William Jackson | Vote for anything but digital

Cybereye | Commentary: States should wait on digital voting until technology standards take root.

You could make a good case that direct-recording electronic (DRE) voting systems are unreliable. Researchers at the University of California recently made such a good case that California’s secretary of state decertified a number of systems from use in the state’s polling places. This raises the question: How should we vote?

It is a question being disputed by voting-rights activists, state and local government officials, academics, and those in the voting and information technology industries. We may never reach agreement on a single method of voting in this country, but it is clear that, at a minimum, we need a standard that the electorate trusts and that allows for a meaningful recount, not just retallying potentially flawed results.

The consensus seems to be that there should be something physical involved — something both the voter and the counters can hold in their hands and lay their eyes on — that is often expressed as a verifiable paper audit trail. A number of states are requiring a paper-based audit trail, but there still is a lot of disagreement over the best form of ballot and how to count it.

In Missouri, a group called Show Me the Vote wants to amend the state constitution to require use of hand-counted paper ballots in all state elections.

The simplicity of the concept will doubtless attract a good deal of popular support, especially in the down-to-earth Show Me State. But it is likely to be greeted with reservation by the officials who would be responsible for managing and counting those millions of paper ballots.

But even if you don’t like DRE systems, hand-counted paper ballots are not necessary, said Christopher Wilson, who has developed his own voting system using digital pens and paper. In his blog on votingindustry.com, he wrote that optical scanning to count paper ballots is a superior method, combining the speed and efficiency of electronic technology with the reliability of paper.

No method is perfect, and hand-counted paper ballots certainly have their flaws. Just look at the 1948 Texas Senate race that Lyndon Johnson won by 87 votes in the infamous Ballot Box 13 from Jim Wells County. Some cynical people say Johnson stole that election. But he bought and paid for those votes, fair and square. Vote buying was an old and honored tradition in South Texas, and if Johnson bought more of them than his opponent, Coke Stevenson…well, Coke had no one to blame but himself.

The point is, we are unlikely to come up with a single method of voting that suits everybody everywhere, but we deserve standards that provide a minimum level of confidence in the results. These standards will have to include stringent requirements for the security and reliability of any technology used — be it electronic, electrical or mechanical — visibility into the development of that technology, and some provision for meaningful recounts. At this point, the best way of achieving such a recount seems to be a paper ballot verified by the voter.

This does not necessarily preclude the use of other technologies, such as the touch-screen terminals that are beginning to fall out of favor, as long as they can be coupled with a reliable paper ballot. But state and local governments should not rush to invest in that technology until it is certified beyond the word of a salesman who asks, “Would I lie to you?”