those weren't. During the recount they had folks trying to figure out just who the heck the morons who'd filled out the ballots intended to vote for.

Funny thing is that when the ballots were messed up in a way that seemed to indicate a vote for Franken was likely then in many cases that'd be counted as a vote for Franken.

Now here's an even funnier thing. When some of the ballots were screwed up in a way that would seem to indicate a vote for Coleman, well wouldn't you know, often they just couldn't figure out for the life of them who the heck that moron intended to vote for and discarded the vote.

I'm all for counting as many votes as possible, just as long as the votes are plainly unambiguous. And by plainly unambiguous I mean can be "told within a split second who the voter intended to vote for." Not "has to be carefully examined and pored over as if divining entrails to figure out voter intent."

If a voter can't be bothered to make sure their ballot is filled out properly then they clearly do not care enough about the election to have their vote counted, especially when the process of determining "voter intent" is opened up to shenanigans by third parties.

Re:What's the margin of victory?

[Weird O'Puns]

The article you linked has couple more points that are in my opinion more important than the problems with the interface:

Additionally, there is risk of a breach of the anonymity of the votes, because the electronic ballot box has been archived with information on who voted and how. The e-voting project had been strongly criticised by Effi from its inception for the lack of transparency both in the process and software.

First, the system is not anonymous. Right people can see from the archives who voted for who. In fact, right after the election they were telling those who asked if their vote had been cast or not.

Secondly, the system is completely closed. There was an audit, but the auditors were forced to sign a NDA. There is also no paper trace. The voter has no guaranties that his/hers vote has been cast to the right candidate.

Re:What's the margin of victory?

[jaria]

I did a little digging, but couldn't find an article in English that mentioned what the actual gap was.

It was zero votes in two towns, for instance. That is, two candidates got the same number of votes, one got to the city council.

That probably means that the gap was greater than 2%, and the various bloggers are afraid that publicizing the fact that the missing votes don't matter might blunt our outrage.

The correct state of affairs was stated here in this thread already. Do your research first before you jump into conclusions.

Re:What's the margin of victory?

[im_thatoneguy]

Except that in every landslide the opponent almost universally concedes (In US politics).

Re:What's the margin of victory?

[dryeo]

That raises the question, how the hell can a ballot be so badly designed that (discounting purposely spoiled ballots) there can be any question about the intent of the voter?
When I vote (Canada) there is like an inch between the boxes and as you say a quick glance is all that it takes to see which box was checked.

Re:What's the margin of victory?

[Anonymous Coward]

Bad design comes in due to the multitude of ballot types in the U.S. For example the punchcard ballots with the infamous "hanging chads" from Florida the 2000 elections. Then there are the ballots mentioned in my post above from the Minnesota election. The "fill in the circle" type. Exactly like what you'd get on a standardized test except they're filled in with a pen rather than the #2 pencil your teachers were so adamant about.

I'm quite sure your Canadian ballots with the boxes an inch apart could be screwed up in exactly the same manner as the circles.

For example someone might make a single mark in Candidate A's box and then completely fills in Candidate B's box. Or they might start to mark Candidate B's box, filling a good portion of it, then realize their mistake and draw a big old X through it and then fill in Candidate A's box properly.

These are all errors on the voter's part that would be easily remedied if any of them were bright enough to just take the ballot back to the voting officials, tell them "I messed my ballot up and need a replacement" and then tear their poorly marked ballot up into several pieces.

Where my objection comes in is in trying to "interpret" the hopelessly screwed up ballots.

From above take the example ballot that has a single mark in Candidate A's box and then completely fills in Candidate B's box. Someone interprets this as being a vote for Candidate B. Seems fair enough. Single mark versus properly filled in box. Can't really argue with that. Intent seems clear.

Take a second example ballot that has a single mark in Candidate B's box and then completely fills in Candidate A's box. Now an impartial person who judged the first ballot as a vote for Candidate B would certainly judge the second as being for Candidate A because it is the same mistake, only reversed. But suppose that the person who judged the first ballot as a vote for Candidate B isn't really impartial and they decide the second ballot isn't clear and won't be counted.

This seems to be the case in Minnesota. In many cases ballots messed up in a way that seemed to favor Candidate B were counted for Candidate B while ballots that were messed up in a way that seemed to favor Candidate A were discarded or counted as votes for Candidate B.

This is of course entirely unethical and wrong. It is selective disenfranchising of voters for Candidate A by a recount process that is supposedly impartial but in reality isn't (for whatever reason.)

This is a corruption of the process that could be entirely avoided by simply discarding all votes whose intent is not clear. Yes, doing this would disenfranchise voters but it would disenfranchise them across the board because they failed to properly fill out their ballots. In effect they have disenfranchised themselves.

In my view this would be better than the current method which makes it possible for the biased to cherry pick votes from ambiguous ballots.

Non-electronic spoilage rate

[profplump]

2% doesn't mean anything unless you know the spoilage rate for non-electronic voting. In the US 2000 elections, 1.94% of the ballots cast were spoiled, and most of those were not electronic. I don't know if Finland usually has similar spoilage rates, but if they do I don't see why this is any more or less a problem than the old method.

Re:Non-electronic spoilage rate

[Kizor]

Where the HELL do you get your ballots? Finnish ballots have a circle. You take a pencil and write a number in the circle. That's it, you're done. Voting is held for one issue at a time. The failure rate of this electronic voting system is several times higher than that of traditional ballots.

Traditional Finnish ballots fail because of unavoidable human stupidity, but here failures are due to technical problems and incompetent design that allows human stupidity to be expressed. The machines apparently allow removing the slotted card used before registering the vote - was there no defensive design at all? The Finns have not gone through the catastrophes you USAns have, and still think that every vote counts. Anything else is a mockery of democracy.

Re:Non-electronic spoilage rate

[raynet]

The Finnish system has very very low spoilage rate, the voting is done by writing the number of the candidate on the ballot and just about everybody does manage to do it correctly. And the margins in small municipalities are very tiny, had I gotten 7 more votes I would have been elected, and I got 3 :)

Re:Non-electronic spoilage rate

[puhuri]

The number of rejected votes has been less than 1% in most muncipal elections.

In Finnish voting, a number of choisen candidate is written in booth by pen on paperboard sheet, that is then folded, stamped by official and put into ballot box. Many of invalid votes can be considered as protest votes (vulgar drawings, names of fictional charactes), but some of votes are rejected because number cannot be clearly identified (like 1 or 7). In larger cities, there are more than 100 candidates, so numbers can be upto 3 digits.

Re:Non-electronic spoilage rate

[thesupraman]

Quite, reading through this reasonably carefully it would appear that 2% of the votes 'Were not counted at all' which from my election experience would mean that 2% of the issued/recorded voters whom entered a booth did not then result in a vote in the system.

