Putting too much stock in the 2006 elections

If something isn't done about electronic voting machines before the 2006 elections, any thought of taking back the House is going to be moot.

A couple of diaries at Kos today reveal some eyebrow-raising voting issues in some 2004 races.

One of them analyzes the squeaker of a gambling referendum in Florida that ended up passing handily once 78,000 absentee ballots were tabluated. It seems that 95% of the absentee ballots were pro-gambling:

As of November 3, the statewide vote margin on Amendment 4 never exceeded 10,000 – ranging from a 6,500-vote edge against allowing slots in South Florida to a margin of 10,000 or so more citizens voting against slot machines.

The amendment looked headed to a razor-thin defeat. Not only that, the 0.1% total state margin would have required a recount of the vote tallies for the constitutional Amendment. (Anything less than 0.5%.) So the vote was 50-50 with 99% of the votes counted, a squeaker.

But a week later, the Amendment was ahead by 93,000 votes. How so?

Well, Broward County found 78,000 absentee votes that had not been counted (absentee votes in Broward are opti-scanned; only the in-precinct Election Day votes are touchscreen). The vendor for Broward is ES&S. Part of the problem was that the ES&S tabulator model used for absentees in Broward could not breach its limit of 32,000 votes without generating a massive numerical error, a newly discovered "counting-backward glitch" (that affected also some North Carolina counties and God knows where else, and that ES&S now has to fix and reprogram for future elections).

Amazingly, in this 50-50 election a miraculous 74,000 of 78,000 new Broward absentee voters voted "Yes" on slot machines. So most voters in the state were split right down the middle, but 95% of the newfound absentee voters were strongly in favor of betting.

Truly, a miracle for betting afficionados and the state Dept. of Education which wanted a dedicated share of the gambling revenues.

Now who was suspicious of these new counts? Well, Paul Seago, executive director of Orlando-based No Casinos, Inc. Also, state rep Randy Johnson (R-Celebration) who was chairman of No Casinos. They maintained doubts about the final approval margin, 119,000 votes, for the pro-slots measure.

[snip]

Johnson checked with statistical experts and he came away believing the disparate vote pattern could not have strayed so far from 50-50 merely by chance, that it would be a 1-in-a-million possibility.

The second aberration was Pinellas County, where a 17,000 vote reversal was recorded and sent off to Tallahassee for final inclusion and certification. The flipped vote-margin remains in the state certified count to this day, even though all parties acknowledge the error. [Translation – Glenda Hood leaned on "regulations" and "deadlines" to forbid the correction of the state number.] The error was not by machine so much as by a paper and pencil reversal by a worker. Staff of Elections supervisor Deborah Clark, a Jeb Bush appointee, learned of the flip on November 4, but 8 days later Clark still sent Tallahassee the wrong results (supposedly approved by 17,000 Pinellas voters instead of rejected by a 17,000 vote margin in Pinellas). "Pinellas elections office knew [Nov. 4] soon after Election Day of a discrepancy in vote totals on a statewide slot machine initiative, but they didn't fully investigate for two weeks." So the false vote totals were maintained knowingly for weeks, until state certification was past.

Uh-huh. And I am Marie of Rumania.

The other one links to this message at Black Box Voting detailing easy hacks to Diebold's optical scan systems that were used in several disputed precincts in 2004.

It's pretty clear that the voting system in this country is broken. The very systems that were supposed to override the kind of partisan manipulation that people had claimed for years are even more susceptible to cheating. And when you have a Republican party that will do anything it takes to win, from destroying the reputation of an opponent to voter suppression, why should we believe that their supporters who make and sell voting machine hardware and software won't manipulate the machines to produce a desired outcome?

When the odds against the 2004 exit polls being so far off are over 10 million to 1, something is terribly, terribly wrong. We can have elections till the cows come home, but if the results are pre-determined, what's the point?