Social Class as a Predictor of Public Opinion (Pre-Recorded)

• In Event: Class, Economic Inequality, & Public Opinion

Abstract

Social class is one of the key concepts of social science. Political scientists have used it to predict people’s voting behavior in democratic elections and their policy preferences. However, the concept of class has come under siege. On one hand, it is argued that class bounds are weakening. One example is that being a member of the working class used to be a solid predictor of voting for a left party but has lost this predictive power in our days. On the other hand, it is argued that other measures of social stratification have grown in importance and may even outperform traditional concepts of class when it comes to political behavior and preferences. There is a large strand of literature—especially from scholars of Political Economy—that investigates policy preferences by replacing class as the independent variable with other measures of inequality, such as risk of income loss, of becoming unemployed, of automation or off-shoring. The majority of these works restrict the phenomena they try to explain to social or redistributional policy preferences; some focus on openness towards immigration.
The paper proposed here takes one step back from that and examines empirically whether it is true that the concept of class lost its predictive power over the last decades. It does so by looking at survey data from a period of almost 40 years and from three different countries with arguably very different class structures: the US, Sweden and Britain. Election studies from these three countries were fielded regularly since around the year 1980 and contain detailed information on respondents’ social background but also on a variety of policy issues.
In a first step, the paper examines whether the absolute differences in preferences have become greater over time or if they have converged. The second step then is to see whether the predictive power of class has diminished over time. In addition, the design allows not only to test different operationalization of the class concept but also to control for other potentially meaningful measures of stratification, e.g. income and education, as proposed by McCall and Manza (2011).
This study makes three important contributions. First, it widens the scope on public opinion. Much of the literature referred to above restricts itself to preferences from one or two particular policy domains and a single country, namely the US. This study uses survey items from a variety of different policies and from three countries over several decades. Thus, it gives a more detailed insight into different classes’ wants and demands, which can help to better understand social conflicts and their changes. Second, it puts the frequent talk about (assumed) growing polarization on empirical grounds: do we indeed see a growing gap between opinions? And if so, how big is it really? Finally, the study helps to decide whether we should abandon the traditional class concepts and turn to newer measures of stratification when studying preferences in the political realm or not.

Author

• Svenja Hense, Goethe University Frankfurt