“Civic engagement” is the cornerstone of democracy, right? Everyone knows we need more civic engagement—everyone, that is, except Ben Berger, whose willingness to argue against convention, particularly when that convention is a key tenet of good governmentism, makes me reflexively sympathetic to his argument. Berger and I both attack foundational assumptions of conventional democratic thought, but we differ in our approaches. I base my arguments on economic theory and quantitative analysis, whereas he bases his arguments primarily on interpretations of canonical theorists, most importantly Hannah Arendt and Alexis de Tocqueville. Ultimately, I am convinced by Berger's argument that civic engagement is a messy concept that cannot be considered an “intrinsic good” for democracy, but like most formal theorists, I needed little convincing because the debate that he characterizes is remarkably similar to one that game theorists held some years ago, and resolved in the same direction. Ironically, then, he may face a more hostile audience from within his own subfield than outside of it.

Berger's argument would be tempting for “good government” types to caricature. While he attacks the assertion that we need more civic engagement, he does not argue that we need less. Instead, he argues that the term itself is poorly conceptualized and, consequently, that normative arguments about our level of civic engagement are devoid of substance at best and misleading at worst. Thus, the most valuable analysis in Berger's book is his effort to conceptualize civic engagement. He argues that the term has been used in ways that encompass too many different forms of political, social, and moral engagement for the umbrella term to communicate very much. Consequently, he argues that the term itself should be abandoned because it is simultaneously too laden with historical baggage and too broadly defined to be analytically useful. I make a similar argument about the term competitive election in my own book reviewed by Berger in this section, and so I am particularly interested in this style of argumentation.

Berger also argues that many types of behavior that others might reasonably call civic engagement can actually be harmful. Consider, for example, Al Qaeda, or the Ku Klux Klan. It is easy to argue that we would benefit from less political engagement by members of these groups, given their objectives and the particular manner in which they engage the social and political systems. Such groups are politically engaged, but lack what Berger calls “moral engagement.” These groups are extreme examples, but they demonstrate three important underlying questions: What are the different kinds of “civic engagement,” which kinds are actually beneficial, and for whom?

After conceptualizing civic engagement, Berger proceeds to ask whether various forms of engagement can be thought to provide “intrinsic” or “instrumental” good. He approaches the question essentially by pitting Arendt and Tocqueville against each other. He derives an argument about the possible intrinsic good of engagement from Arendt's writings, and an argument about the instrumental good of engagement from Tocqueville. Ultimately, Berger finds the instrumental good argument derived from Tocqueville to be less problematic. He argues that it is difficult to build a case for the intrinsic good of engagement without running into internal contradictions that are absent from the more pragmatic Tocquevillian approach.

As I am someone more rooted in game theory than in classical theory, the instrumental versus intrinsic good question brought an entirely different set of scholarly works to mind, and those works made the intrinsic good argument seem almost like a straw man. Game theory is based on utility maximization. We specify each actor's utility function and the structure of strategic interaction, and derive some sort of “equilibrium” in which each actor plays an optimal strategy given the other actors' strategies. So what goes into someone's utility function? According to a restrictive definition of rationality, only factors from which actors derive “instrumental” utility based on pure self-interest should be included in utility functions. According to a more loose definition of rationality, any factor from which people derive even psychological benefits (i.e., “intrinsic good”) can go into a utility function. In fact, this debate played out among game theorists most prominently in a context closely related to Berger's questions about civic engagement.

Consider voter turnout, which is arguably the central form of political engagement. Is it rational to vote? If we follow a strict definition of rationality, then it is rational to vote if and only if the instrumental benefits of voting outweigh the instrumental costs. The instrumental benefit is the product of one's utility for one's chosen candidate defeating the alternative (B), and the probability that one's vote will be pivotal (P). The problem is that in an electorate of any significant size, that probability, P, will be zero. Hence, PB will also be zero. So if there is any cost, C, to voting, it follows that PB < C, in which case voting is irrational.

Alternatively, suppose we adopt a more flexible definition of rationality. For example, what if voters receive some psychological benefit, D, from the act of fulfilling their civic duty? In that case, it is rational to vote if PB + D > C. While P will always be zero, D may not, and so the inequality can be satisfied under reasonable assumptions. This latter modeling trick has been referred to derisively as “the D term” and, in recent years, game theorists have been reluctant to include parameters in utility functions that represent intrinsic rather than instrumental benefits because doing so stretches the bounds of economic analysis. If psychological benefits constitute rational self-interest, then all actions are rational. If so, then rational choice theory predicts everything and therefore nothing.

The claim that political engagement is an intrinsic good because it enriches our lives is not markedly different from the claim that voters derive utility from the psychological satisfaction of fulfilling their civic duty, and the claim that political engagement can only be instrumentally good is not markedly different from the claim that voters derive actual utility only instrumentally from the direct consequences of their actions. Like game theorists, Berger ultimately concludes that arguments built around intrinsic good are unsatisfying, and he winds up making a case that game theorists made years ago using the approach of traditional rather than formal theory. Thus, game theorists may be surprised by the persistence of the debate, but the fact that traditional theorists have not resolved the issue suggests that it is important for a traditional theorist to make the case using the methodology of that subfield.

Attention Deficit Democracy is not purely theoretical, and it does wade briefly into empirical analysis, although more in an effort to debunk overly simplistic, deterministic claims than to test for general tendencies, and Berger raises a number of interesting questions that could be subjected to more systematic empirical analysis than what is presented in the book. Moreover, the fact that he argues so strongly for the “instrumental good” approach to evaluating civic engagement puts a burden on him to consider whether or not various forms of engagement do, empirically, provide any instrumental good. For example, he contends that increased political engagement across the broad spectrum of society is not necessarily important, but that increased political engagement from disadvantaged groups is important for instrumental reasons. However, making that argument in a compelling form requires showing, empirically, the instrumental benefits of political engagement for these disadvantaged groups.

Consider economic circumstances. Different countries have different levels of political engagement by groups with lower socioeconomic status. Are the countries with the highest levels of political engagement by these groups also the countries with the policies most beneficial to these groups, or even with the most upward mobility? In tangible terms, what do such groups actually gain from political engagement? This is an empirical question, and Berger's emphasis on the theoretical importance of instrumental benefits suggests that his argument may actually hinge on the answers to such questions. However, since the empirical analysis only aims to debunk the simplified and deterministic versions of civic engagement claims, such analysis does little to address the question of whether or not increased political engagement by disadvantaged groups is, on average, instrumentally beneficial to those groups.

Ultimately, I am convinced that “civic engagement” encompasses a broad array of behaviors, not all of which are necessarily beneficial. However, I am not ready to abandon the term. Fresh vegetables are healthy. Fast food burgers are not. However, I am unwilling to abandon the word “food” simply because it encompasses both. We simply need to reject the claim that “food is intrinsically healthy.” The term civic engagement encompasses concepts that are not uniformly beneficial for democracy, either intrinsically or instrumentally. Thus, we must recognize that blanket statements about its supposed benefits are overstated at best and misleading at worst, given the ambiguity of the term, and I applaud Berger's effort to add analytic rigor to discussions that frequently devolve into reflexive paeans to civic engagement.