Skip

Voting Scandal
October 10, 2003 9:45 AM   Subscribe

More pesky voting machine problems
posted by sourbrew (36 comments total)

 
I don't think this would have really changed the outcome, just a frightening thought in general.
posted by sourbrew at 9:52 AM on October 10, 2003


You know, Canada has never had any issues like this as far as I know. Do you know how our ballots work?

We get a piece of paper with the candidates' names and using a handy dandy pencil we put an X next to the person we want. Done!

No machinery. No punchcards. No dimpled chads. No "Florida" situations.
posted by Robot Johnny at 9:52 AM on October 10, 2003


You know, Canada has never had any issues like this as far as I know. Do you know how our ballots work?


Actually there were significant problems with the last referendum in Quebec where not-totally-independent counters would reject ballots for reasons such as "the x didn't cover the whole circle". Most elections aren't nearly as close as the last referendum was, so it was never an issue before. But also things like people making a check mark instead of an x (normally to be rejected) can cause a lot of confusion.
posted by Space Coyote at 10:03 AM on October 10, 2003


Actually Quebec had even deeper rooted problems that that for the referendum. I've heard first hand anecdotes of English voters being put through the ringer trying to get registered to vote.

And let's not even talk about what was said after the referendum. Bigots...
posted by jon_kill at 10:12 AM on October 10, 2003


Quoting from the article
As we already know, it is not possible to audit touchscreen machines because Diebold refuse to allow printing of a ballot to be placed in a box as a back up for use in just such an apparent tampering with votes

Oh really ? How absurd is that ? Hell even your mom and pop shop with a computer (yeah their son is good at computers, that geek ! ) know they must print out their accounting not only for legal reasons, but because the computer may have troubles or break. The cost of printing something of paper is minimal , the cost of losing data is realitively enormous (when compared to the cost of printing it even on toilet paper if you so wish).
posted by elpapacito at 10:26 AM on October 10, 2003


How can Diebold refuse to allow the printing of a ballot???

Who the fuck is running the show???
posted by Blue Stone at 10:40 AM on October 10, 2003


If you're interested in effective electronic voting, you should check out Rebecca Mercuri's work. She's written some wonderful papers that are worth putting in front of your local voting commission's nose, if you have the ability to do soo.
posted by mosch at 10:46 AM on October 10, 2003


I realise there are some problems with the machines but certainly a lot less than with punchcards where studies have shown as much as 9% of the people managed to not even vote on wether to recall the guvner or not. It should be noted that even had these folks voted no the outcome would have been the same. Article here

Tinfoil hats aside I really doubt there is an if votes for schmo#4 way less than schmo#1 move votes to schmo#4 statement in Diebolds software. The move to electronic voting machines is a good thing, and they'll only get better. Luddites the lot of you.
posted by zeoslap at 10:57 AM on October 10, 2003


This seems dumb. Why would Diebold have any interest in making sure losers marginally smaller losers? It certainly doesn't appear that the total number of ``suspect'' votes is enough to swing anything.

It seems more likely that one of four things is happening:

One, use of a touch-screen or of optical-scan forms makes it easier to vote for fringe candidates somehow, at least relative to other voting forms (ie, it's easier to vote for Gary Coleman on a touch-screen machine, or on one of those big optical-scan ballots, than it is using smaller ballots or punch-cards).

Two, for one reason or another both are caused by the same exogenous factor -- whatever factors lead to a county buying touch-screen or optical-scan voting machines are correlated with a propensity to vote for oddballs.

Three, candidate order effects. Do the counties with unexpectedly high counts for fringe candidates happen to contain the districts where these candidates would appear relatively high on the ballot?

Four, a simple statistical fluke. With 135 or so candidates, you're simply going to see some positive number of unlikely events, even by random chance. For events that are 95% unlikely, we'd expect to see about 7.

One would expect some of
the 'fringe' candidates to do well in their home
county and then to have a very even distribution
across the rest of the state.


I don't think that's remotely the correct null hypothesis.
posted by ROU_Xenophobe at 11:04 AM on October 10, 2003


zeoslap - I don't think the post was about machine vs. paper, but about a company that defies reasonable assurances of not tampering with their system.
posted by stevis at 11:09 AM on October 10, 2003


ROU_Xenophobe - All four of your points are dead on and probably account for some of the improbabilities, but I think you missed the slipperiness of vote shifting.

