E-Voting Issues At Wired

| 1 Comment

There are a couple of good articles on issues surrounding e-voting up now at Wired:

An Introduction to E-Voting
E-Vote Guidelines Need Work

The Introduction covers a number of things that have been brought up around these parts before.

One of the most alarming things from the "Guidelines Need Work" story is this bit:

"One problem with the Diebold code was that it had large, complex multi-logic statements with no comments (from the designers)," Rubin said. "That wouldn't pass this standard."

I know of few professional programmers that would ever let a serious piece of software out of development without proper commenting. The lack of it indicates sloppy work that was likely done by third-rate teams with more interest in getting out the door than getting it to work. Which means there is more than one significant problem. First off, that the code was allowed into use uncommented means that the writers of the original guidelines were beyond incompetent. Secondly, that Diebold allowed such shoddy work out its door indicates that the committee that is purchasing Diebold's services has chosen the worst example possible of "the lowest bidder". The question of whether or not the code was usable is entirely beside the point. More important is the fact that guidelines are being desigend by woefully uninformed political agents with little incentive to do anything other than "looking like they're doing something".

This strikes me as a political-committee created problem being addressed by yet another political committee. By adding paper-trails to electronic voting machines, you now add in the need to stock paper, reload machines, and, most importantly, require that the printout be easily verifiable by the voter. The original problem was that the ballots -- in a design approved by political agents unrepresentative of the entire population -- were often confusing (the troubles with lining up names and boxes in "butterfly ballots", for example). If the new design has to make both the machine and the new paper easily verifiable to someone for whom the paper ballots were confusing, then you've simply added a new level of useless complexity. Why not spend the money on making better paper ballots? The paper trail would exist, less money would be spent, less staff would be required at polling stations, and the question of hacking computers would be eliminated.

Meanwhile, the continued distrust of computer voting that results will certainly undermine voter confidence the only real leap I can see as worth the effort: the choice to vote online. (N.B.Yes, this is rife with problems, but the reduction in transaction costs, the spread of "voting places" to every library with a modem, while retaining paper ballots for those without access to a computer would achieve, I believe, enhancements in turnout that would be worth the cost of supplying something like unique bar-code printouts for each voter so that a quick recount could be achieved in the case of hacking/fraud. I still think the turnout we have now gives us a very good estimate of the true population, but there is certainly nothing wrong with more people voting.)

1 Comment

The record of government in acquiring software is so bad, I think it might be the biggest reason to question e-voting. Whatever the theoretical advantages, it still has to be implemented by an institution that is spectacularly incompetent to do so.

My favorite example is the computerized records retrieval system for the nuclear waste site at Yucca Mountain. It was abandoned for two reasons: (1) it became clear that the system, on which the government had spent many millions of dollars, could never be made to work; (2) software which would do everything required, and more, could be downloaded for free.

Pages

Powered by Movable Type 5.02

About this Entry

This page contains a single entry by published on July 7, 2005 10:30 AM.

It's about RAND, so I have no comment was the previous entry in this blog.

broadband.google.com? is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.