From my perspective, the recent AWS-3 auction has to be deemed an overall success.  It is hard to say otherwise when it released 65 megahertz of spectrum for more efficient purposes, allocated 1611 licenses to current and prospective wireless providers to expand wireless broadband services, and grossed revenues totaling $44.9billion (net revenues are estimated at $41.3 billion).  Nonetheless, this auction highlighted many important issues and raised quite a few concerns.  Here are some takeaways that will help shape my views as we consider future spectrum policy.

Licensed vs. Unlicensed Spectrum.  This auction clearly demonstrates there is still a critical need for licensed spectrum in our overall spectrum framework.  Although some argue that the future lies only with unlicensed or shared spectrum, this view ignores the fact that our nation’s commercial wireless carriers still seek exclusive spectrum in order to be able to maintain quality of service and network control.  The simple proof of this is the bidding activity, extensive bidding rounds and the final revenue figures for the AWS-3 auction.  Having waited almost seven years since the 700 MHz auction and facing huge projected growth in wireless data usage, participants fought for eleven weeks to win the exclusive right to use the available licenses.  And these entities, along with those that didn’t participate or win large numbers of licenses, are reportedly looking closely at the broadcast incentive auction to acquire additional licensed spectrum. 

Paired vs. Unpaired Spectrum.  As I have said before, I think the Commission, and the American people, lost an opportunity by not finding appropriate spectrum to complement the A1 and B1 blocks.  Specifically, the Commission should have paired this 15 megahertz of spectrum, instead of offering 5 and 10 megahertz unpaired spectrum licenses.  By some accounts, this decision may have cost the taxpayer up to $21 billion in revenues.  Just compare the valuations in this auction: 10 megahertz of unpaired spectrum (the B1 block) garnered almost $2.3 billion in gross revenues, whereas 10 megahertz of paired spectrum (i.e., 5 megahertz uplink and 5 megahertz downlink) auctioned in the same market areas (the H and I blocks) raised approximately $8.4 billion.  Meanwhile, if the 10 megahertz was paired to create a 20 megahertz license, it wouldn’t be unrealistic to imagine that it might have raised upwards of $18.2 billion in gross revenues, which was the price for all of the 20 megahertz paired licenses (J block).    

This auction reaffirms my belief that we should have sought Congressional action to make any necessary changes to the law in order to facilitate an appropriate pairing.  Alas, we did not even try.  At least for the immediate term, the Commission must seek to auction paired blocks. 

Location Still Matters.  This auction confirmed the significance of the value generated from licenses in the largest markets.  As expected, the licenses covering the top 30 metropolitan areas, which account for approximately 10 percent of the licenses offered, brought in 79 percent of the overall gross revenues.[1]  This is nothing against medium- or small-sized markets, but this fact is important as the Commission looks toward the incentive auction and issues raised by potential market impairments.  In particular, the Commission must generate as much revenue as possible from the larger markets in order to make the substantial financial offers to broadcasters that would warrant turning in their licenses.  Accordingly, there should be a recognition that unless absolutely necessary, market impairments – especially in the larger markets – must be prohibited or severely limited.  Auction participants are unlikely to bid meaningful sums in the forward auction if the licenses are encumbered, and this is even more significant given the Commission’s desire to maintain its flawed “unreserved” and “reserved” categorizations.  In other words, if we expect entities to bid high, we can’t offer them cluttered licenses.  Additionally, we must keep in mind the recent results of the Canadian AWS-3 auction where the “reserved” licenses for new entrants were purchased at a significant discount, potentially losing billions of dollars for taxpayers.

Low-Band vs. Mid-Band Spectrum.  The overall revenues of this auction also cast doubt on some estimations of the unappreciated value of mid-band spectrum.  Mid-band is clearly more valuable than some would like you to believe.  Despite the increased cost of buildout – a reality of physics – for mid-band spectrum, bidders still badly wanted it.  These entities knew the expected buildout costs, took them into account when formulating bidding strategies, and still bid almost $45 billion.  It is a fallacy to suggest that low-band spectrum is the key to future network architecture and wireless providers are unable to operate without it.     

Designated Entities.  The Commission must rethink its designated entity policies.  As Commissioner Pai  and others have highlighted, the rules in place for this auction permitted two entities with 85 percent investment from a large, well-financed company to win 702 licenses and apply for a $3.3 billion subsidy to offset part of their collective $13.3 billion in bids.  This company was able to use this structure to outbid not only small businesses and rural telephone companies, but also some of the country's largest wireless providers.  If these bidding credits are granted, which I take no position on at the current time, the rules would allow these licenses to be “flipped” to another company after five years without repaying a penny of the subsidy.  If early analysis is correct, the Commission must thoroughly review the contractual arrangements and coordinated bidding activity.  And, the full Commission should vote on whether to grant the subsidy.  Moreover, this practice should put a final end to any plans to loosen our DE rules.  In fact, the Commission should give serious thought to strengthening these permissive rules and the current buildout requirements.        

Planning Ahead.  To ensure continued growth and innovation in America’s renowned wireless sector, we must identify now the spectrum bands that can be auctioned for exclusive use in the future.  We no longer have the luxury of avoiding the issue of fallow spectrum.  Like AWS-3 and the coming 600 MHz auction, much of tomorrow’s licensed spectrum will have to be reallocated from other uses, such as from the federal government or broadcast licensees.  Clearing spectrum and resolving interference issues take time.  We must start this process now to have a firm plan in place to ensure that we have adequate spectrum in the pipeline for the next demand cycle. 



[1] The licenses included in these calculations are those that cover the 30 largest Metropolitan Statistical Areas according to U.S. Census estimates.  See U.S. Department of Commerce, United States Census Bureau, American FactFinder, Estimates of Resident Population Change and Rankings: July 1, 2012 to July 1, 2013 - United States – Metropolitan Statistical Area; and for Puerto Rico,  http://factfinder.census.gov/faces/tableservices/jsf/pages/productview… (last visited Mar. 25, 2015).  The licenses covering the top 10 (3.7 percent of all licenses) and top 20 (6.8 percent of all licenses) markets accounted for 55 percent and 69 percent of the total revenues, respectively.