If an old lady in front of you drops her purse, there is no economical reason for you to help her out. It is a waste of your time and she can pick it up herself. If she can't pick it up herself, someone else walking the street will help her out. No need to trouble yourself.
However, most people here would help her out because its just the right thing to do.
You might ask how that changes if you and the old lady are alone. You would probably say that you'd still help her out, though I'd say that's just a response trained within a particular environment in which there is a help-old-ladies social expectation.
Unless you view the candidates as equal, or have perfect information about the number and direction of all other votes in your electorate, you vote to increase the probability of an outcome you prefer. Votes are only unnecessary after the fact.
The freakonomics guys are proud of their non-technical background but I find this lack of understanding of a game theory in such an important topic absolutely unforgivably stupid.
The ad hominem attack is unnecessary here. While Dubner is not an economist, Steven Levitt (the other Freakonomics author) is not only an economics PhD, but also a winner of the John Bates Clark medal. In short, he's one of the best economists out there.