Ohio Opts to Put Touch Screen Voting on Hold

Dachannien writes "The AP reports that Ohio's Secretary of State Kenneth Blackwell is asking the Feds for an extension of the deadline for installing touch-screen voting machines at polling places for the 2004 election, citing numerous security concerns. The problems discovered in a review of 57 areas of concern include such gems as "software that permits votes to be counted more than once" and "unauthorized poll workers or others could gain access to hardware that could allow them to perform supervisory tasks, such as closing the polls.""

Comments

About time...

[Ianoo]

Finally they're getting the message about poorly written voting software. In an critical areas like this proper software engineering techniques should be applied lavishly. It seems Diebold in particular has forgotten how to write secure software, or more likely never knew. For example, why couldn't a voting machine do what airplanes do, and have several separate machines processing the same input data, with software written by several different organisations?

Go Ohio!

[HMA2000]

Woohoo! That's my state!

BTW did we ever figure out why a freaking peice of paper and a pencil is no good?

Paper and Pencil:
Auditable: Yes
Easy to verify: Yes
Hard to counterfiet: Yes
Fast Tallycounts: No
Time in use: Predates recorded history (almost)

Electronic voting:
Audiable: depends
Easy to verify: depends
hard to counterfiet: no
Fast Tallycounts: Yes
Time in use: A couple of years

I don't understand what the insane rush to e-voting is all about. (not flamebait) I suspect it has something to do with the hissyfit that certian elements of the political spectrum threw when GWB "stole" the election.

Re:Go Ohio!

[sfjoe]

I don't understand what the insane rush to e-voting is all about. (not flamebait) I suspect it has something to do with the hissyfit that certian elements of the political spectrum threw when GWB "stole" the election.

Having concerns about minorities being removed from the registration lists is hardly a "hissy fit".

Re:Go Ohio!

[cgenman]

Counting votes require quite a bit of manpower and are expensive. E-voting could be auditable, verifiable, hard to counterfeit, accurate, and with a fast, cheap tally.

Of course, in a bit of bait-and-switch, the e-voting machines have been none of the above except for fast, which doesn't help unless they are also accurate. Somehow, I doubt the people elected want total verifiability. Once in a position of power, it becomes almost trivial to fake a vote... Or at least bump your numbers by a few hundred dead people.

Re:Go Ohio!

[llefler]

There have been issues with paper votes too. Mis-counts, "lost" ballots, and a growing demand to have election tallies faster.

There is also no reason why such an important part of our political system can't move forward to newer technologies. When appropriate.

One thing I question is why use touch screen voting? What is wrong with this solution? Votes are cast on a piece of paper, just take a marker and fill in the appropriate bubble. Only one machine is required per location. So you have a fast count, and can always verify the acuracy by counting the paper ballots. And it's cost efficient too.

Re:Go Ohio!

[StenD]

There have been issues with paper votes too. Mis-counts, "lost" ballots, and a growing demand to have election tallies faster.

True, but with paper ballots there's something to recount if there's an allegation of a mis-count. Electronic ballots can be "lost" as well, and speed should not come at the cost of accuracy.

One thing I question is why use touch screen voting? What is wrong with this solution? Votes are cast on a piece of paper, just take a marker and fill in the appropriate bubble. Only one machine is required per location. So you have a fast count, and can always verify the acuracy by counting the paper ballots. And it's cost efficient too.

A device like that can catch overvoting, but not undervoting due to a partially filled in bubble. I think an electronic voting device which printed a filled in bubble ballot would be the ideal. The device would be able to display the ballot in the language of the voter's choice (within reason), and ensure that only valid ballots were printed, reducing spoiled ballots. The printed ballot would still be human readable (in English and the selected alternate language), permitting the voter to verify their vote, and allowing for a manual recount if necessary. I would even go as far as to suggest that once the voter confirmed on the device that the printed ballot was correct, the initial tally could be electronic, with the paper ballots only used for recounts and cross-checking of the electronic tally.

Re:Go Ohio!

[killbill!]

Please someone enlighten me and tell me why paper is wrong!

Europeans countries all use paper vote (one paper sheet per candidate / option), and you know what?

One hour after poll stations close, 90% of the ballots have already been counted. Two hours after, 95% have been counted.
While polling stations typically close at 6 or 7pm, almost-final results are announced at the 8 or 9 o' clock news.

All were hand-counted. And I'm not speaking or Luxemburg, I'm speaking of 60 to 80 million people nations like the UK, France or Germany.

Indeed the US has 3 times as many ballots to count. But it also has 3 times as many people to count them.

Now, optical-scanning devices would sure help cut those two hours to 10 minutes. But anything more than that is a/ unneeded b/ unreliable and c/ costly.

Now if only my state would clue in

[R2.0]

I live in the People's Republic of Maryland, and they are bulling ahead despite the fact that THEIR OWN STUDY said it was insecure.

Basic response: "We're confident that the problems will be fixed in time."

Keep in mind this is one of the only 2 states that have implemented UCITA. Why? "Virginia has implemented it, so we need to implement it to stay competitive."

Physical Security

[BrynM]

From the article:

"These software enhancements will be implemented in all touch screen units deployed within Ohio and the process-related questions are addressed in the Diebold Election Systems training manuals," Radke said.

Why is everyone assuming that software is the only problem. It would seem to me that physical security is just as important. Blackwell points it out, but even most /.ers ignore it. The machines need locking cases. All media needs to be locked up tight. Transportation of these devices is, at least to me, the second weakest link in the security chain (Wireless networking is the top and is to me another physical security issue). During transportation, there is little supervision and scheduled arrival times can be fudged. Just imagine, a laptop with drive image and a power source for the target box are all you would really need. Take a snap of the drive and then see how you could manipulate the software via the wireless connection or even figure out how to fake your own connection. Sure it won't work for the current election, but the next one...

The problem is: physical security takes money and genuine work by the vendor. Software, as we all know, can be made to "seem" to work rather easily.

The

[goombah99]

The score so far is:
california SOS requires verifiable voting
Louisiana SOS wants to replace all their ES&S machines, not sure with what yet.
Nevada SOS has called for a public forum to determine if the public wants verifiable voting, he doesn't think its essential for security but recognizes it may have value for public confidence.
New York Assembly passes bill requiring verified voting. Companion bill is awaiting a vote in the NY senate

Russ Holts bill is now up to 84 sponsors some republican. But it's stuck in a committee chaired by Rep Nye (Ohio--- AKA Diebold) so its unlikely to pass in advance of strong public support. On the bright side, the NY times is now covering this not as a technical issue but as a political issue: Krugman has two columns dedicated to it, and the candidate Kucinich put it on his congressional web site. So its moving into the mainstream at least. Hopefully it can not become tainted as partisan.

Also relevant:
NY City's Council (? not sure what they call it) passed a resolution calling for the senate to pass the bill and the Elections's manager supports it too. I should note that the population of NY City comparable in size to many small states so this is no small endorsement.

So the tide is slowly turning.

The GAO, and the Library of congress have both said the problems are formidable and FEC regulations are not yet up to date.

Marylands SIAC report is an interesting case. The govenor redacted about 80% of it so we dont know exactly what was said. We do know that the Elections officials said the problems are all fixed now, thank you. To which the response from Rubin, the John hopkins researcher that anaylsed the diebold code said: "If they are all fixed then you should release the report", and the Election official had to respond: "oh that would be giving a roadmap to how to break in". What the fuck is going on! are these people blind zealots or retards?