The article
dh mentioned has a serious logical flaw: It
confuses the map with the territory. Here's how: it models an
election with 60% of people voting for Nader and 40% for Browne as an election in which each person flips a 60/40 weighted
coin to choose who to vote for. This is
convinient from a statistical point of view - and if you took only a sample of people, this would be a
valid model. But in an election, every vote counts - and there's no chance (not a vanishingly small chance, but no chance at all) that
Browne's gonna win this one. The law of large numbers (which the article invokes) applies to statistical things, but not to non-random, deterministic things like votes. As a small example, consider the famous case of 2
lions and 1
sheep deciding what to have for dinner. Assuming that the lions (and the sheep) are typical of their
species, the chance of
mutton isn't just near 1, it is 1.
And y'know how there are all these recounts going on in Florida? If it weren't for the EC, they would be happening everywhere - and the votes of people in Florida *would* count. And
Muke's math is bad - it doesn't consider the chance that one person actually will affect the outcome of a state's election (which is zero).
Update: Muke: My logic (above) was confusing you with dh's article (sorry). Here's how your math is bad: You're getting that each
American has more than 1/Xth of the vote, where X is the number of voters. So, you're saying that Americans as a whole have more than 100% of the vote. That's clearly not possible - no matter how you divide up 100%, when you add it back together, you get 100%. I looked around a bit, and found the cause - only 98 million people voted this year (for president). So, the
Californian has now slightly under 1/Xth of the vote, while the
Alaskan has slightly over 2/Xths of the vote.
All the EC does is redistribute the power. I know you wouldn't support simply giving Alaskans 2
votes each and having a popular-vote election - so why is it different when we hide it in the EC? The other thing to consider is that since
Alaska is one of those states that goes Republican no matter what (like
Washington, DC, DC goes Democrat almost no matter what). In this situation, the EC it disenfranchises the minority party members entirely (their vote has no effect at all on the presidential election).
Further update:
I feel the need to clarify that there is no way that, in a system where some group of N people (directly or indirectly) make a decision, that the total decision-making power of those people will add up to greater than 100%. That is, no matter how you slice the political power pie, there's still only one President.