A Tale of Two Papers: How Exclusive Auction-Based Approaches Mean the Demise of Unlicensed Spectrum

Abstract

In 2008 the FCC released two OSP working papers by Bykowsky, Sharkey, and Olson which argued that licensed and unlicensed spectrum were economically inefficiently allocated and for a set of mechanisms to address this problem. This paper analyzes both Bykowsky, et al. papers, concluding that serious problems exist with both. The “tragedy of the commons” approach which they take is examined and rejected as historically unfounded and economically inappropriate for the class of rivalous goods of which spectrum is an example. It demonstrates that Bykowsky, et al. fail to recognize the fundamental differences between licensed and unlicensed spectrum and the ways in which these differences make exclusive-auction mechanisms for the allocation of unlicensed spectrum likely to reduce the economic value of such spectrum and that the market mechanism proposed by Bykowsky, Sharkey, and Olson – auctioning of the rule-making authority over the allocation of licensed and unlicensed spectrum to the highest bidder – is likely to have severely adverse public policy consequences. Bykowsky, Sharkey, and Olson adopt a Coasian approach under precisely the circumstances which Coase himself identifies intractable problems with market approaches. This paper finds that the dependence of Bykowsky, et al.’s model on a very specific auction structure in which firms seeking to acquire unlicensed spectrum outnumber firms seeking to acquire licensed spectrum (and to block the allocation of unlicensed) an unrealistic set of assumptions, renders the model useless for gauging the dynamics of a market in allocation of spectrum allocation rules and of the spectrum itself. Furthermore, their model explicitly violates the two key assumptions of the Revenue Equivalence Theorem. While Bykowsky, et al. are correct about historical provision point and freeridership problems associated with unlicensed spectrum allocation, there is little reason to believe that their model actually provide a remedy. Finally, this paper concludes there is good reason to believe that Bykowsky, Sharkey, and Olson have failed to note developments since 2005 which suggest that incumbents which have traditionally favored licensed spectrum are using significantly more unlicensed spectrum in their operations, particularly for data offloading and underlay sharing, and are now no longer predisposed to continue their historical pattern of opposition to unlicensed spectrum allocation.