The more people who are careful to minimize their energy consumption, the better off we all will be. So we should all try to minimize energy consumption. If many people discarded litter on the streets, the result would be unpleasant. So we should not discard litter on the street. If many people frequently lie, human relations deteriorate significantly. So we should not lie.
These are instances of ethical rules of the form, "If one thinks it would be good if other people did X, then one should do X." This is a rewording of an ethical principle due to the German philosopher Immanuel Kant (1724-1804) . His formulation was, "Always act according to that maxim whose universality as a law you can at the same time will." Another way to say it is, "If you believe that the more people who follow rule R, the better, then you should follow rule R." Yet another statement is: "Adhere to a rule if you believe its widespread observation would be generally beneficial."
One of the best known and useful guides to moral behavior is the golden rule, "Do unto others as you would have them do unto you" . This rule applies to the above example about lying, but it is less general than the Kantian rule; it is limited to relations between individuals, e.g., it does not apply to the above examples concerning littering and energy conservation. It is interesting that a pruning of the wording of the golden rule transforms it into yet another, perhaps the most concise, wording of the Kantian rule: "Behave as you would have others behave."Application to voting
On election day, I wish that other voters would vote for the candidate I consider to be the best in terms of positions on the important issues, and character. Therefore, I will do the same, voting for the candidate I deem to be the best one. If most people voted this way, then those elected would best represent the views of their constituents. This rule clearly expresses the idea of democracy.
Another way to frame this application of the Kantian rule is to say that I wish every voter would vote for the candidate he or she thinks is the best candidate (or best represents his or her views on the basic issues). In all these variations, the result is that the rule motivates me to vote for the candidate I think is the best.
I don't expect to find perfect candidates on the ballot. I consider voting for those who, on the most important issues, have positions acceptable to me. I never vote for a candidate whose position on a critical issue is unacceptable to me.
My choice is based on the issues and on the character of the candidate. I do not consider the likelihood of a candidate winning when I am deciding who to vote for. A vote is not a bet. It is not a "move" in a game. When I cast a vote for a candidate, I feel that I am simultaneously endorsing that candidate's positions on the critical issues. Even if the candidate loses by a large margin, the size of his or her tally is an important indicator of support for the candidate's positions on the key issues. Note that we can't append footnotes to our votes explaining that we really oppose many -- perhaps even most -- of the views expressed by the candidate we just voted for. Every vote for a candidate strengthens support for the positions taken by that candidate.
Elections are the basic mechanism for implementing the idea of democracy, a political system whereby governments respond in a positive manner to the will of the people; more precisely to the wishes of a majority of the people, with some constraints to protect minorities from abuse.
Voters who regularly cast their ballots on a partisan basis, without regard to the actual behavior and positions of the candidates with respect to the issues important to them, thereby surrender their role as citizens of a democracy. The leaders of the party they adhere to are free to ignore them, and to act on the basis other considerations; often obedience to those interests that fund them. Their views can be, and are, ignored by all branches of government. Unfortunately, such blind loyalty to a party, sometimes motivated to a large degree by fear of the "other" major party, is very common.
What about the argument that a vote for a candidate who has no chance to win, might tip the balance and cause the election of a candidate you consider to be terrible? When many thousands, perhaps millions of votes are cast, the probability of such an event approaches zero. But even if that near miracle should occur, the responsibility would be widely shared. If there should be a victory by one vote, then, apart from those who voted for the winner, the "blame" for the outcome would fall equally on everyone who voted for any candidate other than the one who finished second, plus those who did not vote at all. In presidential elections since WWII, on average, more than 40% of eligible voters did not vote .
Yet another important factor is when office holders do bad things, those who voted for them tend not to protest publicly. Even if a very substantial percentage of party members, perhaps even a majority, would like a certain position to be adopted (or not adopted) by the party, the leaders may, and often do, ignore them without fear of consequences, because the great majority of members are unconditionally committed to voting for party candidates. For example, a party ostensibly opposed to militarism can initiate and extend military actions with far less domestic opposition than can a party that noisily advocates such behavior. Objections from their own party constituents will be muffled, and of course those in the other major party will, if anything, applaud. If the more obviously militaristic party did essentially the same thing, many adherents of the other party would protest.
So, even in the short run, we have little, if anything, to lose, by voting for candidates of parties that are currently not contenders, but whose positions on the major issues are in accord with our's. Should this result in those we consider to be the "bad guys" getting elected, their bad acts would result in more opposition than if similar acts were committed by the "good guys." The overall effect is to make it unclear whether "bad guys" in office can do more damage than "good guys." In the long run, voting, in a straightforward way, for those we truly believe to be decent candidates, is the best policy.
Prior to the 2012 presidential election, I wrote an article containing several of the arguments presented here . Readers may judge for themselves how what I wrote then has fared with the passage of time.
 Wikipedia, "Kantian ethics"
 Wikipedia, "The Golden Rule"
 Wikipedia, "Voter turnout in the United States presidential elections"