Descriptive Representation and the Composition of African American Turnout


  • A previous version of this article was presented at the 2005 annual meeting of the American Political Science Association, Washington, DC. We thank Ben Highton for congressional candidate data, Keith Poole for DW-NOMINATE data, and Jim Snyder for campaign finance data. David Campbell, Zoltan Hajnal, Rodney Hero, Cindy Kam, Brian Newman, Gary Segura, and the AJPS reviewers offered very helpful comments. Mallory Brown provided valuable research assistance.

John D. Griffin is assistant professor of political science, 217 O'Shaughnessy Hall, University of Notre Dame, Notre Dame, IN 46556 ( Michael Keane is a Ph.D. student in political science, 217 O'Shaughnessy Hall, University of Notre Dame, Notre Dame, IN 46556 (


Though many studies have focused on African Americans' turnout levels in descriptively represented electoral districts, few have examined the composition of African-American turnout in these districts, compared to districts that are not descriptively represented. This study contends that descriptive representation should conditionally affect African Americans' political participation, given preference heterogeneity among this group. It then examines the extent to which the ideological orientations of African Americans condition the effect of their Representative's race in the 104th House on their probability of participating in the 1996 national election. The study finds that when liberal African Americans are descriptively represented, they are more likely to vote, while moderate and conservative African Americans are less likely to vote. These findings not only help to resolve prior studies' disparate conclusions concerning descriptive representation's participatory effects, but they also show that descriptive representation affects which African Americans' interests are communicated to elected officials through voting.