Why Group-Based Retrospection Matters
Christoffer Hentzer Dausgaard asks how voters evaluate incumbents when economic outcomes are distributed unevenly across social groups. The paper revisits the idea of retrospective economic voting—where voters reward or punish incumbents for past performance—but shifts the focus from national aggregates to voters’ social in-groups. This matters because if voters judge incumbents by how their own group fares relative to the nation, electoral incentives will shape not only overall economic policy but its distribution.
Theory: In-Group Comparisons Versus Sociotropic Judgment
The core argument is that voters evaluate the economy both sociotropically (looking at national performance) and through a group-based lens: they want their social in-groups—defined by geography, age, education, ethnicity, or class—to keep up with or outperform national trends. When in-group outcomes lag, voters are more likely to punish incumbents, creating an accountability channel focused on distributional outcomes rather than only aggregate growth.
How the Study Tests the Idea
The paper brings together two complementary research designs.
- Panel data analysis: The author estimates the relationship between in-group economic performance and incumbent support over time, comparing its size to standard sociotropic effects.
- Survey experiments: Three randomized-information experiments conducted in Denmark and the United States present respondents with manipulated information about how specific groups performed relative to the national economy (groups defined by geography, age, education, ethnicity, and class) to observe causal effects on incumbent support.
What the Data Show
- The panel analysis finds a positive association between in-group economic performance and incumbent support; the estimated effect is comparable in magnitude to sociotropic voting.
- Experimental evidence from Denmark and the U.S. confirms voters respond to information about group outcomes: respondents are less supportive of incumbents when their group is shown to lag the national economy and more supportive when their group is shown to meet or exceed national trends.
- These results indicate limits to purely sociotropic voting: voters care about relative outcomes for the groups they identify with, not just the national average.
Implications for Electoral Accountability and Party Competition
If voters punish incumbents when their social groups fall behind, politicians and parties face incentives to pay attention to distributional winners and losers, not just aggregate growth. This finding reshapes expectations about policy targeting, coalition-building, and how parties frame economic performance to different voter groups.
Author and Venue
The study is authored by Christoffer Hentzer Dausgaard and published in the American Journal of Political Science (AJPS).
