The initial motive for writing “A Theory of the Calculus of Voting” was not the argument that subsequently became the focus of academic discussion, the necessity for positing a D term (“citizen duty”) to render the act of voting rational. Instead, our purpose was to show that people reacted “rationally” (i.e., logically) to an election's competitiveness. Admittedly the data we brought to bear here were weak, but with the addition of D the question arose as to whether we made voting rational only by rendering the concept of rationality a tautology: rational people voted because, with D, we assumed they were rational. However, one response with which I am sure Riker would agree is to note an asymmetry in our thinking about social processes. Few people claim to eat for a reason other than satisfying their own (private) needs as opposed facilitating the (public) objective of diminishing world hunger, and positing a growling tummy as a reason for entering a restaurant is hardly taken as a instance of rationality turned tautology. Yet the argument that the benefits of voting are dominated by a private D term as opposed to some public PB calculation occasions precisely that accusation.

One difficulty with this accusation is that we still do not know what rationality means as a basic concept. The core formal structure of “rational choice theory” is game theory, which is itself an attempt to define “rational.” The notion of a Nash equilibrium in noncooperative games, for example, and the hypothesis that people choose corresponding strategies is an idea about what rationality means in specific interactive situations. In turn, refinements of Nash equilibria that seek to dispense with their common overabundance in models are attempts to refine our understanding of “rational” and “rationality.” In part, then, the survival of this article in the literature and the attention paid to the weight that ought to be given to a PB versus a D term can be attributed to the fact that the search for the meaning of rational—both theoretical and substantive—is a core element of our research agenda.

But that is only part of the explanation. Few if any economists are concerned with why some people prefer carrots to potatoes. Political scientists, though, must focus on the substantive content of choice. Dissatisfaction with the mere addition of D derives, then, from the legitimate view that its determinants require specification before it becomes a meaningful construct. The nature of our discipline dictates that we address in detail the parameters of choice, such as the impact of competitiveness on turnout (P); the implications of negative campaigning on individual beliefs about candidates (B); the incentive to vote strategically as a function of election rules (PB); the impact of ballot forms (C); why turnout varies between presidential and local elections; and why voting and nonvoting are largely habitual acts (D). Thus, the signature expression of “A Theory of the Calculus of Voting,” R=PB−C+D, frames much of our research into elections and voting and casts that research as refinements of the application of the notions of rationality and rational choice.

Of course, one might argue that an essay's decades-long survival marks the absence of theoretical advancement. That may or may not be true, but it speaks to another matter. Of the many formal essays attributed to the Rochester (read: Riker) school, it is perhaps surprising that this article has the longest legs. It suggests that a discipline's seminal contributions are not those replete with notation, lemmas, and theorems; rather they are ones that address basic matters in simple (nonmathematical) ways. The primary challenger to this supposition is McKelvey's (1979) proof of global intransitivities in multidimensional issue spaces, because the “mathematical sophistication” of R=PB−C+D pales in comparison to that proof's content. But it is the general implications of McKelvey's result that capture our interest and not its mathematical elegance. Indeed, it was Riker's (1982) subsequent volume Liberalism Against Populism that revealed the result's full substantive meaning. Perhaps there is a lesson here, then, that goes beyond the subject with which this article deals and extends to how one goes about doing research of lasting value—to activities that do more than add a few lines to a vitae or offer a model of a nonexistent universe. Political science is (or ought to be) an applied discipline, and generally the critical contributions are those that encourage us to fit otherwise disparate empirical patterns into a coherent paradigm. That, I suspect, is the explanation for the citations that this article continues to accumulate.