The Interface of Politics: How Social Groups Guide Political Choice

Dissertation Summary

At the heart of liberal democratic theory lies the notion that governments derive their legitimacy from their ability to represent the interests of individual citizens. Elections serve as the mechanism for this representative claim: through voting, individuals express their preferences and ensure they are equally weighted in the formation of government. Yet this vision contains a fundamental tension. While democratic governments are supposed to represent individual interests, their policy-making does not concern individuals as such. Public policies by definition always a!ect aggregates of individuals who share policy-relevant traits, creating a gap between the personal experiences of voters and the aggregate-level decisions of government. For individual interests to be represented, voters must somehow map their personal interests onto political choices: they must solve what can be termed the “translation problem” between the personal and the political. This dissertation explores how voters navigate this fundamental challenge. 

The dissertation’s main argument is that social groups function as a crucial interface between individual voters and the political system, helping voters solve the translation problem more effectively than relying on personal experience or national
conditions alone. The central research question asks: How do voters use social groups to evaluate policies and parties? The theoretical framework developed here posits that both policy-making and voters’ political interests operate at the group level, making social groups the natural point where citizens and governments interact. When politicians craft policy, they necessarily operate at the level of groups rather than individuals, targeting shared characteristics like age, income, or geography. Similarly, individual interests largely cluster into what we recognize as social groups, as people sharing sociodemographic characteristics tend to have overlapping interests across multiple dimensions. This clustered nature of both policy-making and interests creates conditions under which voters can learn the most about government alignment with their interests by relying on information about their social groups. The dissertation’s primary contribution is demonstrating empirically that group-based political reasoning represents a rational and prevalent response to the informational challenges voters face, challenging prevailing accounts that characterize group-based behavior as purely emotional or tribal.

The dissertation is article-based and consists of a project frame (Chapter 1–4) and four empirical research articles (Chapter 5–8), of which three are solo-authored and one is co-authored. The research articles are self-contained and can be read without reading the project frame. Likewise, the project frame summarizes the entire project by elaborating on the overarching argument and how each article fits within it, discussing existing literature on group-based political behavior, developing a formal theorization of the group interface of politics, discussing cross-cutting methodological challenges, and summarizing key findings and their implications for research and the workings of democracy.

The first empirical paper, “You and Whose Economy?: Group-Based Retrospection in Economic Voting,” investigates whether voters evaluate incumbent performance based on the economic conditions of their social in-groups. Using panel data from the British Election Study and three pre-registered experiments in Denmark and the United States, the study finds that voters systematically assess incumbent performance based on how their social groups have fared economically, especially relative to the national economy. This group-based retrospective voting operates independently of both personal financial circumstances and national economic conditions, suggesting important limits to purely sociotropic models of economic voting. The e!ects are comparable in magnitude to traditional sociotropic voting, indicating that group-based considerations represent a fundamental dimension of electoral accountability.

The second paper, “Who (Else) Benefits?: Group-Based Responses to Distributive Policies,” challenges the conventional assumption that voters respond to targeted government policies primarily through pocketbook considerations. Examining COVID-era stimulus policies in Denmark and the United States, alongside three pre-registered experiments, the study demonstrates that voters’ responses to material benefits are shaped at least as much by perceived in-group benefit as personal economic gains.

The e!ects vary significantly by group identity strength, with positive responses concentrated among groups with strong political identities. These findings help explain the mixed empirical record on distributive policy e!ects and suggest that the electoral returns to targeted spending depend critically on which groups benefit and how clearly this targeting is perceived. 

The third paper, “Elite Rhetoric and the Running Tally of Party-Group Linkages,” co-authored with Frederik Hjorth, examines how citizens form and update perceptions of which political parties represent particular social groups. Using a novel automated
approach with language models to measure group appeals in 1.6 million parliamentary speeches, the study connects party rhetoric to survey measures of perceived party-group linkages over three decades in the UK. The findings demonstrate that group linkages robustly track party elites’ rhetoric, with a one standard deviation increase in positive group appeals associated with approximately 12 percentage point improvements in perceived linkages. The study provides evidence that voters maintain “running tallies” of party-group linkages that respond systematically to elite communication. This challenges views of group linkages as fixed reflections of social cleavages and supports the broader claim that group identities function as the primary interface through which voters interpret party behavior and political representation.

The fourth paper, “Is Pocketbook Voting Sensitive to Policy?,” addresses a key assumption underlying pocketbook voting as a mechanism of democratic accountability. Using a novel approach that links survey panel data to policy microsimulation models,
the study decomposes respondents’ disposable income changes into policy-induced and residual components across a decade of British fiscal policy. The analysis reveals that voters fail to privilege policy-induced income changes over other income fluctuations when evaluating incumbents, instead responding indiscriminately to total income changes. Moreover, policy-induced changes constitute only a small fraction of total income variation, meaning that even indiscriminate pocketbook voting aligns with actual policy e!ects only slightly more often than chance. These findings suggest significant limitations in voters’ ability to hold governments accountable through personal economic experience, reinforcing the importance of group-based heuristics as alternative pathways for democratic accountability.