The Cost of Electronic Voting 158
Wired's Threat Level blog is reporting on an analysis of the cost of electronic voting compared to traditional methods of vote tallying. A group named SaveOurVotes examined Maryland's budget allocations for elections during their switch from optical scanners to touch screens, and found that contrary to official claims, the cost was higher for e-voting (PDF) — much higher.
"Prior to purchasing the touch-screen machines, about 19 of Maryland's 24 voting districts used optical-scan machines. SaveOurVotes examined those counties and compared the cost of the optical-scan equipment they previously used to the touch-screen machines they were forced to buy. The cost for most counties in this category increased 179 percent per voter on average. In at least one county, the cost increased 866 percent per voter — from a total cost of about $22,000 in 2001 to $266,000 in 2007."
Re:Last nail. (Score:2, Insightful)
Why do the U.S. needs machines to count? (Score:3, Insightful)
Cost shouldn't be the biggest issue (Score:5, Insightful)
So it should be able to afford a good voting system. Nothing like the diebold crap.
Manual vote counting and counter-checking can be easily parallelizable. The more voters you have, the more vote counters and observers you should be able to recruit.
It is MUCH harder to tamper with paper ballots. You might be able to do a few areas, but to do it all while the other parties have people watching is hard.
With most electronic voting systems, 3rd parties can't watch the "counting" easily. If you have an e-voting system where 3rd parties can watch easily and it's verifiable, it'll probably cost more in the end.
So what if you have to wait a few hours before you get the results?
Lastly, Elections don't just have to be fair, they have to be _SEEN_ to be fair (enough
It should be cheaper and more secure. (Score:5, Insightful)
The real shame of this is that electronic voting should be cheaper and more secure but Dibold's flawed equipment and business model has given a bad name to the whole concept. While it's true that electronic voting requires more equipment, this equipment should be cheaper. Ten $200 terminals should cost less to purchase and maintain than one specialty machine. Yes, $200 is a reasonable price if free software was used and a free software for voting can easily be written if it's not already available. Instead, Dibold passed on the "commodity" software model, complete with the upgrade treadmill, insecurity and lack of transparency.
Pen and paper still the best (Score:5, Insightful)
No need to worry about educating people on how to use the machine (either for voting or setup), and the paper trail is built in.
Of course, you can still mess with things if the layout of the ballot is inherently flawed (butterfly ballots in 2000, anyone, although with a pen chads aren't a problem), but at least the mechanism itself shouldn't be in question.
Re:It isn't e-voting, it's how (Score:2, Insightful)
Not just diebold (Score:5, Insightful)
Re:Why do the U.S. needs machines to count? (Score:3, Insightful)
Re:Not just diebold (Score:3, Insightful)
I just love statistics (Score:5, Insightful)
I glanced at the article and didn't see any useful data, so I paged through the pdf. There's some stuff in there that I don't understand and could cause some major problems with their statistics.
1) They appear to be comparing projected costs of optical scanners with actual costs of touchscreen machines. The PDF shows a 7 year lifespan of the original optical machine purchase, amortized over the first five years with zero additional purchases for that 7 year period, only warranty repairs. I sincerely doubt that there were zero additional purchases.
2) Can't they hire the same project managers for the touchscreen rollout as for the optical? People management is people management, no real difference.
3) Warehousing costs - aren't they storing the equipment at a state run facility? No reason why there should be a huge capital payment associated with that.
4) Transportaion costs fluctuate wildly on the touchscreen actual costs page, but are unwaveringly cheap on the optical page. The same equipment would always have to be moved to the same place, so I don't see that assumption as valid.
5) Voter outreach is 2x more for touchscreen as it is for the optical assumptions. I don't see how that cost would be different.
6) I don't see a line item for absentee ballot printing on the optical page at all.
7) I call BS on the statement that 10 touchscreens are needed for the job of a single optical scanner. Why would a county be willing to have a single optical scanner during an election? What if it failed? Those people wouldn't be able to vote that day? I think 2-3 is a more legitimate answer to account for quick processing and/or machine failures.
8) What exactly are the optional services that Diebold provides that account for almost $28M. That's a third of the overall total cost. There's no breakdown of what the services are, so there's no way to compare them with line items on the optical scanner costs.
