Black box voting

[ciphergoth]

SoCalDem has done a statistical analysis...on several swing states, and EVERY STATE that has Electronic Voting but no paper trails has an unexplained advantage for Bush of around +5% when comparing exit polls to actual results. In EVERY STATE that has paper audit trails on their E-Voting, the exit poll results match the actual results reported within the margin of error.

Comments

[juliet]

I still fail to understand why they can't just use paper & pen. The US population is about 5 times greater than ours, so even counting at the same rate they could have a result within about 30 hours (& there is, of course, no reason why they shouldn't just get in more people to count).

Actually, I do know why it is: because it would require more people & thus be more expensive, & apparently electoral officials are paid v little in the US. [sigh]

Mind you, I *really* don't know how the fuck states have managed to get away with unaudited black-box voting machines.

[juudes.livejournal.com]

Surely, if the Democrat lawyers could prove anything, they would go to court? Or not, if they were also fiddling figures (just not to the same extent as the Republicans). SURELY they wouldn't be allowed to get away with this, if it was at all provable?

Or am I just very, very naive?

[dennyd.livejournal.com]

Surreal, isn't it?

[dennyd.livejournal.com]

You can't prove anything in a system with no audit trail.

[chillies.livejournal.com]

True. But exit polls aren't truth, either.

Is there a link anywhere to the primary data behind SoCalDem's analysis?

[dennyd.livejournal.com]

Imho it stretches the bounds of credibility to believe that it's a coincidence that a (critical, and similar in each area) percentage of voters changed their minds in all areas which used one particular type of (non-auditable) voting process, where no such shift took place in any of the nearby areas which didn't use said process.

From what I've read it seems that the exit polls match up to the registered voter affiliation too - so when two sets of numbers match, and a third doesn't, which one do you look suspiciously at?

There are some links to data sources in the article which ciphergoth was quoting the description of, I don't know if they're what you are looking for?

[ciphergoth.livejournal.com]

I really want the primary data. The by-county voting numbers aren't too hard to get hold of. Verified Voting is trying to get per-county information on the voting equipment used. I don't know where best to get information on the exit polls.

[lovingboth]

I was wondering how long it would take for this to come up. Will someone have the resources to start suing - and would the courts do anything?

I do note a touching faith in exit polls in several of the articles. Our general election in 1992 should mean they don't trust them so much - a combination of shame reluctance to admit to voting for The Evil One and bad sampling (so much easier to do your exit polls where there are more voters, rather than in the countryside) can easily mean you get the result 5% wrong.

[wechsler.livejournal.com]

In a well-run, properly monitored and fully auditted election, fraud would be comparatively easy to prove, and would end up in court.

However (as the OSCE preliminary report states) this was not one of those. *Many* gaps were created in the process in which improper actions could be taken without a clear chain of proof.

[wechsler.livejournal.com]

Then you'd expect it to 5% wrong everywhere, or effectively random noise. This is why correlations are being noted, not raw percentages.

[juudes.livejournal.com]

Would it not be possible for an independent technician to examine the voting machines and software? I can't imagine that no journalist would be interested in the outcome - if anything were found it would be bigger than Watergate...

[ciphergoth.livejournal.com]

It certainly would be possible, but Diebold refuse all such requests, and indeed even the states that use their equipment have to sign a contract promising not to investigate what's inside.

[ciphergoth.livejournal.com]

I should add that despite all Diebold's efforts to prevent investigation of their security measures, flaws are often found. Googling for "Diebold security" finds gems like this:

http://gort.ucsd.edu/mtdocs/archives/laz/001866.html

http://www.google.com/search?q=diebold+security

Also, it's generally agreed among security experts that there is *no* purely electronic way of having an election that matches the security and verifiability of a paper-based election.

[dennyd.livejournal.com]

I read somewhere that exit polls have been a highly reliable predictor of the outcome of the US elections, up until the last two.

[wechsler.livejournal.com]

You'd think a country that values democracy that highly would be willing to pay for it, wouldn't you?

[lovingboth]

No, it will vary. There's no shame in voting TEO in Redneck County, Nevada: it's what all your neighbours are doing. But in somewhere more split, people will fib to pollsters.

I'm not disputing Diebold may be evil enough to do it, but they'd also have to be spectacularly stupid to just say 'add 5% to the evil vote everywhere we think we can get away with it...'

[zotz]

Wasn't it the prospective polls that were wrong in '92 rather than the exit polls?

From this page:
On the actual day of the election, exit polls carried out by the BBC and ITN both showed there would be a hung parliament, although both of them had the Conservatives slightly ahead. They were both not far from the actual Conservative 43%, and Labour 35%, and if they had predicted using a uniform swing assumption, they would have been very close to the real result. But they adjusted the figures as they were suspicious of the results being so far out of line with the mornings polls.

Funny site that's from, though.

[ciphergoth.livejournal.com]

It will vary, of course, but wechsler specifically mentioned correlations.

And I think we will find

1) that the evidence suggests that Diebold has done exactly that, and

2) that they will not have been spectacularly stupid, but that they correctly predict that any effort to interest the voting public in what will seem like "another hanging chad incident" will fail, and thus that they will totally get away with the fraud they've perpetrated.

The bad guys are often more blatant than you'd ever imagine in your worst nightmares, and they get away with it all the time. Shit, if Bush can get away with blatantly lying to start a war, what can't they get away with?

[mathie.livejournal.com]

Even if you did vote for Bush, would you admit to it? ;-)

[lovingboth]

And there's an argument that they were here, up until 1992. Having been on a campaign that was on the wrong end of a 2% win (exit poll) becoming a 0.1% real loss, I've never been entirely happy with them.

