Simulating Elections in the Context of Voter Behavior and Election Rules

Electoral systems manage complex social phenomena, including representation, consent, social unrest, political stability, economic development, and many others. One enduring problem for scholars and policymakers is determining the optimal design of single-member legislative districts in order to maximize representation and minimize social conflict. This paper contributes to the general theory and practice of election simulations by demonstrating a method for identifying voter ideological ideal points in a microsimulation model using real election data. This model is calibrated with a sample of voters in five state-level legislative districts using real election results from the 2020 U.S. general election. The paper demonstrates that real election results can be replicated by modeling emergent agent preferences using simple agent characteristics and election rules. The value of this replication is then in using the synthetic electorate to test changes in voter behavior and election results under different election rules.