If you wanted Swarzenneger to win and Bustamente to lose, then you could siphon a few percentage of votes from Bustamente to any two dozen of the oddball candidates without raising alarms. It couldn't be too much, but in a close race it could throw the election another way.
posted by stevis at 11:16 AM on October 10, 2003


Well, ROU_, how about a number five: Some hax0rs decided to see if they could really screw with the system. By doing it in an unobtrusive way, that no one would notice (and if they did, no court will ever demand for the inconsistencey to be examined0, any frailties of the system go unnoticed. That way, Deibold it totally unawares if the system is compromised in a big way (say, next November).

Pretty smart if that is what is going down.
posted by jmgorman at 11:20 AM on October 10, 2003


Tinfoil hats aside I really doubt there is an if votes for schmo#4 way less than schmo#1 move votes to schmo#4 statement in Diebolds software. The move to electronic voting machines is a good thing, and they'll only get better. Luddites the lot of you.

Does "tinfoil hat" apply when the concerns are valid? Even if everyone at Biebold is a total freedom-loving Care Bear, it doesn't change the fact that their system for tallying votes is insecure. The problem isn't just the inaccuracy mentioned in this link, it's that votes can be wholly scrubbed, added, or changed after the fact with no paper trail to verify anything.
posted by Ignatius J. Reilly at 11:38 AM on October 10, 2003


zeoslap:
Quoting your posting:
It should be noted that even had these folks voted no the outcome would have been the same

This line of noting or "reasoning" is superficial at best, careless at worst. Even if in THIS occasion the results would have been the same, that doesn't mean that on the next occasion (voting) the difference will not big big enough to actually have an impact on the outcome of the election.

Conspiracy theorist will probably cry on the top of their lungs that Diebold is evil, republican, democrat...let them cry. But given that

1) apparently the Diebold source code wasn't disclosed
2) it is still possible to entirely replace the source code of any machine before the election days, it probably takes a few seconds

any mechanical or electromechanical device or computer can't be trusted as the ONLY source of votes.

A paper ballot may be printed or punched, nobody cares but the voter should always be able to have a paper PROOF of what he voted so that his paper vote is collected, his electronic vote is collected and he also has a recepit of what the machine recorded.

This way you have results VERY quickly (electronic vote) and a double paper trail. One may skip the recepit, but there must be something on paper the vote and the officials can check.

On a tangent: there is really no need to have results in realtime. There is no "void of power" as the president, governor, you-name-it who is still in charge remains in charge until all the votes are casted and scrutinized.

From a cost-saving point of view, I believe that the savings that MAY come from an electronic voting system are BY FAR outweighted by the fact that the person that is -ruling- your life is NOT the person you elected. Imagine what if the guy is the next Hitler.
posted by elpapacito at 12:05 PM on October 10, 2003


Well, that's it. Everybody pack it up, I'm declaring this thread officially Godwined.
posted by monju_bosatsu at 12:13 PM on October 10, 2003


Troutfishing did an excellent post on this subject several weeks ago. There is ample reason to be suspicious of this company without donning a tinfoil hat:

"Diebold, a company which has knowingly - for 10 years - sold security-compromised voting technology, and whose CEO, an aggressive Republican fundraiser, has said "he is "committed to helping Ohio deliver its electoral votes to the president next year.""

No votes without a paper trail for auditing purposes.
No databases without referential integrity.
How hard can a paper trail be?
And using an Access database? Come on! Get real! Everyone knows FoxPro is far superior and could never be hacked like Access! ;-)
posted by nofundy at 12:27 PM on October 10, 2003


Oh, BTW, fuck Godwin. With Hitler's dead, cold dick.
posted by nofundy at 12:28 PM on October 10, 2003


Well, that's it. Everybody pack it up, I'm declaring this thread officially Godwined.

You don't know what Godwin means. You're only making yourself look stupid. Cut it out.
posted by majcher at 12:43 PM on October 10, 2003


nofundy -- Where did you find that they're using Access? God have mercy on us all.
posted by 4easypayments at 12:50 PM on October 10, 2003


Can everyone shut up with the fucking "this thread's been Godwin'd." If you think you're going to shut people up by referencing some rediculous interenet meme rather than engaging in rational discussion you really belong somewhere else.

There's a reason people are frightened of Hitler and there's no reason to think that someone like him would never gain power again.

Sorry for the OT. Carry on . . .
posted by velacroix at 12:54 PM on October 10, 2003


Geez. People sure are sensitive around here today.
posted by monju_bosatsu at 12:56 PM on October 10, 2003


Where did you find that they're using Access?

