| dc.description.abstract | This dissertation studies affiliation and entry in first-price auctions, accomplishing three goals. First, it develops a simple approach to test for affiliation among bidders' private information and applies the approach to the timber sales data in Oregon State. Second, it extends the affiliated private value (APV) model by taking entry into account with symmetric or asymmetric bidders. Lastly, it empirically analyzes the timber sales data in Oregon State by using the structural approach within the APV paradigm with entry and asymmetric bidders and studies several policy related issues by counterfactual analyses. The dissertation consists of four chapters. Chapter One emphasizes the importance of affiliation and entry in the auction study and reviews the related literature. It also introduces the timber sales in Oregon Department of Forestry and the data we use. Chapter Two proposes a simple approach to test for affiliation among bidders' private information using entry behavior. It is based on the insight that affiliation among private information causes affiliation among bidders' entry behavior and developed within the general affiliated value framework. The approach is easy to implement in practice and widely applicable. We apply the approach to the timber sales data and find that bidders are affiliated with a relative small level. Chapter Three studies the relationship between the number of potential bidders and bids. In doing so, we develop a two stage entry and bidding model with symmetric bidders. We find that in our setting, there are three different effects, namely "affiliation effect," "entry effect" and "competition effect," at work. The "affiliation effect" and the "entry effect" have the opposite directions to the "competition effect." Therefore the total effect on bids depends on the relative magnitudes of three effects. The last chapter extends the model in Chapter Three by considering asymmetric bidders. It establishes the existence and uniqueness of the equilibrium of the model. Compared with the model with symmetric bidders, the equilibrium bids do not have closed forms, which complicates the estimation, and therefore a numerical method is required. For the structural analysis, we adopt a two step indirect inference method to estimate the underlying distributions of the model. We apply the model to the timber sales data and examine several policy related issues through a set of counterfactual analyses. Two appendices in the end of the dissertation provide the details about the existence and uniqueness of the equilibrium and the numerical routine to solve the equilibrium bids and entry probabilities. | |
