I agree with Jesse Ventura that this entire fiasco that is going on right now is a direct result of the punch voting systems with their 2-5 per cent margin of error. Folks on the news have said that the cost of updating election systems was too much and that’s why 31% of voting is still done on the old dinosaurs.
My question is this: If optical systems are cost prohibitive, how come every quicky mart on every corner can afford one for the state lotteries???
BTW, I’ve never heard of anyone suing because the optical readers got their lottery ticket wrong…
The quick answer would be that lotteries generate revenue, whereas elections don’t. Election equipment is largely left up to counties, most of which can find more politically-profittable ways of spending money. Likewise, it’s hard for state legislators to impress voters with how much they’re doingfor their constituents by spending money on voting equipment that gets used two days a year by a minority of the public. And error-rates work for/against all candidates–it only becomes an issue in a very close election. It just hasn’t been a high priority.
There are a number of companies that offer “electronic” voting systems (touch screens, non-volatile memory, etc.) However, the cost is pretty high: Anywhere from $400K to $2M to outfit a typical county.
/hijack/
The real question is why don’t we have instant run-off voting?
/end hijack/
I think there is little public pressure to change the voting equipment (that will change, though). WSJ did a piece on the out-datedness of voting technology some weeks ago.
Maybe, just maybe counties will change their policies because of this election. Or maybe the public will forget this whole thing come February sweeps on television.
One thing is for certain: the winner of the election will not make any effort to help change the system that got him elected in the first place. So if you do want change, it’s going to come from the state, county, or more likely city level.
Well, as MysterEcks pointed out, the equipment is used only twice a year. For the other 363 (364 on Leap Years) days, they put the machines on mothballs. I imagine that you need better storing methods to take care of optical scanners than a few needles.
I have no doubt that people will now realize the benefits of getting more adcanced equipment, though not everyone. In fact, I bet that in a few months, people will be saying “Ah, we won’t have such a close election anymore, so why bother?”
Perhaps a better system of voting, not merely counting is in order. John McCain is much more popular than either G.W. or Gore, but he was knocked off due to the peculiar nature of the Super Tuesday Primaries.
A fairer system could be an approval count, where you have a list of cantidates and mark each one yes or no, as if you were answering the question: “Would you trust this person to run the country?”.
There is also a system known as the Borda count, where if there are for example five cantidates, you rate them all in order of preference, five being your favourite and one being Pat Buchanan.
Either system avoids the perils of vote-splitting, and more importantly, can lead to a happy middle ground compromise cantidate.
I hasten to add that these are not my ideas; I’m poaching them from Discover magazine.
I assume everyone is familiar with arrow’s Impossinbility Theorem. Let me summarize, Arrow hypothesized that any voting system should satisfy five
criteria.
A state is simply a unique distribution of preferences for each individual voter.
Dictatorship. No single voters preferences should be decisisive in all states.
Transistivity. Any voting system that ranks A superior to B and B superior C should rank A superior to C.
Independence of Independent Alternatives. Any voting system whcih ranks A superior to B, should still rank A superior B even if option C is added.
Completeness. A result must be obtained from all possible preference states.
Unanimity. The voting system should pick A over B if every individual’s preferences also indicate a is preferred to B.
Well Arrow’s theorem stated that no decision process (voting system) could satisfy all five criteria if there are three or more avaialable choices. Majority rule will satisfy all but #2. The Borda count will satisfy all but #3. Of course, some folks over time have weakened some of the criteria or challenged them. #3 especially is frequently under the most scrutiny, although it is hard to come up with a decisive reason why it is not a good criteria.
Of course the theorem is not meant to say it is hopeless, but simply that there is inherent problems with any system. Pick your poison so to speak.
[nitpick - well sorta]tretiak’s remarks on Arrow’s theorem include
It is important to remember that this refers to any decision process whatsoever that takes into account ordinal preferences only. That is even if there were some disinterested external observer who knew everyone’s preferences there is no system for “adding up” the preferences that would satisfy the axioms.[/nit]
A couple of additional points:[ul][]IIRC Arrow’s “solution” was to note that if preferences were “single peaked” majority voting would be ok. Crudely speaking this means that on the issue of - say - the overall size of govenment where A is smallest, B is middling and C is biggest, no-one’s preferences are A>C>B or C>A>B (or equivalently everyone’s preferences are A>B>C, C>B>A, B>A>C or B>C>A).[]The application of IIA - axiom 3 - is highly problematic in any real-world political situation.It is not clear that voting is a decision process based on preferences as referred to in Arrow (1951). The (rational) ignorance of the voter and the known unlikelihood of influencing the outcome severs the link between votes and preferences.[/ul]
In all the talk about punch cards and hand recounts, there’s one idea I haven’t heard yet. The punch cards should be counted by people who don’t know which hole corresponds with which candidate. After all, once the card is removed from the booklet in the voting booth, the choices aren’t labeled. I know it’s probably too late to apply that for this election, and with any luck we won’t have to put up with this again. But has anyone heard anything about using a blind (and therefore, hopefully, unbiased) hand count of the punch card ballots?
