9411025 Miller This research projects concerns theories of electoral competition between political parties competing for public office in democracies. Electoral politics are dynamic. This research examines the dynamic interactions between election rules, public opinion, and the behavior of political parties. A distinguishing feature of the proposed research is that it employs computer models to discover patterns of political party behavior over a series of elections and across electoral institutions. The models in the project are advanced forms of computer simulations that are new to the social sciences though they have been used extensively in the natural sciences. Among their advantages, these models allow for flexibility and descriptive accuracy in developing theories of political behavior. The current research will continue to develop theories of electoral competition through the use of computers and mathematical modeling, and then test these theories using survey data from the United States and Europe. The research consists of three parts. The first part is the computer modeling, which continues previous research of the collaborative research team. This research has already generated explanations and predictions about the effects of electoral institutions and public opinion on the behavior of political parties, both in the way parties campaign for public office and in the policies they enact once in office. The second part of the project involves using mathematics to substantiate the findings from the computer model. using both techniques, the researchers have so far confirmed the validity of earlier computer findings and made several predictions of political party behavior. The third part uses survey data from election studies in the United States and Europe to test the predictions generated by the computer models. One central theoretical finding from their previous research has generally been supported by survey data from the American Nati onal Election Study 1972-1988. The collaborators intend to undertake more research on real election data as a portion of this research project. The research has two main objectives. First, to understand more fully the way different democratic institutions create incentives for political elites to change governmental policies to appeal to voters. Knowledge from this research may be helpful in the design of constitutions for emerging democracies. The second objective is to extend the use of computer modeling in the social sciences, and to establish computational standards for testing social science models for completeness and accuracy.
