The Federal Communications Commission (FCC) has been auctioning frequency licenses since 1994. To encourage long-run competition in the telecommunication industry, the commission granted bid discounts to the small firms in its spectrum auctions. The primary focus of the dissertation is to evaluate the effect of the program on the number of licenses won by small firms, and on the FCC's auction revenues. It also examines the consequences of access manipulations by large-firm affiliates which used legal loopholes to capture over $7 billion in discounts. The first chapter presents an overview of the FCC auction system and the discount program. It sets up the rest of the analysis. The second chapter of the dissertation applies a non-parametric estimation approach to recover the auction participants' value distribution through its bounds. Auction simulations using the distribution argue that the subsidy increases the share of small firm licenses from 25.9% to 27.6% at a low cost to the commission. They also imply that subsidy access by large firms has no significant effect on FCC's revenues but reduces the program's effectiveness by more than one third. The third chapter extends the analysis by introducing a parametric model that takes advantage of the bidder as well as license characteristics heterogeneity. My findings imply that perfect implementation of the program would triple the number of small firm licenses from 357, when no subsidies are present, to 942. It would also increase auction revenue from $31.89 billion to $34.17 billion. However, the access manipulations led to an increase in small firm licenses by just 266 for a total of 623, and a drop in FCC receipts to $30.21 billion. The results of my dissertation suggest that improving the program's eligibility regulations holds benefits for both small firms and the FCC. Tightening the commission's access criteria would increase welfare.