You know that concern over vote rigging has hit the mainstream...

[jducoeur]

... when it's the lead topic on the Simpsons' Treehouse of Horror episode. (Which featured Homer trying to vote for Obama, and getting tallied for McCain instead.)

I confess, I'm not *very* worried about organized high-level vote-rigging, simply because it would require a well-run conspiracy, and those seem to be fairly rare. Doing something on that scale without big leaks is hard, although not impossible. And heaven knows, if you think that the voting-machine manufacturers are in the tank for the Republicans, it is easy to come up with conspiracy theories. But given the attention on this topic, I'd say it's unlikely.

None of which silences the tiny nervousness at the back of my mind. The actual tally had better line up pretty well with the exit polls, or all hell is going to break loose on Wednesday...

Comments

[talvinamarich.livejournal.com]

I have enough supplies to last a week *minimum* set aside.

I do not consider rioting out of the question if McCain wins. For that matter, victory celebrations may get raucous as well.

They were firing off guns over the World Series, but there are still far more Obama shirts, hats, and buttons on the streets than Phillies gear.

[meiczyslaw.livejournal.com]

The place I'm most worried about is Ohio. They've basically set up a system where people can drive/fly in from out-of-state, register, and then vote that same day. There have been some documented cases of this occurring, and the local government hasn't followed up. ("Some" in this context is a dozen or so out-of-state college students who were stupid enough to post their tactics on blogs. Also "hasn't" could be "can't" as opposed to "won't" -- Ohio is one of the places that's been overwhelmed by bad ACORN registrations.)

This is pushing me to support a National ID card. It would be painful getting them issued -- see how difficult it is to get a passport if you were birthed by a midwife in West Texas -- but once we'd gotten IDs to everyone, we'd get to avoid a bunch of this.

[mindways]

As is so often the case, bradhicks has written a piece on the subject with very interesting points / data and conclusions I don't agree with. :)

There was also something on FiveThirtyEight.com about how exit polls tend to be skewed, but I'm not seeing it under their 'exit poll' tag. Hmm.

[mindways]

Aha - it's in a post about the Bradley effect (or lack thereof) - as reason #3 why the myth of the Bradley effect persists. To quote:

"3. Confusing Exit Polls with pre-Election Polls. Unlike the normal, pre-election polls, exit polls conducted on the day of the election did substantially overstate Barack Obama's margins throughout the primaries. This is something to keep in mind at about 5 PM on November 4, when Matt Drudge and Jim Geraghty begin to leak exit poll results. It is not anything to worry about now, when we are trying to forecast the outcome from pre-election polling.

Nor is it clear the the discrepancies in the exit polls have anything to do with race; John Kerry, somewhat infamously, also underperformed his exit polls. The mechanics of conducting an exit poll are rather haphazard, involving a bunch of college kids and temp workers running around outside a polling place with clipboards and attempting to pass out survey forms to every Nth voter who leaves the ballot booth. This is not much easier than it sounds, and introduces a lot of human error and other forms of sample bias. For this reason, exit polls are not really intended to be used as they so frequently are in the panicked hours before ballot counting begins -- the results need to be calibrated and weighted, and exit polling firms rely on comparing their polls against actual voting results in order to do so."

[meiczyslaw.livejournal.com]

Also, one of the things that Schmot Guys are worried about with this election -- exit polls are based on same-day voting, and not early voting. Now, the Schmot Guys know this, but the media outlets are going to pressure themselves to call the election first.

In a closer election, these forces would produce another "Dewey Defeats Truman" moment.

[meiczyslaw.livejournal.com]

From Forbes (http://www.forbes.com/opinions/2008/11/01/exit-polls-election-oped-cx_kb_1103bowman.html):

In the 1,460 exit poll precincts where Edison/Mitofsky collected both exit poll tallies and actual final vote returns in 2004, the exit poll results overstated the actual difference between John Kerry and Bush by 6.5 points in Kerry's favor.

This problem was due, according to a post-election analysis by Mitofsky, "to Kerry voters participating in the exit polls at a higher rate than Bush voters."

Interesting quote from Fox News (http://foxforum.blogs.foxnews.com/2008/11/03/frj_1103/):

According to varied professional sources with whom I have spoken, there exists a proportionally high number of potential voters who are refusing to be polled or express their opinion publicly. In a historic, high-octane race like 2008, I believe there are more reasons for a McCain supporter to stay silent than for an Obama supporter.

