Models for Two-Party and Multi-Party Elections

Building on work from Gary King and Andrew Gelman, I've modelled two party election systems in the United States and investigated whether the electoral maps as drawn , as well as how responsive a system is to a change in elector preference.

• JudgeIt II: A Program for Evaluating Electoral Systems and Redistricting Plans, with Gelman and King. An R package for modelling two-party election systems across multiple election years, this extends on the original JudgeIt package written in Gauss.
  • For a recent example of JudgeIt used to verify redistricting plans, see Fryer and Holden's 2007 paper on generating compact redistricting plans.
• Partisan Bias of the Electoral College Under Proposed Changes in Elector Apportionment, with Gelman, King and Jonathan Katz, in progress (draft available upon request). [2008] What if electoral votes were divided by congressional district within certain states? We demonstrate that (ceteris paribus) modifications to the current electoral system do not benefit partisan symmetry or responsiveness.
• Measuring The Impact of Independent Redistricting Commissions on State Legislative Elections, with Scott Lesowitz. Working paper [2007]. Independent commissions are associated with higher responsiveness in state legislative elections, but appear to have no impact on partisan symmetry. Note that due to the small number of states using commissions, and the difficulty in finding counterpart states whose legislatures perform redistricting, we cannot yet say anything causal about the implementation of a commission in a state that lacks one.
• Relational data in Political Science. See the relational data page for more information.