The obvious reason for this would be that either a machine or machines were not transfered to the centralised count - something that should stick out in the paperwork like a sore thumb..

OR

2% of people did not understand/complete the voting procedure correctly - which would not be unusual at all.

It is quite common, although rather surprising, to get paper voting papers that have not been marked in any way - one can only guess that the voter got in to the booth, could not find the person/issue/whatever they thought they wanted to vote for, and didnt bother. You also get voters who do quite obviously stupid and incorrect things in marking their paper ballots, like circling the name of the candidate they support, rather than marking the box.

I would not be at all surprised with a 2% missing vote from the combination of people who just didnt finalising their vote on purpose, and people who did not correctly complete whatever the procedure was.

Neither of these is a particularly 'electronic voting' type fault - it happens all the time in paper based systems. If the numbers are much higher than tghe old style voting, then it could mean the new system is not clear or understandable enough.

Of course IF the problem is simply missing voting machine counts, then that is a whole different kettle of fish, and requires investigation.

So long as people can vote in 'privacy and anonymity' then it is damn near impossible to actually get all their votes..

Re:Non-electronic spoilage rate

[Anonymous Coward]

I think you should differentiate between spoilage rate for paper and e-voting.
With the paper ballots you could choose to give an invalid vote to send a political message (after all you *do* care about voting but all choices sucked equally, etc.). On the other hand, with e-voting it's either a system failure or bad design at fault.

Re:Non-electronic spoilage rate

[getuid()]

I don't see why this is any more or less a problem than the old method.

I do. It's because of why the spoilage occured.

In the case of paper ballots, it's easy to imagine where the spoilage comes from: thousands of helpers handle millions of pieces of paper. Either they mis-count, they accidentally destroy bills... whatever. It's easy to imagine that there is some damage, and the most likely scenario is not a systematic error or a manipulation.

In the voting machine case, it's different. There's a machine counting results. It's not supposed to miscount. There's also the machines adding the numbers and doing the math -- it's not supposed to be wrong. If it is wrong, then it's propably not because of 1000 small errors adding uncorrelated small pieces to the spoilage (like it would be in the case of the manual counting), but it's most probably because of one single error or manipulation in the system. Being one single malfunction, it is not at all likely anymore that it has nothing to do with manipulation on purpose -- on the contrary, this scenario is very possible, and more or less likely.

An analogy with weapons, to clarify (not to justify) my points: if a bow+arrow wouldn't be able to hit a 1 foot target at 1000 yards distance, you wouldn't complain. That's the limits what a bow can do, after all... But if a high-tech sniper rifle missed the same target from the same distance, you'd have all rights to complain! The gun is supposed to work orders of magnitude better, there's no room for such a big error there.

Re:Non-electronic spoilage rate

[kaip]

In the Finnish municipal elections 2008, 0.17% of the paper votes were inadvertently spoiled (unclear marking in the ballot ticket etc.) and had to be dismissed. This can be compared with the 2% of the electronic votes lost in three municipalities in which the new voting system was piloted (see Effi's Electronic Voting FAQ, in Finnish).

The total fraction of the spoiled paper votes in the municipal elections was 0.6%. Most of the dismissed paper votes were due to a deliberate action by the voter (votes for Donald Duck - a popular candidate here!, empty ballot tickets etc.). There is no evidence to support the claim that the lost electronic votes were due to a deliberate action by the voters. On the contrary, in addition to the usability problems with the voting machines, there is evidence of system malfunctions which may have contributed to the lost votes (slow response times, freezing of the voting machines during the voting etc.). Additionally, the electronic voting did allow to cast an empty vote.

Re:Non-electronic spoilage rate

[amorsen]

It is quite common, although rather surprising, to get paper voting papers that have not been marked in any way - one can only guess that the voter got in to the booth, could not find the person/issue/whatever they thought they wanted to vote for, and didnt bother.

My guess would be that most of them are protest votes.

Re:Non-electronic spoilage rate

[vuo]

0.5%

Re:Non-electronic spoilage rate

[Anonymous Coward]

> Neither of these is a particularly 'electronic voting' type fault - it happens all the time in paper based systems.

Yes, but only the electronic systems leaves you completely guessing what the fault was.
Obscene drawings on a piece of paper give a quite obvious sign of the intent of the voter, as does a badly written number. A unusually high number of either can be noticed and the election process or political system could be changed to improve things.
A non-registered vote due to someone pulling out the smart-card does not give you any information.

Re:Non-electronic spoilage rate

[SwabTheDeck]

In Finnish voting, a number of choisen candidate is written in booth by pen on paperboard sheet, that is then folded, stamped by official and put into ballot box. Many of invalid votes can be considered as protest votes (vulgar drawings, names of fictional charactes), but some of votes are rejected because number cannot be clearly identified (like 1 or 7). In larger cities, there are more than 100 candidates, so numbers can be upto 3 digits.

Articles in sentences are then discarded to make sound like bad American stereotype of European talking.

Re:Non-electronic spoilage rate

[Anonymous Coward]

You mean, everyone manages to write a number correctly.

By removing all possible redundancy from a system, you lose the ability to detect errors. That doesn't mean your data is now error-free.

If it's almost impossible to cast a spoiled ballot, of course your spoilage rate will be low. That doesn't mean everyone voted the way they meant to.

Re:Non-electronic spoilage rate

[Hatta]

It doesn't matter what the non-electronic spoilage rate is. This is electronic, if it loses anything it is hopelessly broken. Would you be ok with 2% of the deposits at your bank going missing? 2% of your credit card charges going missing? We have the technology to do this right. A single uncounted vote is unacceptable. 2% loss is proof of sheer incompetence, and the results should not be trusted no matter what margin the results seem to indicate.

Re:Non-electronic spoilage rate

[volpe]

So, your opponent got THREE TIMES as many votes (9) as you did (3), and you consider that a small margin?

Re:Non-electronic spoilage rate

[raynet]

Yes, I call that a small margin as that difference is 0.45% of total votes cast, now if I was running in those towns where 2% of votes were lost, I would be furious.

Re:Non-electronic spoilage rate

[caitsith01]

Precisely. There should be 0% of votes 'lost' with an electronic system - where could they go? A computer is not like a physical system operated by humans, with pieces of paper going left, right and centre. It should be utterly trivial to ensure that the number of votes and voters matches precisely.

People are amazingly tolerant/ignorant about these things. Anyone who follows elections closely would be well aware that 2% is more than enough to swing an election in many situations.

