Solving the wrong problem

eVoting machines are the answer... but what is the question?

In the 2000 election, Florida was on the hanging and pregnant chad it list. All eyes were glued to election workers holding ballots up to glaring florescent lights, looking for what the voter intended. Each possibility was examined in the courts: "What if they punched out the chad for one candidate, but only sortof punched out the other one?" "What if the ballot has a dimpled chad, and the punch didn't go all the way through?"

The list went on and on.

Which brings me to the topic of this post. The fiasco in Florida in 2000 was caused by ballots which did not clearly indicate the voter's intent. In other words, the ballot was not clear and consistent. The question we might want to answer becomes "How do we build a voting system that makes sure that the ballots, when cast, will easily and clearly identify the voter's intent?"

Unfortunately, current eVoting machines do not answer that question. They answer the question, "How can we count votes slightly faster?"

A machine that answers the first question is easy to build. Simply have the voter select the candidates they want, review their selection on the screen, and hit the "Print" button. A full-sized optical scan ballot would be printed out, the voter could examine it, and then put it in the ballot box. No hanging chads, no holding the ballot up to the light. No worries about Diebold, or tampering, or anything.

But hey, at least with current machines, we get the results at 9:30pm, instead of 10:30 or 11. Isn't that worth the trouble?