Estimating Seats–Votes Partisan Advantage
20 Pages Posted:
Date Written: December 30, 2021
A new class of metrics is defined that measure whether redistricting plans favor one political party or the other: partisan advantage using a seats–votes benchmark. Ten prominent measures of partisan bias are evaluated against a cross section of nearly three dozen past, present, and hypothetical congressional plans. Only (dis)proportionality and the efficiency gap are shown to reliably measure this across a wide range of statewide vote shares. When winner's bonuses are between one and two inclusive, plans are shown to have acceptable bias with respect to the efficiency gap ideal of 2-proportionality.
Keywords: redistricting, bias
Suggested Citation: Suggested Citation