If you haven't read my recent article on open source voting systems, take a look at it. Since I finished the piece, there have been more than a few examples of serious problems with electronic voting machines across the country. They've been attributed to user error, miscalibration, and a whole host of other reasons. While it may certainly be true that there are people who cannot touch the right part of a screen with their candidate's name, I have a hard time believing that these people can also function in our society to such a degree that they are not only voting, but voting early. If you can't place your finger in the right place on a touchscreen, how could you possibly use a computer, a TV, an ATM, a cell phone or even a gas pump? Some reports are saying that the screens themselves are misleading, and what you appear to be touching on the screen doesn't match up with what you should be touching.
That one blows my mind. It's a touchscreen, for crying out loud. They've been around for decades. I've seen waiters, waitresses, and checkout clerks fly around point-of-sale touchscreens at the speed of light. It's not challenging to use, and it's certainly not challenging to code. I have a really hard time believing that some touchscreen electronic voting machines are so poorly designed that this scenario is possible, much less actually occurring.
[ For more on how technology is reshaping the race for the U.S. presidency, see InfoWorld's special report. ]
As for miscalibration, I suppose that's possible, but miscalibration to such a degree that the vote consistently switches over to the other parties' candidate? Doubtful.
Fortunately, calibrating a touchscreen should take all of 15 seconds, so it can't be a big deal, right? Unfortunately, it appears to be quite problematic.
We have another week until the actual election, and it's going to be a wild ride. I expect to hear all sorts of voting fraud stories, from tales of broken machines to vote flipping to purged voter databases to corrupt memory cards, tornados, locusts, whatever. It'll nicely fill in large portions of the 24-hour news cycle, and possibly even bring elections results into question. Again.
One aspect of electronic voting that I didn't touch on is Internet voting. I've had several people e-mail me about this, so I'll take a moment to discuss it here: It shouldn't happen -- at least not in the foreseeable future.
There are only a few ways to even approach the reality of Internet voting, and they all require national ID cards, and all are open to widespread fraud, or at least the possibility of fraud is much greater. It's one thing to lock down voting machines and have them communicate over encrypted connections; it's a completely different thing to enable users to vote from their laptops in their living rooms.
The only secure scenario that I can envision where this would be possible is if everyone had a form of biometric authentication available, whether it's a retina scanner, fingerprint reader, or whatever. Of course, this would require the government to have that information as well, which isn't desirable for a whole host of other reasons. Also, there's no way to verify that the voter isn't being coerced when they cast their ballot, unless there's a video of the event taken during the voting process. Even then, it's not foolproof.
For these and a whole host of other reasons, Internet voting is not viable, and may never be viable. Not surprisingly, this has less to do with technology than human nature.
The middle ground between mechanical systems and Internet voting is electronic voting, and there's no reason why it can't be viable, secure, and foolproof. The fact that this is even an issue in the United States is terribly embarrassing. We are arguably the most technologically advanced society ever to exist on Earth, yet we can't seem to create something so simple, yet so blindingly essential.