The economic voting literature is vast. With a search on Google Scholar of “economic voting” yielding more than 3,000 articles, one would imagine it quite difficult for a scholar to break new ground. Yet Brad Lockerbie does so in his new book, not by providing a novel theoretical framework but by revisiting classic ones with an expansive examination of the importance of prospective evaluations for federal elections in the United States. As the title of the book reveals, he weighs in on one of the two most important and long-standing debates in the economic voting literature: Do citizens vote retrospectively, looking back rather myopically to past performances of incumbent government officials? Or do they vote prospectively by utilizing a wider array of political information to form expectations about the party that will provide better outcomes in the future? As Lockerbie notes, the answer to this question is hardly trivial but, rather, speaks directly to larger concerns over the meaning of election outcomes, the competence of voters, and democratic accountability. (The other important question, whether citizens vote sociotropically or out of their own personal economic well-being, is tackled, somewhat secondarily, in Chapter 6, to be discussed).

The overall thrust of his results is clear: U.S. voters do look toward the future, evaluating both the Democratic and Republican Parties' ability to provide for future economic success. More often than not, retrospective evaluations operate primarily by influencing prospective ones. While neither of these conclusions is particularly surprising given the existing literature, the sheer exhaustiveness of the author's analyses—he covers all presidential elections from 1956 to 2000 and congressional elections from 1956 to 2002—ensures that they cannot be dismissed as the result of particular eras, elections, offices, or candidates.

The organization of the book is straightforward: Each chapter investigates how retrospective and prospective evaluations influence a particular dependent variable of interest. Chapter 2 explores the bivariate relationship between retrospective and prospective evaluations. Chapter 3 investigates the effect of these evaluations on party identification. Next, Chapters 4 and 5 explore the significance of these evaluations for presidential and congressional vote choice, respectively. Chapter 6 examines the relative importance of sociotropic versus egocentric prospective evaluations. And lastly, the final empirical chapter departs from respondent-level analysis to investigate how successful prospective evaluations are, when aggregated, in predicting aggregate election outcomes.

The literature review is tight and focused, relaying what the author considers to be important contexts for larger normative, positive theoretical, and methodological debates, rather than an exhaustive summary. Those looking for a complete guide to the vast, varied, and often seemingly contradictory economic voting literature, will have to turn elsewhere.

Chapters 3 and 4 are perhaps the most insightful due to Lockerbie's thoughtful efforts to confront the issue of partisan rationalization. The concerns are twofold. First, retrospective evaluations, as many other researchers have demonstrated, are influenced noticeably by respondents' partisan loyalties. Democrats just do not think that the economy has been as bad as Republicans do when a Democrat is president, and vice versa. The second is that expectations about the future may be better predictors of political behavior than retrospective evaluations only because it is easier for partisans to imagine better (or worse) scenarios about future performance of their own (or opposing) party when unconstrained by current realities or recent macroeconomic fortunes. Prospections may just be an opportunity to indulge partisan inclinations.

Lockerbie's individual-level statistical models demonstrate that such partisan rationalization plays an important role. In 1960 and 1980, for example, these “indirect” effects of past party identification are equal to or greater than the total effects (indirect plus direct effects) of either retrospective or prospective evaluations themselves on current party identification. This is clever analysis that demonstrates the potential for evaluations to be colored by partisan rationalizations, while simultaneously showing the limitations of this explanation for an understanding of the full political importance of voters' prospections. Once partisan rationalization is accounted for, not only do these prospections influence individuals' party identifications, but their influence is also significantly greater than the retrospective evaluations so prominently noted in Morris Fiorina's Retrospective Voting in American National Elections (1981).

Chapter 6 may seem slightly out of place in investigating whether voters' prospective evaluations are egocentric or sociotropic. Its analyses convincingly demonstrate that both personal (or familial) judgments and sociotropic ones play a significant role in predicting a variety of electoral behaviors and attitudes—results that may be unexpected in the context of the existing literature. However, by investigating data only from the 1992 elections, this chapter fails to echo the others in their thoroughness and, therefore, generalizability. The book is not without other shortcomings. For example, the author compares the R²s among models with different numbers of independent variables (p. 44). Also, given the significant findings of prospective evaluations for House and Senate elections, I would like to have seen comparable partisan rationalization analysis, since that proved so illuminating in the chapters on party identification and presidential vote choice.

Ultimately, though, these concerns are small compared with the great effort taken by Lockerbie to ensure that most conclusions are generalizable beyond one particular election or political era. For those unsure of the power of retrospective evaluations to explain voter behavior in the United States, this book is quite possibly the most encyclopedic attempt to document their importance in presidential, House, and Senate elections over the past half century.