Social networks, firm behavior, and industry evolution: A study of professional service firms

This dissertation examines the causes of industry evolution: adoption of innovation, mergers and acquisitions, and organizational dissolution. The research setting is the professional service firms, especially the Dutch accounting industry. Data on CPAs' demographic characteristics and their organizational affiliation were collected from CPA directories. The main body of this thesis consists of three empirical studies based on the data. To understand the research setting, I first survey the peculiarities of professional service firms and a partner-associate structure (PA structure). The first empirical study gives attention to the antecedents of the adoption of the PA structure. I distinguish population level factors from firm level ones, and at each level examine the role of "technical efficiency" and "legitimacy" conditions in accounting for adoption. Empirical analysis shows that adoption propensity is positively associated with market signaling, the level of PA diffusion, complementary needs, absorptive capacity, and social networks. The results also suggest that the market signaling, a technical efficiency factor, has a stronger influence during early diffusion periods, while the level of PA diffusion, an institutional factor, has stronger effect during later periods. The second study examines the events that firms founded by mergers and acquisitions (M&As) experienced. Investigation of the history of large Dutch accounting firms shows that most of them were created by M&As. It indicates that M&As have been a route to the emergence of large accounting firms. The multinomial logit analysis shows that firms founded by an M&A of complementary and compatible firms performed better than others. The results indicate that firms that were unable to handle internal variations brought about by previous M&As experienced high dissolution rates. Firms that were successful in dealing with those variations were more likely to utilize M&A specific knowledge by engaging in additional M&As. The third study explores the empirical validity of density dependence hypothesis by controlling for fine-grained organizational factors in explaining organizational dissolution rates. Empirical analysis shows that the density has a strong U-shaped relationship with organizational dissolution even when fine-grained organizational level variables are controlled. The analysis provides a strong support for the density dependence hypothesis. In conclusion, I summarize the findings of the three empirical studies and discuss what we learn from the studies.