[E-voting] Report on the 1% Manual Recount for Special Election Nov. 2005 Los Angeles County

[Catherine Ansbro]

Michael,

A few comments.

Yes this particular report is speaking of optical scanners.  With DREs 
there is /no way/ to know how accurate/inaccurate the counts are 
(especially considering the number of documented backdoors in the US 
models--the current runnning count is over 120.)

Regarding audits and procedures, you're right--and this would have to be 
put in place \i(before) any form of electronic vote-counting (which now 
I would completely oppose).   Ongoing important work in statistics shows 
that it's inappropriate to use an audit of a fixed percentage to 
determine fraud, because it's the sample size that's relevant, not the 
percentage.  In a majority-based system in the USA you'd have to do 
something like a 10% audit to catch fraud with any reliability.  In our 
system, where fraud could be perpetrated using #2 or #3 votes, finding 
fraud could be more challenging.   There is no benefit to using machines 
for counting, when the counting could not/should not ever be assumed to 
be correct.  (And where it's votes in question, we cannot afford to use 
a system that "assumes" ANYTHING.)

In "live" elections using electronic counting--both in Ireland and the 
USA--the public has been systematically lied to as to the "smoothness" 
of the election, problems have been covered up, procedures that were 
supposed to mean something have been ignored (security seals), etc.  
When there's a discrepancy found in the audit it's just ignored, even 
though it indicated the possibility of different winners.  (And in 
Ireland, with no papers to audit, we'll never know.)  The people who 
administer elections have a completely different perspective than that 
of the general public, and we have already seen that in our Irish pilot 
tests: cover-up, PR, don't let anyone know about discrepancies, 
authorize results before the discrepancies are resolved, etc.  (Yet 
these same election officials by and large function with great 
competence when dealing with paper ballots that are counted under close 
public and media scrutiny.)

Relying on procedures which have already shown to have failed does not 
inspire any confidence--nor should it.  We should not be doing 
*anything* with our votes that requires "trust" (and that includes 
transportation of ballot boxes by gardai unaccompanied by election 
observers).

Catherine

Michael McMahon wrote:

> Catherine Ansbro wrote:
>
>> This is an amazing short paper. The small table at the end displays 
>> properly if you go to the link. The implications are enormous--it 
>> shows the weakness of relying on properly carried out audits, since 
>> the results can be ignored, as happened here.
>>
>> It's also interesting to note just how inaccurate the computer counts 
>> were.
>
> The paper seems to be talking about optical scan counting in 
> particular. Also, since mechanical
> counting with a 1% audit/recount has been used in California for many 
> years, I'd say that the
> criticism mainly applies to the particular technology they are using 
> in LA county, rather than to
> computer counting in general.
>
> But I would agree that 1% is too small a sample, and also that 
> well-defined procedures
> need to be grounded in law, to deal with audit discrepancies.
>
> - Michael
>
>
>