| This thesis tackles the question of how organizational structure affects organizational performance from both a theoretical and empirical perspective. From the theoretical standpoint, I develop a closed-form mathematical model that builds on prior work by Sah and Stiglitz (1986) on organizational forms. I extend their original formulation and explore a broad range of decision-making organizational structures that lie between the extreme structures of hierarchy and polyarchy (essentially a flat organizational structure) that Sah and Stiglitz considered. The model adds realism to the work of Sah and Stiglitz by extending it from two to N individuals, allowing for more complex reporting relationships, and predicting not only Type I and II errors (omission and commission errors, respectively), but innovativeness, speed, and expected profits. The model has far reaching implications, as all organizations have to decide which organizational structure they will adopt. One important result is that not all organizations are efficient; some organizational structures should always be preferred over others, leading to the existence of "an efficient frontier in organization design." The implications of the model are discussed in light of the organization design and the innovation literature.;The empirical part of the thesis tests some of the predictions of the model using over 150,000 stock-picking decisions made by 609 mutual funds during two and a half years. Mutual funds offer an ideal and rare setting to test the theory because, as funds are heavily scrutinized, very detailed records exist on the projects they face (possible investments), the decisions they make or do not make (buying or not buying each of these possible investments), and the outcomes of these decisions. The independent variable of the study, the organizational structure of each mutual fund, is coded from textual descriptions of the fund management made by Morningstar. The main dependent variables of the study are measured as the probability of missing good assets (omission error) and of buying bad assets (commission error). The empirical tests show strong and robust support for the model, suggesting that organizational structure has relevant and predictable effects on a wide range of organizations. |
