15 March 2009

Still not ready for prime time

It's been almost 2 years since Elizabeth and I first wrote this article, and it is sadly more topical than ever. Until e-voting has a standard of usability, and open source accountability, this will stay an anti-democratic development. E-voting has the power to increase democratic participation; unfortunately as it is currently being abused it is still not ready for prime time. --Jonathan Arnowitz

Not to mince words: November 2, 2004 was a low point in American technology. Let’s get this straight: There is a current system that is the lifeblood of democracy, voting. There is a new technology, e-voting, that is being deployed, which has not been thoroughly system tested, not checked for security bugs, not even tested with the wide variety of users who must use the system (many of whom are computer illiterate, more of whom are computer skeptical). If this technology fails there is no way to determine the scope of its failure: There is no paper trail to trace down the existence of a problem. And to add insult to injury, one leading system vendor’s machines can be cracked with a secret two-digit code. (Though we don’t think you can call a two-digit code very secret or difficult to crack.)
To make matters worse, the leading vendor of e-voting machines, Diebold, Inc., has been unashamedly activist-conservative. Their methods have been shoddy at best, and at worst, not in good faith. The Cleveland Plain Dealer reported on August 28th that Walden O’Dell, Chief Executive of Diebold, Inc., who became active in U.S. President George W. Bush’s re-election campaign, was quoted saying he was committed to helping Ohio deliver its electoral votes to the sitting president. Several other publications note similar qualms about the Diebold system, including Federal Computer Week, ComputerWorld, Investor’s Business Daily, and others, although an inspection of partisan publishing will, of course, surface opposing views.
Complicate system problems with human decisions and you have, shall we say, an interesting twist. In Santa Clara, California, where voters were given a choice (“paper” or “digital?”), users suspicious of the touch-screen systems opted to cast their ballots in the more traditional physical way. No problem ... until it became clear that these paper ballots were frequently placed in pink “provisional ballot” sleeves, introducing another layer of fragility on the business of counting votes.
How do smart-election officials make bad decisions? Simple, we say: Rush into a technological push before basic security standards, let alone ethical standards, can be developed. If a software company attempted this, the entire development team would be rightfully out of a job.
Yet, this scenario has just played itself out in the U.S. presidential elections. The voices of HCI professionals in this e-voting debacle have been loud, strong, vocal and completely ignored. On reflection: At least we know where we stand.

—Jonathan Arnowitz and Elizabeth Dykstra-Erickson

This is a draft version (preprint) of material which was later published in substantially the same form in my Rant column in ACM’s magazine. The published version is a copyrighted feature of Communications of the ACM. All requests for reproduction and/or distribution of the published version should be directed to the ACM.