The Challenges in Designing the FCC’s Broadcast Spectrum Incentive Auction

By TAP Staff Blogger

Posted on June 17, 2014


Share

In early December 2013, Federal Communications Commission (FCC) Chairman Tom Wheeler wrote in a blog post that he was postponing the broadcast spectrum incentive auction to mid-2015. Originally planned for 2014, Chairman Wheeler stated, “We have but one chance to get the incentive auction right.” While he applauded the work of the Incentive Auction Task Force to this point, Chairman Wheeler pointed to several key components in order for the incentive auction to be a success:


First and foremost, we absolutely must make fact-based policy decisions in an open and transparent manner. Beyond the policy issues, however, we must also exhaustively test the operating systems and the software necessary to conduct the world’s first-of-a kind incentive auction. This includes ensuring that such systems are user-friendly to both broadcasters and wireless carriers who will participate.


The broadcast television spectrum incentive auction will be the first such undertaking ever administered anywhere in the world. At the auction, television stations will consider bids to relinquish their airwaves and either go out of business or be placed in another channel. The spectrum released by broadcasters will then be put up for licensing by wireless carriers.


Professors Ilya Segal and Jonathan Levin, both with Stanford University, are two of the leading experts in auction theory and implementation chosen by the FCC to design the unprecedented spectrum incentive auction. Professor Segal is focusing on designing the reverse auction, the component aimed at creating incentives for broadcasters to relinquish their spectrum licenses. And professor Levin is working to design the forward auction, the section that will present newly available spectrum to broadband companies for bidding.


Professors Segal and Levin graciously shared their expertise and time with TAP in order to explain the significance and complexities of the incentive auction. Though there is a large interdisciplinary team coordinating multiple components to design, build, and implement the auction, this post is focused on the design of the reverse and forward auctions, the components that Professors Segal and Levin are directly involved in.


Why the Broadcast Spectrum Incentive Auction Is Unique


Professor Segal highlights the most significant element:

The “Incentive Auction” is an unprecedented undertaking that differs from all the previous spectrum auctions, or from any other auction run in the world, for that matter. The difference is that unlike all previous spectrum auctions, the FCC now needs to sell to broadband companies what it doesn’t have – the spectrum currently used by TV stations.


The broadcast spectrum incentive auction is comprised of two separate yet interdependent auctions: a reverse auction, which will determine the price at which broadcasters will voluntarily relinquish their spectrum usage rights; and a forward auction, which will determine the price broadband companies will be prepared to pay for new spectrum wireless licenses. The reverse and forward auctions are designed to run simultaneously for the most part. After all, the FCC cannot buy spectrum first from the broadcasters and then sell it to the wireless industry later because it won’t be possible to know if the spectrum bought and sold will generate enough revenue. And by law, the auction is required to generate revenue sufficient to cover certain specified costs.


A second unique component of the auction is the “repacking” process. Repacking involves reorganizing the broadcast television bands so that the television stations that remain on the air following the incentive auction occupy a smaller portion of the UHF band. This will enable the FCC to reconfigure a portion of the UHF band into contiguous blocks of spectrum that are suitable for flexible wireless use. The Spectrum Act requires the FCC to make all reasonable efforts to preserve the “coverage area” and “population served” of television stations involved in the repacking.


A Brief Background on the Incentive Auction


The skyrocketing use of wireless networks has been dramatically increasing demands on both licensed and unlicensed spectrum. Recognizing that mobile broadband networks are not only the foundation of communications services, but also provide infrastructure that supports innovation across entertainment, health care, public safety, and education services, the FCC first proposed spectrum incentive auctions in its 2010 National Broadband Plan. Two years later, the spectrum incentive auction was mandated by Congress as part of the Middle Class Tax Relief and Job Creation Act of 2012 (H.R. 3630) – see Title VI—Public Safety Communications and Electromagnetic Spectrum Auctions.

Note: H. R. 3630 is also referred to at times as the “Spectrum Act of 2012.”


The goals of the spectrum incentive auction are threefold:

1. Make Spectrum Available – the reclaimed spectrum from broadcasters will make additional licensed and unlicensed spectrum available for broadband use.
2. Provide a Financial Incentive for Broadcasters – television stations who wish to release their spectrum rights will receive financial compensation; and broadcasters wishing to remain on the air will be able to do so. However, broadcasters who remain on the air will be subject to being repacked in order to clear contiguous blocks of spectrum in the forward auction.
3. Generate Revenue – the net proceeds are to fund a national first responder network, cover the relocation costs for broadcasters, and the balance is to be used for deficit reduction.

