[EM] First U.S. Scientific Election Audit Reveals Voting System Flaws but Questions Remain Unanswered

[Abd ul-Rahman Lomax]

(To repeat what I wrote in my last post:) Kathy Dopp wrote me off 
list with a helpful suggestion, but I responded as if she had posted 
to the list. I expect that she will not mind if I also post my 
responses here, since I'm seeking broader comment.

At 11:41 PM 5/9/2007, Kathy Dopp wrote:
>Bruce O'Dell has posted several informed discussion of the drawbacks
>of the voting system you favor.  Contact him or google for his
>postings - although I'm not sure if his postings have only been via
>email.  He's with Election Defense Alliance.

I looked and found material about Mr. O'Dell, but nothing on the 
point of making ballot images public. There is some material about 
so-called "ballot images" that are really images produced by voting 
machines from machine votes, they are not ballots.

I'm talking about paper ballots, which are either marked manually by 
voters or, either for all or for some, they are printed by a computer 
printer according to form entry on the computer. This, by the way, 
could also use ubiquitous technology and, as a plus, if vote-buying 
were considered a problem, it could be required that all ballots be 
produced in that way.

In any case, what I'm suggesting is that after ballots are deposited 
and before they are counted, they be immediately serialized, 
presumably with the precinct number and a sequence code. They would 
then be open for imaging by observers. This could be as simple as an 
observer having an opportunity to look at, but not touch, the ballot, 
and to photograph it with a digital camera. In the original idea, the 
observer could scan the ballot as well, perhaps with a hand scanner 
attached to a notebook computer, or feed it through a fax machine.

It should be possible to scan, photograph, or fax ballots much faster 
than they ballots can be cast. I would think that a single precinct 
would not keep a fax machine busy.

The ballots would be under continuous observation at this point. For 
the official count, they could likewise be faxed to a computer, and 
then they would be locked up. They do not need to be seen directly to 
be counted. They exist as physical evidence if there is a challenge 
to the counting process. But with multiple observers taking their own 
images, plus the public process, those ballots are going to be 
actually needed only if somebody has tried to manipulate the images, 
and a credible claim is filed. Note, also, that physical loss of the 
ballots after this point would not lose the information. The 
congruence of multiple testimonies as to the ballot images would 
probably be sufficient legally to prove the information on the lost ballots.

So, really, I'm first of all interested if anyone else has even come 
up with this idea, let alone has written a cogent criticism of it. So 
if anyone knows where such can be found, I will repeat, I would appreciate it.

What this really involves is simply allowing the public to see the 
ballots, and if the public can see the ballots, the public can verify 
the election itself, it does not need to depend on elected officials 
for that. The officials do their job, they report the official 
results. But everyone is looking over their shoulder, not just a few 
selected observers. The selected observers are the ones who image the 
ballots, and the process records what they see, making it possible to count it.

It is not actually a major change. But it could result in a major 
improvement in public confidence.