Introduction

In contrast to other chapters in this volume, the main purpose of this chapter is not to examine how networks are formed. Here, we take social networks as predetermined by nonmarket forces (such as ethnic, religious, and family affiliation) and investigate how market exchange is partly shaped by these social networks. Prior to market exchange, social networks exist but are devoid of economic content: little if any valuable information is available to be channeled through the network, and reputation mechanisms are not in place. As networked agents start trading with each other, however, they gain valuable experience about each other's type. As a result, parts of preexisting networks get activated as business networks and acquire economic value. It is this process we focus on.

The inspiration for this chapter comes from ten years of research on market institutions in Africa. Microevidence from surveys of manufacturing firms and traders in various parts of the world shows that markets do not operate in the manner predicted by standard economic textbooks. Legal institutions are not important, and courts are seldom used to enforce commercial contracts – partly because the amounts involved are small and partly because debtors have no assets to foreclose upon (Fafchamps 1996; Bigsten et al. 2000; Fafchamps and Minten 2001a; McMillan and Woodruff 1999a). Prevention is the dominant solution to the contract enforcement problem; this generates efficiency losses (e.g., [Fafchamps, Gunning, Isaksson, Oduro, Oostendorp, and Patillo,]).