It's a dilemma many voters face whenever there's still a serious third or fourth candidate for president after the primaries. This year, let's say your first choice would be Ralph Nader or Pat Buchanan. But is voting for a candidate with no serious chance of winning just "throwing away your vote"? Worse, is it in effect a vote for the less desirable of the two realistic candidates? Is it wrong to vote for Nader if you would prefer Al Gore over George W. Bush? Or for Buchanan if you would prefer Bush over Gore?
The answer is ultimately subjective, but objective analysis will take you further than you might have thought.
The first useful objective fact is that a vote for Nader is not equivalent to a vote for Bush. A vote for Nader takes a vote from Gore. A vote for Bush would take a vote from Gore and add one to Bush. Voting for Nader is like half a vote for Bush. So in your subjective moral calculus, the downside of voting for Nader is only half as big as you may have thought.
But why vote for Nader or Buchanan at all, given that they can't win? (Yes, it's true: They can't win partly because everybody thinks they can't win. That's ironic, OK? But it doesn't change the fact that they can't win.) The purpose of voting for Nader or Buchanan is to send a message that you are dissatisfied by the two mainstream choices and that you endorse Nader's or Buchanan's views on various issues. By doing so, you increase the possibility that these views, if not these candidates, will be able to win in future elections.
Sending a message is a perfectly valid purpose. But is it worth the cost of helping to elect the wrong guy (from your point of view)? The best argument in favor is that there is no cost. When was any presidential election decided by a single vote? Never. For the purpose of sending a message, every vote adds to the volume of the message. A tally of 8.113 percent speaks a bit louder than 8.112 percent, which speaks louder than 8.111 percent. But for the purpose of electing a president, the result is binary. Either Gore will be president or Bush will be president, and no single vote is likely to change that result. So using your vote to send a message is not irresponsible. In fact, it's a more sensible use of your vote than trying to affect the actual result.
Pretty good argument. But it depends on the premise that your vote is essentially worthless in terms of deciding who becomes president. And in fact, political scientists put the odds of a single vote deciding a presidential election at perhaps one in 100 million. You're more likely to be killed in an auto accident on the way to the polls. By this analysis, voting itself—not just voting for someone who can't win—is an irrational act. Or at best it is an act of patriotic romance and/or civic responsibility.
However, there is a counterargument. It goes like this: Even 1/100,000,000 is greater than zero. And, as Oxford University's Derek Parfit has pointed out, if your vote happened to be the one that made the difference, its worth would be huge. That changes the calculation.
How much would you pay for your preferred candidate to win and be president for the next four years? Not Nader or Buchanan: They can't win. But Bush or Gore. Would you pay $100 a year? If democracy matters at all, it's surely worth at least that much to you as a citizen to have the policies and leadership of one of these gentlemen rather than the other. This includes the spiritual and other nonmonetary benefits, not just your share of a tax cut or Social Security increase. And presumably you believe that most or all other citizens would benefit, too. Perhaps more, perhaps less, but let's say on average roughly as much as you.
Those benefits add up. If roughly 250 million Americans would benefit by an average of $100 a year for each of the four years until the next election, the total benefit from your preferred candidate winning the election is $100 billion. And even if the odds that your vote makes the difference are one in 100 million, that small chance is worth $100 billion divided by 100 million, or $1,000. You can fiddle with the assumptions, but even a result of one-tenth that amount, or $100, would make the value of voting higher than the cost (in time and hassle) for almost everyone.
That still doesn't make voting rational, since the cost is imposed on you while the benefit is spread among all citizens. (And you get the benefit of other people voting whether you vote or not.) But we can also safely assume that you are civic-minded enough to care about what's good for the rest of the country as well as yourself. The problem was that even for the civic-minded, voting to affect the result seemed like an irrational act. But by this logic, it isn't.