2004 election touch-screen voting machines
September 7, 2004 7:35 AM Subscribe
With all this talk about potential touch-screen voting machines, I was thinking, "What, exactly, is the big deal? Worst case scenario, even?" I've heard 1/3 of the senate is up for grabs, along with the entire house of representatives and the presidency (correct me if I'm wrong). So, what would happen if there is some touch-screen SNAFU and they all came back republican? Is there anything anybody could do? Also, with a CURRENT republican majority in the house, the senate and a conservative leaning supreme court (that might not have jurisdiction), what if the probable investigations are stymied until January?
It can't affect all of the House and Senate races for the simple reason that the vast majority of voters won't use touch-screen machines. Most voting will still be done using paper ballots, lever machines, or punch cards.
posted by ROU_Xenophobe at 9:44 AM on September 7, 2004
posted by ROU_Xenophobe at 9:44 AM on September 7, 2004
The worst case would be much the same as for other supposedly secure systems which turn out to be hackable - initial shock and calls for actions, followed by inaction and legal foot-dragging, and finally acceptance as "just part of life."
We sort of accept that today's world includes the eventuality a certain percentage of us will have our identity stolen and our bank account drained by Russian hackers. Ditto for the idea that our inbox will be flooded by worthless spam which we'll never read. Etc.
In the coming years we'll just accept that a certain percentage of votes will be bogus. Since there are too many hackers around the world, and it's too "hard" to just use paper and ballpoints, we'll understand that having the vote 90% right is the best we can do.
That's the worst case.
posted by y6y6y6 at 9:56 AM on September 7, 2004
We sort of accept that today's world includes the eventuality a certain percentage of us will have our identity stolen and our bank account drained by Russian hackers. Ditto for the idea that our inbox will be flooded by worthless spam which we'll never read. Etc.
In the coming years we'll just accept that a certain percentage of votes will be bogus. Since there are too many hackers around the world, and it's too "hard" to just use paper and ballpoints, we'll understand that having the vote 90% right is the best we can do.
That's the worst case.
posted by y6y6y6 at 9:56 AM on September 7, 2004
In the case where every vote comes back Republican (or Democrat) then clearly something has gone wrong, and votes will need to be recast (I presume there are provisions for dealing with completely lost or mistallied votes in most voting laws).
A far more insidious case, and the one you'd want to go with if you were going to rig things, is where just enough votes are changed to give a particular candidate a solid lead (it would not have taken many changes to swing the Presidential vote either way in Florida in 2000, for instance). If the rigger's candidate is already leading then perhaps no votes need to be changed. If necessary then some of the votes for other candidates would be cast for the chosen "winner" instead.
The smart election rigger would probably also make sure that some fraudulent votes were cast for the opponents, even in the case where the rigged winner is already leading. Not too many, mind you, but enough to throw some ambiguity into the mix if anyone manages to have a look.
All of this is possible in any system without a rock-solid audit trail. This could include paper and mechanical ballots depending on precisely what happens after the votes are cast. It's not electronic voting that's inherently bad, it's un-audited voting.
posted by Songdog at 10:12 AM on September 7, 2004
A far more insidious case, and the one you'd want to go with if you were going to rig things, is where just enough votes are changed to give a particular candidate a solid lead (it would not have taken many changes to swing the Presidential vote either way in Florida in 2000, for instance). If the rigger's candidate is already leading then perhaps no votes need to be changed. If necessary then some of the votes for other candidates would be cast for the chosen "winner" instead.
The smart election rigger would probably also make sure that some fraudulent votes were cast for the opponents, even in the case where the rigged winner is already leading. Not too many, mind you, but enough to throw some ambiguity into the mix if anyone manages to have a look.
All of this is possible in any system without a rock-solid audit trail. This could include paper and mechanical ballots depending on precisely what happens after the votes are cast. It's not electronic voting that's inherently bad, it's un-audited voting.
posted by Songdog at 10:12 AM on September 7, 2004
Every election in Oregon is conducted by mail, so I'll be safe. Maybe we can spread that to other states. I actually enjoy being able to research candidates online while I vote at home.
posted by mathowie at 10:13 AM on September 7, 2004
posted by mathowie at 10:13 AM on September 7, 2004
It can't affect all of the House and Senate races for the simple reason that the vast majority of voters won't use touch-screen machines. Most voting will still be done using paper ballots, lever machines, or punch cards.
