According to the paradox, voting is costly and a single vote virtually never makes a difference. So a) why bother? and b) how does this affect appropriate or ethical voting behavior?
According to Kant’s categorical imperative, moral behavior follows this axiom: “Act only according to that maxim whereby you can, at the same time, will that it should become a universal law.”
So not voting is unethical because if everyone did it democracy would collapse. I would also argue that it implies that electability should be a concern in the primaries. Primary populations tend to average out at a different ideological point than the center. So if folks in those primaries ignore electability in stage 1, then they will tend to get a suboptimal outcome in Stage 2, according to their electoral preferences. I guess a certain amount of game theory applies, since the optimal weight on electability turns on what is happening in the other primary. Which is odd, as the categorical imperative usually cuts through such calculations.
Don’t vote. It doesn’t make a difference and it isn’t worth your time. But shut up about it. Don’t discourage others from voting. Because if the costs are still low, the objective benefits of flapping your mouth are lower still. (Not clear really: maybe you want to free up your friends’ time. But you need to recalculate that before opening your mouth. Act utilitarianism involves continual moral calculation.)
Along similar lines act utilitarianism tends to be highly tolerant of lying and deception. Which is why it is mostly rejected in favor of rule utilitarianism, except in circumstances involving 1930s time travel and nuclear exchange.
I think we’re basically back to Kant now. You should choose a voting strategy that provides the best outcomes given that society accepts such rules. Including members of the other party.
My favored approach: a 1% solution
I try to split the difference. My rule is not, “Vote for your favored candidate.” I attempt to respect the underlying empirical reality as well as the civic obligation implied by the Kantian approach. My proposed rule is, “Vote as if you actually held sway over 1% of the ballot”.
Most of the time your vote won’t make a difference. But not all the time. And the margin of defeat and victory will make a difference: you can be sure that political professionals study such things. Mondale’s blowout in 1984 shaped the Democratic party far more than Carter’s narrower defeat. Make one state or electoral district safer and it will free up resources to fight in swing states. If it’s a six-way ballot, think about which candidate you would like to grant an additional 1% margin. And trace the likely consequences, bearing in mind similar cases in the past.
The wiki article addresses positive issues. This thread is normative: with that in mind here is Wiki on the paradox of voting