Everyone who's been canvassing knows people lie to you. It's why it's stupid to ask 'can I count on your support' or similar (and ghod knows why the Tories I've seen in action still seem to do so!) Doubly so when there's an element of shame in admitting your vote.

If you went and asked Americans how they voted now, you'd find that on the surface Bush is missing votes in many places all across the 'Jesusland' bits, because people who didn't, in fact, vote for him will tell you - now he's won - that they did. Equally, in more Democratic places, you'll find more Kerry voters than actually did so.

I'd be very interested to see if the exit pollers this time asked people how they voted in 2000.

[wechsler.livejournal.com]

Being incredibly blatant is one of their core tactics. When accused of something, their response is generally along the lines of "No-one in their right mind would do anything that stupid/blatant/unethical."

And it works, because people believe that.

[wechsler.livejournal.com]

Most Bush supporters were extremely proud of their "patriotic action". Very few are likely to be shy about the admission.

[lovingboth]

Mandelson's been known to talk about a 'get your champagne out' call from the late great Vincent Hannah on election night 1992, based on the latter's knowledge of the exit polling.

The site's an essay bank, isn't it?

[drdoug.livejournal.com]

Presumably you've seen all the stuff about Diebold in the Risks archive?

[ciphergoth.livejournal.com]

Correlation, though.

In any case, you're contradicting yourself. By saying that not even the exit polls provide a guide for what vote we might have expected, you're saying that there's no possible way of catching Diebold at it if they did the thing that you said would be "spectacularly stupid" for them to do.

[zotz]

Mandelson's been known to talk about a 'get your champagne out' call from the late great Vincent Hannah on election night 1992, based on the latter's knowledge of the exit polling.

Yes, presumably based on the polls tweaked to match the prospective ones, if that essay's correct.

The site's an essay bank, isn't it?

Yes, but it matches what I remember. This article (google cache) states that the error in the BBC '92 exit poll was 2%, which is well within the stated margin of error. The problem wasn't the raw result.

[ciphergoth.livejournal.com]

I have now. Scary!

[chillies.livejournal.com]

Sorry ... I though the data on the referenced page was satire e.g. Calhoun County: Republican 741% Democrat-23%. Verified voting is much better.

[chillies.livejournal.com]

yes ... that's the data I can't find. Even if I could I don't have enough statistics know-how to analyse it.

[vvalkyri.livejournal.com]

At this point I figure it's more important to work to the future rathe r than to figure on fighting it. Strange what a couple days does.

I posted a link to this post, though.

It wouldn't be so much 'add 5% wherever we can find it ' as 'tally the votes normally until vote #x, and then flip one to Bush every Y votes.

Dunno. It's all very disheartening, and the fact that it was this close indicates some major change needed in the Dem message anyway.

[jhg.livejournal.com]

(snigger).

As Sir Humphrey once said, 'Ah, one of those democracies'.


J

[jhg.livejournal.com]

"All this was inspired by the principle - which is quite true in itself - that in the big lie there is always a certain force of credibility;"

-Hitler, Mein Kampf

From http://en.wikipedia.org/wiki/Big_Lie

[wechsler.livejournal.com]

Illustrious company Bush keeps. Thanks for that link.

Oh, the shame, the shame

[webcowgirl.livejournal.com]

It's all just too depressing and I feel really disempowered these days. However, in keeping with Tom Tomorrow, might I recommend these scariest Halloween costumes, from my local alternative rag. To some extent it does show that people believe there is a problem ... I don't know why the dems didn't move harder on this. *sob*

[dennyd.livejournal.com]

Those percentages show the change between proclaimed intentions of voters when they registered, and the actual outcome of the vote - they're not absolute figures, they're relative. Did you actually read the page? The numbers are generated by a very simple formula which is explained in a single sentence.

[ciphergoth.livejournal.com]

I don't see a problem with machine-counted paper votes, so long as many randomly-selected bundles of the votes are re-counted by hand to ensure that the machine counts correspond to the correct counts. If those counts show substantial differences, you have to hand count the lot - so best ensure your system doesn't create the possibility of "hanging chad" ballots that your machines will miscount if you want to use them...

[selki.livejournal.com]

See http://www.evoting-experts.com/ (http://www.evoting-experts.com/), an e-voting blog, includes a link to this account of an election judge which shows how easy tampering with DREs could be in real-life situations (Timonium, MD): http://avirubin.com/judge2.html (http://avirubin.com/judge2.html)

Also see http://www.blackboxvoting.org/?q=node/view/78#breaking (http://www.blackboxvoting.org/):

a request filed in King County, Washington on Sept. 15, following the primary election six weeks ago, uncovered an internal audit log containing a three-hour deletion on election night; “trouble slips” revealing suspicious modem activity; and profound problems with security, including accidental disclosure of critically sensitive remote access information to poll workers, office personnel, and even, in a shocking blunder, to Black Box Voting activists.
[...]
The central servers are installed on unpatched, open Windows computers and use RAS (Remote Access Server) to connect to the voting machines through telephone lines. Since RAS is not adequately protected, anyone in the world, even terrorists, who can figure out the server's phone number can change vote totals without being detected by observers.
The passwords in many locations are easily guessed, and the access phone numbers can be learned through social engineering or war dialing.

[vvalkyri.livejournal.com]

Which also brings us to a conversation about why on earth you find people who will vote for someone because he thinks that person will win, rather than because he wants that person. Don't get that one at all.

[chillies.livejournal.com]

I guess that's the dyslexia playing up again, Cap'n.