US Elections: Paper Ballots For California!

In a dramatic late-night press conference, California Secretary of State Debra Bowen decertified, and then recertified with conditions, all but one voting system used in the state. Her decisions, following her unprecedented, independent "Top-to-Bottom Review" of all certified electronic voting systems, came just under the wire to meet state requirements for changes in voting system certification.

Bowen announced that she will be disallowing the use of Direct Recording Electronic (DRE, usually touch-screen) voting systems made by the Diebold and Sequoia companies on Election Day, but for one DRE machine per polling place which may be used for disabled voters. The paper trails from votes cast on DREs manufactured by those two companies must be 100% manually counted after Election Day. DREs made by Hart-Intercivic are used in only one California county and will be allowed for use pending security upgrades.

The InkaVote Plus system, distributed by ES&S and used only in Los Angeles County has been decertified and not recertified for use after the company failed to submit the system source code in a timely manner to Bowen's office. LA County is larger than many states, and questions remain at this time as to what voting system they will use in the next election.

As The BRAD BLOG has been reporting in great detail for the past week since the reports were released, the "Top-to-Bottom Review" had found that all Electronic Voting Systems certified in California were easily accessible to hacking. A single machine, the testers discovered, could be easily tampered with by an Election Insider, Voting Machine Company Employee, or other individual in such a way that an entire election could be effected without detection.

In Bowen's conditional recertification she re-iterated that "expert reviewers demonstrated that the physical and technological security mechanisms" for the electronic voting systems "were inadequate to ensure accuracy and integrity of the elections results and of the systems that provide those results."

The Certification/Recertification documents for each of California's voting systems, including security mitigation procedures and other requirements for use, are now posted on the CA SoS website. The documents, in and of themselves, offer devastating indictments against the security and usability of each of the systems as revealed during Bowen's independent University of California "Top-to-Bottom Review."

Bowen, a Democrat, was elected last November largely on her promise to re-examine the state's voting systems. In an upset victory, she defeated Republican Bruce McPherson who had been appointed as Secretary of State by Gov. Arnold Schwarzenneger. McPherson had been very friendly to the voting machine vendors, allowing for the continued use of virtually every e-voting system submitted for state certification. Several of those systems had been previously revealed to have had grave vulnerabilities and included source code which was in violations of federal voting systems standards and in violation of state law.

The late-night press conference, following Bowen's decisions, was held at the SoS's office in Sacramento at 11:45pm PT, just in time to meet the state law requiring a six-month notification, prior to an election, for changes to certification of voting systems. California's Presidential Election Primary was recently moved up several months by the state legislature to February of '08.

The BRAD BLOG was able to listen in to the presser via a poor cell phone connection out of Bowen's office. As best as we were able to transcribe, these are our notes from Bowen's announcement and the questions from the media which followed...

Direct Recording Electronic (DRE/touch-screen) voting systems decertified and recertified for use by disabled only.
100% manual count for Diebold and Sequoia DREs.
One machine per polling place.

Diebold – Optical scan system: – decertified and recertified only if meets certain requirements.

Diebold TSx DREs – decertified and recertified subject to certain conditions. Only one machine in a polling place for use by disabled voters. Reduces risk of viral attacks that could infect central equipment.

Sequoia – Optec optical scan: – decertified and recertified – subject to conditions

Sequoia AVC Edge I and II (DREs) – decertified and recertified with a number of security requirements including only one machine in each voting location to allow independent access by disabled voters – concern regarding corruption of software and source code

Hart Intercivic – eSlate DRE: Used only by one county – decertifying and recertifying subject to security requirements. Has the least risk of the three systems.

ES&S InkaVote Plus - optical scan: (LA County only) – ES&S ignored my March demand to submit source code. ES&S eventually submitted source code too late for t2b review. Therefore, "I am decertifying the InkaVote Plus without recertification."

Voters are victims of federal certification process that has not done a job of assuring machines are accurate, accessible, secure.

I reject the notion that I should not require changes in systems solely because we already own them. She compares it to a recall of cars....When NASA finds a problem, they don't continue just because they've already spent the money. They scrub the mission and spend the money to get it right. We must do same with elections.

"We have to be sure that our voting systems are secure, accurate and reliable and they shouldn't be used [just because we own them]."

Paper ballots are simply easier to understand by voters.

Had federal testing been done on source code, it could have made a great difference.

(sorry, this is very difficult to hear, doing my best, Emily is now helping...)

I believe that using op-scan systems in the polling place will always be more accessible to the public (because the votes are recorded on paper).

Further down the road the public's clearly expressed desire for transparency.

It is my hope that voting system vendors will, starting tomorrow, start to evaluate the competitive advantage to moving to open source software.

Q [from media] Is there capacity to provide paper ballots for February?

A: I believe so.

Question about why so rushed? Why didn't you start earlier?

On January 8 we began the process of looking at how we could do this.

None of the vendors provided us with the material we asked for within the requested time frame.

Most counties have about 40% voting by mail. The only county that has significantly less is Los Angeles, at 25%. This means counties have optical scan equipment. Some may have to acquire an additional high-speed scanner. Costs much less than DREs.

I tried to be practical about this and find ways that counties could use existing [equipment?] and existing [knowledge?] and so forth.

It leaves us with an original trail that clearly expresses the voters' intent. We can go back and look at it.

Why couldn't you make this announcement earlier?

I probably could have made the general announcement but wanted to provide you with the [legal] documentation to back it up. We had to take all of the info we got at Monday's hearing, work through all of that, we had some very, very good suggestions and thoughts there, and things came in as late as this morning that we were still going through. "It would not be shocking for us to be sued, so we had some pretty good lawyers..."

Q: "You expect to be sued?"

"It wouldn't be shocking."

Q: Should the prior SOS have done a top-to-bottom review?

A lot of the information that made it clear that this was necessary wasn't available then. Talks about Princeton, Johns-Hopkins studies, Florida review last week, etc. Just in the last two weeks two major additional reviews. This is like many other things where info starts to come in slowly, more people get interested and start to look at it, info starts to come in more rapidly.

As it's now 3:30am in Texas, we will likely call it a night from here, but welcome readers to post additional information from press reports and/or the decert/recert documents...

© Scoop Media