POST ELECTION AUDIT A FARCE IN ARIZONA
Not unlike the 18,000 missing votes or counties reporting more votes than citizens or vote flipping in Florida, double votes counted in New Jersey, Republicans impounding all electronic voting machines from 27 counties in Pennsylvania, machines unable to be turned on causing waits hours long in Colorado, Ohio, South Carolina, Indiana, Georgia, and North Carolina resulting in polls held open longer so voters could return to vote, sleepovers of election machines in California and Arizona, and I could go on but I'll refer you to www.BradBlog.com, warning of the problems with electronic voting; the post-election audit law in Arizona to manually count 2% of the precincts was made a farce by Arizona's Secretary of State Jan Brewer.
Perhaps I was too optimistic, perhaps I wanted to believe we had made some progress despite my writing about and knowing that nothing really had changed; that the machines were just as flawed, vulnerable to hacking and vote manipulation, that pre-testing the machines would reveal nothing of bugs, vote flipping programs, Trojan Horses, or trap doors in the secret software. Maybe I wanted so desperately for the message of the dangers of electronic voting to have made an impact that even when I personally observed at our Maricopa County Elections Department the transfer of data between electronic tabulating machines would be performed by "thumb drives" upon which it would be easy to insert a virus to infect all the machines and manipulate votes, that there was no security applied to the custody of the thumb drives before, during, or after the vote counting, that even then, I showed up for the post-election audit skeptical but hopeful that it would happen despite the rigid, nonsensical, arbitrary requirement that a quorum of 72 auditors from each major party participate or the audit could not be performed. Even though I knew that almost every conceivable obstacle to conducting a post-election audit of the electronic voting machines was inserted into SB 1557, Arizona's Election Reform Bill, by the Secretary of State Jan Brewer, a Republican dominated legislature, a small-minded power-mongering Democrat, and elections officials who wanted nothing of this burden added to their already overloaded responsibilities; I thought that just maybe because of the outcome of this election with its the repudiation of an illegal war and of the politics of cronyism, corruption, and greed by a coalition of American voters and not just the party faithful or the base voters, that just maybe, this time, we might have a go at accountability. Sadly, it was not to be.
The first day of the audit was spent scrambling for the necessary quorum to even conduct the audit. Needing 72 auditors from each major party the Democrats fielded an initial 80 and the Republicans only 50 by the initial start time of 1:00PM. To her great credit Elections Director Karen Osborne gave a one hour grace period for each party to beat the bushes (no pun intended) for more bodies. For a moment it appeared as if the operatives from both parties had trolled the local public golf courses for retirees. Then the entire staff of "20 and 30 something's" from the AZ Democratic Party Headquarters paraded in as if displaced from a scene in Animal House. Add to that some mothers with sleeping infants in their carriers, a college freshman here or there, some other late comers, and a wonderful lady whose first vote was probably cast for FDR and VOILA! The quorum was exceeded, there would be an audit! That was the first day.
Returning on the second day well before the 9:00AM drop-dead start time to avoid being shut out from the audit, I had expectations of the time and tedium that would be involved in the manual counting and reconciling of the ballots with the machines. This was going to take several days and be painstaking, boring albeit important work. A statistically significant procedure in SB 1557 for counting the ballots had been negotiated vigorously between the obstructionists and the proponents of real election reform in which I participated. My first hand knowledge came by way of being in support during the negotiations and testifying before the legislature as an election integrity activist and advocate for SB 1557. My experience in dealing with some of the obstructionist legislators, the Secretary of State choosing to manipulate the proceedings insidiously behind the scenes , in negotiating toward a value-added audit procedure can best be described as trying to nail Jell-O ® to a wall! The Republican led opposition (SB 1557 having actually been sponsored by a Republican State Senator) embodied an uncanny ability to deny factual evidence presented right before their eyes and to dismiss statistical methodologies universally accepted by the fact-based scientific community. They stubbornly forced their zero sum agenda of "we win you lose or no deal" into the process. The perceived need to have at least something for the 2006 election led to compromises that should never have been allowed. No deal and public exposure of those entrusted with protecting the public, outing them as election reform denying, self-serving, obfuscating, ethically-challenged politicians would have been better! Still, we got what we got so we could proceed. Then, when all was said and done, the bill passing unanimously in the House, with only three no votes in the Senate (all Republican), and the Governor signing the bill as emergency legislation so it could fast-track through the Department of Justice, Arizona being a Jim Crow state, they manipulated the procedures manual so as to make the post-election audit a travesty.
