This story appeared on Network World Fusion at
http://www.nwfusion.com/columnists/2000/1120bradner.html

'Net Insider:

Next time via the 'Net?

By Scott Bradner
Network World, 11/20/00

For some reason, voting technologies have been on my mind. The voting systems currently used in the U.S. clearly have some shortcomings.

Many of these shortcomings can be overlooked when the voting population expresses a clear preference for a particular candidate, but their impact is magnified when an election is close. Clearly there will be a lot of attention toward election processes and technologies in the next few years. But how seriously should Internet-based voting be considered?

A few people did vote in the 2000 presidential election electronically over a network. Having a few hundred military personnel vote over a closed network may have been an interesting experiment, but there are more than a few problems with developing systems to permit even a small percentage of the American electorate to indicate its preferences over the Internet.

There are obvious issues of scale, reliability and security. Any Internet-based voting system would have to handle millions of simultaneous users. It would have to be designed to prevent Moscow teenagers from deciding the election. The infrastructure would have to be an order of magnitude more reliable than much of the Internet. In addition, the system would have to ensure voter anonymity while at the same time guarantee that people could not vote more than once.

These problems have direct analogues in the current voting system, but there are some other issues that are exacerbated or created with a move to electronic voting. Things as simple as voting hours and the reporting schedule become major issues. What should be considered "Election Day" on an Internet that spans 24 time zones? If it is to be one day (which could get rather hectic), is it 7 a.m. to 8 p.m. in the time zone where the voter is registered? Or should it be some simultaneous window for the whole country (and for overseas voters)? When should the public be able to find out how the voting is going?

Should the current system of staggered reports continue or should we have a "big bang" of simultaneously announced results?

A major problem for some people would be the lack of exit polls - there would be no way to figure out what special interest to target next time. That would break my heart.

Equal access for people from all walks of life would be hard to guarantee.It is probably easier to design confusing electronic ballots than it is to design confusing paper ones, if looking at the Web sites reporting election results is any indicator.

There are some initial attempts at electronic voting, such as the recent Internet Corporation for Assigned Names and Numbers elections run by election.com. But these are halting first steps at best. So there are no panaceas here - we can look forward to this kind of fun for years to come.

Disclaimer: Since Harvard has a law school, the university does not believe in panaceas - they reduce litigation. But the above observation is my own.

All contents copyright 1995-2002 Network World, Inc. http://www.nwfusion.com