« Musical Memory, Dreams, Readership | Main | House of Blues? »
November 29, 2006
Undervotes & Recounts
Faraway, so close...
Looks like we're not quite done with 06 after all. There's still FL-13 where the ghost of Katherine Harris's career is haunting the congressional race. For those of you who are not aware, in Sarasota County - a Dem stronghold - there is 15% undervote on the Congressional race, where the rest of the races average 2.5%. Furthermore, like most of the country, this district voted on DRE machines (aka electronic machines, even though that's a bit of a misnomer since optical scan machines are electronic as well).
I'm not quite sure what happened in FL. One thing they should be doing is examining the image captures in the machines - many of the DREs take a screenshot of the ballot when the voter casts it. But if there was something interfering with the capture of the congressional race data, it likely won't show on the ballot. Nonetheless - worth a look.
The most important thing to remember here is that 15% undervote is an absolute abnormality. 15% of the people did not ignore one of the top tier races in an election year like 06. 3% is the edge of abnormality in down ticket races. Anything higher than that and you've got a problem. In the AG race last year, there were only 55 precincts in the entire state of Virginia which used optical scan & had an undervote on the Attorney Generals race above 5%. The number was far lower for DRE, and this is on a down ticket race in an off year election.
So what happened in FL-13? I don't know. My initial inclination is that the ballot needs to be scrutinized closely. But questions also need to be asked about the calibration of the machines. I would like to take the opportunity to once again remind people of Occam's Razor here - I would consider it very unlikely that malicious code was inserted into all machines in one county affecting only one race. If there was a hacking attempt, it wouldn't have been on this race - the Governor's race would've been a far juicier target. So please leave the tin foil for vegetable storage and mind-beam-repulsing wallpaper. It would be interesting to see if anyone attempts to use Florida's Sunshine Laws to attempt to get the source code of the DRE machines. I'm not a lawyer, however, so I don't know how that would apply, or whether there's really a case to be made.
The one place this election could be lost is the same place the AG race was lost in 05 - in court.
While we're talking about recounts, I wanted to hit on the VA Senate race for a moment. Several friends of mine asked why there was no recount even though the vote was within the margin. Simple reason - A DRE is going to give the same result each time. You're not hunting for abnormalities like you would in an Undervote situation - the count is going to be the count. Furthermore, 8000 votes is not a margin that can be made up during a recount. Hell, 300 isn't a margin that can be made up during a recount. One high level person on the Deeds recount last year said to me, "We're close enough to recount, but not close enough to win a recount" - and that was with a 363 vote difference.
For more on the FL-13 situation, checkout Today's Post.
Posted by MikeSager at November 29, 2006 11:00 AM
Trackback Pings
To send a Trackback, please use:
http://www.carpetblogger.com/cgi-bin/mt-whattocheck.cgi/265
Comments
What happened in Fla happened all around the country in various competitive districts. The diebold machines were set to give the Republicans a small advantage (just enough to win according to the polls) BUT...they failed b/c the margin of victory was greater than expected.
That's why Rove and Bush weren't that worried pre-nov 7th...they thought the bases were covered.
Of course the "lame stream" corporate media is not covering this important issue.
Buzz...Buzz...
Posted by: mosquito at November 29, 2006 12:13 PM
I will refer you back to Occam's Razor - the simplest explanation is likely the correct one. Which is more likely? A massive nationwide conspiracy between several corporations, the White House, & the media? Or a poor ballot designed combined with election worker ignorance on new technology?
Posted by: Mike at November 29, 2006 12:19 PM
I don't know what happened in FL. But it is strange that the undervote was so large.
Posted by: Vivian J. Paige at November 29, 2006 3:15 PM
Stranger still...ALL the mistakes ALWAYS favor Republicans.
REpublican Senator Chuck Hagel was the CHALLENGER and ALL THE POLLS said he didn't stand a chance in Hades...yet he won in an "amazing" upset that defied the polls, the exit polls, etc....funny thing...at that time Chuck Hagel owned a large portion of Diebold and that's the machine they used in the election. On 1996 when Hagel won it was considered the Big Upset of the election...
What's more Hagel lied about his ownership in Diebold at the time....
WHEW...what the corporate media has studiously avoided covereing. Must be that mythical "liberal" bias...
Buzz..Buzz...
Posted by: mosquito at November 29, 2006 5:38 PM
...And there are plenty of places where the voting machines worked fine and the Democrats won. There have been plenty of elections where our candidate has supposedly had no chance, and yet we've gone on to win. Polling is not an exact science, and races can be substantially closer than polls indicate. But I will say that the more we blame the infrastructure and try to claim "They're not playing fair", the more we lose sight of what really matters - talking to voters and winning elections. Are there anomalies in Florida? Absolutely! They need to be investigated. But is there a conspiracy? No, there's not. It's time to stop complaining about the field and to start playing ball.