For example, to pick US presidential elections:

2004: Bush beat Kerry by 2.5%
2000: Bush beat Gore with 0.5% less of the vote
1976: Carter beat Ford by 2%
1968: Nixon beat Humphrey by less than 1%
1960: Kennedy beat Nixon by 0.2%

Although the electoral college obviously affects the impact of the popular vote in those particular elections, 2% of the vote going "missing" would certainly have had the potential to change most of the results.

Re:Non-electronic spoilage rate

[thesupraman]

I couldnt agree more, but again this is nothing to do with electronic voting.

It is the question of should there be a 'None of the above' protest option available - which does actually exist in some electoral systems.

There is one nice system that blocks any party/person who receives less %age of the vote that the 'abstain' vote from being represented, although this is not often used in general elections.

a nice big red button marked 'no confidence in any of these idiots' would be a most sensible addition to any voting system!

Re:Non-electronic spoilage rate

[Anonymous Coward]

You address only one of my points, those "protesting". And already there it is unlikely that any profanity will be allowed - 'None of the above' is not the same thing and some people might not choose it.
I very seriously doubt you will ever find 'no confidence in any of these idiots'.
A free-form field might get close, but compared to paper, there is the "issue"/difference that probably either nobody reads it or it gets actually published.
The other points, like someone failing to use the system correctly seem a lot more difficult to solve though, even if you just aim to record the fact.
These may indeed not be problems of electronic voting in general, but I do think we are talking about electronic voting as of today, not what it could be in 500 years with a single machine costing a few billion dollars (if you pre-order in bulk), I don't doubt you can come close to or even beat paper if you invest that kind of effort - of course not considering what better use there might have been for it.

EFF press release

[martin-boundary]

Unfortunately, around 2% of the EFF's press release was not translated correctly into English. I would like to assure all slashdotters that their comment posts about it will still be accepted for discussion.

2%

[clarkkent09]

I guess this is a question of whether it is possible to have a "perfect" user interface such that 100% of people who use it will get it right. Given the number of nincompoops out there, that is a pretty difficult problem. What is the percentage of mistakes with paper ballots? I bet there are 2% who manage to screw that up too.

I looked at the demo of the voting machine user interface and it seems perfectly sensible. You put your voting card in and press the number of the candidate you are voting for. A message comes up with large friendly letters telling you something like "This is the candidate you are voting for: <candidate details>. Press OK to cast your vote. [OK] [CANCEL] Apparently, at this point 2% of the voters simply pulled their card out of the machine and walked out of the booth without pressing OK. If they didn't have the confirmation screen, then the same people would press the wrong number and vote for the wrong candidate, and then complain that they weren't given a chance to correct it.

Re:2%

[raynet]

It seems that the [OK] [CANCEL] buttons didn't have very good feedback and they didn't work all the time, sometimes requiring multiple clicks to register, which is why some people took their cards after clicking on OK several times.

Re:2%

[Anonymous Coward]

Except, of course that [OK] isn't really Finnish. Why the buttons couldn't have been something like [CAST A VOTE] ([ÃÃNESTÃ]) or [CANCEL VOTING] ([PERUUTA ÃÃNESTYS]) in the first place? And these weren't the greatest quirks.

The interface and functioning are bad usability and that's the crux of the matter.

Re:2%

[Anonymous Coward]

You could always make it so voters can't take out their card without selecting either OK or CANCEL.

Like on the ATM, if you want out, you can't just pull the card. Seems simple enough to me.

Re:2%

[jaria]

There were multiple problems. Bad user interface design, which allowed modes where the votes don't get registered. Machines becoming frozen at the time of the voting process, making it impossible to press the OK button. Instructions which stated to press OK once, when you had to press it twice. And so on.

The most serious issue is that if the machine freezes for several minutes, the voter does not know what to do. If he pulls the card out before the machine returns to life and you can press the 2nd OK, your vote was lost.

No one really knows what happened why the 2% of votes were lost. I presume it is a combination of people simply walking out in the middle of the voting process, machine hangups, and people misunderstanding what they had to do, and possibly some yet unknown problems.

By the, none of the problems described above were in dispute. The court only decided that despite the problems, the result stands.

Re:2%

[clarkkent09]

But who says that people who couldn't handle the simple system mentioned above would actually wait for the card to come out? They would probably just leave the card in the machine. Would that count as a bug with the voting machine too?

Re:2%

[jeti]

Good testing would have found the problem with people pulling the card without confirmation of the vote. The machines should have been modified to lock the card until the vote is cast.

Re:2%

[sa1lnr]

Here's an idea, what if the machine did not release the voting card until a vote had been properly cast. It surely can't be hard to implement in this day and age?

Re:2%

[legirons]

It seems that the [OK] [CANCEL] buttons didn't have very good feedback and they didn't work all the time, sometimes requiring multiple clicks to register, which is why some people took their cards after clicking on OK several times.

OK/Cancel buttons are a disaster-area anyway, since every OS and every application has a different idea on what order they should go in, and people get used to clicking the left/right one for OK without looking at the labels.
 

Re:2%

[legirons]

Here's an idea, what if the machine did not release the voting card until a vote had been properly cast. It surely can't be hard to implement in this day and age?

You're still trusting the software to do the right thing -- I can just imagine this voting machine being like trying to persuade a Mac to spit-out a CD when it's "sure" (incorrectly) that something is using it.

Re:2%

[bentcd]

But who says that people who couldn't handle the simple system mentioned above would actually wait for the card to come out? They would probably just leave the card in the machine.

There would need to be a lamp outside of each booth labeled "voting in progress" and one labeled "voting complete". Then there would need to be functionaries outside the booths ready to stop people who leave booths with the "voting in progress" light on and explain that they are not finished yet.
Alternatively, a door with an automatic lock on it that wouldn't let you out unless you've collected your card but I don't see this being very popular, or safe.

Re:2%

[Anonymous Coward]

I have an even looser criteria: Do we have a high degree of confidence say 99.9% that the winning candidate is the correct winner if the election ballots were somehow "perfect". That is why a recount like in Minnesota is a total waste of time and money. Even after tediously reviewing all the paper ballots there is still enough ambiguity so that each side can reasonably argue that they are the "true" winner. The outcome of a large close election should depend on a simple coin toss rather than judicial whims.

Re:2%

[LateArthurDent]

OK/Cancel buttons are a disaster-area anyway, since every OS and every application has a different idea on what order they should go in, and people get used to clicking the left/right one for OK without looking at the labels.

If you color the ok button green and the cancel button red is there any culture for which that would seem backward? I honestly don't know the answer to that, but the convention of "green=ok, red=pay attention, something's wrong" might be universal.

Re:2%

[ultranova]

I guess this is a question of whether it is possible to have a "perfect" user interface such that 100% of people who use it will get it right. Given the number of nincompoops out there, that is a pretty difficult problem. What is the percentage of mistakes with paper ballots? I bet there are 2% who manage to screw that up too.

