[E-voting] Push to replace voting machines spurs confusion

Catherine Ansbro cansbro at eircom.net
Mon May 9 19:33:02 IST 2005


Anyone know anything about the recent MIT research mentioned here? 
Apparently people were slow to catch errors when comparing a paper 
ballot, but caught more errors when there was an audio playback. (Of 
course, there's no knowing an audio playback would be accurate, since 
you could arrange an audio playback that was the same as voter's intent, 
but different from what was recorded on the paper.) Maybe they used a 
paper ballot that had lots of different elections on it? It would be 
interesting to see the actual study.

At least election officials are worrying because they can now see the 
need for a backup.

Catherine

----------------------------------------------------------------------------

http://www.usatoday.com/news/nation/2005-05-08-voting-machines_x.htm
*Push to replace voting machines spurs confusion*
By Jim Drinkard, USA TODAY

WASHINGTON — Election officials across the nation are scrambling to meet 
a Jan. 1 deadline to replace outmoded voting machines with equipment 
that is supposed to be more accurate. But a controversy over the 
reliability of computerized voting machines continues to cloud their 
decisions.
		Voters use touch-screen machines in Wellington, Fla. Computerized 
systems are to be used nationwide by Jan. 1. 	
Mario Tama, Getty Images

"The people who are trying to get this done at the local level are just 
running blind," said Keith Cunningham, president of the Ohio Association 
of Election Officials. "I hope there's not a 'train wreck,' but that 
term is being used quietly in conversations among election officials."

There is a consensus among state and local election officials that any 
machine that relies on computer technology should have some kind of 
independent backup that voters can use to make sure their votes were 
recorded correctly and that could be used to verify results if a recount 
is needed, said Doug Chapin, director of electionline.org, a non-profit 
organization that monitors election policy.

But election officials are hearing conflicting advice from experts about 
what kind of backup is best. And a new report from a technical standards 
committee, which could be delivered as early as today to the federal 
Election Assistance Commission, takes no position on whether a widely 
favored option — a paper printout — is advisable.

"There is no clarity whatsoever on what the replacement should be," 
Chapin said. For cash-strapped election administrators, "there is 
significant downside risk that you will buy a system that will have to 
be modified later. They are going as slowly as they can, but the 
deadline may wind up forcing their hands."

Adding to the confusion is a new Massachusetts Institute of Technology 
study that found problems with paper backup for electronic voting machines.

In the study, 36 "voters" used electronic machines to pick candidates 
and were asked to double-check their ballots using a paper printout. 
Then they were asked to go through a similar exercise in which their 
vote choices were played back to them by a computer-generated voice 
through headphones. Errors were interspersed in the ballots. Only 8% of 
those using the paper backup caught the errors, compared with 85% using 
the audio system.

Ted Selker, an authority on human interaction with machines who oversaw 
the MIT research, said the study supports observations he made of voters 
using paper backups during recent elections in Chicago and Nevada. "I 
have lost confidence in paper trails," he said.

Ohio, the most closely watched battleground state in the 2004 elections, 
is among those struggling with the decision of what to buy.

In Tuscarawas County, Elections Director Charles Miller is awaiting a 
decision from his elections board by May 24, the deadline set by Ohio 
Secretary of State Kenneth Blackwell.

The county was poised to buy computerized touch-screen machines a year 
ago but delayed the purchase as controversy mounted over their 
reliability. Last year's presidential election was conducted on the 
county's old punch-card machines. Early this year, Blackwell directed 
counties to buy optical-scan machines, which Tuscarawas County proceeded 
to do.

A month ago, the state relented and opened the door for touch-screen 
machines — so long as they have a paper backup — and the county was back 
in limbo.

"They are frustrated about jumping back and forth," Miller said of his 
elections board. "I don't know what they are going to do."

The 2002 Help America Vote Act has provided $2.3 billion to help states 
replace antiquated punch-card and lever voting machines by Jan. 1. Punch 
cards were blamed for many of the problems in the disputed 2000 
presidential election in Florida, giving rise to the term "hanging chad."

But many of the machines remain in use. The problem goes back to a lack 
of federal guidance about what technology works best, said Cunningham, 
who also is election director in Allen County, Ohio.

"The federal government is a year to 18 months behind its own time frame 
but refuses to adjust the deadlines that counties are facing," 
Cunningham said. "We are on the verge of wasting one of the biggest pots 
of money the federal government has ever put forth."






More information about the E-voting mailing list