MONEY AND MIDDLEMEN IN AN ECONOMY WITH PRIVATE INFORMATION

I. INTRODUCTION Money and middlemen are frequently involved in transactions. Both play an important role in facilitating trade. To understand the medium-of-exchange role of money and middlemen, one needs to be precise about the trading frictions which lead to their existence. While many studies focus on the role of intermediaries in matching buyers and sellers,(1) qualitative uncertainty and its impediments to exchange have also been identified as important elements underlying the use of money and emergence of middlemen. For example, Brunner and Metzler [1971] claim that, in an economy where there is uncertainty about the quality of goods, intermediary institutions such as money and specialized traders can reduce the information cost. Many attempts have been made to provide theoretical frameworks of the intermediaries of exchange.(2) Recently, search-based models have been used to formalize the trading frictions and study the role of money and middlemen. For example, Kiyotaki and Wright [1993] show that the absence of double coincidence of wants in bilateral trades can give rise to a role of fiat money as a medium of exchange. Rubinstein and Wolinsky [1987] demonstrate how middlemen help to overcome the search friction. Williamson and Wright [1994] study the function of fiat money in ameliorating the trading friction caused by private information concerning the quality of consumption goods. Li [1998] uses the same informational frictions to motivate the role of middlemen.(3) However, there are no papers that have endogenous roles for these two competing institutions, money and middlemen, and it should be worthwhile to study how they interact. In Williamson and Wright [1994] generally recognizable fiat money can increase welfare in an economy with private information because it leads to agents adopting strategies that increase the probability of high-quality commodities being exchanged. The reason is that the seller recognizes money but may not recognize another commodity and therefore, people have a higher chance to get high-quality goods when they buy with money rather than barter. In Li [1998] I show in a pure barter economy that under certain circumstances, middlemen emerge endogenously to ameliorate the lemons problem in the exchange of goods. Middlemen facilitate exchange by increasing people's incentive to produce high-quality goods and by bringing consumers the trading opportunities which they might not realize without middlemen. Thus, middlemen can improve welfare despite the fact that intermediation takes away the resources that could have been employed in production. While providing explanations for the role of individual intermediary institutions, previous studies leave the following questions unanswered: given that there is only one type of trading friction, if there exists one intermediary medium of exchange, can another intermediary still play some role in overcoming the friction? To be more specific, in an economy with private information, would individuals be willing to engage in indirect exchange involving money even though there exist expert middlemen? To what extent can money reduce the function of middlemen? Can the use of generally recognizable fiat money reduce people's incentive to invest in a quality verification technology and become middlemen? To address these issues, we extend the Williamson-Wright model to study the interaction between money and middlemen in an economy with private information. The model is described as follows. Producers choose to produce high- or low-quality goods, with low-quality goods being cheaper to produce but less desirable for consumption. Agents meet randomly in pairs over time, they carry unit inventories (either high-quality goods, low-quality goods or money) and they trade when it is mutually agreeable. In any meeting, an agent may or may not recognize the quality of goods offered in trade. Hence, producers may have an incentive to take advantage of the lower cost to producing low-quality goods. …