| Basically, my dissertation focuses on License Auctions. Four chapters of my dissertation are theoretical analysis of license auctions. Broadly speaking, I analyze the effects of different auction rules on revenue, efficiency and social welfare.;In the second chapter, I show that if the cost asymmetry between the bidding firms is large enough, then having fewer firms in the market will surprisingly result in higher social welfare. This result is contrast to the common or general case in which "social welfare" will be higher if there are more firms competing in the market.;In the first two chapters, I study the optimal auction designs when the firms bid for only one object. From the third chapter of my dissertation, I extend the analysis to the cases in which some bidders want more than one different object (multi-unit auctions). In the third chapter, I characterize the optimal bidding strategies of local and global bidders for two heterogeneous licenses in a multi-unit simultaneous ascending auction with synergies. In such auctions, some bidders want to win the bundle of licenses to enjoy synergies. So they may bid more than their stand-alone valuation of a license. This exposes them to the risk of losing money since they may win parts of licenses. I determine the optimal bidding strategies in the presence of an exposure problem and show that global bidders may accept a loss even when they win all licenses and moreover, if a "bid-withdrawal" rule is introduced to the auction, the exposure problem disappears, and the simulation results show that revenue will be higher.;In the last chapter, I study the Canadian AWS auction in which 40 percent spectrum are set aside for new firms. I characterize the effect of spectrum set-aside auctions on seller's revenue, consumer surplus and social welfare. In a two-license simultaneous asceriding bid auction model, I show that a spectrum set aside may not only encourage new entry and increase competition in the downstream market, but also under some circumstance, decreases the seller's revenue and consumer surplus. But a spectrum set aside results in inefficient allocation, and this inefficient entry further reduces social welfare.;The first chapter studies the flaw in the design of the 2000 Turkish GSM (Global System for Mobile Communications) auction. In this auction, the Turkish government wants to raise as much revenue as possible and to increase competition in the cell-phone market by selling two licenses to new firms via a sequential auction, but it ends up with only one license sold. I identify this auction design failure. And I also show that if the auction were designed as a "simultaneous auction" , the government would sell two licenses and receive more revenue. |
