Many economic models assume that markets are frictionless - they operate at 100% efficiency and require little interaction between the buyer and the seller. Actual markets, however, face moderate to severe trading frictions, and reuly on various sorts of links between buyers and sellers. Theorists have recently developed new models to deal with these issues, but there is as yet very little empirical knowledge about the effects of networks on market performance.
This project will gather such knowledge using a controlled laboratory market. The PIs will introduce two important market frictions into a highly efficient auction trading procedure. They will create two separate settings. In the first, sellers may cheat buyers on product quality. In the second, buyers may not be aware of all seller prices, so buyers may miss some low cost purchase opportunities. They will then "link" buyers and sellers through a network that is designed to mimic social network links. Buyers and sellers withing the same network know who else in the network is a cheater and who is a low cost seller. The goal is to determine how frictions affect market outcomes and whether networks can overcome these frictions to again achieve efficiency.
Better knowledge of how networks affect market performance will provide a sounder basis for policy reform in several fields. Policy makers interested in increasing the volume of domestic trading in electronic marketplaces or in financial assets will benefit from these results. In addition, networks among a social group have been posited as an explanation for differences in how immigrants assimilate to American society. This study will help to understand how networks could have such effects. A diverse and global community of undergraduates at three different institutions -- Mt. Holyoke College, the University of San Francisco, and the University of California at Santa Cruz -- will participate as research assistants.