They're comparing apples to oranges here with the projected costs of optical. It's simply not a fair comparison. And then not listing what those services are that almost singlehandedly account for the entire difference in cost between optical and touchscreen is ludicrous. If you take that line item out since there is no equivalent line item on the optical sheet, you have $67.5M for touchscreen and $52.4M for optical. Even using the listed number of $95M for touchscreen, that's still a little less than 2x the cost of optical. How exactly did they arrive at a 10 fold increase statistic?
I'm sure that the touchscreens are more expensive than opticals at first. Same thing when companies were first rolling out desktop computers to their workforce a couple decades ago. They understand that it cost a lot of money and a lot of lost productivity, but they also knew that they would reap huge rewards in additional productivity in the long run.
Now that said - let's find some other electronic voting firm to spend our next $100M with instead of Diebold.
Re:Bad hardware. (Score:4, Insightful)
It's new; of course it's more expensive! (Score:4, Insightful)
They had to buy new stuff. And even the article admits some of the money went to training. This isn't necessarily an indication that the higher costs are inherent, just that switching to something new has an initial cost. It would make more sense to see how the costs changed over, e.g., 10 year periods than just after the new technology was introduced.
Personally I think the higher cost would be justified if it led to an increase in democracy. As another poster mentioned, the US is a rich country. If there are demonstrable benefits to the new technology, I would bias in favour of it, even at increased cost.
The big problem, of course, is that the machines are not only expensive, but terrible. They seem to be a step backwards in democracy, not forwards. I live in Canada where we use pencil-and-paper ballots and they work beautifully for our purposes. I can't imagine switching to anything electronic at this point, as it would surely be a step backwards.
Re:Why do the U.S. needs machines to count? (Score:3, Insightful)
Humans are guaranteed to make mistakes, and make them regardless of whether a ballot is well-formed or not. Machines should, in theory, only ever make the same kind of mistakes (so the mistakes should be easily caught, eventually). Obviously, they're a lot faster than people are, and that time costs money. Unless all your vote-counters are volunteers, but then you'll find it very difficult to recruit people who are both A) proficient and B) don't have an agenda.
What the hell is wrong with machine counting?
Heck, with the advances in cryptography, and the ubiquitous network availability, what would be wrong with internet voting (in principle)? We ought to practice this stuff, because the internet also gives us the opportunity for much more direct democracy. The main barrier to having say, a weekly referendum is information availability and communication delay, which the Internet soundly pummels on both counts. I mean, you still need a congress, but why not restructure things to take back some of their power when the technology is available to do so?
Re:Not just diebold (Score:5, Insightful)
Then you get the benefits of electronic input -- like access for the visually impaired, to alternate-language ballots, the ability to correct mistakes, etc. -- without relying on the input device to do all the vote-counting correcting. I expect it would provide a count for quick access to the results, but you wouldn't have to rely on it.
And because the output is computer-generated you can do things to actually improve audibility over traditional hand-written ballots. For one thing, you could print the output onto an optical-scan form, or other machine-and-human-readable, high-accuracy format. You could then buy an optical-scan counting machine from another vendor, and if at the end of the night the numbers from both machines matched up, you could all go home without hand-counting anything. You could also have the machine sign its output so that ballots can be traced back to a particular device, and can be verified as authentic and non-duplicated -- the public could be provided with copies of the ballots to independently verify the results.
Re:Not just diebold (Score:4, Insightful)
Re:It should be cheaper and more secure. (Score:3, Insightful)
Around here we've been using optical scan forms for years, and they work pretty reliably. The only thing that they can't do which the electronic ones can is spit out a receipt.
They provide a built in paper trail, as long as they don't get lost in the mail or in a back room. They can usually be scored in bulk via an auto feeder.
And the cost is significantly lower. As my state switches to an all mail voting process, the equipment is just as useful now as it was when we had to take a sharpy into a voting booth.
Re:Cost shouldn't be the biggest issue (Score:3, Insightful)
1. When you register, you get one "vote card" and a thin envelope. Make the vote card special say with a watermark so it's hard to fake extras.
2. You go into the booth, insert card and make a selection.
3. After it's asked you if you're really really sure, prints it in cleartext and as a barcode (or those better-than-barcode things, I forget).
4. Take the card out and verify your printed vote against the cleartext.
5. The vote should be left on screen until you click "ok, it matches", pretty damning evidence if you stand there with a screen that says one thing and a printed card that says something else.
6. You place it in the envelope, go out and put it in a ballot box as usual.
7. At the end of the day, you pour these into a reader.
8. The reader either removes or scans through the envelope (not perfectly sure how, but quite sure that's doable).
9a. If it can verify watermark and barcode, count and store.
