As the IPv4 address transfer market continues to grow and the Regional Internet Registries deal with address leasing and eliminating needs assessment in the post-free pool era, we thought it would be useful to review the literature and development of another related secondary market: radio spectrum. Spectrum resources and their regulation offer numerous valuable lessons for IP address policymakers. Although radio spectrum is not perfectly analogous to IP addresses, there are enough parallels both in the nature of the resource and in policy to make the comparison useful.
In the United States and elsewhere, the continuing demand for wireless services has driven the development of secondary markets for spectrum over the past decade. As explained by Mayo and Wallsten (2010), the Federal Communications Commission began with a policy statement in 2000 to promote secondary market activity like market transfers and leasing of spectrum. Its logic was that supply and demand conditions change over time, and leasing/transfer policies could “help alleviate spectrum shortages by making unused or underutilized spectrum held by existing licensees more readily available to other users and uses and help to promote the development of new, spectrum efficient technologies.” (FCC, 2000, pg 1) The FCC outlined the necessary framework for a successfully functioning secondary market for spectrum, including:
- Clearly defined economic rights
- Easy market entry and exit
- Full information on prices and products
- Mechanisms for bringing buyers and sellers together
- Many buyers and sellers
To this end, the FCC received comments and adopted two separate Orders , in May 2003 and July 2004, establishing new policies and rules to permit parties to enter into a wide variety of spectrum leasing arrangements summarized in Table 1 below. Similar efforts to develop policies for spectrum secondary markets have occurred elsewhere. E.g., the European Commission issued its Radio Spectrum Policy Programme (RSPP) decision in 2012 which is now being reflected in various domestic policy consultations. Recently Canada has also undertaken a Consultation on Considerations Relating to Transfers, Divisions and Subordinate Licensing of Spectrum Licences.
|License and resource control
|Licensee maintains de jure control of license, de facto control over leased spectrum
|No approval required from FCC
|“De facto transfer”
|De facto control of license transferred
|Approval by FCC immediate, unless following concern(s) raised which stipulate 21-day review: (1) violation of eligibility and use restrictions; (2) foreign ownership; (3) transfers by designated entity and entrepreneur licenses; (4) harm to competition, and (5) other public interest concerns.
Table 1: Leasing Types created by FCC Orders
The Orders moved toward accomplishing the framework, more clearly defining the spectrum usage rights of license holders and facilitating entry and exit. The FCC concurred with the NTIA’s argument that there should be no secondary user registration or certification requirements. Instead, licensees should be required “(1) to maintain on file contact information for any secondary users of their spectrum, (2) to have contractual authority to terminate the lease in response to a violation of law or interfering use by the lessee, (3) to approve any subsequent subleases, and (4) to assist in enforcement of lease conditions and administrative regulations.” Another important conclusion reached by the FCC was that any review, if one were needed at all, was focused on the licensee’s qualifications and/or competition policy concerns.
The cumulative effect of the FCC reforms (along with subsequent developments in the market, e.g., the creation of various trading platforms) has been growth in secondary market trades (Mayo and Wallsten 2010). And while the Commission continues to review its rules addressing competitive harms, the main policy takeaway, one with which economists generally agree, is that the secondary market should be largely unregulated (Hahn and Passell 2013). Any policy should encourage spectrum to be allocated to its highest valued uses, and thus to the party willing to pay most for the right to use it (4).
Parallels and lessons for the IP address market
Arguably a situation similar to the 1990s spectrum environment exists today around IP addresses. There is continuing growth of Internet services and demand for IPv4 addresses (despite the ideal of a transition to IPv6), the free-pool is approaching exhaustion, and a substantial portion of addresses (some estimate as many as 1 billion) are not being utilized, i.e., routed publicly. Subsequently, initial RIR policies facilitating a secondary market have been adopted.
According to our analysis, that secondary market is well underway. From November 2009 to the end of the first quarter of 2013 that there were almost 150 transactions, 312 distinct blocks traded, and a sum total of 9.4M unique IP addresses. These resources have been exchanged under RIRs transfer policies, and are “licensed” by a RIR Registry Service Agreement (RSA) or some undisclosed variation of a Legacy RSA (LRSA) which preserves certain rights of use for the address holder. In addition (and not captured in the above analysis), there are allegedly secondary market transactions where the licensee maintains de jure control of the addresses and subleases them to another party. And finally, there are allegedly transfers of legacy address resources occurring not involving any contractual arrangement with the RIRs. The transfer and use of these latter “unlicensed” resources is governed almost exclusively by the market.
The fact that so many secondary market transactions sidestep some of or otherwise avoid entirely the RIRs transfer policies should be a clear indicator to policymakers that something is amiss. We’ve argued that one culprit is the “needs assessment” requirement in the RIRs transfer policies, and continue to be puzzled by the dogmatic devotion of some to a provision that would certainly be viewed as odd in the spectrum world. What is more confounding is the recurring argument that such a requirement helps prevent anti-competitive behavior.
One need look no further than one of the biggest deals for spectrum in United States history for evidence of how needs assessment fails to address these issues. In initially acquiring a sizable amount of valuable spectrum, cable companies (SpectrumCo) were accused of “hoarding” because they spent a lot of money to get spectrum when they were trying to work together to develop a cross-cable supplier mobile phone service to allow them to compete with the telcos. They never managed to pull it off, however, and so they sat on those frequencies for some time never really using them. This did not seem to be illegal. But it is a good example of how businesses really operate. They did intend to use the spectrum – but they acquired it first when they had a chance and then set about trying to make the (ultimately failed) joint business arrangements to put it to use. Some might say this proves how irrational a market-based system is, but au contraire: the cable operators could have and would have passed a “needs” test. Had they been able to pull together a joint mobile service they would have needed those frequencies, and it would have been easy to show business plans to that effect. Ultimately, the cable companies sold the spectrum to Verizon in the secondary market, in a deal that underwent extensive competition analysis before approval. In that transaction too, nothing in a needs test would have prevented Verizon from buying the spectrum from the cable companies, because Verizon could make use of the spectrum.
Tthe focus of address policymakers should be on using competition policy tools. Something similar to the FCC’s case-by-case approach using a screening process to help identify where the acquisition of spectrum resources provides particular reason for further competitive analysis could certainly be implemented in the address secondary market. The FCC’s screening of spectrum acquisitions initially considers objective changes in market concentration (as a result of the transaction) and is based on the size of the post-transaction Herfindahl-Hirschman Index (HHI) and the change in the HHI. (FCC 2012, 5) Furthermore, it examines the amount of spectrum resources suitable and available on a market-by market basis for the provision of mobile telephony/broadband service. (5) Theoretically, the screening of address transactions would be easier since addresses don’t have the added complications of frequency characteristics or geographic considerations like spectrum. Probably the biggest hurdle to overcome would be pressure from the RIRs and governments to include the geographical constructs of “regions” and states in defining markets. But that is a need to discuss another day.
 Promoting Efficient Use of Spectrum Through Elimination of Barriers to the Development of Secondary Markets, Report and Order and Further Notice of Proposed Rulemaking, 18 FCC Rcd 20604 (2003) (Secondary Markets First Report and Order); Promoting Efficient Use of Spectrum Through Elimination of Barriers to the Development of Secondary Markets, Second Report and Order, Order on Reconsideration, and Second Further Notice of Proposed Rulemaking, 19 FCC Rcd 17503 (2004) (Secondary Markets Second Report and Order).