You lose. According to EFFI, only 0.6 percent of paper votes had to be rejected, and of those, only 0.17 percent were not obviously intentional screwups, such as voting for Daffy Duck etc. Not that Daffy would be a worse politician than the current lot, but that's another discussion ;(.

Re:2%

[Anonymous Coward]

The card should be given to a poll worker to be recorded, and if the entries aren't completed, the voter should be given the choice to discard it and start over. You can't expect a first time user to understand when something works or doesn't work.

It all depends...

[fuzzyfuzzyfungus]

Except in extremely close races, a smallish percentage of lost/spoiled/uncounted votes isn't an issue as long as the lost votes are a representative subset of all the votes. If it is a selective subset, then you have a serious, serious problem.

Same thing as polling. If the people you do poll are a representative sample, you don't actually need all that many of them to get the correct answer. If you get an unrepresentative sample, then your results are worthless. It should be noted, of course, that with elections, unlike polling, you are still obligated to put forth your best effort to count everybody's vote(though, depending on the technology, imperfection is inevitable to some degree).

Re:It all depends...

[Kizor]

Good point. I'd hazard a guess that people from the demographics that have the most trouble with new technology were the most likely to have trouble with electronic voting.

Selective.

Re:It all depends...

[Anonymous Coward]

But they possibly were selective.

Apparently the problems were mostly with unclear UI of the electronic voting machine. As such we have a good reason to assume that they targeted elderly voters more.

Besides, your argument works poorly when it is matter of municipal elections in Finland. We don't have just two parties which need to get about right. We have dozens and each can gain or lose positions on a margin of a few votes. If this 2% translated to the elderly losing 3% of their votes, that could actually mean several spots of loss for the Senior Citzens' party (which actually exists).

Re:It all depends...

[jaria]

Perhaps, like the elderly. But we have to remember that in this case we also had young computer savvy people who experienced problems. (This is all well documented in the court case and not under dispute by the way.)

For instance, a computer science teacher saw the machine crash when he inserted his smart card.

A young student saw the computer froze for several minutes in the middle of the voting process. Had she not waited, pulling out her card before the process was final would have resulted in her vote getting lost.

None of this is demography related. The machine crashes when it decides to do so... (or when networking conditions, timing of key presses, memory leakage reaching a certain point, etc makes it to).

Re:It all depends...

[Anonymous Coward]

> None of this is demography related. The machine crashes when it decides to do so...

Yep, it may very well have _decided_ to do so, e.g. because you voted for a certain candidate (which might also happen due to a bug, e.g. the counted votes are stored on a part of a flash that is failing and causing a timeout). Is it even possible to prove otherwise? The way most voting machines were designed, (not) reviewed and stored the answer is likely "no" and the court just decided to _assume_ it was a random bug, for no particular reason - despite the fact that if there is one thing that computers are not is "random", so you actually already know that the failures had some kind of system to them.
So the question is even more precisely: is it even possible to prove whether the deterministic (but seemingly unknown because nobody cares enough) system according to which the voting process failed is correlated to the vote cast or not?

Re:It all depends...

[dropadrop]

Except in extremely close races, a smallish percentage of lost/spoiled/uncounted votes isn't an issue as long as the lost votes are a representative subset of all the votes. If it is a selective subset, then you have a serious, serious problem.

I would guess that in this case it was mainly older people who did not get it right. I might be wrong, but they tend to be less "computer literate" and would have more usability related problems.

Is this a problem?

[sonamchauhan]

Unless the 2% of lost votes actually matter (e.g., the race is close, or 2% is signifant for proportional representation, or it indicates a deeper systemic problem) why do they really matter?

Re:Is this a problem?

[raynet]

In my municipality, about 80 votes for candidate was about the maximum, 11 got you elected and 5 was enough to get a in as a backup. Also votes are used for dividing up seats in different comissions/boards/positions etc. And if I recall correctly, of about 2000(ish) ballots cast here, one was spoiled, rest were valid.

Re:Is this a problem?

[jaria]

The race was close. In two of the three cities, the last person to be selected was chosen randomly because they got the same number of votes as some one else. In the third city there was just a few votes of difference.

In addition, the lost number of votes alone would have been enough to get someone elected, i.e., 2% would get you to the city council.

Re:Is this a problem?

[legirons]

Unless the 2% of lost votes actually matter (e.g., the race is close, or 2% is signifant for proportional representation, or it indicates a deeper systemic problem) why do they really matter?

Because it shows that their computers can't count. Computers are supposed to be very good at counting - if they fail at something very simple like that, it means they've been programmed to incorrectly count, which would be illegal because that's election-fraud.

Re:Is this a problem?

[ruin20]

Another aspect is how does this compare to the error margine in paper balloting? Is 2% really that much larger then, for example, the number of hanging chads in florida?

First

[Anonymous Coward]

First. It's impossible to spoil an electronic ballot.
Second. If not a ballot problems, what could affect electronic vote??? Only illegal enforced spoilation.

Translation of article

[Novus]

Standard disclaimer: legalese may not be 100% accurate. I am not a lawyer.

Electronic Frontier Finland ry (Effi) is shocked by today's decision by the Helsinki Administrative Court. The court downplayed the problems of e-voting and declined to annul the result of the election. Thus, the elections will not be repeated unless the Supreme Administrative Court overturns the decision. After last year's municipal elections, it was found that 232 voters' votes were lost.

Effi assisted in 16 complaints regarding e-voting in the three municipalities in which it was trialled. Based on the witness and expert statements gathered by Effi, the problems were due, amongst other things, to machines freezing at the moment of voting, inadequate testing, user interface design issues, not fixing detected problems and incorrect instructions. In some trial municipalities, even one vote could have changed the members of the council that was elected.

A central basis for the decision was that "A failure rate of slightly more than 2% can not, as such, be considered to show erroneous activity on the behalf of the election authority... The threshold for repeating an election must also be high even with respect to realising basic state rights."

Lawyer Mikko Välimäki, the complainants' advocate, comments: "The decision does not seem to be well founded. The problems are undeniable, and the election result was incorrect. The Administrative Court's line chips away at the trust in Finnish democracy. What happened to the basic rights of the "slightly more than 2%"?"

The vice chair of Effi, Ville Oksanen, wonders: "I understand that we agree that the election trials had serious problems. Now, however, the Administrative Court has accepted a situation in which it is clear that the result of the election did not correspond to the will of the voters. The last candidates to pass are within the margin of error of the system." Oksanen continues: "Not even the municipalities have denied the existence of problems in the judicial process or the possible effect of the missing votes on the results of the election. Going by the Administrative Court's logic, we could give up recounting votes, because the results don't change by more than a few votes anyway. The constitution guarantees everyone an equal right to vote. It doesn't say anything about 98%!"

