TCS Daily

Nader in '04!

By Keith Burgess-Jackson - August 4, 2003 12:00 AM

The silly season -- i.e., the 2004 presidential campaign -- is upon us, so let me provide a public service by helping you spot (and encouraging you to disregard) a particularly insidious but strangely alluring argument. It is an argument that you will hear and read many times in the next fifteen months. It is an argument that even intelligent, well-educated people find appealing. But it is a bad argument.


The argument is that, by voting for a third-party presidential candidate, such as Ralph Nader or Patrick Buchanan, you are throwing your vote away. Since a vote is a valuable thing (it is assumed), and since one should not throw away valuable things, one should not vote for a third-party candidate.


Sound appealing? Let's explore it. Suppose that, like one of my liberal-arts colleagues, you had the following preference ranking on the verge of the 2000 presidential election: Nader > Gore > Bush. This means that you preferred Nader to Gore and Gore to Bush. Hence, you preferred Nader to Bush. It would seem that if you were going to vote, you should have voted for Nader, since you preferred him (for whatever reasons) to either of the other candidates. Simple, right?


But some people thought you would be throwing your vote away if you voted for Nader. They said that he had "no chance of winning the election," whereas Gore did; and since, by hypothesis, you preferred Gore to Bush, you would have been shooting yourself in the foot (acting imprudently, self-defeatingly) if you had voted for Nader. A vote for Nader, it was said, is a vote for Bush.


My colleague, sadly, fell hook, line, and sinker for this line of reasoning. He thought he was voting strategically. What he was doing was voting stupidly (all the more so because he lives in Texas, which was securely in the Bush camp).


(Note: The fallacy in question gets committed on both sides of the political aisle. Bush supporters used the argument on people with a Buchanan > Bush > Gore preference ranking. They said, in effect, that a vote for Buchanan was a vote for Gore. I am not making a partisan political point in this column; I am making a nonpartisan logical point.)


The problem with the thrown-away vote argument is that it rests on a false assumption. Saying that you throw your vote away if you vote for Nader, Buchanan, or some other third-party candidate implies that you do not throw your vote away if you vote for one of the other (major-party) candidates. The concept of throwing a vote away makes sense, in other words, only if the concept of not throwing a vote away makes sense. This is what we need to explore.


In what sense, if any, does one not throw one's vote away if one votes for Gore or Bush? It might be said that since, realistically speaking, one of them is going to win the election, one does not throw one's vote away by voting for either of them. But this is strange. Does one's voting for either of them make a difference to the outcome? Paul Meehl has estimated that the chance of one's vote making a difference in a national election is one in one hundred million -- about the same as one's chance of being killed on the way to the polling place. (See Paul E. Meehl, "The Selfish Voter Paradox and the Thrown-Away Vote Argument," The American Political Science Review 71 [March 1977]: 11-30.)


While Meehl's point is well taken, it ignores the fact that nobody votes in a national election. There are no national elections in the United States. We vote in states, for presidential electors. But this qualification is not fatal to Meehl's argument. While the precise odds of one person making a difference in the outcome of a statewide election are unclear, they are clearly long. The closest race in 2000, which was one of the closest presidential elections in history, was in New Mexico, which, according to the Federal Election Commission, went to Gore by 366 votes (286,783 to 286,417). The Florida vote was almost as close in absolute terms (and significantly closer in comparative terms), with Bush winning by 537 votes (2,912,790 to 2,912,253). Nader received 21,251 votes in New Mexico and 97,488 in Florida. Buchanan received 1,392 votes in New Mexico and 17,484 in Florida.


Suppose you're a Floridian and that you voted for Gore rather than your preferred Nader because you sensed that the election would be close. Your vote did not matter. Read that sentence again. Close your eyes and repeat it to yourself. Had you voted for Nader rather than Gore, Gore would have lost by 538 votes instead of 537. Nor was there any significant chance, antecedently, that your vote would matter. The only way it would have mattered is if, without your vote, Gore either lost by one vote or tied, for then your vote would have either given him a tie or put him ahead by one vote. Either outcome, I hope you will agree, is vanishingly unlikely. As depressing as it is to contemplate, your vote affects only the final digit of the vote total. That is the extent of your control.


This point bears emphasis, for I have heard it said many times, even by intelligent people, that individuals vote in blocs, not as individuals; and if a large enough bloc votes for candidate A rather than candidate B, it could make a difference to the outcome. But this is just the fallacy restated. It conflates the individual voter with the collection of individual voters. I, Keith Burgess-Jackson, have one vote. I wish I had more than one vote, and I certainly deserve more than one vote, but that's all I have. The question is what to do with it. Should I cast it for A, for B, or for C? Maybe I shouldn't vote at all. But this much is clear: What other people do with their votes is up to them, not to me. Even if we contracted among ourselves to vote for a particular candidate, it would still be up to me whether to honor or breach the contract when I enter the voting booth. Stop thinking of blocs; that way lies madness (and fallacy). Start thinking of your one vote and what to do with it.


In the sense in which one throws one's vote away by voting for a third-party candidate such as Nader or Buchanan, therefore, one throws one's vote away by voting for either of the major-party candidates, Gore or Bush. There is almost no chance that one's vote matters, however it is cast. Since the thrown-away vote argument implies that some votes do matter, i.e., are not thrown away, it is unsound. Either every vote is thrown away or no vote is thrown away.


My own view is that no vote is thrown away, even in a statewide election. (Local elections are different. In a local election, such as an election for school-board membership, voting has a nonnegligible chance of making a difference.) I vote. I always have. I take it seriously. But when I vote, I don't flatter myself by thinking that I affect the outcome. I have no effect on the outcome. I vote for Kantian reasons, because I cannot universally prescribe that people refrain from voting out of self-interest, ignorance, or apathy. It would be the end of democracy, which I value as the least-bad form of government. I vote for the same reason that I refrain from taking shortcuts across the grass of my campus: not because my steps make the difference between the grass living and the grass dying (they don't), but to express and implement my values.


That, in the end, is the best reason to vote. Voting is an expressive, symbolic act. It is the act of a citizen (homo politicus), not of a self-interested utility maximizer (homo economicus). It is an exercise of autonomy and ultimately an affirmation of one's dignity as a person. It registers and signifies one's values. If I had the Nader > Gore > Bush preference ranking, I would vote for Nader even if it were clear to me, from polls and such, that (1) he were vastly less likely than Gore to be elected, (2) the contest between Gore and Bush would be close, and (3) I much preferred Gore to Bush. I would vote for Nader because he is my preferred candidate. I have only one vote, after all, and I know, having thought critically about it, that it won't make any difference to the outcome. Indeed, wouldn't I be throwing my vote away if I did vote for Gore in these circumstances? I would be expressing a preference for someone (Gore) whom I prefer less than an alternative (Nader)! That is beyond silly; it is downright perverse.


Keith Burgess-Jackson, J.D., Ph.D., is Associate Professor of Philosophy, Department of Philosophy and Humanities, The University of Texas at Arlington. Burgess-Jackson has voted for Republican (1976), Libertarian (1980), Democrat (1984, 1988, 1992), and Green (1996, 2000) party presidential candidates. None of his votes affected the outcome.


TCS Daily Archives