Firms' boundaries and the division of labor: empirical strategies

A large theoretical literature focuses on the question: What determines firms' boundaries? Recently, Garicano and Santos and Holmstrom and Milgrom have proposed theories in which firms' boundaries reflect the division of labor across individuals. This paper discusses strategies for generating and testing empirical propositions from this emerging class of theories. We propose that variation in the returns to specialization is extremely useful to test and quantify the effect of the division of labor on firms' boundaries. We discuss the use of two shifters to the returns to specialization, and relate how we exploit one of these, market size, in preliminary work on the determinants of law firms' boundaries in the United States