PLUS: Why Machine-Printed Ballots and So-Called 'Paper Trails' are ALWAYS a No-Go if You Wish to Have Verifiable Elections...
By Brad Friedman on 1/29/2008, 2:54pm PT  

Our friends over at BuzzFlash are featuring an exclusive interview with David Earnhardt, filmmaker of Uncounted: The New Math of American Elections, today.

Disclosure: Though we appear in the film, and Earnhardt advertises it here at The BRAD BLOG, we do not make money on the sale of the film. BuzzFlash, however, does, if you purchase it through BuzzFlash's sales link, which we recommend you do, in case you've yet to see this seminal film. They and Earnhardt deserve the support.

We hope you'll read the interview, as well.

We've got a small, if friendly, nit to pick with both Earnhardt and BuzzFlash's Mark Karlin in regard to one section of the interview in which they discuss the notion of ballots being printed out by machine, as opposed to marked by hand by the voter.

Machine-printed ballots can never be verified, after an election, as actually reflecting the voters' intent, and thus --- except for cases of disabled voters who wish to vote on such devices, as necessary, in order to vote privately and independently --- there is no legitimate reason to use such voting machines. Ever. That is, of course, if such machines even work at all. When they don't, voters can't even cast their votes.

The quick, scientific-study-based bullet points on why such machine printed ballots or "paper trails" can never be known to reflect voter intent follow below for those who'd like to know...

NOTE THESE IMPORTANT POINTS, about DRE voting machines (usually touch-screen), and other similar computerized ballot-marking devices which produce computer-printed "paper trails" or paper ballots, as opposed to hand-marked paper ballots...

  • A number of studies by the CalTech/MIT Voting Technology Project show that most voters (as many as 80%) don't bother to check them before casting their vote. (Diebold's machines even put an opaque door over the "paper trail" display, which most folks don't know to lift up. If they do lift it up, they must then view a tiny cash-register style tape through the magnifying glass --- you read that right --- that Diebold supplies with each machine.)
  • According to a Rice Univ. study, more than two-thirds of voters who do bother to confirm their ballots on the machines at the end of the voting process don't notice vote-flips on the ballots which appear right in front of their face.
  • According to studies by NYU's Brennan Center for Justice, the U. of Calif. and CalTech/MIT, due to the above, it is very easy to hack both the internal numbers and the so-called "paper trail" printout,, such that no count or audit of those "paper trails," even up to 100% of them (if anyone bothered to do such a check), would reveal that the election had been hacked.
  • According to a recent study of DRE machines commissioned by Cuyahoga County, OH after their 2006 election, on at least 10% of the machines tested, the internal numbers and the "paper trails" did not match up.
  • Finally, even if voters bother to check their paper trails (when Caltech/MIT says most don't), and even if they happen to notice any votes flips on then (when Rice U. says most don't), and even if the machines and paper trails are not gamed (as Brennan and U of C says they can be), and even if the DREs manage to count every vote 100% accurately as cast (as the Ohio study says they don't), there is absolutely no way to know any of that after an election on such machines.
  • Therefore, a paper ballot, hand-marked by the voter, is the only way to be relatively certain after an election that one is actually reviewing the actual voter intent, in the event that anybody ever bothers to examine any of those so-called "paper trails" or computer-printed ballots (which almost nobody, in the entire country, ever does anyway).