The recent automation of the American stock market has replaced floor intermediaries with trading algorithms, calling into question the sociological claim that markets are structured by networks of intermediaries. Our study examines the social nature of markets in automated settings with an inductive, qualitative study of the automation of the NYSE during the period 2003-12. It proposes the concept of blended automation to denote an automation design that preserves the social structure of a market. Our analysis of the Flash Crash of 2010 suggests that such design offers greater resilience to economic shocks. Our study contributes to the literature on technology in organizations by characterizing a novel automation design that reconciles technology with social relations, and contributes to economic sociology by outlining how automated markets can remain socially structured, pointing to role of politics, ideology and design in market automation.
