Here is my letter on to Congress on ways to improve spectrum policy. I submitted this as part of the House Energy and Commerce Committee’s process to update the Communications Act.
Essentially I advocate that federal agencies pay for their use of spectrum, just as they do for any other input they need. To that end, I support the creation of a National Spectrum Service. I believe that we should maximize the revenue from spectrum auctions and conduct them in fair and transparent fashion. This means that no parties should be excluded from participating in the auction, regardless of their existing spectrum holdings. Those who value the spectrum most will put it to the most productive use, and they will pay the most for it. It is in the public interest to maximize the revenue of auctions so revenues can be raised for essential public projects. For example, First Responder Network FirstNET, a national communications network for public safety is key objective for which spectrum license revenue will be focused.
April 25, 2014Hon. Fred Upton Chairman Energy and Commerce Committee US House of Representatives 2125 Rayburn House Office Building Washington, DC 20515 Hon. Greg Walden Chairman Communications and Technology Subcommittee Energy and Commerce Committee US House of Representatives 2125 Rayburn House Office Building Washington, DC 20515
Re: Spectrum Policy
Dear Chairman Upton and Chairman Walden:
I am an American who studies the economics of the internet and telecommunications at a European university. This position allows me the opportunity to reflect on spectrum policies of other countries compared to the US. The views in this letter are my own.
It is undisputed that the US leads the world on a number of important mobile and wireless measures, including the number of mobile broadband subscriptions, diversification of mobile technologies, the number of 4G/LTE smartphones sold, and the proliferation of mobile applications. CTIA the Wireless Association observes that key achievements have been made even in just the last five years, including growing from zero to some 50 million 4G/LTE subscriptions. Smartphone subscriptions have increased from 41 million to more than 150 million. The iPad didn’t exist in 2009, but 220 million have been sold since. Meanwhile the number of apps has increased from 150,000 to 4 million. Mobile penetration increased from the already high 89 percent to 110 percent. The amount of SMS and MMS have doubled. This short list of accomplishments doesn’t begin to describe the advancements being launched in entirely new industries of m-health, m-education, and m-transportation.
The market-oriented spectrum policy reforms adopted by Congress and operationalized by the FCC over the past two decades have generated enormous benefits for consumers, and are one of the main reasons the U.S. now has the world’s most advanced mobile wireless services. Market-based spectrum allocation has allowed spectrum to flow away from inefficient uses to more highly valued ones and thus made possible the explosive growth of mobile broadband.
But more needs to be done. America’s global leadership in wireless rests on the effective optimization of one asset above all: spectrum.
The US has taken advantage of technologies to improve the utilization of spectrum, but relying on efficiency enhancement alone is not enough. The supply of spectrum is fixed, and it needs to be allocated and utilized more efficiently.
A suboptimal approach to spectrum management may “satisfice” for the moment, but it is not strategic for the long term. The US faces an exploding demand for mobile data. Wireless carriers don’t even have 16 percent of the airwaves that are best suited for mobile broadband. The government has the lion’s share and is undisputedly an inefficient user
This situation of squandered spectrum is a great concern to the nation and a threat to future economic growth and global competitiveness. Citing the National Telecommunications and Information Administration’s Office of Spectrum Management, the President’s Council on Advisors for Science and Technology explains the situation.
Federal agencies have exclusive use of 18.1% (629 MHz) of the frequencies between 225 and 3700 MHz (traditionally referred to as the “beachfront frequencies”), while non-Federal users have exclusive licenses to 30.4% (1058 MHz). The remaining 51.5% is shared, with Federal use primary and private sector use secondary. Approximately 80% of the shared allocation—or 40% of the total—have a “dominant” Federal use (e.g., radar, aeronautical telemetry) that under the current coordination regime effectively precludes substantial commercial use of those bands. In other words, nearly 60% of the beachfront frequencies are predominantly allocated to Federal uses.
Therefore I applaud your committee’s leadership to take action on the important issue of spectrum.
The Office of Science and Technology Policy (OSTP) also deserves commendation for its request for information on behalf of the White House Spectrum Policy Team to solicit ways to provide greater incentives for agencies to share or relinquish spectrum.
Federal spectrum holdings are assigned to some 60 federal agencies which don’t necessarily have the information or incentives to steward their use of the resource. Given the importance of spectrum to the nation’s economic health and security, a rational spectrum policy to recover unused and underutilized spectrum is in order. A Consumer Electronics Association study suggests there is a $1 trillion business opportunity in converting some $62 billion worth of spectrum. Mobile telephony is just one of many areas where high value use can be substituted for low value use, bringing greater efficiency and economic welfare.