Based on trends, I'm pretty sure McCain's going to over-perform all the polls -- including the exit polls.

It's gotten me thinking -- with how the media's been biased towards Obama, I'm wondering if the polls have all been wrong because McCain's supporters don't trust the media enough to talk to them. (And this would also mean that they see pollsters as agents of the media.)

[jducoeur]

I'm skeptical on a couple of different levels.

First, the "the media's in the tank for Obama" thing is rather overdone. It's true in some cases, wildly false in others. Fox News' influence seems to be disproportionately high, and it is *far* more biased towards McCain than most of the rest is towards Obama. (Sometimes to an almost laughable degree.)

I don't doubt that there is some personal bias, but there hasn't been nearly the degree of bias of *coverage* that the conservatives would like to believe. All the stories that they claim have been ignored were anything but: they simply got played until the public seemed to be losing interest, then dropped, which is pretty much exactly the way they've treated McCain. Far as I can tell, they wouldn't be happy unless the news was all-anti-Obama, all the time, so I look at this claim very acidly.

Is it possible that a huge pro-McCain cohort *thinks* the pollsters are in the tank for Obama, and refuse to co-operate? I suppose, but I find it very strange. In most of the US, it is still more politically correct to be conservative than liberal. Hence, it seems strange that people would be so reticent to proclaim their conservative views -- if anything, I'd expect to see it go the other way.

More important, though, is the fact that *all* of the reasonably neutral polls have a strong lead for Obama. There's a huge range in those polls -- but that range is between a 2% and a 10% lead for Obama. Yes, you can say that the 2% end is within the margin of error, and McCain is basically hanging his hopes on that. But I'll still be astounded and suspicious if Obama doesn't win at this point: it would require many millions of people, following a very strong statistical bias, lying to the pollsters consistently.

Could it happen? Maybe. But this is essentially the "silent majority" argument, which has usually historically been used as a technique to lie to the public...

[meiczyslaw.livejournal.com]

I suppose, but I find it very strange.

Sadly, it's not so strange here. When enough Obama supporters casually throw around the idea that only racists would choose not to vote for him, who would want to admit their opposition to the man in public?

(Also, my assumption is not simply that 6% of the electorate has been hiding their preferences. I'm also taking into account the fact that the polls' "likely voter" models have consistently had a 3-4% Democratic skew since 1980, and I see no reason to believe they've finally corrected that.)

[hudebnik.livejournal.com]

I'm not *very* worried about organized high-level vote-rigging, simply because it would require a well-run conspiracy, and those seem to be fairly rare. Doing something on that scale without big leaks is hard, although not impossible.

I am worried about it, precisely because it would not require a large, well-run conspiracy: half a dozen people in the right places at the right times could shift lots of votes, in many cases undetectably, through voting-machine manipulation.

Have you seen "Hacking Democracy"? In the climactic sequence, a computer-security expert gets hold of a removable memory card for a voting machine, analyzes its contents, changes them a little, returns it, and we watch on camera as a test election produces massively skewed results, undetectably unless there were a manual recount of paper ballots. This would require some access to the memory cards before the election, but since the machine manufacturer swears that the cards contain only passive data, and they're checked for "zero-ness" before the election starts, security is typically much looser before an election than after it.

So imagine you live in a swing state which is using that sort of voting machine. A few days or weeks before the election, you break into wherever the removable memory cards are stored, and you spend an hour or two burning your modified program into a few thousand of them. On the morning of election day, those cards are inserted into a few thousand voting machines, your program dutifully reports that it is working correctly and the totals are zero (although they're actually, say, +30 for your candidate and -30 for the other), and they go into action. Each individual vote is correctly added into the total, so voter-verification doesn't catch it. The recorded vote differs by only a few percent from the polls, so nobody suspects foul play... but you've shifted 100,000 votes, which could change the outcome in a swing state.

(Of course, the details will differ depending on the model of voting machine and what hacks are applicable to it; the above scenario is for the particular hack illustrated in "Hacking Democracy".)

If you happen to live in a state without paper ballots at all, it's even better: even if somebody does suspect foul play, there's nothing they can do about it, because the only source of data for a recount are the memory cards themselves.

Of course, if you happen to work at a voting-machine manufacturer, things are much easier, since you don't need to break into anything.