The auditors were divided up in to several rooms in tables of three persons each. The tables alternated two D's with one R or two R's with one D. The atmosphere was collegial and friendly regardless of party affiliation. These were patriots assembled for fact finding not fault finding or partisanship. After taking an oath administered by the Director of Elections everyone heartily agreed that we were no longer members of individual parties. Every person gave over their time and energy knowing that they were to attend for the duration of the counting for it to proceed without being prematurely terminated producing, in effect, no audit. Frankly, I was proud to be among these good people. I don't know, maybe, that's partly why I am so angry that our time and energy was expended in a wasted effort. We showed up to give assurance that this past election was accurately and honestly recorded and, if not, made good. What we got was a deceitfully manipulated counting procedure that so diluted the sample plan as to call the statistical validity of the count into serious doubt.
According to Dr. Tom Ryan, PhD, an engineer and a principal negotiator of the election reform bill from Tucson Arizona, "The Audit Procedures Manual is evidence of a lack of support and interest from the Secretary of State. This manual is supposed to make the procedures clear, removing ambiguity in the law. While it did remove some ambiguity, it did so in a way that would minimize the size of the test, and in other aspects, it added confusion and is just plain wrong." My understanding of the counting procedure as negotiated by Dr. Ryan and the other "conspiracy theorists" and "anarchists" for election reform was that 2% of the precincts, which was rounded up to 24 individual precincts in Maricopa County, would be selected for counting four of the same races in each precinct. Seems simple enough. This was designed to provide enough samples, or ballots, such that a statistically significant number of ballots would be obtained without counting every race and every proposition. If any discrepancies were uncovered, everyone could be assured that the anomalies were not just "glitches" or isolated incidents but actual events or trends requiring attention and remedy. Dr. Ryan confirmed this saying, "The 24 precincts in Maricopa County would be statistically significant if done correctly." Theoretically, then, each table of auditors should have been assigned a precinct and the four races that would be manually counted and compared to the machine generated tape of the results. Yeah, I know, how do I know the software in the machine wasn't manipulated to reconcile the ballots with the tape regardless of the final count! I don't. Nobody does because the counting software is secret ergo one of the fallacies of the post-election audit. Still, ya gotta start somewhere, right? What we got, however, was significantly different than what I expected.
Instead of counting four races selected from one federal race, one statewide race, one congressional race, and one proposition; what we received was one congressional race in one precinct. That was to be our entire assignment and the elections officials were gleefully talking of being done with the entire audit by the end of the day. Clearly, this didn't fit with my perception of reality as we started into our assignment, but, I figured we would receive the other races as we finished this first assignment. As time passed completing the counting I noticed the room becoming visibly absent of the same number of people as began in the morning. We completed our "first" race while others had suspended to have lunch. When we broke for lunch I asked the Elections Director about the sampling plan. That's when I learned the ugly truth of the day. Bumfuzzled by the tedium of counting with extreme attention to accuracy and a bit slow on the uptake in general, I didn't quite comprehend what I was hearing. In my mind something was wrong with what I was being told because it just didn't compute with everything I knew was to happen. So, I asked around to some of the less arithmetically challenged election integrity activists and confirmed what I was told. The sample plan had been radically altered resulting in significantly lower ballots being sampled than had been the intent of the legislation to produce!
Instead of counting four races in each of 24 precincts, the precincts were divided into four subsets of six precincts. Then, each precinct was assigned a single, not four but only one, race to audit selecting from the federal, congressional, statewide race, or proposition. This dilution of the sample plan by 75% had the effect of making a 2% audit a 0.5% audit and statistically insignificant. Slipped into SB 1557 was an additional reduction in the auditing of early ballots at only 1% of those ballots cast. This dilutes the sample plan even further which adds to the farce!
