Local Income Inequality and Citizens' Demand for Redistribution

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

Abstract

Income inequality has risen in most developed democracies over the last decades, and this has in turn prompted concerns over its consequences for individual well-being as well as social and political cohesion. Crucial for understanding the extent of the negative ramifications of income inequality, is the question of whether it is of a self-correcting or a reinforcing nature through its effects on citizens’ demand (preference) for redistribution. Prominent theories in the field yield opposing predictions. On the one hand, the influential Meltzer-Richard model from political economy predicts that citizens would demand more redistribution when faced with higher inequality. Conversely, the system justification theory from social psychology predicts that citizens should be content with—and even endorse—high levels of inequality to rationalize their own situation.

Reflecting both theories’ plausible nature ex ante, the supporting empirical evidence, while substantial, is inconclusive. Further, previous analyses suffer from a number of interrelated problems, which question the validity of existing evidence in the first place. First, previous studies have studied income inequality at exceedingly aggregate contextual levels, often using countries as contextual units. Given lessons from neighboring literatures, it is highly questionable whether such aggregate indicators in any meaningful way reflect people’s actual experiences of inequality. Second, partly following from being studied at highly aggregate contextual levels, it is difficult to tease apart effects of income inequality from other aspects of that context such as the sociodemographic and ethnic composition. Third, given their cross-sectional nature, most previous studies have been unable to rule out self-selection (i.e. individuals sorting into more or less unequal contexts based on factors also related to political preferences) as an alternative explanation of an observed relationship.

Against the shortcomings of existing studies, we examine the relationship between income inequality and citizens’ redistribution preferences using a much more rigorous research design. More specifically, we link several large georeferenced survey datasets containing citizens’ redistribution preferences with individual-level administrative data on, inter alia, income and geolocation of the full population in Denmark. This enables us to flexible create measures of income inequality in respondents’ immediate residential environment down to within meters of their place of residence. Consequently, we can examine whether and how locally experienced inequality is associated with a demand for redistribution, while simultaneously controlling for other related aspects of the context. Further, some of the survey data have a panel structure, which enable us to study the relationship between inequality and redistribution preferences within individuals over time, thereby bypassing any time-invariant confounding. This considerably strengthens our ability for drawing causal inference compared to previous analyses, generally relying on cross-sectional data. Collectively, our approach enables us to conduct a highly rigorous analysis of the consequences of locally experienced income inequality for citizens' redistribution preferences.

Tentative empirical results indicate that income inequality has a negative—that is, a reinforcing–effect on demand for redistribution. Individuals living in more local unequal residential contexts, tend to support redistribution less. This result is independent of one’s own economic situation (i.e., not different for rich or poor individuals), and primarily driven by economic inequality between high-earners and median income earners. Further explorative analyses find evidence consistent with the system justification theory as an explanation for the observed negative relationship between local income inequality and redistribution preferences.

Authors

• Matias Engdal Christensen, Aarhus University

• Peter Thisted Dinesen, UCL/University of Copenhagen

• Kim Mannemar Soenderskov, Aarhus University