PRESS RELEASE from Ida Briggs www.invisibleida.com
Sunday, November 07, 2004
For immediate release
NEW HAMPSHIRE COMPUTER VOTE TALLIES STRAIN BELIEF, SAYS EXPERT
Computer-counted vote tallies in New Hampshire for Tuesday’s election tend far out of the range of common sense, while counties without the computers follow a more plausible pattern, reports a statistical expert who analyzed the data. Voting activists and respected scientists are reporting similar disparities in other states that used computer counting. In response to the suspect patterns, Ralph Nader’s campaign formally requested a hand recount in the state.
“It’s not uncommon to find wards [precincts] here and there that fall outside the trend by a few percent one way or the other,” explained Ida Briggs, a software consultant with 20 years of computer experience including statistical and database consulting for major corporations. “But it defies common sense that votes in computerized counties – be they north, south, urban or rural – would be out of trend 78% of the time.”
Briggs determined the statistical trend (that is, an equal or increased percent differential between Bush and Kerry as compared to Bush and Gore) by taking into account each ward’s voting history and increase in voter turnout, as reported on the New Hampshire Secretary of State’s website. The trend generally holds true in wards where ballots were counted by hand: only 8% of ballots that tallied out of trend came from hand counted wards, a figure within expectable margins, even though 21% of the state’s ballots are counted by hand.
In comparison, 78% of ballots from out of trend wards were tabulated by Diebold’s AccuVote system. Since AccuVote is used mostly in large wards, it is used to tabulate two-thirds of New Hampshire ballots, which means that over 200,000 votes – nearly a third of votes cast in the state – were cast in the AccuVote wards with questionable, off-trend counts.
The Optech system, from Election Systems & Software, tabulates 12% of the state’s ballots, but accounted for 14% of the ballots from off-trend wards, for a total of 35,839 questionably-counted Optech votes.
“This is not a complex formula from some ivory-tower statistician,” said Briggs, who lives and works in Michigan. “Anyone who sits down like I did and carefully compares the vote totals between 2000 and 2004 can see the pattern.” Briggs emphasized that her study is not based on comparing vote tallies to exit polls.
Briggs joined with thousands of voting activists to press presidential candidate Ralph Nader to request a hand recount to verify the reported tallies match the ballots. The New Hampshire Secretary of State is considering the request. Briggs chose to analyse New Hampshire first because of its small sample size, but is coordinating with other respected statisticians on similar studies in Ohio, Florida and other states.
“It’s a good thing New Hampshire saves individual paper ballots,” Briggs remarked. “Anyone who goes to the supermarket knows that when a can of peas rings up at the wrong price, you go back to the shelf to check it. In my consulting business as well, when report numbers look screwy we go back and check the raw data. Here they can audit the actual ballots to see whether the counting software is accurate or not.”
The unexpected vote tally patterns do not correspond to statements such as, “Republicans brought out a lot more rural voters than ever,” or “Republicans apparently won over far more Democratic wards than ever.” Those aren’t the patterns that appeared.
Instead, the differences between credible and incredible totals fall out most markedly along lines of vote counting systems, with old-fashioned hand counted paper ballots coming out on top as the most accurate.
The analysis is available at: www.invisibleida.com
# # #