Larry Bartels and Chris Achen have written a path-breaking book, which synthesizes several decades of research on the shortcomings of democratic politics. The book is unique in our discipline, due to its ability to bring together empirical research covering a broad range of topics in voting behavior and electoral politics and normative research about the foundations of democratic politics.

Achen and Bartels critique two standard explanations of democratic politics, which they label as the folk theory of democracy. The first approach is the spatial model of voting, which goes back to the contribution of Anthony Downs. This model of voting presupposes that citizens with well-articulated preferences choose representatives that best represent these preferences. Bartels and Achen also critique retrospective explanations of voting, which presuppose that citizens sanction the poor performance of incumbents, while selecting more competent politicians.

Bartels’ and Achen’s most provocative critique of the model of accountability is presented in Chapter 5. Here Achen and Bartels analyze the electoral consequences of a shark attack that occurred on the New Jersey shore in 1916. Achen and Bartels document that the shark attack contributed to a decline in the electoral support for President Wilson of over 10 percent, even though no politician could have prevented such attack. Bartels and Achen interpret such punishment as evidence of blind electoral retrospection, which is defined as the decision of voters “to kick the government when they are in pain … so long as they can justify doing so with whatever plausible constructions are available to them” [page 118].

The second part of the book articulates an alternative theory that may account for the sources of voters’ loyalties. The authors stress that the primary sources of partisan loyalties are social identities and group attachments, not abstract policy preferences. The authors illustrate the importance of group alliances as the driving force of voter preferences by examining how changes in group alliances precipitate partisan alliances at critical junctures in American political development. Drawing on original research and synthesizing a large body of existing research, they show that different ethnic groups responded to the New Deal in different ways, that Catholic and Protestant voters responded differently to the nomination of JFK, and that the salience of the “white southern identity” explains variation in partisan realignment among Southern voters during the 1960s.

One possible objection to the overall construction of the book is that its two parts are only very loosely integrated. Achen and Bartels formulate the group-based theory of voter’s preferences only in the second part of the book. The argument that group memberships shape and inform the beliefs, preferences, and political behavior of democratic citizens unfolds only there. It is quite difficult to understand how the propositions developed there affect the findings presented in the first part of the book. One wishes the authors had integrated the findings from the different parts of the book better.

Consider some of these tensions that follow from the incomplete integration of the two parts of the book. While the analysis of the electoral fallout of shark attacks is an extremely powerful metaphor, I wish the authors had provided more evidence that attempted to tease out the factors that mediate the intensity in voters’ anger. Given the centrality of this type of example for the argument advanced in the book, I wish the authors had followed up other more recent examples with similar electoral properties that could have allowed them to tease out the factors that might account for the intensity in voters’ anger and the variation in the ways in which voters attributed responsibility for such events. Bridging to the second part of the book, I wish Bartels and Achen had discussed whether and how differences in group identity matter for cases of “blind retrospection.” Does group membership constrain blind retrospection? If so, how strong is such constraint and what are its implications for electoral accountability?

Similarly, one wonders how the insights about the salience of group identities inform the results of responsiveness during the events of the Great Depression, which are presented in Chapter 7. The group theory of politics that is proposed in the second part of the book implies that voters’ response to the Great Depression was likely to vary across different groups of voters and that the pro-Roosevelt swing in political support of African American occurred at a different point in time as compared to the shift in support of support among voters from other ethnic groups. Some of the variation in the differential responsiveness of different groups of voters may be affected by the policies adopted by the Roosevelt administration. Such differential group realignment comes into tension with the argument of myopic retrospection faced by Bartels and Achen, yet the authors do very little to address such tension between the two arguments they are advancing.

Consider also some of the comparative examples used in the book. At different points Achen and Bartels use different explanations for the same political outcome, but offer no discussion on how to reconcile these competing explanations. Consider how Achen and Bartels use the framework of the book to explain one of the most momentous cases of democratic failure: the election of Adolph Hitler in Weimar Germany. In Chapter 7, the authors present this case as an example of myopic retrospection and present aggregate results linking decline in economic output and support for the National Socialist Party. At other points, the authors note that support for the National Socialist Party varied tremendously across different subgroups of voters. In addressing these findings, historians of the period have discussed the importance of different cultural millieus, an explanation which is consistent with Achen and Bartels’ group-based theory politics. Bartels and Achen do not resolve the tension between myopic retrospection and a group-based explanation of Nazi vote. If the voting calculations are so imbued by group attachments, can one still speak of myopic or blind retrospection? In short, it is not entirely clear how the two parts of the explanation work together in illuminating this case.

These are small quibbles that should not detract us from recognizing the monumental importance of this contribution. I cannot think of any other book that can engage so many different scholars working in so many subfields. And even if one does not agree with all of its conclusions, the book challenges and forces us to provide stronger justifications for many assumptions we take for granted in our research.