Did George W. Bush steal last week's presidential election? It was a given that if he won, the accusation would be made, and it has been. And there's some evidence, at least of the circumstantial variety, to support the claim. While the media falls all over itself to explain why last Tuesday's exit polls were "wrong," the fact remains that in states with auditable vote counts, the exit polling matched the counts to within 0.1%. It was only in states that used electronic voting machines, unencumbered by any "paper trail" which could be checked, that the results inexplicably tilted in Bush's favor by about a 5% margin over the exit polls.
The toothpaste probably can't be put back in the tube. The lack of auditable results didn't just make theft possible -- it makes it untraceable unless some independent corroboration (testimony from those involved in the fraud, for example, or the recovery of physical evidence of tampering) can be found. We will never know who legitimately won last week's election.
What we do know, however -- or at least what I know -- is that votes were, in at least one case, systematically "disappeared." I know it because I watched it happen; if the reader considers my testimony unreliable or non-objective, so be it ... but it is true nonetheless. Eppure si muove, the demands of the Inquisition notwithstanding.
The victim of the vote theft which I witnessed was, as one might easily guess, Libertarian presidential candidate Michael Badnarik. I witnessed it only because I had the distinct privilege of working on Badnarik's campaign team, and specifically on election night, following incoming precinct returns. Here's what happened:
The election night watch party was held at Legends, a sports bar in Austin, Texas, near the campaign's headquarters. While other Libertarians gathered to celebrated and watch "the election results" on the bar's 30 televisions, a few of us undertook an operation of a type which has probably not been possible for Libertarian presidential campaigns in the past. The Associated Press charges $15,000 for a live feed of incoming results ... an expense the campaign was not prepared to pay.
However, the Internet, like 9/11, has changed everything. We set up a whiteboard on which to track Badnarik's performance by state -- his total votes "in the bank," the percentage of the return so far that those votes represented, and the percentage of precincts reporting. And we set up four computers to follow the incoming results via Yahoo!'s display of their AP feed, as well as by pinging the sites of the various Secretaries of State directly. In this manner, we were able to track Badnarik's progress through the evening. Additionally, one Libertarian had created a spreadsheet program to keep running track of overall "banked votes," and to project the final total -- the formula being, of course, (V/PR)*100, where V is votes and PR is precincts reporting. If Badnarik had 100 votes with 10% of precincts reporting, the projection would be for him to have 1000 votes with 100% of precincts reporting.
It is by no means apparent that such a projection formula is automatically accurate for any particular state. A candidate may do well in rural precincts and not as well in urban precincts; one or the other type of precinct may be more likely to report early; he may do better in early/absentee voting, and so forth; and therefore a candidate may be seen losing or picking up votes versus the projection as the vote total approaches 100%. But while the projection formula may not be accurate for any particular state, it probably is accurate "in the round" -- different kinds of precincts report at different times in different places, some states count absentee ballots first and some count them last, etc. On average, the formula is sound. A candidate might see his percentage go down in one state, but it will go up in another, if the same criteria are at play.
What we saw instead was this: As each state hit 20-25% of precincts reporting, Badnarik's "votes added" began to descend. This descent seemed -- I didn't keep running track of each state as the trend didn't become apparent for awhile -- to be fairly uniform across the country, but with a somewhat higher attrition in Arizona, Nevada and New Mexico, states where he seemed likely to affect the disposition of electoral votes. Since the polls across the nation close in one-hour increments by time zone, the effect was staggered. It was probably about 9 p.m. Central time when Badnarik began dipping below 1% in Massachusetts; and 11:30 or so when his projected totals began their nosedive in California.
As of the first precinct reports from the Pacific time zone, Badnarik was tracking at a projected total of about 1.2 million votes. This projection was already beginning to descend; the key time marker in my mind is 9:30 p.m. Central, when campaign manager Fred Collins asked for the numbers, preparatory to the evening's "thanks to everyone and here's how we're doing" speech at 9:45. At that time, the projection was still well over one million.
The descent was rapid, steep and formulaic. Someone, for some reason, was systematically "disappearing" Badnarik's votes from the later precinct counts with two specific goals in mind: To ensure that he affected the disposition of no electoral votes, and to "norm" his vote total to about 400,000. There's really no other way to interpret what happened.
I expected -- and my expectations were not disappointed -- rejection of this hypothesis by some whom I initially shared it with. They pointed out that such a phenomenon seems to imply the existence of a massive conspiracy, the cooperation of hundreds or thousands of individuals, and a sophisticated organization which it would be difficult to keep secret. They were right to point these things out. In reply, I can only quote Galileo again -- Eppure si muove -- and in addendum the words of a character as fictional as Badnarik's reported vote count: "When you have eliminated all which is impossible, then whatever remains, however improbable, must be the truth."