Jari Arkko, who complained about the elections in Kauniainen, intends to appeal the decision: "We will study the court's decision in the next few days, but we have previously considered the matter to be so important in principle that we have reason enough to appeal to the Supreme Administrative Court." In Vihti, complainant Tero Miettinen agrees: "A badly implemented system should not decide who is elected to the council of my home town. The margin of error in the electronic voting was many times that of the traditional election system. It is hard to understand why the Administrative Court does not consider this an indication that the system has failed."

Handling of the problems was even a bigger mess

[jaria]

We can never have a 100% perfect system. Paper ballots lose about 0.5% of votes in Finland. But 2% is way too much. We spent a lot of money on that system, and it gives worse results than the almost free paper ballot system (the votes are counted by volunteers).

The reasons for the mess include incompetence on the part of the ministry organizing the elections and completely ignoring the feedback from external experts prior to the election. For instance, minister of justice, Tuija Brax, painted the worries as "science fiction" just before the elections.

But I am even more stunned by the handling of the problems after they were discovered. Normally, if you get problems you try to deal with them and rectify the situation. But many of the government officials, voting boards, etc. have focused on blaming the users, explaining that 2% isn't a big deal, and attempting to avoid discussion of the actual technical problems that were discovered.

And it gets worse. My city voting board actually blamed the votes for purposefully misusing the machines so that they would appear unreliable. Conspiracy! Normally it is the crazy citizens who suspect the government of conspiracies, but this time the government thinks the citizens are conspiring against them. Maybe the officials involved should be re-allocated for JFK murder investigations...

More information here:

http://www.arkko.com/vaalit/evoting.html

Re:Handling of the problems was even a bigger mess

[SLi]

We can never have a 100% perfect system. Paper ballots lose about 0.5% of votes in Finland. But 2% is way too much. We spent a lot of money on that system, and it gives worse results than the almost free paper ballot system (the votes are counted by volunteers).

Volunteers?

I assure you I got paid in the order of 300 euros for essentially the single voting day. Ok, only 92 (IIRC) euros of that were for the counting itself.

Re:Handling of the problems was even a bigger mess

[fgouget]

We can never have a 100% perfect system. Paper ballots lose about 0.5% of votes in Finland.

A link to the paper finding that manual counting has a 0.5% error rate in Finland would be nice...

This is not the final decision in this matter

[jaria]

There will be an appeal to the highest court.

At least *my* vote was counted

[blind biker]

Luckily, I am one of those who voted with pen and paper. From what I've heard, the electronic voting system was fairly complicated, and my guess is that I could have fallen victim of it.

The candidate I voted for didn't get through. I think I'll blame it on the fucking electronic voting (I'm sure it had nothing to do with the fact that he promoted a rabidly anti-car and pro-cycling agenda).

WTF?

[EdIII]

Does this mean 98% of votes is enough to figure out how the other 2% voted?

Of course NOT. ABSOLUTELY NOT. The answer here could never be yes if you are going to even begin to pretend you have a form of democracy.

I can accept that my own vote may not be counted due to any number of errors in the voting process. I WILL NEVER ACCEPT that somebody uses math to determine how I *may* have voted and use that to determine who will represent my interests in government.

If the 2% does not make a difference in deciding the vote, then the error rate does not affect the outcome. If that 2% could change the outcome, than the ONLY solution is to vote again.

To not have another vote is accepting the outcome of a flawed process. It may be practical to do so in some situations, but it does not change the fact the process is *FLAWED*.

Of course, I don't know a single government that is not deeply flawed anyways.

Re:WTF?

[Anonymous Coward]

Using the 98% to guess the remaining 2% is stupid, but probably gives you the correct answer.
But only of you are absolutely sure that that 98% is correct. Missing 2% because someone "lost" them means that you can't thrust the other 98% anyway.

Re:WTF?

[justinlee37]

That's all true, but you're thinking of this in the wrong way. If I count up the votes and candidate A receives 70% of the vote, and candidate B receives 28% of the vote, then it really doesn't matter how the remaining 2% voted. Mathematically speaking, they can't change the outcome of the election.

Re:WTF?

[EdIII]

I think you missed my point. I even stated that if the 2% made no difference in the vote, than they would not matter.

What the article was suggesting was "figuring out" how the 2% voted. You can't figure it out or make any assumptions as to which way it voted. It's an error. Math cannot apply here since it is a vote. It's wrong to mathematically determine the probability of what an error is and then using that information.

It's like saying, "Yes Mr. X, we had an error with your vote. However, our mathematicians have determine you probably meant to vote this way". Of course they can't ask Mr.X what his vote was, but that's what I feel the question was asking.

No, let's have an ATM

[Anonymous Coward]

That gives you up to 2% less money than you asked for. I would expect it to be shut down, if not trashed by vigilantes.

You can make a better UI

[Anonymous Coward]

The machine should "eat" the card and not return it until the vote was cast.

1. you put the card in
2. press the number of the candidate
3. ok, cancel
4. if ok get card else return to 2

If you come to vote you need to vote. If people come to vote without knowing who they'll vote for and then want to decide not to vote at all, they should've stayed at home in the first place. Or allow not casting a vote, but the user would be required to get one of the staff to help with that. Cancellation would require a key or something. This would have two good sides. One: user can't cancel unknowingly. Two: comming to the election uninformed should put them on the spot (shame as deterrent).

The card should be something the user needs outside of voting. Like the state's ID card, driver's license etc. Something that identifies you digitally.

In the opening where the card is put in you could have a small hole, so if the machine dies you can get the card out with pincers.

Re:You can make a better UI

[andy.ruddock]

You'd better make sure that one of the options is "None of the above". I can care enough to cast my vote, but currently if I feel that none of the candidates is sufficiently qualified to do the job then my only option is to spoil my paper.
If I stay at home I get counted as a "didn't care enough to vote", if I spoil my paper I may get counted as "idiot who didn't understand the process", but I've made my mark.
You can't get staff to help with not casting a vote, the process has to be anonymous.

Just not getting it...

[Max Romantschuk]

I can deal with 2% lost votes in one election. But why they want to use electronic voting machines in the first place I'm not getting. And if they plan to use them after this I'll be voting by pen and paper as long as it is an option.

The Finnish system is so simple that you can't really make it any better. You get a piece of paper, with a circle designating where to write your number. A line shows which way is down.

In the voting booth there are writing models of every number, and a list of all candidates with their numbers. Most six year olds have mastered pen usage enough to copy the model numbers so that there is a 99% chance they won't be ambiguous. They are, of course, designed to be non-ambiguous in the first place.

Counting is done by hand, huge amounts of eyeballs, usually representatives from each party. You'd have to corrupt an insane amount of people to make a dent in the system.

