Date of Award

1-1-2011

Document Type

Campus Access Dissertation

Department

Sociology

First Advisor

Shane R Thye

Abstract

The use of majority rule voting in group decision making is pervasive within democratic societies. Groups, often acting on behalf of a population, must choose between competing alternatives and use majority rule to decide which alternative to support. In fact, classic social choice theory is predicated upon and motivated by the democratic tenant that group choice, social policy, or collective action should be based on the diverse preferences and information of its constituents and the most prevalent decision making rule for doing so is majority rule voting (Miller 1983). However, the earliest and most notable model showing the superiority of group over individual decision making using majority rule, the Condorcet Jury Theorem, assumes group members vote independently. However, CJT and its primary assumption stands in sharp contrast to social psychological accounts of real groups that tend to display correlated behavior that depends on members' characteristics relative to others. These observations raise serious questions regarding the applicability of CJT to real world collective decision settings. This research constructs a two stage logit based voting game to mathematically model how correlated behavior can arise in groups based on the social structures that emerge within them. In particular, the model shows how the level of correlation between group member's votes is determined by the social hierarchy and diversity within a group. Newly designed laboratory experiments as well as existing field data from a national election will test the conclusions of this mathematical model and provide evidence as to how social diversity impacts the efficacy of group decision making when groups employ majority rule.

Share

COinS