Voting Machines Banned by Dutch Minister

5heep writes "Dutch Government Renewal Minister Atzo Nicolai has banned the use of one type of computer voting machine in national elections next month. The turnabout came after a group called We Don't Trust Voting Computers protested the vulnerability of electronic voting to fraud or manipulation. The reason for this ban is the radio signals emitted by the machines which can be used to peek at a voters' choice from several dozen meters away."

Comments

Hey Slashdot

[isaacklinger]

How about writing to the people responsible to show some support?

Better late than never, I guess...

[truedfx]

According to my local newspaper, these voting machines have been used in the last two elections.

Re:Better late than never, I guess...

[FST777]

More information:

There have been roughly two types of voting machines in use in the Netherlands: one produced by Nedap and one produced by Sdu. The latter is the one that is banned a few days ago, because they could be spyed on from a distance. Curiously enough, the platform "we don't trust voting machines" proved that voting machines can be spyed on a few weeks ago, but they proved that with the Nedap machine.

The platform never had a chance to test this problem with the Sdu, since they only had access to the Nedap machine (unofficially, they never had permission or anything but just received one machine from a muncipality when asked for it). The ministry of Government Renewal has not yet made any public comment on the problems with the Nedap machines.

The muncipalities that used the Sdu machines are now forced to arrange Nedap equipment (which will certainly not be possible on such a short term, the elections are on 22 november) or use the old paper-and-pencil method (which will need massive restructuring of organization to arrange that on such a short term). Amsterdam is one of the muncipalities that already declared to go with the latter option, and it is already clear that there are virtually no Nedap machines to spare.

This could become interesting. I predict chaos, but not more than chaos. The counting will take considerably longer (since I think there is not enough manpower to count on short term) but I still believe the results will be correct.

Give me something I can Count!

[Zarniwoop_Editor]

Paper ballots... you can count them... You can check them, you can verify them.
Computer Ballots don't leave the average Joe with any sense that they can be verified.
Too much potential for problems with Electronic voting from a voter perception perspective.
I like putting my little X on the ballot.

Re:Give me something I can Count!

[MeltUp]

Why can't they make these things simple and thrushtworthy for everyone?
It ain't so hard:

- Enter voting office
- Receive a "voting paper"
- Enter voting booth
- Insert voting paper into machine
- Push the button for the candidate you want. (Machine's critical components are covered in faraday cage, to stop any readable transmissions)
- Vote is printed on paper
- Check the print and fold the paper along the prefolded line, so text is no longer visible
- Publicaly put it in the urn (where they just fall in a disorderly stack, so order can't be traced)

Counting (when the election is over, earlier is cheating off course)
- Manually get all papers, and feed them into a counting machine
- Machine sorts and counts all votes. returns each type of vote paper in easily countable stacks. Result is immediatly made public.
- Count a large part of the votes by hand.

Any problem with this type of voting? It offers all benefits of electronic voting, and none of the drawbacks (or so I'd think).

Re:Give me something I can Count!

[quigonn]

The only justification (at least in the Netherlands and Germany) for voting computers is cost reduction - adding a voter-verifiable paper trail would completely totally destroy this "advantage" (which is very questionable, anyway).

But, in fact, there is no reason to reduce cost in this process. Cost shouldn't matter here, since secret, equal, free elections are a crucial process within democratic systems. Besides that, the pen and paper method is the most simple method you have, everybody understands it. In fact it's so simple, everyone can audit the whole process. Contrary to that, audits of computer-based systems can only be done by a few experts (and a complete audit goes from a security audit of the software down to as far as checking the hardware for possible modifications).

Re:Give me something I can Count!

[quigonn]

How can this save costs? Do you know how expensive those machines are? Compared to simple pen and paper, they amortize after about 20 years of operation. And that's the _maximum_ operation period (for XP-based machines like the SDU voting computers probably even shorter). Faster counting? How relevant is that whether the election officials can go home one or two hours earlier? You shouldn't sacrifice something as crucial as _voting_ to getting home earlier. And regarding your "changing your vote" argument: how is that supposed to work? That would only work if the vote was associated with some kind of unique user ID, and that would be totally against the provision of secrecy. So I see no advantages for voting computers, but a lot of bugs in so far _all_ existing implementations, versus a well-known system that just works.

What's the big deal adding a paper trail?

[Panaqqa]

I know more than a bit about electronics and electrical engineering, having spent part of my early career designing various gadgets. I have to wonder why all this fiasco with these electronic voting machines when it would be so very easy to simply build a small printer which uses a roll of paper (think cash register) inside to create a paper polling record.