There is no box to mis-tick, no buttons to mispress. Why fuck the system up by complicating it? Making voting cheaper is not an argument. I'd rather pay more taxes and trust that my vote counts whey they spend the rest...

Re:Just not getting it...

[drsmithy]

The Finnish system is so simple that you can't really make it any better. You get a piece of paper, with a circle designating where to write your number. A line shows which way is down.

Heh. Now that's an interesting little cultural difference. As an Australian I would expect an arrow like that to be showing which way is up. :)

Re:Just not getting it...

[ZygnuX]

I think it's not an arrow, but an horizontal line which designates the "floor".

Re:Just not getting it...

[Max Romantschuk]

I think it's not an arrow, but an horizontal line which designates the "floor".

Exactly. :)

Re:Just not getting it...

[SLi]

The Finnish system is so simple that you can't really make it any better. You get a piece of paper, with a circle designating where to write your number. A line shows which way is down.

Heh. Now that's an interesting little cultural difference. As an Australian I would expect an arrow like that to be showing which way is up. :)

Well, that's not surprising, because you live down there and there's really nothing that is more down except penguins :)

Let me get this straight...

[abigsmurf]

The EFF says the system is flawed because it requires people to verify their vote once they selected it? It would be far worse if there wasn't a verification.

It's not like this is unusual behaviour in any electronic system. You don't take your bank card out of an ATM or chip and pin machine until you're told to and most ATMs require a 'yes I'm sure' for any actions that would cost. You also don't yank out your card until you're told to.

A 2% spoilage rate although higher than the typical rate, isn't incredibly high. At some point you really can't make the process any easier without your actions actually making mistakes and confusion more likely. Of course there's the question : do you really want people who can't handle incredibly basic instructions voting?

Re:Let me get this straight...

[Weird O'Puns]

The EFF says the system is flawed because it requires people to verify their vote once they selected it?

No, they criticize that it was possible to stop the voting process unintentionally. This could have been solved using card readers that took whole card inside the reader, so it wouldn't have been possible pull it out without voting. Other solution would have been to display clear message that the voting process was interrupted and no vote was cast.

You also don't yank out your card until you're told to.

ATM won't give you any money if you yank out the card too early. That means that there is a clear feedback when the transaction has happened correctly. With voting machines there was no such feedback.

A 2% spoilage rate although higher than the typical rate, isn't incredibly high.

Why spend money to switch to a system that is worse?

Re:Let me get this straight...

[abigsmurf]

ATM's don't let you yank out the card but there are scores of Chip and Pin machines that will.

Who says the system is worse? Spoilage rate and costs are only 2 factors. There's staff levels, level of monitoring needed, speed of voting, speed of counting etc.

2% spoilage isn't huge and it's probably safe to assume that it'll improve in the next election. All the advantages with only a slight difference in spoilage would make it a sensible choice.

Re:Let me get this straight...

[Ornedan]

It's voting. That means spoilage has weighting ratio relative to all other considerations of +infinity.

Re:Let me get this straight...

[Joutsa]

Actually, ATMs around here let you pull out the card during transaction. Unlike with voting machines, you can easily tell that the transaction failed by the fact that you have no cash.

Re:Let me get this straight...

[corsec67]

you can easily tell that the transaction failed by the fact that you have no cash.

The bank would quickly jump on the issue if it was possible for a transaction to fail and the user to have cash.

Or too much cash, what if someone got $50s instead of $20s? For e-voting that would be like people voting more than once.

Re:Let me get this straight...

[Anonymous Coward]

It's not like this is unusual behaviour in any electronic system. You don't take your bank card out of an ATM or chip and pin machine until you're told to and most ATMs require a 'yes I'm sure' for any actions that would cost. You also don't yank out your card until you're told to.

That's only part of the issue. What if the voter simply leaves the card in the machine and walks away at the same point in the process? Nothing is idiot-proof.

Now, before you say "almost nobody walks away from the ATM without taking thir card." Well, almost nobody is still not nobody. In the case of ATMs, the ATM user also owns the card. People are far less likely to leave behind something they own (an ATM card) as opposed to something they only possess for a few minutes (the voting smart card).

While I say that no system is completely idiot-proof, here's how I would implement the system:

1. Electronic voting is optional.

2. Every legally registered voter gets a smart card of their own that serves as a voter ID card.

3. For people electronically voting, the card is presented to an official who verifies the identity of the voter. The official inserts and removes the card from a machine that records that that individual has been allowed to vote. This machine does not communicate with the voting machines. This facilitates comparing the total number of voters vs. the total number of votes cast.

4. The voter goes into the booth and inserts the card there. The card is locked into the machine.

5. The machine beeps at you if you do not move forward with the voting process in a reasonable amount of time.

6. The voting machine records the anonymous number of the card with the vote.

7. The voter is given a paper receipt with a vote serial number on it and their voter number. The machine does not record the voter number with the serial number. Guess you have to trust it.

8. Only then does the machine give back the card.

9. The voter should be able to look up his/her individual vote in the combined voting database for the election using the serial number on the receipt.

10. This all still doesn't guarantee that all votes are properly counted. It does minimize errors at the local level and provides the voter with some level of confidence that their local polling place isn't screwing up.

11. Voters are asked to keep their receipts until after the election is certified. In the event of a dispute or extremely close race the voters can be asked to voluntarily bring in receipts to verify the results.

Now 10,000 people can freely poke holes in this, as I know there are some.

Re:Let me get this straight...

[Anonymous Coward]

We (in Finland) have had a good system for decades - there's simply no need to change it. Volunteers from all political parties take part in the counting process and scrutinize each other very hard. Voting speed isn't significantly slower and speed of counting doesn't IMNSHO matter at all - we've always gotten the results by midnight on election day. There's absolutely no need to get them faster. You talk about advantages but there just doesn't seem to be any.

Re:Let me get this straight...

[Anonymous Coward]

For #6, I meant to introduce the vote serial number here, called it the anonymous number. Oops.

Before anyone tells me that a paper receipt makes it too easy to sell a vote, consider these things:

1. In some states in the U.S. there are mail-in ballots. You don't hear about rampant vote purchasing even though it'd be pretty easy to show the buyer your ballot.

2. Rampant vote purchasing would not go unnoticed.

3. Why is it a problem to sell a vote anyway? I should be able to liquidate any asset I have. Besides, it's not like a huge organization would want to buy individual votes anyway when they can just buy the politicians themselves.

We can fix this.

[Richard Kirk]

I remember reading about 'foolproof' paperless voting machines in the 1960's. In fifty years, nothing seems to have changed except for the technology. You don't have a full record of the votes. People vote for a day, and at the end of the day the total does not tally, but you don't know what went wrong.

