The Only Surefire Means of Fair Voting

Having worked in computers for my entire adult life, I am of the opinion that computers are not the means by which we can be confident that our elections are fair.

There's been a lot of discussion in the opinion pages about electronic voting lately. Some think it's the best way to facilitate elections, while others are of the opinion that electronic voting is fraught with peril. I happen to belong to the second camp.

Not only can computer voting be affected by fraud, things can also go wrong that no one even thought of, leaving people without a means of voting or people voting for the wrong candidate because the computer programmer(s) made a mistake. I ought to know; I'm a computer programmer. I've written voting systems, and I know how easy it would be to make a mistake in the program as well as to defraud the voters by manipulating the results.

The only way to ensure confidence in our elections is to have paper balloting. It takes a lot more time to count the votes, but in the long run it is well worth it. The paper balloting that we have used in the past, however, is still somewhat prone to fraud. This process should be enhanced.

As it has happened in our recent past (until we decided to get all fancy while lining Diebold Corporation's pockets by paying for their untrustworthy counting machines) each ballot can be individually numbered.

In addition to providing uniquely numbered ballots, it should be the requirement of all voting precincts to
  • provide a carbon copy of each voter's ballot, including the unique ballot ID, to the voter.
  • post in a conspicuous place, such as on the internet, the voting results of each ballot, so that each voter can verify the accuracy of his or her votes, and so that it is easy to determine that all votes cast add up to the tally results that are reported at the end of the election.

In this way, everyone could know how everyone voted, but there would be no way to relate each unique ballot number to the person who cast that ballot, except that each voter would have a copy of how he or she voted.

Every voter would be able to compare his or her vote(s) with the official ballot postings to determine if they were represented correctly. If they were not, the duplicate copy of the ballot could be furnished as proof that a mistake had occurred.

With a great deal of effort, computers can be made fraud-proof in this way. I suspect that all of the necessary measures have not been taken in this regard. Our experience with them thus far has given us a bad taste in our mouths. Paper-ballot voting instills much more confidence in voters that the process is sound.


  1. This could be a much more informative comment if I searched my old documents first to find my research paper from 2001, but I agree that computers are no guarantee of fair elections.

    I did a research paper right after the election of 2000 on the subject of technology and voting and learned some very interesting stuff. My conclusion then was that computers don't necessarily make voting less secure either, but the real key to fair elections lies more in better trained voters and poll workers than in better technology.

  2. Making a unique identifier for each record shouldn't be all that hard considering that it's the cardinal rule of database design. I think you're definitely on to something. With the possibility that any one person doing a verification could expose elections fraud, it would be a strong enough deterrent from trying to do so.

  3. Computer counts are more accurate than manual paper counts so long as the counting algorithm is correct. I like the current Utah system with electronic voting and a paper audit trail. I'm satisfied with a system that lets us use statistical sampling of paper ballots to verify the integrity of the vote counting programs.

    The voting system you propose seems only feasible with computer support. The Scantron pencil voting could facilitate this, but scantron can leave questions about voter intent. Maybe even better would be a computer system that allows the user to vote on-screen, prints a completed scantron ballot that they verify and drop in a box. Counts could be done in both ways: the internal computer tally and a scantron tally using a totally separate system.


Post a Comment

Thank you for commenting. If you have a Google/Blogger account, to be apprised of ongoing comment activity on this article, please click the "Subscribe" link below.

Popular posts from this blog