Electronic voting machines

[damont]

I haven't said this in my blog yet... I firmly believe that (a) electronic voting machines are truly the wave of the future; BUT (b) we have GOT to design a better surfboard. The paper trail for voting can be anonymized... and thus I see no realistic objections to having a real, dead-tree, paper trail. Today I got this URL from non-LJ friends, and I have to agree with them that Bruce Schneier says all this much better than I can. Read this, then lobby your local/state/federal governments accordingly.

The Problem with Electronic Voting Machines

And in case you were wondering, I live in the precinct in Fairfax County cited in his article. My vote was very likely one of those affected.

Comments

Simpler is not better

[sue-n-julia.livejournal.com]

I found his assertions that fewer steps translates into a more accurate tally to be somewhat disingenuous. While I agree with his basic premise - that we need to develop a technology to more accurately and more quickly count ballots - I find his serious undervaluing of human error to be quite alarming. When performing a repetitive task for hours on end, humans lose focus and make mistakes. This too can lead to a very serious miscount in an election.

If a new democracy (say Iraq) were to hold an election in the style that the US uses, it would be decried worldwide as inaccurate. Voting precincts not opening on time, machines that break down (and in areas that historically vote for one party), poll workers with misinformation, campaign workers that deliberately misinform -- the list of non-voting machine problems goes on.

Re: Simpler is not better

[montuos]

When performing a repetitive task for hours on end, humans lose focus and make mistakes. This too can lead to a very serious miscount in an election.

Speaking as someone who spends 6+ hours a day correcting other people's mistakes, I can tell you with absolute certainty from empirical evidence that the more steps there are in a process, the more opportunities there are to introduce error, and that errors which build on earlier errors can become egregious indeed. This is true regardless of whether humans or computers are used.

Yes, humans counting ballots by hand can make mistakes (and as a former staff member of an association, I've done that too, both counting and making mistakes!), but they won't be on the same scale as computer error.

Humans counting ballots is a process that can be repeated as many times as you like for confirmation; if you get the same answer multiple times, it's pretty certain to be right. But how do you confirm results when there is no audit trail?

Also, the honest errors from humans counting will be evenly distributed. Compare that to touchscreens which fail to count one candidate's votes because the sensor was detecting in the wrong area. Humans will not count in negative numbers, and they will certainly not count three times as many votes as there are ballots.

As for dishonest errors? Compare how many votes a few "bought" humans can er, adjust, with how many a single, widely-distributed, hacked program can do.

[Ugh! I meant to make a simple comment, and have written a book!]

Re: Simpler is not better

[sue-n-julia.livejournal.com]

I wasn't suggesting no audit trail. As someone who has worked in the computer industry documenting and testing software, I am only too aware that errors happen. Myself, I prefer optical scanning. When performing a recount it only makes sense to use a method different from the first to tally the totals.

Re: Simpler is not better

[dr-zrfq.livejournal.com]

If a new democracy (say Iraq) were to hold an election in the style that the US uses, it would be decried worldwide as inaccurate.

Seems to me that's the very problem he's trying to help fix -- though he's concentrating on one particular aspect of the problem. I'm not sure I've ever heard of a large-scale real election that couldn't be decried as similarly or even more inaccurate.