Disentangling Constituency and Legislator Effects in Legislative Representation: Black Legislators or Black Districts?


  • *I thank Dick Fenno, Stephen Gent, Bill Hixon, Gail McElroy, Dick Niemi, Valeria Sinclair-Chapman, Harold Stanley, and Antoine Yoshinaka for comments and suggestions. Data and coding information are available on request.

Direct correspondence to Christian R. Grose, Assistant Professor, Vanderbilt University, Department of Political Science, Calhoun Hall, VU Station B #351817, Nashville TN 37235-1817 〈christian.grose@vanderbilt.edu〉.


Objective. Does the race of a legislator or does the black population of a district best predict legislative roll-call voting in the interests of African Americans? Due to methodological limitations, no prior study has found that both the race of the legislator and the black district population are significant predictors of congressional roll-call voting. Drawing on post Shaw v. Reno/Miller v. Johnson congressional districts (with greater data variance), I examine the effect of these two racial representation variables on roll-call voting in the 104th–106th Congresses.

Methods. Linear regression with random effects is employed in two statistical models.

Results. Even when the black district population and party are considered, the presence of an African-American legislator leads to greater substantive representation of black constituents.

Conclusion. Districting plans that maximize the election of black legislators and Democrats are the most important for the aggregate enhancement of liberal voting in Congress, while districting plans that maximize black district populations and Democrats are the most important for the aggregate enhancement of civil rights voting records in Congress.