From: Phil deHaan (dehp@calvin.edu)
Date: Wed Dec 04 2002 - 14:35:00 EST
December 4, 2002 == FOR IMMEDIATE RELEASE
Voting systems have received significant scrutiny in recent elections,
including the controversies in Florida in the last Presidential election. So
Calvin College senior Matt Post had a lot to dig into when he made computerized
voting systems the focus of his senior project as a computer science major.
And after months of research he has come to some conclusions about the
prospects for major changes to the ways in which Americans vote. He will share
those findings on Thursday, December 5 at 3:30 p.m. at Calvin in North Hall
253.
"The many problems with voting systems in recent years have given rise to a
great pressure to replace traditional mechanical systems with electronic ones,"
says Post, a native of Marinette, Wis., located about 50 miles north of Green
Bay. "Proponents of electronic systems hope that using computers to manage
elections will eliminate voter fraud, lost or miscalculated ballots, and other
errors, and thus capture the true will of the people, electronically."
Post designed a system that could be used for electronic voting. And he will
demonstrate it in his Calvin presentation. But, says Post, the problems
surrounding electronic voting are myriad and complex. "There are many issues,
social as well as technological, which make achieving this hope (nationwide
secure electronic voting) very difficult," he says.
Even his own system, he says, could not actually be used.
"I focused," he says, "on one piece of the election process: ballot storage on
public networks. In order to actually be used, my system would have to be
integrated with secure means of registering voters, ensuring their identity,
presenting them with the ballot and transmitting those ballots to the storage
servers (where my project comes into the picture). Furthermore, my system is
still vulnerable to many kinds of attacks. However, I believe that these
vulnerabilities are inherent in computers and networking and are not specific to
my system."
One of the benefits often touted for electronic voting is the ability for
people to avoid a trip to the polls by voting from their home or their
workplace. Post calls this the "voting in your underwear" dream. And he says
it's a fantasy not likely to come to fruition anytime soon.
"The internet," he says, "is too insecure. It would be relatively trivial for
an attacker to release viruses that spread via email and infect people's systems
before elections, perhaps preparing those computers to alter votes, or redirect
them to a malicious server which then intercepts the vote and changes it or
blocks it from being cast. This could be done behind the scenes without the
voter ever knowing."
The dangers are not quite so bad with polling-place Internet voting, where the
Internet is used for the underlying ballot delivery structure, but votes can
only be cast from designated polling stations. Yet even there, Post notes,
there is no guarantee that what is being presented on the screen is what is
actually being stored on the computer when a vote is made. Says Post: "You
can't be sure that your vote will actually be stored and counted correctly." He
says the same is true of paper ballots, since someone could later remove a
ballot from the box, but the problem with computers is that this can occur on
such a large scale, especially with computers on a public network.
Finally Post says that while there is a need to reevaluate confusing voting
systems such as butterfly ballots and punchcards it is just as easy to make a
bad and confusing interface with a computer.
"People often seem to believe in technology as a panacea for all of society's
woes," he says. "A well-designed electronic voting system that maintains
privacy, contains an open specification for the computer's code, and maintains
an audit trail of printed paper ballots reviewed by the voter could well bring
benefits to the voting process. But the current trend isn't in that
direction."
Contact Post at 285-9255 or via e-mail at mpost89@calvin.edu
-end-
This archive was generated by hypermail 2.1.4 : Wed Dec 04 2002 - 14:35:17 EST