If at the end of the day, the machines logged who voted, which way, and when, then everyone would be able to check that their vote was logged correctly. However, this might allow others to know or guess the way you voted, so the ballot would not be secret.

Suppose your voting paper had a unique random barcode generated at the time your ballot paper was printed. The machine takes the candidate number of your vote and adds it to the total. It also adds your candidate number to your barcode number, and puts that in a public database. The public database would contain a set of apparently random numbers. However, if you keep your ballot paper with the number, you or someone at the voting booth ought to be able to find the number that corresponded to your vote, and check that the machine correctly tallied it.

This is a crude proposal. There are probably much better ones out there. I bet ATM software doesn't put up with a 2% error rate.

Re:We can fix this.

[bentcd]

Suppose your voting paper had a unique random barcode generated at the time your ballot paper was printed. The machine takes the candidate number of your vote and adds it to the total. It also adds your candidate number to your barcode number, and puts that in a public database. The public database would contain a set of apparently random numbers. However, if you keep your ballot paper with the number, you or someone at the voting booth ought to be able to find the number that corresponded to your vote, and check that the machine correctly tallied it.

This would make it possible to purchase votes and have a receipt to check that the service was delivered. Or bully people for votes or whatever. It is generally considered an undesirable feature.

This is a crude proposal. There are probably much better ones out there. I bet ATM software doesn't put up with a 2% error rate.

An ATM doesn't have the onerous requirements that a voting process does: it doesn't need to not give receipts and it doesn't need to be ignorant about who is using it. This makes auditing very much easier. It is also not a disaster if someone tries to use it, fails to and walks away thinking they have received cash when they have not. Moreover, it is legal to get direct assistance in using an ATM if for whatever reason you can't figure out how.

Re:We can fix this.

[nedlohs]

For god sake, every single time.

Secret voting means you can't connect a vote to a voter. In your case someone who gets your receipt, say your boss or union rep who demands everyone hand them over after the election, can determine who you voted for.

Re:We can fix this.

[Aerion]

This is a crude proposal. There are probably much better ones out there.

Yes, quite a few. They tend to run along the same lines, but with different approaches for ensuring, with high probability, that your ballot ID can't be tied to you.

See ThreeBallot (and variants), for example. Others include Scantegrity II, which has gotten a fair amount of attention recently.

Re:We can fix this.

[Richard Kirk]

This would make it possible to purchase votes and have a receipt to check that the service was delivered. Or bully people for votes or whatever. It is generally considered an undesirable feature.

I disagree. The simple scheme is just to cover the 'uh-oh - I don't think it took my vote' case. The piece of paper with their number should not leave the polling station.

An ATM doesn't have the onerous requirements that a voting process does: it doesn't need to not give receipts and it doesn't need to be ignorant about who is using it. This makes auditing very much easier. It is also not a disaster if someone tries to use it, fails to and walks away thinking they have received cash when they have not. Moreover, it is legal to get direct assistance in using an ATM if for whatever reason you can't figure out how.

Do voters need a permanent receipt for their vote? Surely, this allows for the very vote selling and bullying you were arguing against? Quite apart from that, I don't think it helps. Once the voter is satisfied their vote has gone into some distributed robust database, it should be impossible to lose except by some sort of widespread fraud. It should be impossible to lose it by resetting the local voting machine or pulling the power on the polling station.

Re:We can fix this.

[bentcd]

I disagree. The simple scheme is just to cover the 'uh-oh - I don't think it took my vote' case. The piece of paper with their number should not leave the polling station.

This requires you to keep the voters separate not only in the voting booth but also when moving from the booth to the functionary to cast the ballot. Which is just a matter of logistics I suppose, if a bit inconvenient to arrange. Also, if I understand the proposal correctly, it is undesirable that the functionary be allowed to learn what vote you cast as it, again, opens up for verifiable purchase of votes. This can presumably be kept under some sort of control though if a record is kept of which functionaries double-checked how many votes and perhaps even what those votes were (not who cast them though).

Re:At least *my* vote was counted

[Anonymous Coward]

> From what I've heard, the electronic voting system
> was fairly complicated, and my guess is that
> I could have fallen victim of it.

Insert ID card.
Type in candidate's number.
Press OK.
The screen shows the candidate's name and OK/Cancel buttons.
Press OK. (<- IQ test! Potential for epic fail!)
Take out ID card.

Similar to but easier than taking money out of an ATM.

You can also ask for assistance if you cannot manage the above process, e.g. due to a disability.

Finnish Government

[Anonymous Coward]

Finland's government has some interesting ideas of democracy. There is the case, for example, between Helsinki and Sipoo and the annexation of land to Helsinki. Sipoo held a referendum according to Finnish law - 98% voted against the forced joining with Helsinki. The Finnish government ministers including prime minister and the minister for justice declared the vote irrelevant. There are lots of other "strange" incidents relating to this particular incident too - Wikileaks might be the place to go.

Anyway, the Finnish justice system ignoring the constitution and "reinterpreting" laws to serve themselves is quite normal here. Sad thing is that NOONE seems to actually care.

Note: Finland is the ONLY democratic government with no separation of powers between the legislature and justice dept.

If you dig deep then there's a huge amount of corruption in Finnish politics, most of it centered around Helsinki City and its incestuous relationship with the current ruling party (Keskusta).

Last year, after ignoring the Sipoo referendum result and forcing certain interpretations of the law, minister Brax (Justice) asked the question why people were disillusioned with Finnish politics and democracy. Ministers then went on to vote the keep their expenses private. There were other "corruption" incidents too but these have also been kept private.

2% error in voting...acceptible in Finland

Re:Finnish Government

[SLi]

Well, you can't also hold an election in your household about paying taxes and expect the government to honor it when you say that 0 % here wants to pay taxes.

The problem with Sipoo was that they were total asses, not accepting that people who work in the neighboring big city may actually have the right to housing there. No, if they had it their way, they wanted to be a plot of wheat fields inside a metropolis or something. It doesn't take a genius to see that won't work.

Really, the municipal borders are human-drawn. There's no special reason to respect them. They are there for convenience of everybody. Perhaps in the perfect system you would have more say to things that happen closer to where you live. In that case I'm sure the result would have been the same. But if you start to act in ways that make the life for everybody around you inconvenient, you really have no right to expect them to treat you kindly.

Who Cares?

[Anonymous Coward]

I'd be less concerned about the lost 2% and more concerned that the other 98% weren't just "made up" by the voting machine vendor.

Did each voter get a human-readable receipt that was then deposited in a secure box for later verification? A subset of the paper votes needs to be pulled and counted by hand and if the sample result does not equal the population result within an acceptable margin of error (t-scores, etc.) then count the whole thing by hand. In fact, multiple samples should be taken from each precinct and they should be matching within the margin of error.

