A Release by Michigan Election Reform Alliance
(Lansing, January 8, 2014) — The credibility of Michigan voting results is endangered by a system relying on aging machines utilizing unreliable technology. These machines produce tabulation error rates large enough to change election outcomes, according to a report from the Michigan Election Reform Alliance (MERA).
“Facing Michigan’s Election Cliff” cites widespread machine breakdowns along with MERA-conducted random audits showing significant tabulation error rates. It recommends the state transition to a “more transparent, accurate, and verifiable tabulation system” for future elections, possibly even returning to manual vote counts.
As an interim measure, MERA urges “implementation of a program of random hand count audits to verify the accuracy of machine-produced results.”
“The costs of doing nothing are high, while the costs of addressing the problem are small in comparison, with a high rate of return. Investing in a dependable, efficient vote tabulation system would result in savings of real dollars, an enhanced public perception that Michigan’s elections are fairly and equitably administered, and the certainty that our elected officials are indeed the ones for whom the majority voted.”
Michigan voting relies on optical-scan tabulators. Physical limitations of the technology make the system’s accuracy unreliable, subject to paper jams and misreads of ballots.
MERA has conducted sample audits in both the 2008 and 2012 general elections. Both were presidential elections which have the highest voter participation. The 2008 audit of state election board results in 17 precincts found machine error rates of 0.09% to 0.48%, with an overall average of 0.26%. Three precincts from the November, 2012 general election showed discrepancies of 0.33% to 0.45%. (More than a dozen races at the sate and county level were decided by a margin of 1% or less.)
“This is not a partisan issue,” noted MERA Statewide Coordinator Jan BenDor. “Everyone has a stake in having the most accurate election counts possible. The current system simply cannot be trusted to tell us the true winner in close races.”
The report also documents large numbers of machine breakdown reports during elections. In 2012, there were at least 783 documented service requests for tabulators. Each service request adds expense and uncertainty to the process. Adding to the confusion and expense has been the introduction of electronic poll books which have replaced printed precinct voter registration lists.
“The 2012 voter hotline reports included numerous complaints about delays caused by failed” electronic poll books, the report notes.
Current voting systems were purchased utilizing federal funds under the Help America Vote Act at a cost of more than $43-million. Replacing the old equipment would cost an estimated $65 million (including maintenance contracts).
MERA reports that “private meetings on replacing election equipment have been taking place among the Bureau of Elections staff and a small group of hand-picked Clerks. These meetings have included visits to “warehouses” where participants are given sales
demonstrations. Such meetings do not pass the smell test for an equal opportunity procurement procedure, and can hardly be considered a public and transparent decision making process.”
MERA urges the state instead institute an open, public process to considering new vote counting approaches. The system must combine accuracy, security against fraud and manipulation, generate an audit trail, and allow for manual recounts.
MERA suggests a return to hand-counted elections would be more accurate, and also cost significantly less than replacing expensive equipment with newer machines.
The Michigan Election Reform Alliance (MERA) is a non-profit, non-partisan, pro-democracy, grassroots organization dedicated to the realization of election processes that consistently uphold the principles of democracy to ensure the confidence of voters and maximize representation of all citizens of the United States of America.
The full Election Cliff report, is available at: