Rick, I question the need to ‘let passions cool’ before (re-)raising this issue. Even in the midst of the five-week Battle for Florida, passions never ran so high here that they precluded rational argument. And since we were still in the thick of things, it was easier, then, for those of us involved to locate and link to the thread where we’d just responded to an idea, rather than having to either do extensive searches, or restate our ideas from scratch.
Also, I think you’ve got not one question here, but three:
-
Should we do everything reasonably possible to minimize the opportunity for voter and machine error in the future?
-
Once we’ve done so, will there ever be any situations where hand-examining ballots to discern the ‘intent of the voter’ is better than not doing so?
And, inevitably:
- Did it make more sense than not to consider the intent of the voter in the election just concluded?
Responses:
-
is pretty much a given: of course we should. We should, as you have suggested, have universal standards of reliability for voting systems’ ability to translate voter intent into machine-readable marks. If voters overvote or undervote, the machines they feed their ballots into should inform them of their likely error, and give them the opportunity to correct it.
-
If we do what we seem to have the political willingness to do right now, and fix the mechanics of our voting systems nationwide, it goes without saying that we should drastically reduce the number of elections that are in any sort of doubt by the following morning.
But as long as there is any potential for a combination of human and machine error, there will be ballots that are unclear to the machine. In some cases, the apparent margin of victory will be substantially smaller than the number of such ballots. Should we not look at them with human eyes, to see what they say?
For instance, here’s my ideal voting system for 2004:
At the voting booth, you find a monitor, mouse and keypad, and laser printer. The monitor offers you a choice of candidates for each office; after you indicate a choice with the mouse or keypad, it asks: “You are voting for Joe Shmoe for U.S. Senate. Are you sure this is who you want to vote for?” If you don’t want to vote for anyone, it’ll let you proceed, but only after asking you once if you really don’t want to vote for anyone; if you vote for multiple candidates where that’s not allowed, it reminds you that you can only vote for one, and so forth.
Once you’ve given your voting preferences (or declined to vote) in all contests and referenda, it laser-prints you a ballot with the offices, and the candidates you are voting for, in bold caps. You can look that over, and if it’s not satisfactory, you can start over again. If you’re happy with what it says you’ve done, you feed it into the optical character-reader (OCR) machine, which is the action by which you actually vote.
Specially marked booths would have pictures of the candidates, and give instructions in a variety of languages. Others would have a keyboard to allow write-in votes.
Will this screw up often? No.
Will it screw up sometimes? Yes.
Will there on occasion be something that the eye can make sense of, but the machine can’t? Sure. A printer might print weakly, the machine might fail to read it, but a person might find the printing faint but clear. In this hypothetical case, the “intent of the voter” can clearly be seen, just not by the machine.
But such a system would reduce, by a couple of orders of magnitude, the number of elections worth contesting. And we’d probably never again have a Presidential election hanging on chads, or their latter-day equivalent. With any luck, the system would be reliable enough for us to go to direct election of the President, if we choose. (I leave for other threads the issue of whether we should, or shouldn’t, for other reasons.)
But one can never exclude the possibility of having human inspection of individual ballots being advantageous in terms of increasing the accuracy of an election result, IMO.
- Did it make sense to consider the intent of the voter in this past election? I sure think so, and I think I argued my case pretty convincingly. For instance, in Palm Beach County, we now know that the Gore/Buchanan overvotes voted something like 10-1 for the Democrat in the Senate race. The problem with the ‘intent of the voter’ business was its restriction to a one-ballot-at-a-time approach, and its resulting failure to take the statistical measure of the intent of large groups such as this.
I think I’ll just leave it at that one example, rather than reopen the Battle for Florida. Hopefully, no major election in this country will ever again turn on the sorts of circumstances that our recent Presidential election did.