Array The Basic Questions We Ask |

The Basic Questions We Ask

Are You Sure Your Vote Will Be Counted?


Consider These facts:

  • Connecticut voting machines are programmed in secret by unidentified employees of voting companies.
  • Each election, each race, and each precinct is programmed separately – each providing a unique potential for error and an opportunity for fraud.
  • It is impossible to review or test computer code to prove it will accurately count votes.
  • The only way to detect errors and deter fraud is to count enough of the paper.
  • Audits passed by the legislature in June 2007 areinsufficient to detect most errors or deter fraud. Many races and questions will not be audited and there are loopholes.
  • We need to audit more races, more quickly, with stronger criteria to trigger full manual paper recounts.

Why should we be concerned with optical scan voting machines?

Our concerns with Connecticut’s new voting machines and audit law are similar to the ones articulated in reports from the University of Connecticut, reports commissioned by the Secretary of the State of California, and elsewhere:

  • The Diebold AccuVote-OS machines can be easily hacked via the memory cards at any point in the process – altered invisibly to do anything conceivable.
  • The weakest point in the process is in the programming of each election, where those that have full access to the cards can do almost anything.
  • Programming of each of our elections is outsourced to Diebold and its distributor LHS. (Note: Diebold’s name is changing to Premier)

Computer scientists everywhere agree:

  • No matter what we do, a computer system cannot be proven to be accurate or tested to be accurate.
  • The only way to assure an electronic election is correct is post-election audits that are effective.
  • Connecticut’s new law, PA 07-194, audits a minimum of three or 20% of races.
  • No referendums or ballot questions will be audited.
  • The maximum probability of detecting an error or fraud in a state wide race is 20%- 30%.
  • In races, such as state representative, state senator, or small to mid size municipal races, the probability of detecting an error fraud is just 2%-4%.
  • In the case of a close or contested statewide race, there will be no random audit in the entire election. This is a loophole.
  • Criteria for counting discrepancies and triggering recounts creates barriers that are too high.

History shows that no matter how secure procedures are on paper, they are often violated, even in Connecticut. In fact, they were overlooked by voting officials and violated by LHS, with no penalty in the Nov 2007, 2nd District recount. An LHS employee substituted an insecure voting machine in his car for one sealed by the state and tested by the town.

Should we vote entirely on paper?

Voting all paper has several advantages. It is simple, which means that we can all understand the process and the equipment involved. It can easily be made transparent; procedures can be put in place so that the counting and voting can be viewed and verified publicly.

Registrars and long term election officials argue against only paper. With technology they can close the polls and go home in an hour or so. They point out that paper does not have such a sterling record and that the other democracies that vote on paper have much simpler ballots with less to count.

Counting all the paper would be safer than optical scan and insufficient audits. Optical scan followed by 100% audits would be quite safe, however, we can be protected with optical scan followed by sufficient random audits.