Measuring Group Alignment in Public Opinion

Issue polling typically reports group opinion one item at a time. However, democratic governance demands that groups act on agendas and form coalitions across multiple issues simultaneously. This paper introduces a non-parametric measurement framework for group alignment: the strength of a particular group's cumulative support across a basket of policy proposals. The framework yields several complementary metrics centered around cumulative alignment, each capturing a distinct facet of within-group opinion cohesion. Applying the framework to the Cooperative Election Study (2008-2024), I find that cross-issue alignment is remarkably low even in an era of sharp partisan sorting: while individual policy items command majority coalitions of 60–90% support among co-partisans, the share of cumulative majorities across every item in the core battery is nearly zero. Democrats exhibit modestly higher alignment than Republican, and Republicans experienced a decline in alignment during the first Trump presidency. Across racial, generational, and educational groups we similarly identify more- and less-aligned groups which are robust to alternative issue baskets and nonresponse assumptions. The framework offers researchers a transparent and interpretable method for summarizing intra-group opinion structure, tracking cohesion over time, and identifying the fault lines along which political coalitions may fracture.