It would not add substantially to the cost, and the small rolls of paper that resulted would be perfect in cases where a recount was demanded or required. Why the resistance? Make it too difficult to steal an election?

Re:What's the big deal adding a paper trail?

[ben there...]

If the paper were inside, it would be just as useless. All you'd need to do is hack the system to display a different vote than it prints.

Any electronic voting machine should print a ballot that you stuff in a box. Electronic tabulation of votes could be used for preliminary results, but the printed ballot that the voter can read and verify should be the final word.

Re:In a related story ...

[Phantom of the Opera]

The FCC has banned computerized trading

And the FDIC has banned computerized banking

And the FAA has banned computerized flight control

What next? Seriously, why is voting any different from these other very important uses of computers? Doesn't it make more sense to fix the problem rather than ban the machines?

Think about it. Your examples revolve around money or human life. The manufacturers of those machines __must__ get them right, or there is immediate finantial fallout.

Those examples do not produce hidden results; if your atm gobbles up your deposit without crediting it, you find out. If a plane crashes, you hear about it.

The average voter has no way of knowing if a voting machine is doing it's job. What is the penalty if the manufacturer 'unwittingly' messes up? There is not as much incentive for accuracy in this case.

There is no "fix" for voting machines...

[thrill12]

Voting machines can be constructed in any way possible, but never completely exclude the fact that you can commit fraud with them.
 
One solution often presented is the XBOX-type of security - encrypted links between hardware, redundancy etc etc - but as *we* know this type of security is breakable. You only have to do this once to break the security of all voting machines.

Apart from this, some people mention the use of a paper trail. This trail itself has to be counted fully then, irrespectful of the outcome that the machines themselves produce, to verify a correct vote has been cast:

The voting machine in itself can still not accurately or thrustworthingly tell the outcome of the elections and becomes a nice "exit-poll".

Voting also brings with it the right for secrecy: this is something that does NOT occur in your examples. While the data is compartimentalized to certain groups of people, the data is still available on multiple sites and can be cross-verified. Voting machines store the data on 1 place (with or without redundancy) and when the vote has been changed, you can no longer cross-verify whether the voter actually did vote what he appeared to have voted...

WHY?

[Virgil Tibbs]

exactly what is it that the dutch and the americans find so hard about putting an x on a piece of paper is beyond me next they will be telling us that only 99.9% of them are illiterate!

Re:WHY?

[aadvancedGIR]

Years ago, I heard about a big election in an african country in which a large proportion of the population was illiterate. The solution was easy, they simply ask every party to choose a distinctive sign. On the ballot, they put those signs in front of the names of the candidates (plus their photos) and it worked rather fine.
Conclusion: illiteracy is absolutely not an excuse for not having correct voting procedures.

Re:In a related story ...

[esme]

Seriously, why is voting any different from these other very important uses of computers?

Voting is different for a number of reasons:

• Voting is done in secret, with the only way of knowing the results being the voting machines. And the makers of electronic voting machines are against the only decent way of double-checking them (voter-verifiable paper trail).
• Voting is only done a few times a year, rather than continuously, year-round.
• Voting is administered by the people who have the largest incentive and opportunity to cheat.
• And most importantly, unlike the other examples you list (banking, trading, flight control), electronic voting machines have not been shown to be more reliable and accurate than humans.

This last point is a little fuzzy, because I'm sure electronic voting machines are better than poorly-designed punch-card ballots, and maybe some other flawed mechanisms. But the best system available right now is optical-scan paper ballots that can easily be hand-audited and hand-recounted. They are easy to use, require only a very circumscribed use of technology, and can easily be verified by people if there are any problems or a very close result.

Doesn't it make more sense to fix the problem rather than ban the machines?

Sure -- I don't think anyone is saying we should never use computers for voting. Fix the problems, and then use them for voting. Advocates of electronic voting seem to be saying we should do it the other way around, which is insane.

The current round of voting machines are insultingly under-engineered, considering the problems I listed above. There are many types of threats to the integrity of voting machines, and Diebold et. al. aren't interested in addressing them. They're more interested in shutting down debate and research about them, in fact, which is very worrying to me.

-Esme

playing chess on a voting-machine

[SiggyRadiation]

In the PDF document they tell that they were challenged by the builder of the original voting-machine to turn it into a chess-computer.

Which they did. :)

Re:And once again

[quigonn]

Actually, banning _is_ the solution. Voting computers make voting less verifiable, less auditable, more expensive (although the Voting machine producers claim otherwise), so why use them? What reason justifies switching from a proven, working, easy-to-use, easy-to-audit system (pen and paper) to a new technology of questionable quality?