Higher costs appeal to voters: implications of expressive voting

Brennan and Lomasky (1993) provide the most complete analysis of the expressive voting model and its applications. Tullock (1971) contributed the first explicit mention of expressive voting, although he didn’t use the term expressive voting. The term may have been used first by Brennan and Lomasky (1993), although Brennan wrote that he is not positive that is the case in correspondence with one of the authors. More recent treatments of expressive voting include Caplan (2007), Brennan (2008), Lee and Clark (2014) and Lee (2015).

There are many reasons people derive utility from voting. The utility function in (1) gives only the expressive utility a voter attaches to voting for a policy he/she believes serves a moral purpose, net of the negative utility effect of the expected financial cost of that vote. Also, we recognize that it is conventional to put variables that generate utility in utility functions, but not their cost. But there are no variables under the voter’s control in (1). The only relevant decision the voter can make that affects his expressive utility is whether to vote yes or no given the variables that are exogenous to that decision.

The greater sense of morality from the increase in X could result in the voter feeling that he is making a greater personal sacrifice regardless of the amount of good he believes will be done and/or the belief that more good will be done. A belief that a larger X means more good will result may be correct, but it can easily be wrong. It is the belief, not the reality, that is relevant.

Even with M′(X) > 0, in a small-number election that is expected to be close the probability P can be sufficiently large to reverse inequality in (4). The voter may continue to vote yes in this case, but he will receive less utility from doing so. But obviously our result is more likely in elections with large numbers of voters, as are the general results of the expressive voter model.

An increase in the number of voters cannot be ruled out since the additional utility from voting yes may motivate some potential yes voters to become actual yes voters.

Even if a credible measure of expressive utility were available for individual voters, it would be difficult to use it by arranging an experiment in which the same person is able to vote for a particular policy at one cost and then vote again for the same policy at a higher cost.

This comment is consistent with Caplan’s (2007, Chapter 5) stress on the importance of “rational irrationality” in which people are easily capable of having convenient beliefs that aren’t true if they suffer no negative consequences from doing so. Caplan sees “rational irrationality” as the important distinction between his view of expressive voting and that of Brennan and Lomasky (1993), who argue that a majority might vote for a war, the horror about which they have no illusions, and would not vote for it if their vote were decisive. While we agree with Caplan that “rational irrationality” is an important concept and a common feature of voting, we have no problem with Brennan and Lomasky’s broader application of expressive voting. After all, expressive voting is all about the importance of emotions, and anger is a strong emotion. If, as we well know, anger can motivate a person to make a decisive decision that he knows is socially and personally harmful, why should we expect such decisions to be made in the voting booth where they are almost sure not to be decisive?

Note that in these two examples we have not mentioned the Environmental Protection Agency (EPA), despite the important role they play in supporting and enforcing environmental regulation. The reason is that we are not sure whether to list them as bootleggers or Baptists. For obvious reasons, they appear to us to be a hybrid of the two.

We do not mean to imply by our examples that all government programs motivated by noble aspirations fail to provide social benefits in excess of their social costs. But we are not reluctant to argue that most are less successful in this regard than their political supporters believe and/or claim.