To be sure, it can be difficult to get agencies to relinquish spectrum that they don’t use. To that end, Congress needs to pursue a carrots and sticks approach. The carrots can include incentive auctions. Sticks can be reclaiming the spectrum. However a viable middle road is getting agencies to pay fees for spectrum.
The key theoretical notion underpinning agencies paying fees for spectrum is that federal agencies procure other resources through the market and competitive processes. There is no justification that spectrum, one of the most valuable inputs, should not be part of that process.
Federal agencies use the General Services Administration (GSA) to procure their inputs of land and capital. They go to labor markets to hire employees. Agencies already have experience using markets, and this suggests that a GSA-like agency could also manage the allocation of spectrum. As government agencies do in the UK and Australia, American agencies can pay annual fees for spectrum, like any other inputs. An additional benefit of this process and the establishment of such a GSA-like entity would be to create transparency with a centralized database of all spectrum.
One suggestion by the Technology Policy Institute is a Government Spectrum Ownership Corporation (GSOC) would own and administer federal spectrum through annual fees. Surplus spectrum could be sold or rented to the private sector, as well as additional spectrum purchased.
Lessons from the UK, New Zealand, and Australia suggest that spectrum fees can work. Though new regimes take time to develop and require audit and revision, within 2-3 years of launch, these countries were able to realize that government agencies paying annual fees for spectrum.
In any event, spectrum held by agencies that is not being used should be put up for auction as soon as possible. The academic theory introduced by Herzel, formalized by Coase, and demonstrated successively with auctions, is that those who value spectrum most will pay the most for it and thereby put it to the most productive use. Americans should be interested in maximizing the revenue of spectrum auctions because revenue can be used to purchase and invest in important social goods. For example, auction proceeds could be used to fund FirstNet, a nationwide public safety network.
Following are comments to some of your questions to inform the development with spectrum policy. Thank you for your commitment to the important endeavor of modernizing America’s Communications Act.
Response to questions 1 and 10
The move to make government more efficient through consolidation and re-conception of agencies is supported broadly by Americans and is hardly unprecedented. Other countries have undergone a similar reorganization of their telecom regulatory authorities precisely to improve the delivery of government services, maximize regulatory efficiency, and achieve cost savings.
Communications regulation needs to be transitioned from the current silo-based, sector specific paradigm to a modern, technology-neutral, competition-oriented approach. The functions of the Federal Communications Commission are duplicative of functions performed by other agencies. Functions and resources should be rationalized effectively and redeployed to the appropriate agencies, or bundled into a specific and perhaps new agency for the management of spectrum.
To be sure, the FCC has valuable expertise in managing spectrum allocation. However the agency, perhaps inevitably, is subject to politics and regulatory capture. This political pressure leads to tinkering with auction rules to favor some parties over others. This lack of transparency, predictability, and standardization does not make for a fair auction and does not serve the interest of the American people.
As I discussed in the first submission for Communications Act Update in cooperation with scholars affiliated with the American Enterprise Institute, the FCC and NTIA are developing duplicative competencies. Spectrum management is a discrete function which belongs in a stand-alone agency, perhaps combined with the government spectrum functions currently performed by NTIA. Congress should consider different forms for this agency, including a semi-autonomous entity with sufficient authority to reassign underutilized spectrum from government to private sector use, the aforementioned GSOC for example.
A single agency with jurisdiction over allocation of spectrum for both commercial and government use could help to correct the current over-allocation of spectrum to lightly-used and technologically stagnant government systems. A national spectrum service in the spirit of the GSA should have the power to reallocate spectrum from government to the commercial sector, to conduct auctions, and to perform other functions currently executed by the FCC or NTIA to manage spectrum in the public interest.
Question 3 on Spectrum Sharing
While sharing has a role in spectrum policy, the US should certainly not give up the valuable efforts to auction relinquished spectrum for licensed use. Indeed the United Kingdom realizes 84 percent of its spectrum being traded, and where necessary, the government has seized spectrum from uncooperative government agencies.
