October 25, 2003
Houston's electronic voting machines
Harris County has had electronic voting machines for some time now. Not the crappy Diebold machines, but that doesn't mean that people haven't been voicing concerns about them since before their adoption. Recent events have not done anything to change skeptics' minds.
Birnberg said some voters don't trust the machines because there is no way to prove to them that the vote counts are accurate and confidential.
"If you have ever hooked up a computer, you know they come with glitches," he said.
In the 2002 election, about 25 local Democratic voters complained that they used eSlate to select a "straight party" vote for every Democrat on the ballot, only to find on the summary screen that the machine had not recorded a vote for Democratic U.S. Senate candidate Ron Kirk, Birnberg said. The voters were able to go back and mark a vote for Kirk before pressing the "cast ballot" button, but suspicion was born.
Election workers impounded the eSlates in question, checked them and found no sign of a technical problem, Kaufman said.
She said the voters may have been confused by campaign literature urging them to vote for Kirk and all other Democratic candidates. Choosing the "straight party" option and then marking a single candidate might make the machine "de-select" that candidate.
Birnberg surmises that voters accidentally pressed a different sequence of buttons that somehow canceled some of their votes, meaning the machines -- not the voters -- made a mistake.
The Democratic chairman acknowledged that because of the Florida voting debacle and more recent controversies such as the drawing of new boundaries for U.S. House districts in Texas, Democrats across the nation are on edge about voting systems chosen by GOP election officials.
In contrast, Harris County Republican Party Chairman Jared Woodfill said voters seem satisfied with eSlate and that he has heard no complaints.
I suppose when you're winning all the elections it's easier to feel comfortable with the equipment. Be that as it may, there are nonpartisan questions as well.
"The burden is not on me to say it is flawed. The burden on them is to prove it is safe and they have not done that," Rice University computer science professor Dan Wallach said of electronic voting system manufacturers, including eSlate's.
Election officials test voting machines to see if they have recorded votes correctly, but Wallach said the machines can be programmed to report they have worked correctly when in fact they have not. So when local officials said they have found no evidence of tampering or error, that is not evidence that such a thing never took place.
Like some of his counterparts at other colleges, Wallach says the best voting system would use machines like eSlate only as sophisticated printers that produce paper ballots, which would then be tabulated by a computerized "optical scanner" like the ones used to produce standardized academic test scores. Creating a "paper trail" would increase voters' confidence that their selections are recorded correctly and provide an independent backup record for all votes, Wallach said.
Full disclosure: Dan's a good friend of mine. He's also absolutely right, and I think his quote about the eSlate machines should be the mantra of every skeptic across the country. All they have to do is make their code available to outside experts. I have no sympathy to their wailing about their competitors possibly seeing it. This is too important for that.
Posted by Charles Kuffner on October 25, 2003 to Technology, science, and math
I've got a degree in computer science myself, and the world I work in (go NASA!) is one of anal-retentive flaw avoidance. (And even then, it's never enough).
Diebold's machines are scary. The only non-conspiracy explantion that makes any sense is to assume gross incompetance through all levels of development. Unpaid college interns would have put together something more solid...and safer.
It's insane. (Using Access? ACCESS? Access has it's hands full trying to keep a private CD collection straight....jesus!)
Eslate isn't as well researched, but I think a paper trail is something that needs to be pushed. There's a bill languishing in the House (HB 2339, I think. It's on my blog somewhere) about voter accountability that should be pushed. But it's nice to see this stuff getting front-page (even if below the fold) coverage.
To be honest, I've been bothered by electronic voting because of the lack of a paper trail. I would like if it in addition to the electronic ballot, the machine would print out a paper ballot that could be placed in the traditional ballot box in case of a recount. I don't think my idea is all that original, but I wonder why none of the companies (to my limited knowledge) who make the machines have thought of this.
They have. Diebold made it's money on ATM machines. It's not like "receipts" are new to them.
And the general idea behind a paper trail is exactly as you describe it: The machine prints out a receipt, one the voter can read. The voter checks it, verifies it's correct, then marks the "voting complete" button and places the receipt into a locked bin.
If a recount is requested, the paper receipts are checked against the electronic tally. The House Bill was talking about requires random audits (0.5% of all precints each election) to verify no "silly buggers".
I think it's a good idea, simply because it's difficult to manage large-scale ballot manipulations when you have to stuff actual paper. One precints or a dozen, maybe. An entire state? You'll get caught.
With electronic voting, however, large-scale manipulation becomes very feasible, and very difficult to catch.
This would be a non-issue had we spent as much time educating people how to use a punch card as we do telling them about how to use a condom, how to consult with the HISD appointed Planned Parenthood folks in the schools, how to diversify, how to fight the pledge, how to fight school prayer, and everything else that has absolutely nothing to do with reading, writing, arithmetic, science, and other classes.
Both systems are easy. Both systems are safe. Although, I do understand the resentment in losing. It leads to finger pointing at Big Brother. It seems that liberals are troubled with the Big Brother scare. Yes, the very same Big Brother that they want more and more of.
I feel sorry for those folks who were too busy drinking beer and smoking pot in school to receive a decent education. Now all they want are handouts and entitlements.