The Seattle Times quotes an internal e-mail from Diebold: "Right now you can open [the voting system's] .mdb file with MS-Access, and alter its contents. That includes the audit log." The voting system is -- no kidding -- just a pretty interface on top of Access. How horrifying is that?
posted by waldo at 12:57 PM on October 10, 2003


elpacito my point was that these machines are better than what is currently in place i.e punch cards. You can just as easily lose a few boxes of paper votes as you can hack into Diebolds systems and change things that way. I don't really care if it's Diebold or not, sounds like a shoddy operation from what I've read but it's still a step in the right direction

However the article linked to in the post is purely tinfoil-ville, there is no way that the results in this election were being skewed to bump up the votes of 'low order' candidates.
posted by zeoslap at 1:47 PM on October 10, 2003


Where did you find that they're using Access? God have mercy on us all.

Here's the oringial investigative piece that began this flap. How telling that it came from New Zealand media. At least someone is concerned about the fate of American democracy.

However the article linked to in the post is purely tinfoil-ville, there is no way that the results in this election were being skewed to bump up the votes of 'low order' candidates.

Frist, why is there "no way?" Isn't an assertion better countered with an actual argument? You off-hand dismissal isn't too convincing. You may not like the main link of the FPP, but in looking at Bev Johnson's work, how can you call it "tinfoil-ville?" The only thing I could figure out would be if you think she's making the whole thing up. Is she lying? If not, can you draw a better conclusion from her data?
posted by Ignatius J. Reilly at 1:52 PM on October 10, 2003


You're right the article has uncovered the shocking truth. Try as they might to cover their tracks the truth wants to be free. Diebold is deeply embroiled in the vast Coleman-wing conspicacy. Turns out that Walden O'Dell sold his soul to the devil aka Gary Coleman over two decades ago while watching a particulary funny episode of Diff'rent Strokes. Witnesses recall him exclaiming 'Ah'd sell my soul to the divil if Gazzer aint the funniest kid around'. Unluckily for Walden, Coleman wasn't the funniest kid around, that honor lay squarely on the shoulders of Ikzak Moojonkawitz the famed Polish ventriloquist. Walden and Colemans paths were not to cross again until the perfect political storm produced the California Goobernutorial Recall, it was time to pay the piper and leverage his companies shoddily cobbled together voting machines to push Colemans votes up to the low teens... This is just the beginning.... Stay tuned kids...

On a more serious note, I have no idea what the numbers mean, all I know is that the author clearly had an agenda before writing the piece and stats can play tricks on you just as easily as the divil....
posted by zeoslap at 2:21 PM on October 10, 2003


As someone who lived in Tulare County, CA for 9 years (Visalia specifically) it wouldn't surprise me if that really was how they voted. The people living there are nuts. You know, the kind of people who'd take the trouble to register and go down to the polls on Election Day and vote for some no-name candidate just so they can tell everyone they did.
posted by fishbulb at 2:31 PM on October 10, 2003


zeoslap:
No desire for substantive engagement of the issue? Have you even read any of the Bev Harris pieces? Even if the California data is meaningless, it ought to be looked at in the context of the greater body of evidence about Diebold. Your desire to ridicule is outpacing your actual intellectual curiosity ten-to-one at this point, ya know.

But in any case, thanks for replying to the actual points. Your little skit sure does debunk the stacks of independently verified research done on electronic voting over the last several years. I, for one, will sleep easier now that I have laughed away all the facts thanks to your enlightening Gary Coleman diatribe.

On a more serious note, I have no idea what the numbers mean, all I know is that the author clearly had an agenda before writing the piece and stats can play tricks on you just as easily as the divil....

That is a useful observation. I think that the reason why a lot of people still consider the California numbers worth looking into (which is different than knee-jerk acceptance, by the way) is that it could be much more easilly understood when considered along with what we already know about Diebold. And even if the analysis in the linked piece is off, that doesn't dispell the other nasty Diebold shit, does it?

And you still haven't said why it is a priori impossible that votes would be siphoned away from a big candidate and distributed piecemeal among lesser candidates to cover the tracks. Unlikely? Crazy? Yeah, but impossible? Basically, this software is a well-documented election-stealing tool, but you are saying that worrying about it being used to steal elections is silly. How does that work?
posted by Ignatius J. Reilly at 3:10 PM on October 10, 2003


I think I found the cause of this oddity.

