Voting Behavior between Class, Geography, and Structural Economic Changes

Two main perspectives have animated political science research on the relationship between economic conditions and voting behavior in democracies over the past decades: class politics and the economic vote. We recognize the value of these two toolboxes, which at the same time have shortcomings that we address constructively. The class politics literature tends to use an overly static approach: the focus is on the political consequences of being in a position in a distribution (e.g., of income or skills) at a given point in time, rather than conceiving income and labor market trajectories as a major factor shaping life experiences, and consequently political behavior. When choosing how to vote, it might not matter just where a given voter is, but where they perceive to be going. We discuss dynamics that lead to the collective decline of the standard of living and the labor market outcomes of a given social group. The economic vote literature, conversely, has not assigned a sufficiently important role to the heterogeneity of experienced economic conditions, both in terms of position in social stratification and in terms of economic geography. In other words, the perception of whether the economy is “doing well” or not depends on who one is and, importantly, on where they live. Two important insights on these themes emerge from the discussion. First, vulnerability to structural changes is a fundamental feature of the experience of individuals in terms of the economy. A definition of class that is useful to understand voting behavior, particularly in the current historical moment, needs to incorporate this vulnerability in its analysis. The second insight is that the political-economic geography dimension is key. Importantly, trajectories over time in an individual's income or labor market outcomes are affected in a substantial---and often inescapable---manner by place, and so are political stances and choices. We draw on an empirical innovation that we have introduced to study vulnerability to technological change, by which individual exposure depends on the occupational profiles an individual is predicted to have based on one’s demographic and human capital characteristics, and the occupational patterns prevailing in the labor market of their region of residence. This predicted exposure, we argue, better captures a “class-based” effect on political behavior than actual current occupational characteristics of the individual. We provide a more theoretical grounding to this approach, linking it to its precursors in the empirical work on occupational risk in Rehm (2009) and on the “personal unemployment rate” of Ansolabehere et al. (2014), but also making explicit its connection to classic ideas of class as related to collective destinies and objective conditions: ultimately, class has more to do with what is reserved to people like oneself than with where one currently is. At the same time, economic dynamics, albeit heterogeneous, are highly consequential for vote choice, as per the economic vote approach. We outline an argument that links economic drivers and voting and illustrate a set of empirical predictions for voting behavior in response to globalization and technological change.