To say that my blood pressure increased significantly, that I literally saw red for the first time I can recall, and that I wanted to effect some corporal punishment on the nearest election official would be an understatement. If not for the circumstance that the audit was being held in the Maricopa County Sheriff's Training Facility and that we were under armed guard, I might not have acted out physically but you can bet I would have found someone in authority to verbally execute some "extreme prejudice" on! To make matters all the worse, I discovered that those people assigned to audit the small number of touchscreen machines were released almost immediately after beginning the audit because by and large the touchscreen voting machines were not used.
Arizona uses the ES&S optical scanning machines that read connected arrows on paper ballots for regular voting. That's a good thing. Our highly partisan Secretary of State used her extremely suspect interpretation of the 2002 Help America Vote Act to mandate the purchase of one electronic voting machine in every precinct statewide even though HAVA makes no such mandate for the use of electronic voting machines. I don't even know how many of the touchscreen machines were audited or what rationale was applied to garnering any statistical significance. It's a moot point with a touch of irony, anyway. Arizona's Secretary of State may have signed off on upwards of $30 million in taxpayer money to supply each precinct with demonstrably suspect touchscreen machines over numerous citizens objections. Here's the irony: The touchscreen machines were almost never used! In the last primary of the 1,142 precincts in Maricopa County only about 100 ballots were cast on the touchscreen machines with only about 150 ballots cast statewide. If the count doubled for the general election, that roughly calculates to about $100,000 per ballot cast by a disabled voter. I can think of a lot of ways to help a disabled voter to vote in private that won't cost $100,000 a person. But wait that's not all!
Election integrity activists in Pima County, Arizona's second largest county with it's second largest city Tucson, battled the emasculated sample plan successfully with its Board of Supervisors, County Administrator, Elections Director, IT Director, etc. and so on, to force the County Attorney responsible for elections to relent and allow an adjusted sample plan approaching statistical significance to be used in Pima County. While I applaud the successes of the proactive activists in Tucson, I can't help but wonder what additional dilution and confusion to the statewide audit was added by having two different counties count the ballot with two different protocols? Even their more accurate procedure which reflected a better picture of reality didn't meet the parameters of the legislative intent from the original legislative language contained in SB 1557.
Oh, and one other itsy bitsy little quirk in the law. Counties can opt out of even performing the audit! Six of which, who didn't care enough about the integrity of elections apparently, chose to do so. Good grief!
So, there I was around 2:30PM with nothing to do and no one to abuse having turned in our first auditing of 469 ballots with an added triple counting of just the ballots to be sure we reconciled with the tapes. My team of three that audited and counted the one precinct with one race, took approximately 5 hours (not including lunch) to complete. If the original sample plan had been adhered to, if the Secretary of State's office had not handed down the edict of a much abridged protocol, I estimate it would have taken no less than 20 actual work hours, not accounting for breaks, lunch, and trips to the sandbox, to perform a statistically significant, value-added audit. From beginning to end, the audit should have taken 3 ½ to 4 days, minimum! Figuring the auditors in Maricopa County were paid $10.00/hour and assuming only the minimum of 72 auditors were employed (there were several more on this day) that would have worked out to $18,000 just to compensate the auditors, not accounting for county personnel, the sheriff's deputies, or facilities usage. Since the total time the longest serving auditor served was less than 10 hours, if all the auditors only worked 10 hours total that equals $7,200, or a cost savings of $10,800. And we all know the language of management is money!
The post-election audit of the 2006 general election in Arizona was indeed a farce! All those good people who volunteered their time, energy, and sincerity were abused, wasted, and effectively disenfranchised of their right to verify the election. The law was abused by a Secretary of State who fought election reform tooth and nail from the get-go (although she delights in taking credit for "The most comprehensive election reform in Arizona's history under my watch!" during her campaign for reelection) in order to prevent the audit of the election, the only check on electronic voting machines available, from ever happening. Well, now we know. Now we know what we need to do to get ready for 2008, what to work for in a change in the law to prevent this abomination from happening the next time. Better yet, let's just throw the machines out and return to hand counted manual ballots. The time and effort to hand count the ballots would be less than all the work going into trying to "test in" veracity. Let's replace the highly suspect, hackable, secret software computers that will never be trustworthy, the redundant audit, and the corrupt politicians and corporations controlling our votes with an engaged electorate who will honestly and accurately count our votes. It might take a little more time to get results but it would be a helluva lot simpler and less expensive!