> Internet Voting in Public Elections
>In recent months interest has been growing in systems for online voting.
>The Secretary of State of California recently appointed a task force to
>study the issues, and its report is available at
>This talk will summarize the technical issues addressed in that report
>and various others: security, privacy, failure tolerance, and availability.
>Dr. David Jefferson is a Senior Member of the Research Staff at Compaq
>Systems Research Center in Palo Alto, CA, where among other things
>he has been doing research on the use of the Internet in the conduct of
>public elections for seven years. He is a Director, and former Chairman
>of the Board, of the California Voter Foundation (www.calvoter.org), a
>nonprofit, nonpartisan organization dedicated to a more informed and
>engaged California electorate, especially through use of the Internet.
Dr. Jefferson began his talk by saying that even before the catastrophe in Florida had caused a flurry of interest in voting systems, the Internet had caused a lot of discussion of the issue. To understand the complexities of voting systems, it helps to know that ballot laws are done on a State by State basis, and in most States they are administered by the Counties. Exceptions are places like the State of Rhode Island, which could be considered to be a big County. Everywhere in the U. S., anonymous ballots are the linchpin of our system.
He then distinguished two basic types of Internet voting systems. The first is one where you vote at a specialized machine, from which the Internet is used to relay the results to the County. The second type, which is the one most people focus their energy on, is the one where you allow people to use their own computers to vote, submitting the results to elections officials from there over the Internet directly.
There are many things that going to the polls give you that are hard to duplicate over a distributed system. One thing is that you have a guarantee that nobody but the actual voter really knows how s/he voted. This is completely different from using the Internet for commerce, in which both the buyer and the seller have receipts and can describe exactly what happened. Protecting of anonymous voting with distributed terminals from large scale interference by software viruses, Trojan horses or worms is a challenge that has not been resolved yet. Another question is what to do when somebody demands a recount. Asking the same query of a database of votes does not seem to have the certainty that checking the ballots gives you.
Another question that Jefferson discussed was registering voters. The way things are now, you have to sign a form you fill out with your information to register. It turns out that there is no government data-base to cross reference registrations submitted electronically against as a check for efforts to stuff ballot boxes. Without a way to be sure that people registering exist, allowing registration by computer is far too vulnerable to fraud for it to happen. Because of people's desire to protect their privacy, he doubts such will become possible soon.
Looking forward, Dr. Jefferson expects that Internet voting from kiosks, both attended by a voting official and unattended, will likely happen over the next few years or decades. However, because of the rapid change in home systems, and the need for voting technology to support all of them or none of them, he is dubious that will become a viable option any time soon.
Fun sound bite: in 2000, in Florida, 82 people voted electronically.