November 02, 2003

On voting machines

The Register-Mail published an AP article today about voting machines. I wrote the following letter in response (no word yet on whether it'll be published):

As indicated in "Voting machines debated" in Sunday's paper, touchscreen voting machines are not the panacea that many people seem to think, and may not even solve some of the confusion problems we're trying to solve---differing viewing angles may cause people to try to touch one name, but be recorded as touching another. (Ever been to an ATM where you weren't sure which button the "press this button for withdrawal" arrow was pointing to?)

Optical-scan voting systems seem to be the best answer to a number of requirements---legibility, ease of use, lack of confusion, and recountability. As much as we value the convenience that computers provide, these systems are probably a better bet (and a better buy).

If you are interested in computerized systems, however, you should be looking at touchscreen voting aids. These would be machines that provide a computer interface to the process, but in the end, print out a paper ballot that the voter can read---and verify---before submitting it for an official tally. Of the companies currently marketing computerized voting systems, only Avante produces anything meeting these vitally important accountability requirements.

In this sort of system, computer security concerns are minimized, because even if someone cracked into the system to tamper with it, the voter would notice the discrepancy before handing in their ballot. Furthermore, the existence of a paper ballot enables a recount to occur later if the result is in dispute; recounts are simply not possible with most companies' systems, including those from Diebold and ES&S, two major players in the industry.

What if the system has a bug in it? Many of us can't even trust our computers to stay up without rebooting for more than a day or two. It is a scary thought that we might blindly entrust the entire democratic process to machines whose performance can't even be verified. If we at least require our voting machines to leave a paper trail, voting fraud---not a new phenomenon, just in a new guise---can be detected, the offenders prosecuted, and the results corrected.

I would direct interested readers to a much more complete analysis by Dr. Rebecca Mercuri on the problems of electronic voting, along with some proposed solutions, at the following URL:

On The Time Machine (the 2002 movie): "It starts out slightly silly, then gets annoying, followed by predictable, unintentionally humorous, ridiculous and meaningless, followed by long periods of bad. Then it wraps itself up in blatantly lame with spouts of mere badness and disgust and misery." --The Self-Made Critic

Posted by blahedo at 11:03pm on 2 Nov 2003
Valid XHTML 1.0!