3 ms·

> If you want to maximize voter's voting power, the absolute best way to do it is to randomly select a single vote to be the deciding one. Hold on. How is this

> If you want to maximize voter's voting power, the absolute best way to do it is to randomly select a single vote to be the deciding one. Hold on. How is this true? I thought that the theorem was maximizing all voters' voting power, not just one. Am I misunderstanding?

btilly 14y ago
He defined a voter's voting power as the odds that YOUR vote decided the election. My claim is that the odds that you individually will be selected by the lottery that I describe and then wind up with decisive power exceed the odds that the election will be so finely balanced that it comes down to one vote, in one state, and that vote is yours. This is just the logical extreme version of dividing voters up into blocks called "states", and then dividing up politics so that one state is likely to decide it. Back to the state version, in a million coin flips, a standard deviation is of size 500, and it is virtually impossible to deviate from the true average result by more than 1500 votes. Therefore unless the popular vote in the deciding state is almost exactly balanced, and all other states are balanced, it is very difficult for any vote to truly be deciding. The closest that the USA has ever come to what he describes as ideal is Florida in 2000. However when it came down to a handful, was that good for democracy? Not really. There were endless arguments about who meant to vote for whom. It is certain that an illegal voter purge changed that election result. It is also certain that correlations between voting machine problems and socioeconomic status did as well. (Please tell me a legitimate reason for a voting machine to have a switch - set differently in different precincts - where one setting immediately tells you if your ballot is spoiled so you get a second chance to vote, while the other silently accepts the ballot and discards the vote. And then it was set to the "silently not count you" in black precincts which therefore had a much higher incidence of spoiled ballots.) This is all ancient history at this point. But none of it is good for democracy.
- saraid216 14y ago
  > This is just the logical extreme version of dividing voters up into blocks called "states", and then dividing up politics so that one state is likely to decide it. Erm. There isn't a logical extreme to dividing voters up into blocks. Both extremes are simply a popular vote. Unless you're talking about infinitely sized blocks, in which case I'm not sure he's the one whose connection to reality is more dubious. > Therefore unless the popular vote in the deciding state is almost exactly balanced, and all other states are balanced, it is very difficult for any vote to truly be deciding. It seems like you disagree with his definition of a voter's voting power. How would you define it instead? > It is certain that an illegal voter purge changed that election result. How is that relevant to the question of an electoral college? I'm not questioning whether or not it happened, but you haven't actually demonstrated a link. I mean, of course voter disenfranchisement is harmful to democracy; isn't that sort of obvious?
  - btilly 14y ago
    His definition of voting power is, What is the probability that one person’s vote will be able to turn a national election? That's the same as my definition. His theorem is that, in the presence of an overall preference for one candidate over another, voting power diminishes rapidly. But the diminishment is significantly lower in a districting system. (Of course this increase in voting power is directly tied to the possibility that a district could send the election to the loser of the popular vote - which most of us regard as a bad thing.) My point is that, even with significant overall preferences, the lottery system that I describe keeps voter power constant. And in the presence of a noticeable national preference, that constant value is much larger than districting provides. If you want voter power (your ballot could wind up meaning something), the lottery system provides it. As to the question of illegal voter purges and the electoral college, whenever a small, identifiable, subset of the population is seen to determine the election, then incentives say that all possible means of swaying the election will be focused on that group. The fact that advertising is focused on swing states is obvious. The fact that dirty tricks are likewise is less obvious, but no less true. And the more you increase the relative importance of those states, the more true this becomes.
    - saraid216 14y ago
      Interesting. I'll assume that your math is correct, since I'm admittedly too lazy to actually run the numbers and check it. My question is this: in the question of whether or not we should have the electoral college (or more broadly, in the question of what kind of system we should use for a national election), is voting power an ideal measure? Your OP suggested that it was not. Do you have an alternate suggestion?

btilly 14y ago

He defined a voter's voting power as the odds that YOUR vote decided the election. My claim is that the odds that you individually will be selected by the lottery that I describe and then wind up with decisive power exceed the odds that the election will be so finely balanced that it comes down to one vote, in one state, and that vote is yours. This is just the logical extreme version of dividing voters up into blocks called "states", and then dividing up politics so that one state is likely to decide it. Back to the state version, in a million coin flips, a standard deviation is of size 500, and it is virtually impossible to deviate from the true average result by more than 1500 votes. Therefore unless the popular vote in the deciding state is almost exactly balanced, and all other states are balanced, it is very difficult for any vote to truly be deciding. The closest that the USA has ever come to what he describes as ideal is Florida in 2000. However when it came down to a handful, was that good for democracy? Not really. There were endless arguments about who meant to vote for whom. It is certain that an illegal voter purge changed that election result. It is also certain that correlations between voting machine problems and socioeconomic status did as well. (Please tell me a legitimate reason for a voting machine to have a switch - set differently in different precincts - where one setting immediately tells you if your ballot is spoiled so you get a second chance to vote, while the other silently accepts the ballot and discards the vote. And then it was set to the "silently not count you" in black precincts which therefore had a much higher incidence of spoiled ballots.) This is all ancient history at this point. But none of it is good for democracy.

saraid216 14y ago
> This is just the logical extreme version of dividing voters up into blocks called "states", and then dividing up politics so that one state is likely to decide it. Erm. There isn't a logical extreme to dividing voters up into blocks. Both extremes are simply a popular vote. Unless you're talking about infinitely sized blocks, in which case I'm not sure he's the one whose connection to reality is more dubious. > Therefore unless the popular vote in the deciding state is almost exactly balanced, and all other states are balanced, it is very difficult for any vote to truly be deciding. It seems like you disagree with his definition of a voter's voting power. How would you define it instead? > It is certain that an illegal voter purge changed that election result. How is that relevant to the question of an electoral college? I'm not questioning whether or not it happened, but you haven't actually demonstrated a link. I mean, of course voter disenfranchisement is harmful to democracy; isn't that sort of obvious?
- btilly 14y ago
  His definition of voting power is, What is the probability that one person’s vote will be able to turn a national election? That's the same as my definition. His theorem is that, in the presence of an overall preference for one candidate over another, voting power diminishes rapidly. But the diminishment is significantly lower in a districting system. (Of course this increase in voting power is directly tied to the possibility that a district could send the election to the loser of the popular vote - which most of us regard as a bad thing.) My point is that, even with significant overall preferences, the lottery system that I describe keeps voter power constant. And in the presence of a noticeable national preference, that constant value is much larger than districting provides. If you want voter power (your ballot could wind up meaning something), the lottery system provides it. As to the question of illegal voter purges and the electoral college, whenever a small, identifiable, subset of the population is seen to determine the election, then incentives say that all possible means of swaying the election will be focused on that group. The fact that advertising is focused on swing states is obvious. The fact that dirty tricks are likewise is less obvious, but no less true. And the more you increase the relative importance of those states, the more true this becomes.
  - saraid216 14y ago
    Interesting. I'll assume that your math is correct, since I'm admittedly too lazy to actually run the numbers and check it. My question is this: in the question of whether or not we should have the electoral college (or more broadly, in the question of what kind of system we should use for a national election), is voting power an ideal measure? Your OP suggested that it was not. Do you have an alternate suggestion?