3 ms·
Anyone who hasn't seen it should read Peter Norvig's Election FAQ, particularly the section "Is it rational to vote?" [1] Short version: "Yes. Voting for presi
by crntaylor 14y ago
Anyone who hasn't seen it should read Peter Norvig's Election FAQ, particularly the section "Is it rational to vote?" [1]
Short version: "Yes. Voting for president is one of the most cost-effective actions any patriotic American can take."
Slightly longer version: "The value of not voting is that you save an hour of your time. The value of voting is the probability that your vote will decide the election (1 in 10 million if you live in a swing state) times the cost difference (potentially $6 trillion). That means the expected value of your vote is $600,000."
[1] http://norvig.com/election-faq-2012.html#rational http://norvig.com/election-faq-2012.html#rational
- mseebach2 14y agoThat makes sense only if one of the outcomes is worth $6 trillion more than the other and you live in a swing state.
- the_gipsy 14y ago1. It is 1 in 10 million if you live in a swing state. 2. You expect $6 trillion, that figure is incorrect as mseebach2 points out. 3. The probability of winning the state lottery might be 1 in 10 million. So I'm far better off buying a lottery ticket, since in the improbable event of winning I would actually make money, in contrast to hoping that the improbable event of my vote tipping the balance and creating a stronger economy would improve my income (the contrary might be true).
- mdemare 14y agoNonsense, only in California are there more than 10 million voters. In Virginia, there are 3.8M, 1.9M per party. If the candidates are 2% apart, there's about a 1/100K chance of casting a deciding vote. Also: http://what-if.xkcd.com/19/ http://what-if.xkcd.com/19/
- the_gipsy 14y agoI only cited the number from OP. What is nonsense, is valuing one vote $600,000.
- yummyfajitas 14y agoBut the value of your vote isn't actually 1 in # of votes. It's actually P(your state undecided) x P(nation undecided without your state) x (Value of your candidate winning - Value of other candidate winning). For example, say your state has 10 million voters and the election is a statistical dead heat (i.e., each voter has a probability of p=0.5 for voting for candidate A vs B). Then P(your state undecided without your vote) = (5 million choose 10 million) x pow(2^-10 million) ~= 1/(sqrt(2 pi) 2.5 million) = 1.6e-7. And that's assuming the best case where your state's voters could go either way. The probabilities become exponentially smaller (literally: exp(-n^2)) if your state actually has political leanings. I'm also not sure how we could even approach a $6 trillion gap under plausible assumptions. Does anyone really believe that the gap between each candidate is $6 trillion (note: total economy is $15B) or $20k/person?
- goldfeld 14y agoIf you take compound interest and the legacy for generations to come into consideration, the difference from decisions made today between the candidates would easily surpass $6 trillion, and possibly while you're still alive. Besides, things like education, over time, have far more intangible and invaluable implications than a figure could ever show.
- yummyfajitas 14y agoIf you take compound interest and the legacy for generations to come into consideration... And also don't use a discount rate... Besides, things like education, over time, have far more intangible and invaluable implications than a figure could ever show. Could you explain this claim?
- mdemare 14y agoThe Iraq war, at $4 trillion? These things add up. Also, over 4 years, the U.S. GDP is $60 trillion, $6 trillion is 10% of that. How much better do you think that your preferred candidate will do? 5% better? 1% better?
- 14y ago
- scottjad 14y agoThe only way your one vote has any chance of determining a US presidential election is if you sit on the US Supreme Court. Anything within a few hundred votes will be decided by the courts. See Gore v. Bush. "It's not the people who vote that count. It's the people who count the votes." (Joseph Stalin)