The successes of the Federal Communications Commission (FCC) auctions for electromagnetic spectrum licenses have been trumpeted far and wide as a case study in the successful application of game theory to public policy. Finding the evidence for such assertions lacking, this dissertation instead seeks to approach the spectrum auctions as a case study in the economics of science. Chapter 1 peruses existing accounts of the FCC spectrum auctions, finds a lack of consensus both on the role of game theory and on the performance of the FCC auctions, and establishes the central task of the dissertation as evaluating the role of academic mechanism designers in acting as 'consulting engineers for the market economy.' Chapter 2 reviews the evolution of the communications industry up to the period immediately preceding the FCC auctions, for the purpose of understanding the policy approaches mechanism design would replace, who supported the new approach, and why. Chapter 3 reviews the field of mechanism design, and finds three discrete and to some extent incompatible versions distinguished by modeling strategy and, importantly, by intended user of the expertise. Chapter 4 studies both the published accounts of the FCC mechanism design process as well as the unpublished materials in the FCC archives, and finds that the strategic imperatives of a handful of large telecommunications companies displaced the scientific imperatives of mechanism design. Chapter 5 reviews the empirical evidence for the performance of the spectrum auctions, and finds that policy imperatives, too, fell by the wayside. Chapter 6 concludes.