A Reasonable Bias Method For Redistricting: A New Tool for an Old Problem

• In Event: Racial and Partisan Gerrymandering: New Approaches for the Next Decade?

Abstract

Despite several decades of persistent efforts by academics, reformers, and a few political officials, there is still no consensus in the US about how best to measure and judge the partisan fairness of any proposed districting plan. We argue that advances in modern computing and a “reasonable bias” standard for judging partisan fairness are the best hope for solving this problem. A “reasonable bias” approach accepts that political fairness has to be traded off with other important redistricting values. It identifies the lower boundary of fairness, excluding plans that are below the threshold value but allowing for plans that are as good or better than the floor value. In this paper, we will first compare our reasonable bias standard to previous attempts to define political gerrymandering, and argue that our approach allows for a jurisdiction to determine how much it wants to tradeoff partisan fairness with other political values. Secondly, we will introduce the Tam et al algorithm and show how it can generate a set of constitutionally feasible plans to provide a comparison set for judging fairness, and an analytical tool to examine the effects of different redistricting standard. Thirdly, we illustrate the value of this approach by applying it to Minnesota Congressional redistricting. We show that the partisan bias measures that have been proposed do not overlap very much, similar to the problem with different compactness measures. Finally, we conclude by suggesting ways that this approach could be tried out in the next round of redistricting.

Authors

• Bruce E. Cain, Stanford University

• Wendy K. Tam Cho, University of Illinois at Urbana-Champaign