9b. If it can't, return for a manual count
10. Spot checks to verify the machiens aren't printing one thing in cleartext and another barcode. Would be pretty damning evidence.
Plan B, if the macine isn't working/printing/whatever:
1. Have manual ballots ready. Triple-warn on vote card, ballot and computer that they're only to be used if the computer's not working properly.
2. Fill out manual ballot (these can be plain paper, so no big cost to print up) and put it and your vote card in the envelope
3. Since it lacks any barcode, it'll get returned for a manual count in 9b) above.
Results:
1. Your vote *exists*, it has a paper trail.
2. You can be quite sure that your selection == printed in cleartext == printed as barcode.
3. No vote "reciept" which is a bad thing.
4. Optical counts.
5. Hand recounts or recounts by 3rd party optical machine are possible.
The only possible cheat I see here is that the counting machines can swap out votes, though it'd require a very custom design with extra votes inside and it'd be easily detectable by running the same pile twice, for example. All in all I think that's the closest thing to bulletproof I can think of. Then again, I don't see what's wrong with the old way.
Re:Carroll County 22k to 260k! (Score:2, Insightful)
Re:Slavery is more expensive. (Score:4, Insightful)
I think Diebold probably made a LOT of money on it - initially. My guess is that they probably lost because they were forced to re-examine, re-implement, and re-certify the crap that they tried to pass off as secure voting machines. Now that the cat's out of the bag, it's understandable that Diebold would want to distance itself as much as possible.
Re:Not just diebold (Score:2, Insightful)
Really? I agree that it's more expensive and probably more prone to failure, but I would argue that a system should provide better, faster, AND more verifiable results.
Better: If you have a "voter marks a ballot, machine counts ballot" system, that will have recognition errors. These can be upwards of 99%, but there are important elections where the margin is smaller than that. A computer voting system should have NO error. The computer won't occasionally add 257 + 1 and get 258. (Bizarre quantum effects and energetic particles hitting the RAM notwithstanding; and you could always have it do every calculation twice if you really want to worry about those.) There are still other sources of inaccuracy and fraud in election, but why not remove one part?
Faster: It should be virtually instant. Even assuming that the machines aren't connected to an outside network (which is how it should be), precincts should be able to report almost instant vote totals. For instance, at election close, someone at each precinct calls the statewide election office and reports the total for each machine (perhaps in encrypted form). Mutual authentication ensures that the person calling is the designated representative. I can imagine several other schemes where perfectly accurate (assuming subsequent audits are clean) statewide results can be available within 5 or 10 minutes of the close of elections. None of this waiting several hours for Cleveland to count their ballots to even get the first number.
Verifiable: A paper trail provides essentially as much verification as any other system. Because it would be printed by the computer, quality control could ensure that the paper ballots are clear in their intention and all valid. It would be impossible to create a paper ballot that had two votes for the same office, and squabbles about voter intention should all but disappear.
I think a much better argument would be that the "better" result is a tiny part of voter inequity and isn't worth the extra money, and faster isn't really a worthwhile goal.
Re:A real open source solution? (Score:2, Insightful)
Everything else is just insecure: Even if electronic voting machines use open source software, how do you know the code you check earlier is the same that the computer use during the election ?
Jeez... We use this SIMPLE and EASY paper voting system for years, why the hell do we have to search for a more COMPLEXE alternative ?
Why look at the cost (Score:2, Insightful)
The e-voting machines have resulted in a second term for the world's most visible terrorist, and they've wasted countless man-millenia as everyone discussed, debated or idly witnessed the chaos surrounding voting fraud. Hell I don't even LIVE in the U.S. and I watched a "documentary" about how easy it is to screw with the Diebold counting machines. That's 90 minutes of my life I won't get back, all because of one messed up government and its conveniently incompetent equipment contractor.
If you really want to tally the cost of something, you have to look at _everything_. The up-front dollar amount is nothing compared to the thousands of people that had to deal with these broken machines and learn how to use them, along with the millions who had to waste yet more brain cells on this dead-end gadget. How about the increased difficulty to implement a working e-voting solution due to voter reluctance ? That's a tough one!
cp reality speculation
vi speculation
diff reality speculation
Yeap, not easy to estimate the net impact of any change on your whole concept of reality. The e-voting fiasco's true cost cannot be quantified, though in the grand scheme of things it's a small line-item.