You and Whose Economy? The Neglected Role of In-group Economic Perceptions

Retrospective evaluations of the economy play a critical role in democratic accountability. It is among the most established findings in political science that voters reward incumbent politicians when the national economy improves and punish them when it declines. However, macroeconomic trends rarely affect everyone the same way. Political parties have different economic priorities, and as they enter government and implement their policy platforms, certain groups benefit while others lose out. Despite this fact, we know very little about whether and how voters react to changing conditions of their social in-groups based on e.g. class, ethnicity, gender, age and locality. The retrospective voting literature has largely ignored social groups, focusing on voters’ responses to changing national or pocketbook conditions instead. To address this shortcoming, this paper proposes a theory of ‘in-group economic voting’. Drawing on theories of social identities in politics, the paper argues that voters may care about group-level economic conditions either as a heuristic to assess how well the incumbent serves their interests or because of solidarity with in-group members. In addition, the paper argues that voters may have reason to care more about the relative than absolute conditions of their in-group, as this provides a stronger signal of the incumbent’s concern for the group. The paper then tests the theory with two empirical studies. The first study uses data from a large British survey panel to estimate effects of changing perceptions of group-level conditions within individuals across ethnic, class and geographical groups. To strengthen causal identification, the second study implements an original pre-registered and replicated survey experiment on large samples of Danish voters. The experiment uses true economic information about social groups to experimentally manipulate respondents’ group-level economic perceptions. By randomly varying combinations of negative signals about 22 different social groups and positive signals about the macro-economy, the experiment tests whether perceived changes to in-group economic conditions impact various outcome measures. Across both studies, the paper finds consistent and robust evidence that voters hold incumbents accountable for the plight of their social in-groups, independently of their own pocketbooks and national economic trends. Additionally, the paper finds that voters appear to be more sensitive to relative changes to in-group conditions than absolute changes across both studies. This is most obvious in the experiment, in which voters do not respond positively to national economic improvements if their in-group does not benefit. These findings suggest that the role of social groups in voters' evaluations of the economy has been underrated and raise questions for further research on this topic.