Posted by: Mike at November 29, 2006 5:57 PM
First off, hello cousin Mike! Just found your blog, and thought I'd take a look around.
Now on to the meat...
Occam's razor does dictate that the simplest solution is usually the correct one, but I wonder if you're applying it in a reasonable manner here.
In every case, literally without exception, discrepancies between 2006 election exit polls and the results spat out by the DRE machines favored Republicans. The chances of that happening by random chance are ridiculously small.
Furthermore, the work required for a successful election hack isn't nearly as complicated as you'd think. Those machines are electronic security nightmares. A single technician with the right access could install any code desired using a memory card.
Worse, the actual data from the elections is *still* stored in a Microsoft Access database on Diebold machines. This was a scandal when that was discovered three years ago, but it still hasn't changed. Post-election, anyone with access to those machines and Microsoft Office could literally change the totals to whatever they want them to be. Some of the machines Diebold produces were found to have MS Office already installed during an audit.
On top of all that, there is a program called Portable Apps that allows you to fully install an application on a flash memory device, like those used in the voting machines. It's certainly illegal to use it to make a copy of Microsoft Office you can take with you like that, but legality isn't going to be a barrier to someone who wants to hack the election.
With the impossibly low likelihood of these results happening by chance, their absurdly partisan bent, and the simplicity and ease of changing the election results untraceably, Occam's Razor ceases to dictate that a hacked election is the least likely cause of the discrepancies. In fact, I would argue that bungled election thievery - certainly a staple of American politics - is the *most* likely solution now. Every other solution relies on random chance resulting in a statistical impossibility.
Posted by: Damon Kaswell at December 1, 2006 3:19 PM
Damon - good to see ya!
I'm not sure which exit polls you are referring to, but exit polls onto themselves are inherently unreliable. There were many times they were incorrect before DRE had massive penetration.
You pointed out a tech with the right access - the right access being the key. The people who have access to the current machines are the same people (for the most part) as the ones who had access to the lever machines & punch cards & optical scan ballots. The election employees are often times geriatric, and unable to operate VCRs and Toaster Ovens. These people are the only ones with access - they won't be capable of hacking these things before they're deployed. If it's the supervisor of elections carrying this out, why weren't they doing it with other technology? Furthermore, a person trying to fuck with a voting machine during an election is going to be quite obvious in most circumstances - even to said geriatric poll watchers.
Plus, once you copy the file down you're going to take it to a computer with Access and change the results, and then you or second person go in to replace the valid data?
To me - there are too many barriers to hacking said machines, despite their obvious security flaws (Access), that it is no longer practical to attempt massive fraud like this.
Additionally - if the vote were being stolen, why would they carry out extreme measures to attempt to suppress the vote in other ways? The harassing robocalls, police parked in front of precincts in Democratic areas, misinformation campaigns. These methods are far easier to carry out, and frankly probably more effective.
I am personally a full supporter of a paper ballot on top of DRE, but I don't buy for a second that all over the country there are election workers hacking the machines or people illegally gaining access. It requires too many things to fall into place to be effective.
There are simpler ways to steal or swing an election.
People are always suspicious of new technology. We should be suspicious and vigilant when dealing with our voting equipment. But we must also stay grounded. The more we blame the infrastructure, the more people will say "Fuck it - my vote doesn't count anyways!" And then, our people won't turn out - and we will truly lose.
Posted by: Mike at December 1, 2006 3:37 PM
Machines can also be compromised at the factory, at the elections commission inspection, before or after certification, and so forth. There are lots of opportunities to compromise a large number of machines at once.
And who says you need to copy the file down at all? Just swap it for one you brought on your own USB key; the database isn't even encrypted or password protected. Or worse, bring in a script that automatically turns one out of every ten votes for a Democrat into a vote for a Republican? A corrupt technician could do that in plain sight, in front of all those geriatric poll workers, and never even be noticed.
However, I don't actually think there was any massive attempt to hack the election machines this year. A few machines here and there would be all it would take in a tight election, but of course, this turned out not to be a tight election, thank goodness.
In this case, what I think happened is that a few machines did get hacked, and in a couple of isolated places it was enough to sway local elections, like in FL-13. But the sheer number of Democrats getting out the vote overwhelmed most of these attempts.
The other measures - robocalls, et al - you're talking about were the bigger, more organized ones, of course. And more blatant, pathetic, and desperate. Like I said, this wasn't a close election, and the Republican party was certainly aware of that. But why make it even slightly possible to sway a bunch of DRE machines at once? Until all of them produce a paper trail, I'll consider any and every election that uses them suspect. In the old days, you had to stuff ballot boxes and produce physical forgeries. Now all you'd need is a thumb drive.
Posted by: Damon Kaswell at December 1, 2006 5:11 PM
Post a comment