{"id":{"repo_id":"gmu","oai_identifier":"oai:MARS:1920/15232"},"canonical_url":"https://search.dev.ndltd.org/etd/gmu/oai:MARS:1920/15232","repository":{"repo_id":"gmu","name":"George Mason University","base_url":"https://mars.gmu.edu/server/oai/request"},"display":{"title":"Agent-Based Simulation of Election Reforms in U.S. Congressional Elections","abstract":"One enduring problem for scholars and policymakers is to determine the optimal and fair design of election rules for single-member legislative districts. A series of election reforms have been proposed to address gerrymandering and political polarization. This dissertation develops an agent-based model to explore simulated voting behavior under current and proposed election rules. A base model demonstrates the interaction between candidate strategies and the distribution of voter ideology. Certain strategies become more or less successful based on the particular balance of voter ideologies, and these same parameters contribute to higher or lower voter turnout. A machine learning model then demonstrates an innovative method for identifying and quantifying the NOMINATE-based political ideology of legislative candidates using legislative speeches and campaign statements. The resulting model can be readily used in an ABM. These candidate ideology values are combined in the next step with voter ideologies probabilistically assigned according to the demographics of specific congressional districts. These values are developed to replicate the 2020 U.S. House elections in Virginia at the congressional district level. Each district produces a combination of parameters that replicates the real results of the 2020 elections in each respective district. The interaction between strategy choice and distribution of ideology among the district’s voting population in context continues to affect election results and voter turnout. The successful election replication supports using the empirically-grounded agent-based model to simulate the effect of election reforms. A final model demonstrates this opportunity by simulating the top-two election reform using Virginia’s 2020 congressional districts. The result of these simulations are shaped by the candidate strategies and the underlying ideology of voters in the district, as well as the tolerance for centrist candidates. The models developed for this dissertation can serve as a framework for expanding into the future study of other types of election reforms, such as ranked choice voting. These efforts collectively set the stage for the next major task of simulating more complex voting systems and contribute to computational social science as a discipline by developing a reliable, empirical model to advance the study of an important social process (voting) and policy area (elections). The model has implications to the study of how to redraw congressional districts for the goal of balancing districts ideologically.","abstract_html":"One enduring problem for scholars and policymakers is to determine the optimal and fair design of election rules for single-member legislative districts. A series of election reforms have been proposed to address gerrymandering and political polarization. This dissertation develops an agent-based model to explore simulated voting behavior under current and proposed election rules. A base model demonstrates the interaction between candidate strategies and the distribution of voter ideology. Certain strategies become more or less successful based on the particular balance of voter ideologies, and these same parameters contribute to higher or lower voter turnout. A machine learning model then demonstrates an innovative method for identifying and quantifying the NOMINATE-based political ideology of legislative candidates using legislative speeches and campaign statements. The resulting model can be readily used in an ABM. These candidate ideology values are combined in the next step with voter ideologies probabilistically assigned according to the demographics of specific congressional districts. These values are developed to replicate the 2020 U.S. House elections in Virginia at the congressional district level. Each district produces a combination of parameters that replicates the real results of the 2020 elections in each respective district. The interaction between strategy choice and distribution of ideology among the district’s voting population in context continues to affect election results and voter turnout. The successful election replication supports using the empirically-grounded agent-based model to simulate the effect of election reforms. A final model demonstrates this opportunity by simulating the top-two election reform using Virginia’s 2020 congressional districts. The result of these simulations are shaped by the candidate strategies and the underlying ideology of voters in the district, as well as the tolerance for centrist candidates. The models developed for this dissertation can serve as a framework for expanding into the future study of other types of election reforms, such as ranked choice voting. These efforts collectively set the stage for the next major task of simulating more complex voting systems and contribute to computational social science as a discipline by developing a reliable, empirical model to advance the study of an important social process (voting) and policy area (elections). The model has implications to the study of how to redraw congressional districts for the goal of balancing districts ideologically.","abstract_has_math":false,"creators":["Hammer, Michael"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025","date_published":"2025","updated_at":"2026-07-27T19:51:54Z","subjects":["agent-based modeling","computational social science","elections","politial science","political behavior","voting"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["hdl:1920/15232"],"render_values":[{"text":"hdl:1920/15232","href":null,"code":true}]}]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null},"metadata_groups":[{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2025"]},{"key":"dc:type","label":"Dc