A number of economists and engineers have observed the downsides of spectrum sharing. Faulhaber and Farber estimate that sharing can reduce the value of a spectrum by 60 percent. Cooper suggests that a sharing requirement made the 700 MHz band D block spectrum so unattractive that no commercial actor would take it up. Moreover, in a seminal analysis of spectrum auctions in 25 countries, Hazlett and Munoz conclude that auctions overwhelmingly support consumer welfare, greater than other methods of spectrum allocation, including sharing. They estimate a lost opportunity of $67 billion in consumer welfare over 6 years for the failure to include an additional 30 MHz in the C block auction in 1996.
Both sharing and relinquishing spectrum to market mechanisms are two important paths that the US needs to pursue. Sharing is seen as a solution to working with reluctant agencies that won’t relinquish spectrum. However other countries, particularly the UK, New Zealand, and Australia, all with similar legal traditions to the US, have developed national markets with relinquished spectrum. The recovered spectrum is auctioned, traded, and leased. Compared to the US where some 60 percent of prime spectrum is held by government agencies unavailable to private users, in the UK over 75 percent of spectrum is available to all comers. Of this, 46 percent is occupied by private users and 29 percent is shared by private and public users. Public actors occupy just 25 percent. There is no reason why the US cannot and should not develop this toolkit of capabilities. No agency should be able to handcuff the wireless future and supersede the American citizens it serves
In instances where sharing is at play overlay licenses are a possible solution to some of the challenges mentioned. An overlay license is a flexible-use license which encourages the new service provider and incumbent to find voluntary settlements to the shared spectrum. The license is awarded in an auction where the new entrant wins primary rights with the incumbent holding secondary rights. There is generally a deadline in which the incumbent needs to vacate the band. For further discussion see “Reclaiming Federal Spectrum: Proposals and Recommendations.”
It can take 6-10 years to reallocate spectrum. Spectrum is needed today, and so there is an imperative to add more to the pipeline immediately. Congress has pursued a number of important activities, but executive orders may be necessary to expedite the process.
The command and control approach has the advantage of removing political pressure and temptation for any political actor to influence the process to relinquish spectrum. The Base Realignment and Closure (BRAC) project facilitated the difficult process of closing bases in phases following the Cold War. The US needs to take the same approach with spectrum, also known as BRAC the spectrum. A helpful discussion of this is available in “Getting Away from Gosplan: A BRAC like effort is need to repurpose federal spectrum.”
Spectrum is a scarce, valuable resource which should not be free. It should be reclaimed quickly with a minimum of fuss and priced in the market. The drawn out process of engaging with agency stakeholders is not productive. Dr. Phillipa Marks, key architect of the UK and New Zealand policy, has observed that the US has been too lenient with agencies and “too incremental” in its approach to spectrum. Other countries have used executive power to force the parties to give up the spectrum.
Sometimes a hegemon is needed to bring order for the greater good. The Federal government works this way to organize the 50 states, and a similar discipline can be applied to Federal agencies. One success story for Congress was the 1993 Omnibus Budget Reconciliation Act, which in addition to introducing the legislation that allowed competitive bidding for spectrum, reinforced the role of the federal government to ensure a national telecommunications market. Mobile operators were able to take advantage of one set of systems and processes to serve the entire country, rather than having to roll out state by state. Had the states taken the lead, it is likely that there would have been 50 different, potentially conflicting, sets of regulatory obligations. The US would probably be in the situation where the European Union is today, with a fragmented market, limited scale, and few global internet companies. The EU with 28 nations, 17 languages and 11 currencies, is hardly a physical single market, let alone a digital one.
America’s de facto single market is one of its key assets. Bringing spectrum into the national market can only improve competitiveness.
Price is a valuable signal, and it should not be ignored. Expectation of revenues, provided that it is a reliable figure, is important information to create a transparent auction. Also, it is important to remember that maximizing revenue from the auction serves the valuable goal of raising money for the government which it can use for a variety of important public programs.
While the economics and politics to share and relinquish spectrum are important, policymakers should not forget the engineering elements in designing optimal spectrum policy. An excellent paper on the “Technical Principles of Spectrum Allocation” offers valuable guidance on this front.
In certain instances, there are logical opportunities for reorganization and pairing which could substantially increase revenue. As an example of a way to reduce costs and eliminate the need to displace non-federal incumbents, NTIA in a letter to the FCC expressed their support of the Department of Defense vacating the 1755-1780 MHz spectrum so that it can be paired with the 2155-2180 MHz band (AWS-3 band) – maximizing revenue potential during the upcoming AWS auction[CH1] and marking progress in the effort to free up government-held spectrum for commercial use. The Department of Defense could then be allocated the 2025-2110 MHz band. Putting the two bands together “would yield substantially greater auction revenues than if those bands were auctioned separately”, a group of senators wrote in a letter to Defense Secretary Chuck Hagel, Commerce Secretary Penny Pritzker and former acting FCC Chairwoman Mignon Clyburn. It is important that the Defense Department work with spectrum stakeholders to effect this transition in time for the February 2015 auction.
