Flattened resource allocation, hierarchy design and the boundaries of the firm.

摘要

The general theme of my dissertation is to understand why and how institutions matter in the efficiency of organizations. Empirical results suggest that multi-segment firms underperform stand-alone firms in resource allocations by underinvesting in high-profit segments and overinvesting in low-profit segments—a phenomenon I called flattened resource allocations (henceforth FRA). In the first two chapters I provide a theoretical rationale for FRA and examine its relationship to hierarchy design. The third chapter changes focus to explore the role of the boundaries of the firm in the absence of property rights.; In the first chapter - “Flattened Resource Allocations, Separation of Ownership and Control, and Diversification of the Firm”, I present a model of FRA. This model is more robust than existing FRA models because in the existing literature. I show that FRA is a negative externality existing in multi-segment firms but not in stand-alone firms. Also, consistent with the facts from empirical studies, the inefficiency of FRA increases in both the agency cost and diversification of the multi-segment firm. There can be both underinvestment and overinvestment in the total resource available to the firm. These results hold when management ownership is either exogenous or endogenous. In the latter case, the pattern of FRA becomes less clear-cut. More interestingly, the CEO can obtain more than he would receive in a stand-alone firm, explaining why CEOs might prefer managing multi-segment firms. Finally, the total resource available to the firm is lower when management ownership is endogenous than when it is exogenous.; In the second chapter - “Hierarchy Design with Flattened Resource Allocations”, I compare the efficiency of flat and tall hierarchies from the perspective of FRA. I allow the possibility of divisionalization—grouping elementary business segments into a fewer number of divisions and transforming flat hierarchies to tall hierarchies. I characterize the equilibrium of this model and the inefficiency of FRA is shown to be decreasing in management ownership. Most importantly, FRA is aggravated by divisionalization, implying tall hierarchies are necessarily less efficient than flat hierarchies. This result suggests that U-form (undivisionalized) organizations, represented by flat hierarchies, can actually outperform M-form (divisionalized) organizations, represented by tall hierarchies, in resource allocations—a result opposite to Williamson's “M-form hypothesis.”; In the third chapter - “The Boundaries of the Firms in the Absence of Property Rights” it is shown that the boundaries of the firm can continue to matter in an environment where there is no physical asset and hence property rights do not play a role. I assume that the boundaries of the firm work as a form of information barrier: once parties enter the boundaries of the firm they work together behind ‘frosted-glass windows’ and hence their outside options get averaged out. I show that mergers can induce first-best investment. This new view provides a rationale for the large observed investment in adjusting firm boundaries in industries where physical assets are of minimal importance.