Tulare county ballot

let me cut the ballot down for you

Notice anything?

Kunzman, Sprague and Palmieri all got unusually high vote totals

and all of them were next to more prominent candidates.
posted by RobbieFal at 3:46 PM on October 10, 2003


zeoslap:

I also doubt that if you dissasembled TODAY any of the binaries you can find in the Diebold machines you would find IF dem > rep THEN move vote from dem to rep, or from rep to dem or whatever political party one may prefer.

I mean that would be absolutely idiotic for them to leave such tracks in the very code. TODAY ..or yesterday.

But what I care for is tomorrow Anybody with enough skill and motivation (money,political reason,plain old insanity, name the reason you like) will just need to click a couple buttons and change that source code ; poof, one second ANY proof of the voting is gone , because _there is no paper ballot_ and any kind of other paper records (for instace votes lists)can be much easily altered.

Now, you say that computer voting machine is better ? What do you think is the most difficult, changing a few lines out of a database or concealing, destroying and changing any -significant- number of paper ballots ?

Yeah somebody can steal a box or two of paper ballots, so what ? Try stealing 500.000 without having an HUGE scandal ....or try blaming "an evil hacker " or " a computer glitch" or an "evil virus" or a "destabilizing terrorist"...which is easier ?

Yeah somebody may also steal all the paper ballots or give fire to them or whatever, that's why they should be adequately protected.

So again, why why why do you think computer voting machine is better ? Care to share your vision with us or you're going distract me with Luddites pseudoinsult and tangents on Gary Coleman ?
posted by elpapacito at 3:52 PM on October 10, 2003


Hows this for starters.... with regards Tulare county that got Mr Miller in a tizzy in the first place

"Still, county results showed that some residents may have unintentionally voted for the wrong candidate.

Democratic candidate Ronald J. Palmieri, Independent candidate Jerry Kunzman and Republican Randall D. Sprague each tallied a significant number of votes.

But the vote totals for the three candidates may have come at the expense of confused voters who intended on selecting either Schwarz-enegger, Bustamante or McClintock.

The names of Palmieri, Kunzman and Sprague were placed next to those of the three frontrunners on ballots. Statewide voting results did not show the same trend. "

From here

It should also be noted that Tulare County uses the Diebold OptiScan and not the Accuvote-TS (touch screen) which was the subject of Wallachs paper.
posted by zeoslap at 4:24 PM on October 10, 2003


And Elpapacito, computer voting, specifically touchscreen systems are better because it makes it much easier for people to generate a valid vote for the candidate they actually want, and Tulare is a perfect example of this.

It's not the OptiScanner having it's wicked way with the votes it's the fact that the paper ballot it scans confuses people, same with the old punchcard that resulted in 10% of votes not even being valid in LA.
posted by zeoslap at 4:48 PM on October 10, 2003


RobbieFal - didn't see your comment. Precisely...
posted by zeoslap at 4:49 PM on October 10, 2003


I think it is truly frightening that the state of the American mind is such that it can not handle a simple paper ballot.
posted by five fresh fish at 5:32 PM on October 10, 2003


zeoslap: I may agree that a touch screen may be a better interface for voters, but then again I see a lot of people having a lot of trouble with ATMs and video rental machines that employ the touchscreens.

If the touch screen is deployed on the grounds that people don't know how to make a cross on a circle, oh well it seems like touch screen is more confusing then making a cross for me. I mean, even illiterates and people that didn't attend k-12 know how to make a cross.

Anyway, as long as there is a paper ballot for each vote that is casted AND the voter receives a paper ballot printed by the machine , I may agree on a double voting system that allows

1)electronic votes to be counted realtime
2) paper votes to be counted by hand or OCR

So we have two good voting system with an "apparently" redundant paper trail which makes everybody happy, recounters included.

But I think the trend is for a complete touch-screen or no-paper voting system, which is totally insane.
posted by elpapacito at 5:33 PM on October 10, 2003


Robot Johnny:
"Optiscan" means optical scanner. Most of these counties were using the "X-marks the choice" system, and they were being counted by diebold equipment.
posted by delmoi at 5:40 PM on October 10, 2003


Ah the sweet sound of Ignatius' silence. ;)
posted by zeoslap at 9:13 AM on October 11, 2003


« Older ten years of matt's life   |   Fun with the Constitution! Newer »


This thread has been archived and is closed to new comments



Post