It is often said that American voters are “irrational.” But what exactly does the suggestion of “irrationality” mean, and to what extent does it pose a problem for those analyzing politics? There are two kinds of “irrationality” in voting behavior. The first is when a voter votes contrary to his tangible interest, expressed in economic, social or military terms. For example, a middle-class American voting for candidates who support tax cuts for the rich, or a woman voting for candidates who are against legal abortion, or a young man voting for candidates who support a potential draft.

Some analysts begin with the supposition that it doesn’t make sense for voters to support such policies in their voting habits. So these analysts look to external factors to explain this “irrationality.” Perhaps there were no candidates that represented the voter’s tangible interest, or maybe the voter had insufficient information regarding the candidate’s policy, or it could be that the voter misunderstood the application of such policy to his own interest. And for some individuals these external factors do a good job of explaining their voting habits.

But when applied to society as a whole, this kind of analysis has always failed. In fact, it is bound to fail because the initial assumption that undergirds this entire class of analysis is misguided. The assumption is that given perfect knowledge and a viable candidate, a voter will always vote for the candidate who supports his tangible interest. Yet anyone who thinks about his own experience will recognize that we humans often do things contrary to our tangible interest. Whether humans are ultimately motivated by self-interest is another question for another column. But it seems clear that voters often take into account a conception of justice, of equality, of the common good, solidarity, civic duty or some other higher concern that overrides their tangible interest.

The lesson: There is an internal point of view to human deliberation that must be considered in any attempt to explain or understand voting habits. But this brings us to the second kind of “irrationality” in voting — when a voter votes for candidates who advocate policies contrary to the voter’s stated preferences. Is there any way to explain this “irrationality”? Presumably, the “higher concerns” referenced to explain the first form of “irrationality” have informed the voter’s policy preferences, so there seems to be little possibility to explain the second “irrationality” with respect to the “internal point of view” of voter deliberation. And though increased knowledge and voter alternatives go a small way toward eliminating this “irrationality,” there remains in society a strong strain of “irrationality” of this second kind.

Though we lack a clear language to explain this “irrationality,” we are loathe to admit any arbitrariness in the voting habits of the populace. Thus, we utilize imprecise terms such as integrity, sympathy and command to approximate the way in which candidates resonate despite policy preferences contrary to our own. On this page yesterday, Rachel Bayefsky theorized this dissonance as a conflict between “policy” and “character” as competing paradigms for vote determination. She argued that voters ought to support candidates who share their same policy preferences, leaving considerations of character on the periphery.

Bayefsky is right that the “character” paradigm is subject to serious abuse by the management of candidate presentation and the intermediary institution of the media. But theorizing the dissonance of this voter “irrationality” as a conflict between “policy” and “character” ultimately obfuscates the problem at hand. Analysis of candidates in the “character” paradigm is a necessary consequence of the inadequacy of the “policy” paradigm. It will therefore do no good to exhort voters to select the “policy” paradigm over the “character” paradigm. The cat is already out of the bag, as it were; only the voter who has found a candidate perfectly matching his policy preferences can avoid the intangible realm of “character.” (That only happens when the voter is the candidate.)

Once we are in the realm of “character,” “policy” has lost its luster. Consideration of this realm does not boil down to the virtues of candidates considered in the abstract. Rather, candidates resonate within a context, in a particular historical moment, given the demands of the time, the nature of the office, the ethos of the people, the prospects for the future and the spirit of the age. If the determination of a candidate is to have the character of finality, its preceding deliberation must inhabit this realm.

Peter Johnston is a junior in Saybrook College. His column usually runs on alternate Wednesdays.