Should You Vote for the Best Candidate?

Stephen H. Unger
August 7, 2012

In the upcoming presidential election, a large number of voters plan to hold their noses as they cast their ballots. They intend to vote for a candidate who is far from representing their views on most important issues, even tho there may be one or more other candidates on the ballot representing their views quite well, and who they would prefer to see in the White House. The reason is that their favorite has no realistic chance to win, and they fear that, if they do not vote for the least obnoxious of the two contenders, the election will be won by a candidate who they consider to be the worst of the two.

On the surface, their reasoning seems soundly based on the idea that we should act so as to produce the greatest good and the least harm. So, goes this argument, if we believe that the election will be won by either of two candidates, we ought to vote for the better of the two, even tho we believe both of them are truly awful.

I have the greatest respect for many holders of this view, who I know are very intelligent and are motivated by the best intentions. But I think they are making a terrible mistake.

A Tough Choice

Let's consider what might appear to be an analogous situation. Suppose you are driving a big truck at 60 MPH, and suddenly you see two people 10 yards in front of you. There is no way you could stop in time to avoid hitting them. A steel railing on the right prevents you from steering around them in that direction. In front, to your left, is a group of six people, and still further to the left is another steel railing that prevents you from steering to the left around the six. So, you have to choose between exactly two actions: keep going straight and kill two people, or swerve to the left and kill six people.

This is truly a case where you have to choose between two actions, both of which are terrible. Tragically, the best you can do is to jam on the brakes, keep going straight, and almost certainly kill two people.

But is the election case really similar to the truck case? I think not. The election case is far more complex in ways that suggest a very different course of action.

Effects on Issues

Unlike the truck driving case, we have more than two options in the election case. There are almost always more than two candidates on the ballot for president, and there is always the write-in option. Apart from the matter of who is elected, the number of votes cast for non-contending candidates can be significant in a number of ways. Voting for a candidate is very different from betting on that candidate.

Votes cast for a minority party candidate can signal support for the positions advocated by that candidate. This can influence both the legislative and executive branches of government. E.g., in the past, such matters as child labor laws, and women's suffrage were first effectively promulgated by candidates of minority parties such as the Socialist and Bull Moose parties.

How to Become Politically Invisible

Any segment of the electorate that demonstrates that they will always vote for the candidates of a particular party, regardless of what members of that party do when in office, will have essentially no political influence at all. Neither their chosen party, nor any other party would have any reason to take their views into account. In particular, this is true for the party in power.

Votes Are Party-Building Bricks

New political parties do not usually become contenders overnight. If each of us decides to continue supporting what we perceive to be the least pernicious of the two very bad contending parties until a decent alternative party magically appears with many millions of supporters, we will continue to sink deeper and deeper into the political sewer. Unless a lot of people vote for a non-contending party, it can never grow to become a contender.

If we act as tho the upcoming election is the last one that will ever be held, ignoring the effect of current votes on the building of organizations to affect future elections, it will never be possible to escape from the trap in which we are told that our only choice is between candidates of two parties, both of which are hostile to our interests.

Cart Pulls Horse

Even in the short term, voting for a candidate that carries out pernicious policies is a bad idea, even if we assume that the other possible winner might behave somewhat worse. This is because, after we vote for a winning candidate, we tend, perhaps unconsciously, to justify our vote by defending what that candidate does in office. Thus, many people who strongly opposed certain policies of an administration they considered "bad", turn around and support the same, very similar, or even worse, policies, when they are furthered by the candidate they voted for. So, paradoxically, the outcome may be worse when the candidate considered "not quite as bad" is elected, because there will be much less public opposition to bad policies.

Voting for, and advocating that others vote for, a candidate hostile to one's views on most important matters is likely to lead people less committed on the issues to discount the sincerity of your arguments on those issues.

Votes Are Unconditional

When you vote for a candidate, you thereby endorse both the candidate and the policies he or she is associated with. You can't put an asterisk next to your vote pointing to a footnote explaining that you really intensely dislike both the candidate and the policies, and that your vote was only an effort to prevent a worse candidate from winning.

After the votes are counted, supporters of the bad policies will point to the election results as objective evidence of public support for those policies. "Clearly", they will say, "since very few people voted for the candidates who campaigned against those policies, the public has, in effect, ratified them". In every respect, your vote will have exactly the same effect as the vote cast by an all-out supporter of both candidate and policies.


It is often the case that the simplest, most straightforward approach to a complex problem leads to the best solution. Engineers refer to this as the KISS principle: "Keep it simple, stupid". For the reasons sketched above, I believe that voting in political elections falls into this category. If everyone voted for, and otherwise supported, candidates they believed best represented their views on the important issues, we would all be better off. This is the naive assumption behind the democratic concept of elections. When people start playing games in which they decide how to vote based on how they think other people are going to vote, the system breaks down. Rather than worrying about what other people are going to do, we should each simply express our own views directly in votes that we won't be ashamed to tell our grandchildren about.

Some Related Matters

I have kept the above arguments in general, rather abstract, terms, since they are not simply about the details of our current political situation. But they clearly apply very directly to what is going on now. I have written two essays showing the connection. [Unger-Wake][Unger-Heads].

In addition to the problem treated here, there are a number of other problems that are seriously eroding our political system. These include campaign finance, e-voting systems, the plurality voting system (which greatly exacerbates the problem treated in this article), the electoral college, and gerrymandering. These are briefly discussed in another article. [Unger-Fix]

[Note, added 9/9/12. A brief, well written, article, by Michael Ossipoff, with some additional arguments, arrives at a conclusion similar to that presented here [Ossipoff]]


Michael Ossipoff, "Making Best Use of Plurality, Until We Get Better", Democracy Chronicles, August 25, 2012

Stephen H. Unger-Wake, "Wake Up America! 2012 is Coming!", Ends and Means, July 27, 2011

Stephen H. Unger-Heads, "Heads They Win, Tails We Lose: Our Fake Two-Party System", Ends and Means, January 5, 2011

Stephen H. Unger-Fix, "Fixing Our Broken Democracy", Ends and Means, October 22, 2009

Comments are welcomed and can be sent to me at unger(at)cs(dot)columbia(dot)edu

Don't forget to replace (at) with @ and (dot) with the symbol .

Return to Ends and Means to see other articles that you might find interesting.

free web stats