Update 3/16/2008 Please see the 1st comment from George Cody correcting my impression that he was against the bill.
Yesterday, I testified in favor of HB 5888 recommending some suggested changes. With some improvements the bill would be a big step forward providing an Independent Audit Board, more effective statistics based audits, removes some of exemptions in the current law, and provides for “Hot Audits” close to election day.
The bill text is close to the recently passed New Jersey bill. It can be made clearer and stronger with some additions and revisions from proposed Federal “Holt Bill” HR 5036. It also needs some tweaking in its customizations for Connecticut. It should also remove all of the loopholes and exemptions from the current law passed last year. <my testimony>
HB 5888 also includes a Voters’ Bill of Rights and a provision calling for manual counting in recounts:
Sec. 2. (NEW) (Effective from passage) Notwithstanding any provision of the general statutes, if a recanvass is required by law in a municipality that uses marksense voting tabulators, such recanvass shall be conducted by hand count.
George Cody, representing the Registrars Of Voters Association, Connecticut (ROVAC) spoke later in the day. As expected George spoke in opposition to the bill. Mentioning in particular statistical provisions and the hand counted recanvass. Here is what I said yesterday:
We all know that people can count accurately – that manual recounts can produce results that can overcome any machine counting errors and accurately determine voters’ intent.
It is the burden of those proposing machine recounts to come forward with a detailed plan. It is their burden to prove that their detailed plan is workable. It is their burden to prove that their plan would be reliably accurate.
Although I doubt it, perhaps there is a reliable and workable machine recount plan possible. I have not seen a written detailed plan proposed. What I have heard is speculation of what might be done. Speculation and ideas that are incomplete – Speculation that is inconsistent and unworkable.
I wish we did not have to do manual audits and manual recounts, they are inconvenient. However, the promise of democracy and the necessity of election integrity are more important than wishes and inconvenience.
I will have more to say about hand and machine recounts in the next few days.
George Coty cited <Dr. Shvartsman’s slides> from his forum on Monday 3/10 to the GAE Committee as indicating in Coty’s opinion that our machines are statistically accurate. Coty and others may have seen Dr. Shvartsman’s answer to a question from the committee, aired on CT-N, stating that machines could be used to recount votes.
What was not covered was my conversation with Dr. Shvartsman after the forum, discussing the obvious problem that if an identical memory card is used, programmed from the same source as the first, that any errors or fraud in the original memory card would likely be repeated in the second card. I don’t have an exact quote, but Dr. Shvartsman agrees with me on this point. I pointed this out to the committee chairs, with Dr. Shvartsman confirming.
Also in questioning by the committee , Dr. Shvartsman agreed with the value of 100% independent testing of memory cards, after programming, and before shipment to registrars.