Type","values":["Dissertation"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["agent-based modeling","computational social science","elections","politial science","political behavior","voting"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["hdl:1920/15232"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.other","label":"Dc Description Other","values":["One enduring problem for scholars and policymakers is to determine the optimal and fair design of election rules for single-member legislative districts. A series of election reforms have been proposed to address gerrymandering and political polarization. This dissertation develops an agent-based model to explore simulated voting behavior under current and proposed election rules. A base model demonstrates the interaction between candidate strategies and the distribution of voter ideology. Certain strategies become more or less successful based on the particular balance of voter ideologies, and these same parameters contribute to higher or lower voter turnout. A machine learning model then demonstrates an innovative method for identifying and quantifying the NOMINATE-based political ideology of legislative candidates using legislative speeches and campaign statements. The resulting model can be readily used in an ABM. These candidate ideology values are combined in the next step with voter ideologies probabilistically assigned according to the demographics of specific congressional districts. These values are developed to replicate the 2020 U.S. House elections in Virginia at the congressional district level. Each district produces a combination of parameters that replicates the real results of the 2020 elections in each respective district. The interaction between strategy choice and distribution of ideology among the district’s voting population in context continues to affect election results and voter turnout. The successful election replication supports using the empirically-grounded agent-based model to simulate the effect of election reforms. A final model demonstrates this opportunity by simulating the top-two election reform using Virginia’s 2020 congressional districts. The result of these simulations are shaped by the candidate strategies and the underlying ideology of voters in the district, as well as the tolerance for centrist candidates. The models developed for this dissertation can serve as a framework for expanding into the future study of other types of election reforms, such as ranked choice voting. These efforts collectively set the stage for the next major task of simulating more complex voting systems and contribute to computational social science as a discipline by developing a reliable, empirical model to advance the study of an important social process (voting) and policy area (elections). The model has implications to the study of how to redraw congressional districts for the goal of balancing districts ideologically."]},{"key":"dc:title","label":"Title","values":["Agent-Based Simulation of Election Reforms in U.S. Congressional Elections"]}]}],"canonical_facts":{"dc:date.issued":["2025"],"dc:description.other":["One enduring problem for scholars and policymakers is to determine the optimal and fair design of election rules for single-member legislative districts. A series of election reforms have been proposed to address gerrymandering and political polarization. This dissertation develops an agent-based model to explore simulated voting behavior under current and proposed election rules. A base model demonstrates the interaction between candidate strategies and the distribution of voter ideology. Certain strategies become more or less successful based on the particular balance of voter ideologies, and these same parameters contribute to higher or lower voter turnout. A machine learning model then demonstrates an innovative method for identifying and quantifying the NOMINATE-based political ideology of legislative candidates using legislative speeches and campaign statements. The resulting model can be readily used in an ABM. These candidate ideology values are combined in the next step with voter ideologies probabilistically assigned according to the demographics of specific congressional districts. These values are developed to replicate the 2020 U.S. House elections in Virginia at the congressional district level. Each district produces a combination of parameters that replicates the real results of the 2020 elections in each respective district. The interaction between strategy choice and distribution of ideology among the district’s voting population in context continues to affect election results and voter turnout. The successful election replication supports using the empirically-grounded agent-based model to simulate the effect of election reforms. A final model demonstrates this opportunity by simulating the top-two election reform using Virginia’s 2020 congressional districts. The result of these simulations are shaped by the candidate strategies and the underlying ideology of voters in the district, as well as the tolerance for centrist candidates. The models developed for this dissertation can serve as a framework for expanding into the future study of other types of election reforms, such as ranked choice voting. These efforts collectively set the stage for the next major task of simulating more complex voting systems and contribute to computational social science as a discipline by developing a reliable, empirical model to advance the study of an important social process (voting) and policy area (elections). The model has implications to the study of how to redraw congressional districts for the goal of balancing districts ideologically."],"dc:identifier":["hdl:1920/15232"],"dc:subject":["agent-based modeling","computational social science","elections","politial science","political behavior","voting"],"dc:title":["Agent-Based Simulation of Election Reforms in U.S. Congressional Elections"],"dc:type":["Dissertation"]},"updated_at":"2026-07-27T19:51:54Z"}