Company that made this software

[Anonymous Coward]

The company behind this electronic voting system is called Tieto (ex TietoEnator). It keeps screwing up more and more, and has even been on TheDailyWTF-article ( Tieto was behind this Finnish bank's software transition: http://thedailywtf.com/Articles/Sampo-UhOh.aspx )

I can read news every month about their software being delayed, costing too much or even not working at all. Still they keep getting every software development job government needs to be done, I wonder why..

Re:Company that made this software

[Anonymous Coward]

It's probably do with them knowing how to compete for goverment jobs. You promise software too cheaply and then overcharge for the eventual changes. And there is allways changes in software project. Some building companies do that with building sites to get the job. Other companies just doesn't know how to play a ball or they are more honest on their offers. Bottom line is that in capitalist way of thinking cheapest is allways best.

Re:Company that made this software

[SLi]

( Tieto was behind this Finnish bank's software transition: http://thedailywtf.com/Articles/Sampo-UhOh.aspx )

No it wasn't. TietoEnator was responsible for the old, working system. The new, broken system was Danske Bank's own.

Only 232/157/N votes rejected by e-evoting

[Anonymous Coward]

Electronic voting was piloted in three (3) small municipalities totaling:

- 34 062 total eligible voters
- 20 912 votes casted
- 157 votes rejected

Not really a national catastrophe, only few test sites.

All of Finland had:

- 4 191 949 total eligible voters
- 2 554 319 votes casted
- 16 281 votes rejected

Rest of the country collected the old-skool paper votes. That way it was still possible to vote D. Duck, or the multi rower upside-down rowing boat, which is also marginally customary.

Re:Only 232/157/N votes rejected by e-evoting

[Anne Thwacks]

That way it was still possible to vote D. Duck, or the multi rower upside-down rowing boat,

Either of which would have run the banks better than the present regime.

Re:Only 232/157/N votes rejected by e-evoting

[jaria]

Not really a national catastrophe, only few test sites.

Of course. But this case is important, because if this test had seen as a "success", the next time the system would be used for real in the whole country. I suspect the government might have actually done so, even with this mess, if there had not been a public outcry. Luckily the issues have been publicized widely, and I think we won't see a similar system in the near future. The minister of justice has talked about re-evaluating whether the whole thing makes sense, and if they go ahead, about the need for open source implementation and paper trail. This is a good thing.

Two Percent Lost? OMG

[kenh]

From the Slashdot article:

Does this mean 98% of votes is enough to figure out how the other 2% voted?

I can think of at least two folks in Minnesota that would have a problem with that many votes lost...

http://latimesblogs.latimes.com/washington/2008/11/franken-coleman.html

Re:Two Percent Lost? OMG

[Locke2005]

It's not getting it to lose 2% of the votes that is important... the trick is getting it to lose the right 2% of the votes!

Yes, if all lost votes being cast for one candidate or the other would make no difference in any of the outcomes, then the results should be certified. If there was any possibility that it would make a difference (as was clearly the case in the Minnesota senate election) then they should hold another special election.

Medical attention for modern /. ?

[Anonymous Coward]

From what I've heard, the electronic voting system was fairly complicated, and my guess is that I could have fallen victim of it.

Yes Virginia, I am also confident that none of the modern /. readers would be able to figure out what to do with the big green OK button.

The people who's electronic votes did not get counted, did not figure out the green OK button.

It's a bug,

[martas]

NOT a feature!

Moot Anyway

[qreeves]

a) It only matters if the voting margin was within that 2% otherwise it's pretty easy to determine the winner.
b) There is probably a higher rate of error getting humans to transcribe votes, with less of a chance of being picked up.

Think about it.

Re:Moot Anyway

[Anonymous Coward]

The error margin of traditional voting has been carefully studied in Finland, and found to be 0.5%.

Would I be correct...

[ibm1130]

Would I be correct to assume that the usual collection of smelly Socialists won the overall election? If so I can hardly wait til it is their ox that gets gored.

Re:Would I be correct...

[jaria]

Not really, it depends on the city in question. In my city they are not the winners, can't remember if they have even a single seat. In the other two cities they are more significant. As a rule of thumb, Finland's political power is divided between the three roughly equal size parties: the left (Social Democrats), center (Keskusta), and right wing (Kokoomus) parties.

Re:Well my God! A Communist country VOTING?

[Anonymous Coward]

Are you American by chance? Did you look at this before posting? http://www.ilmatar.net/~np/hate/america.gif

Oh my...

[spazdor]

OMG...
Did you just...

If it is a first past the post system this is a problem.

OH NO YOU DI'INT!

Re:Oh my...

[dov_0]

Hehe. Unintentional I'm sure...

Statistically valid.

[Dputiger]

" Does this mean 98% of votes is enough to figure out how the other 2% voted?" A properly done statistical analysis points the opposite direction; a small sample, properly randomized, can figure out how the other 98 percent voted. And if the margin is, say, 70/30, I don't think we care how they voted anyway. Now, in a 51/49 split, sure. Those don't happen very often.

Still better automated

[Anonymous Coward]

It's better to lose a random 2%, than the traditional 2% that the counters 'miscount' because they oppose their political view.

Summary is wrong

[Phasma Felis]

According to the Electronic Frontier Finland article, the troubled system was used in areas representing 2% of the electoral roll, but actually affected only 232 voters total. Finland has a population of 5.5 million. Depending on how many of those voted, less than 0.01% of all votes were actually lost; a much better margin of error than most paper ballot systems.

Only the Turre Legal blog mentions the "2% of all votes" figure, and its source is...another Slashdot summary, which is also wrong.

It's still a problem, especially at the municipal level, but it's a much smaller problem than it's made out to be.

Puuh

[Anonymous Coward]

Technically EFFI is right as they keep noise about this matter and about flaws in current generation of finnish e-votes. This good thing and should be done. And in matter of fact finnish justice ministry has put electronic voting on hold and if it is going to be tested again in some elections (bear in mind number of municipals in the e-voting test was quite small) then for example open source systems will be considered.

In all, good thing. Maybe in future systems will become better to use. Hell we do have ATM's and they are used, how hard it can be to do simple voting machine ;). (With publicly audited code if needed!)

However what comes to this lets redo do the elections because of this, well ... **ck it. In Finland local municipal elections are more or less joke where old and proven candidates usually win because of younger generation does not really feel like voting. It is not problem of e-voting but in system itself.

Has this been done in parliament elections then situation would be more problematic in some sense but same rules still apply.

We Finns have saying: "Onko kiire vai tehdÃÃnkà tietokoneella?" (roughly in english that "Are we in hurry with this or shall use the computer with this?"). This is extremely applicable in this situation :)