UC Berkeley

This dissertation is comprised of three studies that investigate the implications and determinants of firms' choice of organizational form.

In the first study, I present a model of a negotiated sales production process with two variations, whether the firm has a parallel or serial allocation of tasks, and whether they have a process for accounting for customers' return to the system. I predict, first, that a hierarchal sales process allows firms to capture profitable low valuation sales that are ignored by firms with a parallel process. Second, I predict that an information tracking process will allow firms to capture additional value from transactions that would have been completed anyway. I find support for these predictions in a dataset combining data on organizational details collected from a survey I conducted of 500 US auto dealers and transaction-level data on auto sales at those dealerships.

The second study investigates the allocation of control rights by firms. I present a model of a multidivisional firm faced with a choice relating to its divisions. The managers of those divisions have more information about the most productive choice for their division, but there is value to coordinating the choices. I predict that tasks with high coordination values will be more likely to be centralized and that for tasks with lower coordination values, delegation is more likely when the manager has a greater information advantage, which manifests in the volatility of the environment.

The third study proposes that vertical integration between manufacturers and lessors can generate externalities that improve the competitiveness of competing independent lessors. Often, manufacturers provide warranties in the sales market to inspire confidence in their customers. Because they are unable to observe the identity of customers, however, these same warranties can be used to recondition cars returned to independent lessors from leases shorter than the warranty. The ability to free-ride on maintenance cost in this segment of the market allows independent lessors to overcome some of the captives' informational advantage. I find support for this proposal in a dataset of 200,000 leases from 1997-2002.