E-voting has gone from curiosity to production in less than a decade, with the most recent push coming from the Help America Vote Act -- the legislation intended to help fix the counting crisis that followed the 2000 election, in which the U.S. Supreme Court effectively cast the deciding vote.
Next year's election marks the debut of e-voting in prime time. Among the many flavors of new election gear are direct recording electronic (DRE) voting machines -- some are touch-screen units and others act more like ATMs.
Their growing use has attracted increasing scrutiny of their integrity. A vocal chorus of critics warns that the risks inherent in electronic voting place our very democracy at risk.
Indeed, public trust in the electoral process rests on accurate, auditable tabulations of all votes. One that issue, there is general agreement. The point of departure is whether DRE machines can do the job.
Critics contend the answer is no -- not now, not ever. They argue e-voting is riddled with attack points, and the only way to make electronic voting machines safe for democracy is to print a paper copy of every digital transaction.
As critics see it, the paper record is the only record that really counts -- it is a voter verified paper trail in what they call computer-assisted elections.
At the risk of parsing words, computer-assisted balloting is apparently what happens to e-voting when people refuse to get over their love affair with the physical artifact.
For their part, the companies that make e-voting technologies hold out the successful (if limited) administration of elections in nearly 40 states during the last decade. With variation among competing products, manufacturers also point to built-in audit trails, encryption and their system's ability to print receipts. Importantly some systems keep the electronic ballot whole (without a physical artifact) that can be subjected to automated and manual recounts.
Source code for one voting system, which appeared earlier this year on the Internet, was purported to provide smoking gun evidence of e-voting's shortcomings. The documents behind last July's public spat between a researcher at John Hopkins University who analyzed the code and Diebold Elections Systems are a remarkably dissatisfying read. Diebold was "shocked and disappointed" by the allegations but failed to fully refute them. The John Hopkins report was subsequently sullied by the researcher's failure to disclose a financial interest in a Diebold competitor.
Expecting Too Much?
The debate about the future of the polling place ignores how a growing majority of American voters actually cast their ballots -- through the mail. Absentee voters trust their ballots to the U.S. Postal Service, and those ballots are caught up in the flow of the 200 billion pieces of mail delivered by the USPS each year -- much like the Internet routes messages regardless of type.
The Internet analogy is important here. Vote-by-mail is to e-voting what NetFlix is to video-on-demand: They are both future-leaning models that use the conventional mail as a surrogate for a network. Where vote-by-mail is concerned, the USPS delivery network is trusted because it is familiar -- not because it is free of attack points.
Regrettably, without perfect results next year, e-voting itself is at risk. These recently certified systems bring with them unrealistic expectations of perfect accuracy, even as they replace decertified punch-card machines taken out of service because their growing rates of lost votes can no longer be tolerated.
Is 2004 too early for e-voting? Perhaps. Is it a delicious target? Probably. A useful scapegoat if things go wrong? Absolutely.
Should an epitaph for e-voting become necessary, we could use a throwaway line from a recent decision of the 9th Circuit Court of Appeals: "No voting system is foolproof, of course, and the Constitution does not demand the use of the best available technology."
