There has been a good deal of discussion recently among libertarians and conservatives over whether to vote in the upcoming midterms. Those among us (I confess to being one) who had hopes that what passes for the political right among American politicians might actually seize the moment and institute real change are sorely disappointed. It is obvious that is not going to happen and that there was no intention that it would.

I’m not going to presume to tell anyone whether or how to vote, but I do have observations about the rationality of voting and the reasons we have, or don’t have, for voting in elections generally. These form my response to the kinds of comments I have seen about the instrumental rationality of voting for the “lesser of two evils.”

Voting is almost never instrumentally rational because we control only one vote, our own, and in an election of any significance, our own vote will make no difference in the result. By “instrumental” rationality, I mean taking a reasonably efficient means of accomplishing one’s ends. It is true that in local elections with a very small electorate, it is conceivable that our vote matters, but in congressional and statewide elections, our votes have only an insignificant impact on the outcome.

Furthermore, voting involves some cost. If we vote a party preference, all we need to know is the party of the candidates, but there is still some small cost we incur. Since we know that our vote is insignificant as to the outcome, even the small cost of going to the polls and voting is not instrumentally rational. Taking time to weigh candidates or parties against one another involves incurring the cost of acquiring information and devoting time and energy to considering the alternatives, which is even less instrumentally rational. Being an informed voter is, in a word, irrational. We owe these insights about the rationality of voting to economist Anthony Downs, whose “paradox of voting” demonstrates that, generally speaking, voting is not instrumentally rational.

This relates directly to the claim that it is rational to vote for the “lesser of two evils” because we don’t want the greater of the two evils to win. However one defines evil, no one wants more of it, and where we have a choice, we have the chance to vote for the lesser of two evils, even if both alternatives are undesirable. Therefore, the reasoning concludes, it is rational to vote for the alternative that we regard as the lesser of two evils, and it is irrational to fail to do so. As Downs would tell us, this argument is unsound, because it neglects the empirical fact that our vote, the only one we control, is almost certainly going to be insignificant in determining the result of any but a small local election. The outcome of the lesser evil prevailing is a collective or public good, and our contribution to the production of that good is insufficient to make a meaningful difference in whether it is provided. Consequently, incurring even the very small cost of voting for the lesser of two evils is not instrumentally rational.

While the “lesser of two evils” argument is easily disposed of, its appeal to people remains powerful. I suspect there are at least three reasons for this. One reason is that most people do not analyze these questions carefully, so the prominent fact that our votes are insignificant does not occur to them. Another reason is that the “lesser of two evils” argument is a rhetorical device that partisans use to encourage weakly partisan voters to back their preferred candidates by appealing to the alleged rationality of doing so. Don’t let the commies win! Your vote matters! These kinds of exhortations are designed to appeal to fear, not reason, and thus are the ordinary stuff of politics. A third reason stems from the system itself, and here I mean broadly the system of organized coercion that employs elections as a means of legitimating itself. The state apparatus and the interests, both state and non-state, that benefit from its operations need the ruled to participate in elections to create the veneer of representation that democratic liberalism insists is the chief legitimating feature of any political system. Those who operate the system and those who benefit from it need people to participate to avoid a crisis of legitimation, which looms in the rising populism we see in western politics today, as many people become more aware that democratic politics serves elite more than popular interests. Thus, elite interests have a powerful incentive to encourage participation, including the offering of baseless appeals to rational self-interest in voting for the “lesser evil.”

Building on this last point, consider a substantive rather than formal reason why our votes don’t matter. Donors, not voters, set the agenda for political elites. If you ask why elected officials betray their voters, it is because voters are akin to a herd of livestock from which politicians subsist but do not take direction. It is the donors and powerful financial interests who determine what issues are seriously considered and acted upon. The money on which politics runs provides for advertising, travel, staff, and political favors to other politicians in return for support, and opportunities for profitable investments arising from information and connections obtained from being inside the governing structure. Elites prevail in the game of power and wealth by serving donors and financial interests, not those of voters.

Considering all this, do we ever have any reason to vote? While it is not instrumentally rational to vote, we do have other kinds of reasons for action besides instrumentally rational ones. A reason can be understood as an objective norm that underlies our actions, one that we could employ to explain or justify our actions to others. Instrumentally rational reasons are reasons because they refer to objective means to attain our ends that we can explain or justify to others as means of attaining our ends.

Instrumental rationality is not a reason to vote, but that does not exhaust the kinds of reasons we may have for voting. We may feel an attachment to the system of organized coercion that relies on our participation for legitimacy, so we may vote out of a sense of civic duty. In Orwell’s classic novel Nineteen Eighty-Four, Winston Smith realized in the end that he loved Big Brother. This is a reason we could articulate to others, though some of them might wonder at the placement of our affections. A more palatable reason might be that we endorse the parties or candidates for whom we cast a vote. This, while not instrumentally rational, is in fact a reason to vote, and its rationality is unassailable. Here, we are voting to express our values in a choice we have made, and we certainly have a reason to do that.

We can have a reason to vote, but the “lesser evil” is not one of them. If our reason to vote is to endorse a party or candidate because the party or candidate represents our values, that party or candidate is not a lesser evil, but instead is a positive good. We cannot have a reason to endorse that which we regard as evil, lesser or otherwise. Thus, we never have a reason to vote for the lesser evil. It is not instrumentally rational to do so, and we cannot have a non-instrumental reason for doing so, except perhaps the Winston Smith variety. If you have a reason to vote, it is because you have found an opportunity to express your values with your choice, and that is a good thing. For many libertarians and conservatives, it has become a rare opportunity, but if you find one, you have reason to take it.