The study of economic voting is one of the most well plowed fields in political science. After decades of debate and refinement, and scores of books and articles, one often wonders if there is anything new to be said on the subject, at least in broad strokes. In The Economy and the Vote, however, Wouter van der Brug, Cees van der Eijk, and Mark Franklin have managed to offer a genuinely new and insightful perspective on the topic. Their work, while confirming much received wisdom in the field, highlights serious deficiencies in existing scholarship and calls for a reassessment of theoretical conceptualizations and empirical modeling strategies. This book moves the study of economic voting forward in a significant way, both breaking new ground and resolving old puzzles.

The central premise of the book is that political considerations, including economic conditions, are translated into electoral choices in a two-stage process. Drawing heavily on Anthony Downs's (1957) An Economic Theory of Democracy, the authors argue that voters develop a “propensity to support” each party in a political system based on the interaction of their individual characteristics, preferences, and judgments with the characteristics, histories, and issue stances of the parties (which may include strategic considerations). They then vote for the party for whom they have the highest support propensity (measured directly in the studies that the authors use). In this model, it is important to note, an issue, event, or evaluation can have significant effects in the first stage without ultimately shifting an individual's vote choice. A strong Labour partisan in Britain, for example, might have a “baseline” support propensity (on a 10-point scale) of 9 for Labour, 2 for the Tories, and 4 for all other parties. Because of economic difficulties under a Labour government, he or she might drop the support propensity for Labour to 6 but, assuming that all other parties remain stable, would still be expected to vote Labour. Conventional economic voting models, the authors argue, would count this voter as unaffected by economic considerations, because the vote choice did not actually change. Such a conclusion, however, masks a genuine shift in preferences, and the stability in voting behavior results not from indifference to the economy but from characteristics of competition within the party system and the individual's preexisting partisan preference structure.

This important insight, and the authors' accompanying decision to use respondent-party pairs as the unit of analysis, allows them to diagnose significant problems with existing work and to shed light on inconsistencies in previous findings. Their analyses are based on data from the 1989, 1994, and 1999 European Elections Studies, allowing them to examine the dynamics of economic voting under coalition governments (the norm in Europe). They are able to demonstrate compellingly that the traditional dichotomy between government and opposition parties, with its accompanying expectation that the former would suffer and the latter benefit from bad economic times, is a vast oversimplification of reality, accounting for many null and inconsistent findings in previous studies. Government versus opposition status is only one of many variables shaping how parties are affected by economic conditions, with party size, ideological position, attributional clarity, and competitive situation within the party system being at least as important.

As the authors acknowledge, none of these insights, taken individually, is new. This study, however, is the first to “integrate the notions put forward by others in the context of a research design that would permit the implications of these ideas to become evident, and simultaneously to take account of all these variables in a single model” (p. 115). Moreover, their respondent-party analytical frame allows the authors to speak to both individual and aggregate-level studies in an edifying and coherent way. Chapter 6 specifically provides a clear demonstration of the ways in which very similar relationships between economic conditions and political attitudes across countries can produce very different electoral outcomes, depending on the electoral context and the structure of individual partisan preference orderings. Reading this chapter produces some real “aha!” moments for those who have puzzled over the inconsistent findings of previous studies.

At the same time, however, the book tends to gloss over certain potential complications for the theory. To begin with, it completely ignores the possibility of rational abstention as a response to decreasing partisan utility differentials. Given the Downsian roots of the preference model, this is a somewhat surprising omission. The authors are quite right that if an individual's “support propensity” differential between Party A and Party B decreases from 8-5 to 6-5, that individual should still rationally vote for Party A. Anthony Downs, however, would argue that the individual in question would be less likely to show up at the polls, because the strength of his or her preference for one electoral outcome over the other is markedly less. Clearly, differential turnout levels can significantly shape electoral outcomes, but the authors completely ignore this possibility. This omission is not fatal, and is indeed common to analyses of economic voting, but it should at least be noted in any “Downsian” account of the relationship between economic conditions and electoral behavior.

Another question to which the authors could give more attention is the potentially problematic relationship between partisan support propensities and accountability. They express doubts about using measures “whose link to normative concerns, such as government accountability, is questionable” (p. 193). Party support propensities, however, could be described as just such a measure. A governing party cannot observe shifts in voters' “support propensities”; it can only observe changes in actual votes (or shifts in turnout, as discussed previously). Thus, citizens who do not change their actual partisan vote choice have done nothing to create accountability in the system, regardless of how much their underlying support propensities may have shifted. One could argue that in the real world of politics, it is votes that matter, and that observing changes in support propensities is a mere abstract exercise. This disconnect is reflected by the authors' contention that, in at least one of their models, “all parties gain from unemployment” (p. 87). In a zero-sum electoral system, it is impossible for all parties to gain in a tangible way from any political development. On balance, the benefits from incorporating the idea of support propensities into the model of economic voting outweigh the costs, but it should be acknowledged that the concept is not unproblematic.

A final issue is that the book, on a couple of issues, speaks much more definitively than the data permit. For example, the authors dismiss the idea of pocketbook voting, despite having very poor measures of the concept. Because of their aversion to using self-reports of economic conditions, they are forced to look for pocketbook effects through the very rough proxies of unemployment and retirement. Because the jobless and retirees do not appear to respond to aggregate economic data differently than do other voters, they conclude that there is no pocketbook voting in the sample. This is a shaky foundation on which to base a strong assertion. In addition, they dismiss the idea of heterogeneity on the basis of sophistication, despite having what they admit are weak measures of the concept (education and interest, as opposed to actual political knowledge). Some caveats and restraint in these claims would be welcome.

While these concerns are not trivial, they should not detract from the overall message that The Economy and the Vote is an excellent and important work. It is rare that a book comes along that truly demands a significant reconceptualization of a long-standing field of inquiry. Van der Brug, van der Eijk, and Franklin have made just such a contribution.