The FCC defines an incentive auction as a voluntary, market-based means of repurposing much-needed spectrum for flexible use, including mobile services.
The effort for incentive auctions should be applauded, and Congress has intended for there to be an open auction. But the original good idea has been marred in a few recent occasions. It is not possible to have a bona fide auction if arbitrary and capricious conditions are added to the auction (not allowing certain players to bid, restricting participating etc). Such practices distort the information and incentives of the agencies that are foregoing the spectrum. Without having a true reflection of the market value or the buyers interested in the spectrum, agencies can’t get a clear sense of the value they are relinquishing and what returns they can expect in future. The spectrum auction has to be held in good faith and with transparency in order to work. H.R. 3674 had this in mind. 
Use it or lose it requirements are a good idea. The FCC might also explore rewarding longer license life as an incentive for providers to achieve certain goals such as build out in rural areas. For example, providers could receive an additional 10 years for their spectrum license if they agree to serve remote areas. Both the operator and the regulator can calculate the net present value of the spectrum versus capital investment (capex) needed for additional investment.
Thank you again for the opportunity to comment.
Center for Communication, Media and Information Technologies
Frederikskaj 12, 3rd Floor
Copenhagen, Denmark 2450
 President’s Council on Advisors for Science and Technology, “Realizing the Full Potential of Government-held Spectrum to Spur Economic Growth,” July 2012. http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast_spectrum_report_final_july_20_2012.pd; Karl Nebbia, Director, NTIA Office of Spectrum Management, presentation to the Commerce Spectrum Management Advisory Committee (CSMAC), Dec. 9, 2009.
 Ofcom, “Spectrum Management Strategy: Ofcom’s approach to and priorities for spectrum management over the next ten years”, 2013 http://stakeholders.ofcom.org.uk/binaries/consultations/spectrum-management-strategy/summary/spectrum_management_strategy.pdf
 Faulhaber, Gerard R and David J. Farber. “The Open Internet: A Customer-Centric Framework”. International Journal of Communication 4 (2010). http://ijoc.org/index.php/ijoc/article/viewFile/727/411
 Cooper, Seth L. “Sharing Licensed Spectrum with Government Lessens Prospects for Wireless Broadband”, The Free State Foundation, March 4, 2013, Vol. 8, No.7. http://www.freestatefoundation.org/images/Sharing_Licensed_Spectrum_with_Government_Lessens_Prospects_for_Wireless_Broadband_030413.pdf
 Hazlett, Thomas W. and Roberto E. Munoz. “A welfare analysis of spectrum allocation policies”, RAND Journal of Economics Vol. 40, No. 3, Autumn 2009 http://mason.gmu.edu/~thazlett/pubs/Hazlett.Munoz.RandJournalofEconomics.pdf
 Skorup, Brent. “Reclaiming Federal Spectrum: Proposals and Recommendations”. Mercatus Center, George Mason University, May 2013. http://mercatus.org/sites/default/files/Skorup_FederalSpectrum_v1%5B1%5D.pdf
 Skorup, Brent. “Getting Away from Gosplan: A BRAC like effort is need to repurpose federal spectrum” Regulation, Winter 2013-2014. http://object.cato.org/sites/cato.org/files/serials/files/regulation/2014/1/regulation-v36n4-7.pdf
 Bennett, Richard. “Technical Principles of Spectrum Allocation”, TPRC 41: The 41st Research Conference on Communication, Information and Internet Policy 2013 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2240625
“Lawmakers have urged Defense officials to work with spectrum stakeholders in order to vacate the 1755-1780 MHz band so it can be paired with the 2155-2180 MHz band, also known as the AWS-3 band, which is required to be auctioned by the Federal Communications Commission by February 2015. Pairing the two bands ‘would yield substantially greater auction revenues than if those bands were auctioned separately,’ a group of senators said in an Aug. 1 letter to Defense Secretary Chuck Hagel, Commerce Secretary Penny Pritzker and former acting FCC Chairwoman Mignon Clyburn.”