Mathematically Quantifying Gerrymandering and the Non-responsiveness of the 2021 Georgia Congressional Districting Plan

Abstract

While gerrymandering has been widely suspected in Georgia for years, it has been difficult to quantify. We understand a plan to be a partisan gerrymander if it produces election results whose outcomes and structures are not typically seen in maps drawn only with non-partisan considerations. In particular, our use of the term gerrymandering does not make reference to any concept of proportionality and includes how the spatial distribution of the state's electorate interacts with the redistricting process.Using historic voting data, we compare the Georgia congressional redistricting plan enacted in 2021 with a large collection of randomly generated non-partisan maps 1 (see Section 1.1 for details). Similar methodologies have been employed in various states, including North Carolina [9, 5], Ohio [13], Pennsylvania [10,14], Virginia [3], Maryland [8], and Wisconsin [6]. In each state, the methods are adapted to account for the state's specific redistricting requirements. We continue these explorations by examining the partisan behavior under Georgia's enacted 2021 congressional map and a collection of maps that were sampled without partisan consideration. Our main findings are as follows: