Specialization without Comparative Advantage

    Using an evolutionary model of coalition building among independent social actors, this paper demonstrates a case where the value of information is ambiguous.  This study serves as a caveat to scientists who wish to model information indiscriminately as an asset to its users.  It suggests that the value of information to anyone depends on what information others possess and how they choose to apply it.  In the real world, where it is rare to have perfect insight into what others know and how others make decisions, acting ignorantly can be strategically advantageous, our model suggests.  As a model of social coalitions, this model has not been tested empirically and thus it is uncertain to what extent it describes the behavior of political parties, social classes, ethnicities, and races.  At this point, it is simply a model of the cooperative and competitive behavior of independent agents who share a common heuristic, but who based their decisions on different sets of knowledge.  I will describe this project in four parts.  First, I will describe the building and functioning of the artificial society on which this experiment was carried out.  Second, I will discuss what parts of the model were controlled and varied for this experiment.  Third, I will summarize the results.   Lastly, I will interpret the findings, address unanswered questions, and speculate on how the finding relate to social reality.  I would like to thank Susanne Lohmann, Nick Gessler, John Zaller, Tom Schwartz, James Honaker, James Tong, and Mark Sawyer for support and inspiration.


Jeffrey K. Shih
University of California in Los Angeles
Department of Political Science
jeff_shih@yahoo.com