Computerized Elections: Appropriate use of Technology?

Today I have a round-up of a few of the many recent news items about elections, technology, problems, and trust — from around the country. But first I want to put them in the context of what I think is a fundamental question in common to these and many other items of election tech news: what is the appropriate use of computing technology in elections?

Even stated that simply, this question falls afoul of buzzword storms. I just have to try to cut through some of that. The buzzword that has been making the rounds most often most recently is “digital voting.” You’d think that we’d be pleased about that, given that is it half of the name “Open Source Digital Voting” of the Foundation that supports the TrustTheVote project. But the problem I have is that “digital voting” has a sort of high-tech, 21st century feel to it, that people connect to cutting edge technology like registering to vote using an iPhone.

So let me use a different term — computerized elections — with a more 20th century feel that’s appropriate because today’s voting and election tech really is 20th century stuff: old fashioned databases for voter registration, decades old software for scanning paper ballots, and touchscreen voting machines running on hardware that in some cases was new at the time when “video game” meant Pac-Man in a student lounge around the corner from where I was programming DEC Vax computers.

This is the technology for computerized elections, which the TTV project is working to provide a much improved upgrade — but not an alternative means of running elections. We talk to elections officials who use computers to count ballots or manage election data, and want to continue doing so in pretty much the same way, but with vastly improved tools, and without vendor lock-in and brutal service contracts.

With that in mind, back to the question, modified: given that many election officials already run computerized elections, and that some of computers don’t always work so well, what is the appropriate use of computers in the conduct and administration of elections? Today’s round-up provides several good examples of this question in many forms:

• NYU Brennan Center’s Larry Norden on NPR: election tech glitches keep happening and “can shake people’s faith in their voting systems” – an example of creaky technology undermining trust in the election process.
• NYT’s Ian Urbina Fraudulent Voting Re-emerges as a Partisan Issue reports on the kind of ugly politics that arises in part because of lack of transparency of creaky voter registration technology and process.
• MIT’s Charles Stewart quoted by USA Today’s Richard Wolf: “there’s probably another million lost votes to be had by making ballots less confusing and by making sure that the boring stuff of election administration is taken care of” shows how creaky election administration technology may be reducing voter participation.
• Nevada voting machine glitches rouse suspicion of incumbent candidate pre-selections, and Texas voting machine glitches spark claims of “vote-flipping” and incorrect election results, while other glitches in NC spark claims of incorrectly programmed voting machines favoring one party. All of these stories show how creaky voting machine technology reduces trust in elections, whether or not these vote-stealing claims are actually true.

Again, all these news items are about really trailing-edge “computerized elections” technology rather than today’s experimentation or imagination with futuristic “digital voting” technology. What we’re up to in the TTV project is developing updated and open technology for these same functions of computerized elections today, in order to help election officials increase trust, without changing the way that they want to conduct elections.

— EJS