Does class still matter? As we noted in Chapter 1, the 'class' question has long been central to sociological research on politics and political behavior. Part of the explanation for this emphasis is that many sociological studies of voting have started from the assumption that all social cleavages are in large part a product of material inequalities. Scholars of U.S. politics have generally agreed that, outside the South, the New Deal party system rested in large measure on economic class divisions. This conventional wisdom also asserts that since the 1950s, electoral class divisions--alongside the New Deal party system itself--gradually began to erode with postwar societal affluence. This decline is often viewed as accelerating in the 1960s as a result of intense social conflicts over issues such as civil rights, gender equality, and the environment. The idea of the 'Reagan Democrat' (working-class supporters of Republican presidential candidates) and the 'new class' liberal (educated middle-class Democrats) gained currency in both journalistic and scholarly accounts of this period.
In this chapter we re-examine such widely held assumptions about the role of class divisions in American electoral politics. We begin with a historical overview of class politics. We pay particular attention to the different ways in which class has been conceptualized in earlier research. These concepts, as we will see, usually lead to the finding that class divisions are either insignificant or declining in importance over time. Next, we present an alternative approach for studying class voting. This approach builds from recent research on conceptualizing class in other social science subfields. In contrast to almost all of the existing political behavior literature, we employ a structural, multi- category class scheme that is not based on simple income groups, a blue-collar/white-collar distinction, or subjective class identification. We then discuss the statistical models and indices we use to measure overall and group-specific trends in class voting.
Following these preliminaries, we turn to our two main empirical questions. First, to what extent has the class cleavage declined over the