This is a book about the political choices that British voters make. More precisely, it is about the decisions that they make in successive general elections. In each election, voters have two choices—first, whether to cast a ballot, and second, which party to choose. Since the first British Election Study (BES) some four decades ago, the emphasis has been very much on the second of these decisions, party choice. The prior, turnout decision has received very little attention. In the 1950s and 1960s, neglecting turnout seemed reasonable, or at least acceptable, since the vast majority of people regularly voted in general elections. That has changed. In 2001, less than three in five members of the eligible electorate went to the polls, thereby clearly signalling that the turnout decision has become an important aspect of electoral choice in contemporary Britain. Turnout and party choice both merit attention.
Viewed generally, the literature on electoral choice is dominated by two theoretical perspectives—what we call the sociological and individual rationality frameworks. These two frameworks subsume a wide variety of explanatory models. Historically, most studies of voting behaviour in Britain and elsewhere have adopted, and then strongly advocated, a single model within one of these two frameworks. Ensuing empirical analyses typically, and predictably, demonstrate the power of the preferred model. We contend that this research strategy imposes undesirable theoretical costs because the explanatory contributions and potential superiority of rival models are ignored. These costs may not be recoverable because national election studies are a very scarce commodity. Almost always there is only one study per research community per election cycle. When used to guide the construction of the survey instruments, the 'single model' approach means that the data needed to investigate the utility of a range of alternative models are often unavailable.
In Political Choice in Britain, we specify and test several rival explanatory models of electoral choice. As principal investigators of the 2001 BES, we explicitly designed the survey instruments with this goal in mind. Since electoral participation had been largely ignored in previous studies, we constructed substantial batteries of new questions to generate the data needed to test alternative models