[meiczyslaw.livejournal.com]

Yeah, but if DIEBOLD is in on it, they've been doing a really crappy job of keeping Democrats from getting their hands on the system in Ohio, don't you think?

[hudebnik.livejournal.com]

I didn't actually mention Diebold, its new incarnation (whose name I've forgotten), or any other specific manufacturer. True, the particular hack demonstrated in the movie was applicable to a particular model of Diebold machine, and I suggested how a few people (no vast conspiracy) could take advantage of that hack; the details will vary for different machine models, different vulnerabilities, and different hacks.

Ohio went Republican in 2004, probably more because of "caging" and enormous waiting lines, especially in low-income and minority precincts, than because of voting-machine hacking (although the testing of those machines was grossly inadequate, so we'll never really know whether voting-machine hacking was also a factor). And it was a swing state, so it's possible it would have gone Republican even without those biasing factors. In Florida 2000, I think voting-machine hacking was probably a factor, if you believe reports that that Al Gore got negative numbers of votes in some precincts.

2006 was a Democratic tidal wave almost everywhere in the country. If Democrats won in Ohio, that doesn't disprove the conspiracy theory that Diebold tried to swing the election for the Republicans; it only proves that, if so, they were unable to swing it far enough without being blatantly obvious and getting caught. I honestly don't know whether to believe in that conspiracy theory, so don't assume I'm defending it.

[jducoeur]

Ohio went Republican in 2004, probably more because of "caging" and enormous waiting lines, especially in low-income and minority precincts, than because of voting-machine hacking

Oh, sure -- *this* is clearly true, and the Republicans have pulled some particularly awful dirty tricks this time around. (Eg, calling people and telling them that Republicans vote on Tuesday, Democrats on Wednesday.) But there's nothing subtle or hidden about any of that: it's pretty overt attempts to game the system and confuse the populace...

[jducoeur]

It's possible that it could be done, but it's dangerous at best. Yes, it could be done with half-a-dozen people, but they would have to be exactly the *right* half-a-dozen people in terms of skills and access; in practice, that means you need a significantly larger conspiracy in order to round up the right folks. And to really swing an apparently non-close election like this, you'd need this to be repeated several times across the country.

And while I have been mildly suspicious of Diebold, I'd be very surprised if they aren't keeping their noses scrupulously clean this time around. While it's conceivable that they are fanatical enough about the right wing to try something, normal corporate behaviour is to support the winners, and in no respect are the Republicans looking like winners this time around -- at *worst*, the Democrats are going to dramatically increase their control over Congress. In such circumstances, no one sane does something that invites Congressional inquiries that could destroy the company. (And again, it would require a non-trivial conspiracy, albeit well-contained inside the company.)

Hence my "not very worried". It could happen, and if things look truly suspicious tomorrow I might be more open-minded about the possibility, but it's hard to see getting away with it on a substantial scale without being caught...

[hudebnik.livejournal.com]

Yes, it could be done with half-a-dozen people, but they would have to be exactly the *right* half-a-dozen people in terms of skills and access....

It requires one person who can do some security hacking, decompiling, etc. You and I probably both know people like that. It requires, perhaps, one person who can break into lightly-secured warehouses; we might both know people like that too. Other than that, it requires people who can stick a memory card into a card reader/writer and push a button, over and over again.

I'd be very surprised if [Diebold] aren't keeping their noses scrupulously clean this time around.... normal corporate behaviour is to support the winners

True, but that doesn't mean one or a few employees at one of the voting-machine manufacturers won't try to swing things their way (whatever that way may be).

it's hard to see getting away with it on a substantial scale without being caught...

A voting-machine-manufacturer employee might be caught by his own company's QA people. Failing that, it'll take a manual recount to prove that things have been hacked. If there is no independent record to be manually recounted, that can't happen. If there is an independent record, but a recount is only ordered under rare circumstances, the hack might still not be discovered.

[jducoeur]

It's moot at this point, but seriously: conspiracies don't generally break because of something like QA or recounts -- they break because they are made of *people*, and people, by and large, can't shut up. The old saw about three people being able to keep a secret if two of them are dead is mostly true, except that even *one* person is often pretty bad at keeping a secret.

Hence my skepticism. Really, I think it unlikely that you could pull something like this off without having at least a dozen people involved at *some* level (as part of the networking to get the right people, if nothing else), and a dozen people is already a large enough number that keeping it silent is difficult unless they are unusually focused, disciplined and well-organized...