Essays in Regional Economics and Natural Resource Management

My dissertation consists of three essays that examine the empirical issues in Regional Economics and Natural Resource management. The first essay proposes a solution to the problem of using small datasets to predict regional fiscal conditions. I formulate a "Model-Averaging" framework via Bayesian methods and test it on a Vector Autoregressive Model (VAR) using a fiscal dataset for Nevada counties. I show that informative priors from neighboring counties are effective in lowering the variance of the posterior distributions of the individual counties. Furthermore, I show that the "Model-averaged" predictive posteriors via the marginal likelihoods of each individual model enhance the prediction accuracy. I find that region-pairs with similar revenue and expenditure levels generate better predictions than region-pairs with fiscal gaps.;The second essay investigates the "Jump bidding" behavior in the National Wild Horse and Burro Internet Adoptions in response to the problem of unsatisfactory adoption rate and climbing maintenance costs that the Bureau of Land Management (BLM) is facing. A Poisson model based on the number of aggressive bids each adopter submits reveal that "Jump-bidding" is related to the bidding quota assigned by the Bureau of Land Management (BLM). Adopters of higher allowance tend to bid more aggressively than others. The results also indicate that only certain types of animals are receiving aggressive bids, and that the level of competition is intensified by aggressive bidders. This eliminates the possibility that aggressive bidding discourages competition and concludes the auction early as previous literature suggest. I also show that "Jump-bidding" is a major booster of the animal selling price, this contradicts the previous findings that jump bidders have negative impacts on auction revenue.;The third essay explores "Late bidding" behavior in the National Wild Horse and Burro Internet Adoptions. I examine the relationship between late bidding and its impact on the level of competition and adoption price. I show that the number of bids each bidder submits drives the late bidding occurrence, this indicates animals receive higher levels of bidding competition attract more "Snipers". I find that bidders of larger bidding quota are less motivated to be "Snipers", and the colors that draw the interest of "Snipers" are similar to colors that draw the attention of "Jump bidders". I also show that the final selling price is inversely related to the submission time of the winning bid. Animals that are adopted at a later time are sold at higher prices. This implies a symmetric distribution of the valuations for the wild horses and burros that are adopted via the Internet.