Revisiting Economic Action and Social Structure: the Role of Embeddedness in the Age of Amazon
Abstract: Network scholars posit that networks provide reputational enforcement superior to market reputation (i.e. the reputation that would emerge from a sample of market participants). In this view, reputational signals are local to social network neighborhoods: they are stronger in the presence of shared 3rd parties and decay with network distance, as a result of two mechanisms: (1) it is costly to identify and communicate with socially distant others; (2) there is little reason to trust them. Consequentially, reputational sanctions and norm enforcement are also local to the network. The rise of reputational aggregation platforms (RAP) such as Amazon and eBay calls these mechanisms into question. This paper identifies an additional mechanism driving network-based reputational enforcement, that places a boundary condition on both network and market reputation-based reputational enforcement. It argues that the informativeness of a reputational signal from A to B about C is increasing in B’s belief that A used the same norms B would in making her evaluation of C. The closer A’s norms are to B’s, the more informative the signal. Though some exchanges rely on widely shared norms, many exchanges rely on locally shared understandings between exchange partners. While such norms can and do differ across actors, thereby diluting the informativeness of reputational signals, there are good reasons to believe that norms, and thus reputations and reputational enforcement are local to social network neighborhoods. The more local the norms, the more local the reputation and the less informative is market reputation (and thus, RAPs). The more universal the norms, the more informative is market reputation, and the greater the ability of RAPs to enforce norms.
