The Help America Vote Act of 2002 (HAVA), Section 301(a)(5), requires that the vote-counting error rate of each voting system used in federal elections must comply with the 2002 Voluntary Voting Systems Standards. UPDATE, May 5, 2008. It has come to our attention that the 1/500,000 error rate previously quoted here applies to the testing process. For operation in an election, the allowable error rate is 1 error in 10,000,000 ballot lines (0.00001%).

Here are some examples of voting system error rates logged in elections since HAVA was enacted, yet no federal agency has taken action to stop the use of these systems. In fact, all these systems are still in use in the United States.

Maximum legal error rate:	0.00001%
Diebold (Premier) GEMS 1.18.19 error rate: Humboldt County, CA. Nov. 2008 64,161 ballots; data for 197 ballots deleted from system.	0.31%
Sequoia 400C optical scanner error rate: Palm Beach County, FL. Sept., 2008 102,746 ballot; 756 errors in 1 contest, with 2 ballot lines.	0.37%
ES&S InkaVote Plus ballot-marking error rate: Los Angeles County, CA. June, 2008 1 ballot; 4 errors in 13 contests, with 35 ballot lines.	11.43%
ES&S iVotronic (e-vote) error rate: Faulkner County, AR. East Cadron B. May, 2008 57 ballots; 57 errors in 1 of 5 contests.	11.11%
Sequoia AVC Advantage (e-vote) error rate: Camden County, NJ. Pennsauken Dist. 6. Feb. 2008 One machine, 21 votes, 20 ballots.	0.68%
Sequoia AVC Advantage (e-vote) error rate: Bergen County, NJ. Engl.Cliff.Dist 4. Feb. 2008 One machine, 106 votes, 105 ballots.	0.12%
Premier (Diebold) GEMS v.1-18-24.0 error rate: Butler County, OH. March 2008 105 entire ballots uncounted.	0.08%
ES&S Optech IIIP Eagle scanner error rates: Milwaukee City, WI, Ward 43. Nov. 2004 Milwaukee City, WI, Ward 44. Nov. 2004 Milwaukee City, WI, Ward 98. Nov. 2004 Entire ballots uncounted in each case.	0.43% 0.10% 0.43%
Diebold AccuVote OS scanner error rate: Germantown Village, WI, District 1. Nov. 2004 Eleven entire ballots uncounted.	0.11%
Hart InterCivic Ballot Now scanner error rate: Yakima County, WA. Precinct 3301. Nov. 2004 One machine, all five contests in one column uncounted on 24 ballots.	0.18%
Sequoia AVC Edge touch screen error rate: Bernalillo Co, NM. Precinct 558-early voting. Nov. 2004 More votes than voters (phantom votes) in all 37 contests.	29.49%
Sequoia Optech 4C scanner error rate: Dona Ana Co, NM. Precinct 106-absentee. Nov. 2004 More votes than absentee voters in 16 of the 22 contests on the ballot. (For the precinct as a whole, a total of 65 more presidential votes than the number of voters registered in the precinct.)	23.57%

Many other news stories tell of equally shocking error rates. Though we have not calculated the error rate per ballot line (choice on the ballot), it's clear that these miscounts are far beyond the legal limit of one error in 10,000,000 ballot choices.

For example:

	Phillips Co, AR. 2006. 432 of 2,011 ballots had "mistakenly been counted as Republican ballots, effectively nullifying them." (Thanks to ES&S programming of the iVotronic touch screens.)
	Pottawattamie Co, IA. 2006. A machine count of the absentee ballots showed Sciortino 79, Duran 99; but a subsequent hand count gave the correct count: Sciortino 153, Duran 25. (Thanks again, ES&S.)
	Johnson Co, KS. 2002. Results were misreported in six races. "The system miscounted hundreds of votes, and a re-count was ordered." (Diebold admitted it was a software error.)
	Sarasota Co, FL. 2006. ES&S touch screen voting machines reported 18,000 fewer votes than voters in a US Congressional contest won by less than 400 votes. Wait! My mistake. It's impossible to tell whether this was a miscount or not since the electronic evidence has long since vanished into thin air and cannot be recovered.

The list goes on and on and on ... Every election, federal law is being violated in hundreds, maybe thousands, of polling places, and no federal agency takes action.

Worse yet, democracy is being violated.

Furthermore, HAVA Section 301(a)(3), which requires that each poll site provide a means by which voters with disabilities can vote privately and independently, has been the driving force behind the widespread purchase of equipment that DOES NOT COMPLY with the very law it purports to satisfy.

The official report from the California "Top to Bottom Review" of voting systems shows that much of the "accessible" equipment doesn't even meet common-sense standards, and stories from poll sites are heart-wrenching. Yet, the fact that "accessible" equipment is inaccessible is being ignored by virtually all officials. Certainly no action has been taken.

Voting system vendors are taking billions of tax-payer dollars and, in return, giving us inaccurate, inaccessible, unauditable, unreliable voting equipment that counts our votes in secret.

You Can Take Action. It's time to call for accountability in the machinery of our electoral system. Please sign the petition in support of VoterAction's call for Congressional investigations into blatant and destructive commercial fraud in the electronic voting machine industry.

Electronic elections
have electronic records,
which require electronic oversight.
~ Jim March