|
From: | Karen May |
Subject: | [Swarmfest2007] paper submission for Swarmfest 2007 |
Date: | Thu, 31 May 2007 12:22:21 -0700 |
(Additional authors: Travis Coan, Mirya Holman, Myungjae Kwak, Rita Chiang, Wen-Yang Chang)
ABSTRACT:
This paper presents an Agent-Based Model (ABM) of the political process using the NetLogo platform to investigate the dynamics of information and preference exchange among voters and politicians, and the effects of these dynamics on policy outcomes. The model focuses on two broad agent types, voters and politicians, and three levels of analysis—voter interaction, politician interaction, and policy votes. Voters are divided into six categories, depending on information levels and preferences. As interactions occur, voters are influenced and can change both preferences and voter type. The median voter position influences politician preferences, which are also affected by personal beliefs and a measure of power. Policy “votes” take place at regular intervals and this information is fed back to voters, who decide whether to vote the incumbent out. Equilibrium behavior is largely consistent with empirical evidence that mainstream party voters comprise the largest segment of the voting public, and rates of incumbency are relatively stable. Additional scenarios such as polarized policy environments, sticky preferences, and impediments to information are also explored.
[Prev in Thread] | Current Thread | [Next in Thread] |