Speaker:  Cyrus Hettle, Georgia Tech

Title:  Mathematically Quantifying Gerrymandering in

          Georgia’s Congressional Redistricting

Abstract:

While gerrymandering has been widely suspected in Georgia for years,

it has been difficult to quantify. We generate a large ensemble of

randomly generated non-partisan maps that are sampled from a

probability distribution which respects the geographical constraints

of the redistricting process. Using a Markov chain Monte Carlo process

and techniques involving spanning trees, we can quickly generate a

robust set of plans.



Based on historical voting data, we compare the Georgia congressional

redistricting plan enacted in 2021 with the non-partisan maps. We find

that the 2021 plan will likely be highly non-responsive to changing

opinions of the electorate, unlike the plans in the ensemble. Using

additional spatial analysis, we highlight areas where the map has been

redrawn to weaken the influence of Democratic voters.



This talk is based on joint work with Swati Gupta, Gregory Herschlag,

Jonathan Mattingly, Dana Randall, and Zhanzhan Zhao.