Re: Measurement Error and the 2000 US Election
Topics: Democracy
14 Nov 2000
From: Ervan Darnell
At 01:14 AM 11/13/2000 , Vincent Kargatis wrote:
>On improved voting systems: instant-runoff/ranking voting seems to me superior to any of the others, better than approval, and way better than 1 person/1 vote.
Mote on that later....
>On Ervan's "futility of voting" post - I think the only ridiculous post I've ever seen from Ervan, imo. :) It seemed tantamount to arguing that a grain of sand is weightless because it takes a gazillion of them to weight a ton. The only important facts to me are this:
It's more subtle than that, at least it was meant to be. If you aren't familiar with the prisoner's dilemma, you should check out the links I offered to understand the context of the argument. It is a paradox of sort, if everyone agrees to make the same decision you reach an opposite conclusion than if you evaluate your purely local circumstances. Your analogy to a "ton of sand" sees only the "cooperate" half of the prisoner's dilemma.
The relevant variable is the prevailing preference towards either candidate. If it's any perceptible amount above 0, it's completely pointless *individually* to vote. In the case of FL, it was dead on 0. The mistake I'm confessing to was assuming that 0 was an unrealistic bias in a real race. The rest of the analysis holds. Is there some part of the math you care to disagree with?
I'm sure everyone who failed to vote in FL is now kicking themselves. But ultimately, that doesn't make much sense either. Even with a dead heat like this, the chance of your vote mattering in just FL is around 1 in 1000. Maybe that's enough to matter. If the outcome of the election is worth $10K to you (a plausible, but I expect high figure for most people), then your expected value is $10. If you knew there were a $10 bill lying on the ground at the polling booth and you had to wait in line 30 minutes to pick it up, would you bother? Most of us would say "no", yet most of us vote any way.
For me in CA with a 4% bias in favor of Gore, my vote was not even a grain sand. My vote was not even one virtual photon binding one electron in one silicon atom in one grain of sand. That's assuming one looks at the "defect" side of the prisoner's dilemma. As I wrote originally, it all depends on whether you think are you bound to an agreement to cooperate with other people. It's a dilemma, not an absolute prescription.
But my real point is about the cost of information and how it applies to making complex decisions in a democratic system where you vote is very unlikely to matter. Even if people vote, no matter how irrational that may be, the frictional cost is low enough that the duty part of it herds people into voting booths. When it comes to making rational decisions the problem surfaces. Consider prop 38 in CA, school vouchers. There are lots of things to consider here, case studies of other voucher plans, the details of this one to see if it would achieve the same result, the question of what happens to taxes (which go up in the short run to cover students now in private schools, but may go down in the long run). The reality is you have people like me with no kids, who don't really care except that I'm grossly overtaxed, and have no time to think about these issues, nor any incentive to bother even if I had the time. Why do you want your kids' fate left up to me? Even parents w
ho have their own children's best interest at heart don't know about problems in other districts, and certainly don't know much about the global economic issues, and probably didn't study other voucher plans. Nearly everyone voted blindly on either pre-conceived philosophy (as I did) or even worse, based on the preposterously misleading advertisements placed by either side. My point is that ultimately such poor decision making mechanisms are rational for democratic decisions because your expected value in the outcome is nearly zero.
>1) Votes (plural) matter, since without votes there would be no election.
>2) All votes are (systemically) treated equally.
Those don't add up to a complete argument since decisions are individual while "votes (plural)" are collective.
>On the Electoral College, I'd like to hear justification for having geography (States) trump individual votes in affecting the Executive Branch (unlike the Legislative Branch, which was explicitly created to address geographicy).
If you are referring to something I wrote, I'm not sure which bit it was. I've equivocated on this point. If the choice is between a strictly popular vote to choose electors and a strictly popular vote, I don't see much reason to choose one over the other. "fairer" because each vote gets the same weight doesn't mean much to me. "fairer" at what? The current system is that occasionally 49.9% of the people get to screw 50.1% instead of the usual case of 50.1% screwing the other 49.9%. I don't see either situation as appreciably better.
I'd prefer the old system where the state legislature really chooses the state's electors in the interest of the state. Up to some reasonable limit, I think it makes sense for states to compete against each other for better legal and tax systems. If states truly had their interest represented in the executive, I think we'd see less of the sort of nonsense that goes on with the feds imposing ordinary criminal laws (e.g. .08% for DUI) via the highway fund, and more latitude for states to experiment. A second reason is that as little as I think of politicians, I think even less of the common voter. It's certainly subject to verification, but I expect state legislatures to pick smarter and more centrist candidates on average than voters. That makes for more federal stability and at least better policy in a technical implementation sense. That's really only a tiny improvement though. I think these arguments make more sense for repealing the amendment allowing direct electi
on of senators.
-------------------------- eGroups Sponsor -------------------------~-~>
eLerts
It's Easy. It's Fun. Best of All, it's Free!
http://click.egroups.com/1/9699/4/_/220122/_/974225055/
---------------------------------------------------------------------_->
Home