Polarization Alters Descriptive Representation: Democrats Reward Women and Minority Members

Why This Matters: Descriptive representation—the idea that voters prefer representatives who share their race or gender—has long been a foundation of research on representation. Anna Weissman shows that rising partisan divergence in attitudes toward marginalized groups can disrupt this pattern: when one party grows more favorable toward outgroups while the other does not, partisanship can matter more than voters’ own ascriptive traits in shaping evaluations of legislators.

What Weissman Asks: Do changing attitudes about race and gender alter the usual advantage that co‑racial or co‑gender representatives enjoy? The paper tests whether positive affect toward marginalized groups among Democrats, combined with stable or less positive attitudes among Republicans, shifts constituent responses to members of Congress by identity.

Data and Design: The study analyzes Cooperative Congressional Election Study (CCES) survey responses from 2008–2020. Weissman leverages a difference‑in‑differences design to compare changes over time in constituent approval of members of Congress who are women or from racially marginalized groups, testing whether these changes differ by respondents’ partisanship and by respondents’ own race or gender.

Key Findings:

• Democrats—regardless of their own race or gender—now report higher approval of Congress members from historically marginalized groups.
• Republican respondents’ approval is not meaningfully related to representatives’ race or gender over the same period.
• Democratic respondents also show greater tolerance when women and minority representatives deviate from partisan ideological expectations, giving those members more ideological leeway.

These results indicate that partisan shifts in attitudes about race and gender can overturn the standard expectation that voters most favor co‑group representatives.

What This Means: Polarized changes in group attitudes reshape how descriptive representation operates in practice. For scholars, candidates, and parties, the findings imply that who benefits from shared identity depends on partisan context as much as on voters’ own demographics, with implications for candidate recruitment, messaging, and theories of representation.