Gerrymandering the Swing Voters

About this event

Gerrymandering is the strategic redrawing of the boundaries of congressional and state legislative districts. We develop a new theoretical model of gerrymandering. We embrace the fact that there is a difference between the total population of a state and its voting population by allowing for heterogeneous turnout among voters. We show that parties have incentives to exploit these differences to their advantage, for instance by combining low turnout opponents with high turnout supporters, and isolating their lowest turnout supporters in safe districts. The presence of safe districts in favor of the party that controls redistricting cannot be explained easily by existing models of gerrymandering, though we see them in reality. To understand if the mechanisms in our model are important in practice, we collect geospatial data for proposed maps from the 2020 redistricting cycle. This most recent redistricting cycle offers a new opportunity to compare proposals from Democrats and Republicans in the same state. With this novel dataset, we find support for several key predictions from our theoretical model. For example, we see that precincts are treated differently in Republican and Democrat redistricting proposals, depending on the precinct’s (predicted) turnout rate.