A minority of electronic votes can change any election, local or national. All it requires is that an election be a close one.
Also, some communities have been doing electronic voting for decades. Touch-screen voting is only the latest paradigm of electronic voting, although you are right: mechanical voting still dominates.
In the coming years we'll just accept that a certain percentage of votes will be bogus.
That this should already be something we are all aware of, even before electronic voting, is slowly dawning on the new voting activists and the people who are reading the warning messages about the hackability of EVMs. Voting fraud is common, ordinary, and a factor in many elections in the United States and has been since voting began here. The problem has always been, as with electronic voting, in proving it. Now that we have electronic voting, people want a higher level of accountability than they've ever had. The Mercuri method is not the *same* level of accountability we have now, because we don't have a foolproof system now. It's *more* accountable. Remember how the recounts in Florida came up different each time they were done? Also, there are a thousand ways to steal an election and many of them happen before the ballot ever reaches the ballot box (purging voter rolls according to unfavorable profiles, for example).
So although I am in agreement with the Mercuri method, I think we have to shoot for statistical accuracy rather than vote-for-vote accuracy. As long as the tendency for error can be shown not to favor a specific candidate, then that error should be considered moot. Again, though, the trouble is in the proving.
It reminds me of the beginning days of Internet advertising when the advertisers said (and are still saying), "Finally! We can have true and accurate information about who is seeing our ads, which ads work best, the value of branding, etc., etc." Which led to the pay-per-click, "click farms" in India, the "hit the monkey" ads, and the false Windows error banners. So the results are still inaccurate. You get clicks, but what do they mean?
It's also similar to the recording industry's response to the digital realm. They think, "Wow! Now we can get paid every single time one of our songs is played!" And the movie industry thinks, "Holy moley! Now we can make sure we get paid for every single viewing." But it's a fool's mission. Not only is getting paid for every single instance of any action logistically impossible, it is economically unviable.
This is because all systems can be gamed. All. You have to accept that and build in those losses, real or imagined, into the system. The solution--and it's been said before by smarter people--for elections (and music and movies and ads) is that you need the equivalent of break-away traffic signs: when hit by a vehicle, they are prepared to fail, and they fail elegantly. They reduce damage to the more important components, the vehicle and its passengers, while still doing their job.
A good voting system must be built to expect failures, must be built to handle those failures, and must still report statistically accurate results, even when it fails.
posted by Mo Nickels at 12:23 PM on September 7, 2004
A minority of electronic votes can change any election, local or national. All it requires is that an election be a close one.
Also, some communities have been doing electronic voting for decades. Touch-screen voting is only the latest paradigm of electronic voting, although you are right: mechanical voting still dominates.
In the coming years we'll just accept that a certain percentage of votes will be bogus.
That this should already be something we are all aware of, even before electronic voting, is slowly dawning on the new voting activists and the people who are reading the warning messages about the hackability of EVMs. Voting fraud is common, ordinary, and a factor in many elections in the United States and has been since voting began here. The problem has always been, as with electronic voting, in proving it. Now that we have electronic voting, people want a higher level of accountability than they've ever had. The Mercuri method is not the *same* level of accountability we have now, because we don't have a foolproof system now. It's *more* accountable. Remember how the recounts in Florida came up different each time they were done? Also, there are a thousand ways to steal an election and many of them happen before the ballot ever reaches the ballot box (purging voter rolls according to unfavorable profiles, for example).
So although I am in agreement with the Mercuri method, I think we have to shoot for statistical accuracy rather than vote-for-vote accuracy. As long as the tendency for error can be shown not to favor a specific candidate, then that error should be considered moot. Again, though, the trouble is in the proving.