The bandwidth being targeted for this auction is the low-frequency airwaves, the 600MHz band that is currently occupied by the higher UHF television channels. There are 2,247 television stations that qualify to participate in the reverse incentive auction. The lower frequency is desirable by wireless carriers because signals can cover greater distances and travel more easily through physical barriers.


The design of the spectrum incentive auction involves a large interdisciplinary team, which includes:

  • engineers (to calculate interference),
  • computer scientists (to calculate feasible retuning of stations subject to the interference constraints),
  • optimization scientists (to calculate near-optimal assignment of stations to channels given various criteria),
  • economists (to design the rules of the auction), and
  • programmers (to implement the algorithms and design interface).

The Reverse Auction


The reverse auction will determine the prices at which broadcasters will voluntarily relinquish their spectrum usage rights and the amount of spectrum available in each market for the forward auction. Broadcasters will compete against one another by picking the prices at which they would be willing to give up some or all of their spectrum rights. The reverse auction can be considered as the supply side of the market for repurposed broadcast television spectrum.


The FCC’s Broadcast Television Spectrum Incentive Auction Summary outlines the four options available to broadcasters:

1. Participate and bid to give up all rights to their channel and go off the air.
2. Participate and bid to give up all rights to their channel but move from UHF to VHF and remain on the air.
3. Participate and bid to give up all rights to their channel but share a channel with another broadcaster after the auction.
4. Do not participate in the auction and stay on the air on the same or another channel in the same band, as determined in the repacking process.

Professor Segal points out that “a statute passed by Congress stipulates that these licenses can be acquired from those stations only on a voluntary basis, through the incentive auction.” Stations will be provided monetary “incentives” to relinquish their licenses.


Challenges with the Reverse Auction


First, spectrum is sliced differently for broadband use than for broadcast.

  • Geographic Differences – Television stations have geographic areas in which their signals are protected from interference. These areas look roughly like circles centered around the station’s antenna. Broadband companies are interested in much larger areas; they are interested in assembling areas for national coverage.
  • Frequency Differences – TV channels use a 6MHz band. Broadband stations require at least 5MHz for uplink and 5MHz for downlink with a band separating the two.
  • Interference Issues – having TV and broadband signals in geographic and frequency proximity to each other will more than likely create interference. For both broadcasters and wireless, a partition is required; referred to as guard bands, they separate TV from mobile spectrum in order to prevent interference.

Each of these differences need to be taken into considering in designing the reverse auction so that the resulting TV channels that remain continue to serve their customer populations, and the broadband spectrum available for sale is desirable for wireless carriers.


Second, eliminating the possibility for a TV station “hold out.”

Professor Segal makes the following analogy:


The problem of buying TV stations is similar to the problem of assembling parcels of land in order to build a highway. In this type of situation, land owners or TV stations could have an incentive to “hold out” for a higher selling price. One seller holding out for a larger piece of the buyers’ offerings could jeopardize the whole project.


“Fortunately,” says Professor Segal, “the incentive auction deals with this issue by stipulating that TV stations that do not relinquish their licenses can be retuned to other channels so that contiguous spectrum could still be cleared for broadband use.” This provision ensures that few stations would be “indispensable” for clearing a given amount of spectrum.


Third, the complexities of retuning broadcast stations.

Retuning (“repacking”) the broadcast stations that decide not to participate in the auction will be very important for creating a competitive auction; however, it also presents several challenging computational problems:

  • The FCC is required to make all reasonable efforts to preserve the “coverage area” and “population served” of television stations involved in the repacking. This requirement imposes many constraints on which stations can be assigned the same channel or adjacent channels.
  • Stations can bid not only for going off the air but also for going to a lower band (e.g. from a UHF channel to an upper or lower VHF channel).
  • Determining the most efficient way to pay stations for voluntarily going off the air necessitates a method for determining the optimal number of stations that need to be repacked.

On this latter issue, Professor Segal and his colleagues working on designing the reverse auction have proposed a “heuristic” approach that determines which bids are accepted and which bids are rejected. The approach can be implemented by means of a clock auction: initially generous offers are quoted to stations for releasing their spectrum. But as the desired spectrum becomes available from relinquishing stations, offers are reduced to the remaining stations. For more information on this heuristic approach, see “Deferred-Acceptance Heuristic Auctions,” written by Professors Ilya Segal and Paul Milgrom.


Of special note: Paul Milgrom is the Ely Professor of Humanities and Sciences in the Department of Economics at Stanford University. He was one of the key people who designed the initial FCC spectrum auction and is leading the team of auction experts in designing this incentive auction.


The final challenge is that the auction must have an easy-to-use interface.

This challenge applies to the forward auction as well.


