Computerized (DRE) Elections Are Not Trustworthy, p. 2D
II. GENERAL DISCUSSION -- HOW COMPUTERIZED ELECTIONS CAN FAIL.
D) THEY WOULD NOT DO THAT, WOULD THEY?
There is some indication that a variant of the above scheme to miscast electronic votes has already been used:
The electronic voting machine provides balloting information for well known (important) offices and issues. These would be ones which the voter would immediately notice and would complain if the item did not appear. On the basis of a voter's previous selections, the machine may or may not provide
a less well known office or issue. That is, if the voter is expected to vote in the "approved" manner, then the item would be displayed to the voter, and the voter would make a choice. However, if the voter was expected to vote in a "not approved" manner, then the item would not be presented, and the voting machine could select the "approved" choice, but not tell the voter.
If this cheating were done on a random basis, some voters on both sides of the issue would remember voting without a problem. Even if the voter asks to review the ballot at the end, hidden choices might not be presented. If the voter specifically asks to return to the selection area, the voting machine could then allow the previously hidden item to be viewed and voted upon. Some voters would catch the cheating, but they would have to admit they ultimately got their choice. Some voters would NOT catch the cheating, but they would later have only a hazy recollection of not voting on a specific item.
On a long or complex ballot, this method of cheating would be particularly successful. People would forget which way they voted on a specific item. In this case, even a voter-verifiable "printed receipt" version of the ballot (then kept by the voting machine) would be long and complex enough to successfully cheat a significant number of voters. If the voter had a "sample" ballot, previously filled out by the voter, which appeared identical to the one printed by the computer, then the voter would have a chance of catching errors. However, at least some California electronic ballots have been substantially different than "sample" ballots. But this would be only a partial cure -- the issue of a proper "secret" ballot is still an open problem.
Last modified April 09, 2007 05:34 PM