It reminds me of the beginning days of Internet advertising when the advertisers said (and are still saying), "Finally! We can have true and accurate information about who is seeing our ads, which ads work best, the value of branding, etc., etc." Which led to the pay-per-click, "click farms" in India, the "hit the monkey" ads, and the false Windows error banners. So the results are still inaccurate. You get clicks, but what do they mean?
It's also similar to the recording industry's response to the digital realm. They think, "Wow! Now we can get paid every single time one of our songs is played!" And the movie industry thinks, "Holy moley! Now we can make sure we get paid for every single viewing." But it's a fool's mission. Not only is getting paid for every single instance of any action logistically impossible, it is economically unviable.
This is because all systems can be gamed. All. You have to accept that and build in those losses, real or imagined, into the system. The solution--and it's been said before by smarter people--for elections (and music and movies and ads) is that you need the equivalent of break-away traffic signs: when hit by a vehicle, they are prepared to fail, and they fail elegantly. They reduce damage to the more important components, the vehicle and its passengers, while still doing their job.
A good voting system must be built to expect failures, must be built to handle those failures, and must still report statistically accurate results, even when it fails.
posted by Mo Nickels at 12:23 PM on September 7, 2004
Response by poster: Nice brainstorming.
But seriously, what if most of the results come back republican? What if the house is suddenly almost all republican, and the senate jumps from 51 republican senators to 70?
Are there any laws to protect us? If so, will there be anybody to enforce them?
posted by taumeson at 1:22 PM on September 7, 2004
But seriously, what if most of the results come back republican? What if the house is suddenly almost all republican, and the senate jumps from 51 republican senators to 70?
Are there any laws to protect us? If so, will there be anybody to enforce them?
posted by taumeson at 1:22 PM on September 7, 2004
1) Is election fraud a crime? Yes.
2) Is the justice system perfect? No.
3) Is there a law against electing a Republican senate? No.
4) Is an investigation automatically triggered if the Repubs win by a landslide? No, it's triggered when evidence of faulty tampering or fault is found.
5) If anyone were to steal an election, would they ever be stupid enough to fix the voting machines so that 100% of the votes went their way? Yes, but only Saddam Hussein.
6) Is anyone going to save us from the Bush administration? No, probably not.
posted by scarabic at 2:08 PM on September 7, 2004
2) Is the justice system perfect? No.
3) Is there a law against electing a Republican senate? No.
4) Is an investigation automatically triggered if the Repubs win by a landslide? No, it's triggered when evidence of faulty tampering or fault is found.
5) If anyone were to steal an election, would they ever be stupid enough to fix the voting machines so that 100% of the votes went their way? Yes, but only Saddam Hussein.
6) Is anyone going to save us from the Bush administration? No, probably not.
posted by scarabic at 2:08 PM on September 7, 2004
Welcome to the New America, taumeson. Talk like that will get you arrested.
posted by rafter at 2:10 PM on September 7, 2004
posted by rafter at 2:10 PM on September 7, 2004
As I recall, there were some election irregularities around Diebold machines in the California recall election (some districts with improbably high numbers of voters for one side, that sort of thing). The problem with the Diebold machines is that there's several ways not only to alter the vote totals without being detected, but also to do it without leaving any kind of auditable trace later (once it's done, it's done).
The odd counts in California wouldn't have been enough to change the outcome of the vote even if they had been wrong, so not much was done. The state AG is apparently going to sue Diebold for making fraudulent claims about the machines, however.
posted by Dipsomaniac at 12:16 AM on September 8, 2004
The odd counts in California wouldn't have been enough to change the outcome of the vote even if they had been wrong, so not much was done. The state AG is apparently going to sue Diebold for making fraudulent claims about the machines, however.
posted by Dipsomaniac at 12:16 AM on September 8, 2004
This thread is closed to new comments.
Is there anything we can do? Become voting-integrity activists. Demand paper trails, transparency, that sort of thing. If a vote comes in looking suspect (as has happened before), though, I don't think there's much everyday citizens can do, unless we physically catch election officials or voting-machine makers tampering with the vote. Recounts are obviously impossible.
posted by adamrice at 8:25 AM on September 7, 2004