The chief aim of the design team is to make the auction very easy for bidders to participate in. The complexities of the actual design and implementation will be hidden “under the hood,” while bidders will have a simple interface in which the current clock prices are displayed and as these prices change, bidders can switch their preferred option or choose to exit the auction.


Professor Segal emphasizes, “Our auction is very likely a one-time event and we must therefore get it right the first time.”


The Forward Auction


The forward auction will identify the prices that broadband companies will pay for new flexible-use spectrum licenses. The FCC is familiar with forward auctions in the spectrum context, and has been conducting them for nearly two decades. The forward auction can be considered as the demand side of the market of the repurposed broadcast spectrum.


The typical process for past spectrum auctions involves the FCC first determining a band plan. This creates licenses that give rights to a particular segment of spectrum over a particular geographic area.


Professor Levin states that, “in the case of the forward auction, the FCC has a lot of experience running auctions to sell licenses to wireless companies. This shouldn't be all that different from the prior auctions.” He goes on to say, “The twist in this case is that we won't know exactly how much spectrum there is to be sold until we see all the bids.”


Challenges with the Forward Auction


Professor Levin summarizes the issues with the forward auction:

Many of the challenges in making this auction work are similar to those the FCC has faced in the past: making sure the design is relatively simple to explain, understandable to the bidders, provides good incentives for bidders to be competitive, is perceived as fair, and importantly (and this is the key to the reverse auction as well) encourages companies to participate. But there are challenges that are unique to this auction.


First, it is unknown how much spectrum will be available for bidding.

The forward auction requires information regarding what spectrum rights were tendered in the reverse auction, and at what price. However, this information won’t become available until the reverse auction is underway.


Professor Levin explains that the forward auction is “likely to start with a band plan that targets the most spectrum clearing that is feasible given broadcaster participation. Then if necessary the auction will reduce the target depending on how the revenue that can be raised from the wireless company bids compares to the amount needed to compensate the broadcasters relinquishing their licenses and to meet the other revenue requirements of the auction.”


To address this unknown element of what spectrum bands will be available for bidding, Professor Levin and his colleagues on the forward auction design team, Ilya Segal, Larry Ausubel, and Paul Milgrom, have proposed a variation on the traditional ascending auction that the FCC has used successfully over the last two decades. This proposal has the potential to run faster than the traditional FCC spectrum auctions which will be quite advantageous given that the forward and reverse auctions need to be coordinated carefully.


Professor Levin explains this plan:

In our proposal, the FCC would define "products" (e.g. a 10 MHz paired license covering a particular geographic area) and there would be a price for each product. In each round of the auction, the FCC would say how much the price of each product was going to rise, and the bidders would say how many of each license they wanted to buy at the going prices, or whether they wanted to reduce their demands. When the number of licenses demanded drops enough that it equals the amount available, the FCC stops the auction, and checks to see if the revenue is sufficient to pay for clearing the spectrum. If it is, the auction is over. If not, the FCC has to be less ambitious and reduce the target spectrum to be cleared.


Of special note: Lawrence M. Ausubel is Professor of Economics in the Department of Economics at the University of Maryland and a key member of the forward auction design team.


Second, the auction must generate revenue.

Congress mandated in Section VI, Subtitle D ‘Spectrum Auction Authority” of the Middle Class Tax Relief and Job Creation Act of 2012 (H.R. 3630) that the proceeds from the broadcast incentive spectrum auction must generate certain “minimum proceeds” -- the cost of the spectrum procurement and funding for broadcaster retuning expenses in the reverse auction. Congress further directed that any additional net proceeds from the broadcast incentive auction are to be deposited in the Public Safety Trust Fund to fund a national first responder network, and the balance is to be used for deficit reduction.


Finally, the auction must have an easy-to-use interface.

As discussed above, this challenge applies to both the forward and reverse auction design and implementation.


Running the Reverse and Forward Auctions in Parallel


The reverse and forward auctions are being designed to run, for the most part, simultaneously. The FCC cannot buy spectrum first from the broadcasters and then sell it to the wireless industry later because it won’t be possible to know if the spectrum bought and sold will generate enough revenue.


This post covers only two of the many components of the FCC’s broadcast spectrum incentive auction. In terms of the number of participants (thousands of TV stations), monetary amounts involved (tens of millions of dollars), and the computational complexities, this auction truly is unprecedented.


** **


TAP graciously thanks Professors Ilya Segal and Jonathan Levin for sharing their expertise and time. They were instrumental in the development of this post. If after reading this, you are awed by the magnitude of the complexities inherent in this undertaking or have gained even a slight understanding of what is involved in the creation of the spectrum incentive auction, it is solely because of the professors’ ability to break very complicated issues down into digestible components and patiently explain the challenges this auction poses.

 


Share