Director, Center for Information Policy, University of Maryland
July 16, 2001
Today the U.S. national domain, .us, is the home for tens of thousands of state and local agencies. It has few private users because private users must register under localities and their respective states, so that IBM is ibm.armonk.ny.us rather than ibm.com. This legacy system has been untouched for years. But the .us is about to change in a land rush for premium space under what may become the world's premium top-level domain -- just as U.S. currency is the world's preferred medium of exchange.
On June 13, 2001, the National Telecommunications and Information Administration, acting through the National Institute of Standards and Technology, issued a solicitation ("Request for Quotation") for a contractor to operate the registry for a greatly expanded .us. (1) The expanded .us is to be virtually unrestricted very much like .com, which, as the most favored real estate on the Internet, has drawn millions of registrations. Demand for .com has developed over fifteen years, accelerating as Internet users began to understand the value of domain names for expression and recognition. However, .us will now be thrown open all at once to speculators, trademark owners, and ordinary Internet users.
The solicitation treats .us as another generic top-level domain like .com, even though the United States has a distinct interest on its own national domain. NTIA's stated motivation is to increase use of .us by making an unlimited number of second-level names available:
The general absence of less hierarchical registration opportunities in the usTLD has limited the domain's attractiveness to users. It has been suggested that this more "generic" space would greatly increase the utility of the usTLD. Therefore, this acquisition also encompasses functions that will allow, on a competitive basis, for the registration of second level domains directly under usTLD (such as example.us). (2)
This justification merely states the obvious - that availability increases use.
To be sure, there are benefits to unlimited availability, but today names are hoarded rather than used. The experience with .com shows that an unrestricted first-come, first-served regime enriches intermediaries and speculators who add little if any original value to the inherent value of the names - while it maximizes collisions with trademark holders. Yet NTIA embraces this model without further explanation and solicits bidders to set up a tiered regulatory regime for .us to be sustained by cascading transaction costs and the sale of lottery tickets.
Two Solicitations in One
At first glance, an open solicitation appears more conventional and more fair than the "calling forth" of the nonprofit Internet Corporation for Assigned Names and Numbers (ICANN) in 1998 followed by a sole-source contract. (3) But this is no ordinary solicitation for a well-defined set of tasks. It is a solicitation for a private operator to set up a self-sustaining regulatory regime for .us in which it can charge for services to recover costs and earn a profit. The government will pay nothing. Within very broad parameters, bidders are invited to propose policies, implement and manage systems pursuant to those policies, and to earn money by providing services under those policies and systems.
The basic business model is unrestricted open registration just as in .com, with two exceptions: Registrants must have a U.S. nexus of some sort and there must be a sunrise policy, i.e., an advance name reservation or registration process limited to trademark owners. It should be noted that conflicts between domain names and trademarks are already addressed by the Uniform Dispute Resolution Policy (UDRP), in which new registrants must agree to arbitration, and by anti-cybersquatting legislation passed by Congress in 1999. (4)
The requirement for a sunrise policy follows policies set for the newly authorized top-level domains by the Internet Corporation for Assigned Names and Numbers (ICANN). However, sunrise policies in the newly authorized .biz and .info top-level domains show that these policies may be a significant new source of revenue for registrars and registries. Unlike the UDRP, the sunrise policies differ, each opening up a new front between domain name applicants and trademark owners while allowing administrative intermediaries to benefit from another tier of processing fees.
The solicitation gives no indication as to how conflicts among trademark owners are to be resolved, including whether common-law and state-registered trademarks will be equal to federal registered trademarks. It does not consider the gaming that may result from the trademarking of generic terms in contexts where they are not generic (e.g., registering "business" as a trademark for guns). However these issues are resolved, under NTIA's vision of an unrestricted self-financing operation it is clearly in the interests of the operator to maximize the number, scope, and scale of revenue-generating activities - including any special reservation or advance registration processes. Conflicts, whether among owners of trademarks like "United" or among competitors for generic terms like "loans", will have to be settled by round robins or some other form of lottery, in which the real winners will be the ticket sellers.
The solicitation requires the operator set up a system of competing registrars, like the formal separation of registrar and registry functions developed for global top-level domains under ICANN. However, that system was set up specifically to remedy Network Solutions' government-granted monopoly on both registration and registry maintenance. NTIA offers no reason for setting up this complex system from scratch, or for letting the registry operator take the lead in setting it up.
Most significantly, the solicitation will combine the monopoly registry function with policy development and oversight, making the contractor a regulator of both its own registry and the complementary registration services to be performed by registrars. The first two bars in Figure 1. show how the remedy for the NSI monopoly was implemented as responsibility for policy was moved from the IANA (5) to ICANN. The last bar illustrates how the solicitation adopts the split between maintenance of the registry and registration services for the expanded .us -- but ties together policy development and registry services!
The bundling of policy and registry functions is analogous to bundling the functions performed by ICANN and by Network Solutions for .com, .net, and .org and putting them out to bid as a package. This cozy arrangement violates basic policy principles by contracting out development of policies and procedures to a monopoly operator who is to be subject those same policies and procedures - and is to stand at the center of a new regulatory regime that it is charged to implement.
This extraordinary plan has been undertaken without consultation with
Congress and flies in the face of NTIA's past policies on domain name management.
The Department of Commerce has labored long and hard to make ICANN's decision-making
accountable, democratic, and transparent. Putting a private registry operator
in charge of policy and administrative decisions for .us is equivalent
to letting the operator of the .com registry, VeriSign/Network Solutions,
assume control of ICANN.
|ICANN||NSI (now VeriSign)||competitive (supervised by ICANN)|
expanded .us under NTIA/NIST solicitation, 2001
||competitive (supervised by contract operator)|
Figure 1. Allocation of policy, registry, and registrar functions
A Limited Competition?
As a practical matter, the solicitation is not as open as it might appear. Not only does NTIA disclaim knowledge about the actual scale and costs of operations under .us, but the agency demands that bidders "[d]emonstrate clearly, concisely and accurately, in written narrative form, the Offeror's understanding of the current state of the usTLD domain space." Only one company, VeriSign/Network Solutions has a deep understanding of the costs and problems of administering .us by virtue of its past nine years of experience overseeing .us, including its contacts with subcontractors, delegees, and users. (6)
Network Solutions has not been required to report on its administration of .us. The expenses have been covered (at no cost to the U.S. government, of course) by internal cross-subsidies from the company's monopoly registry operations for .com, .net, and .org. (7) I.e., all .com registrants, whether from the U.S. or other countries, have been and still are taxed indirectly for the operation of .us registry. It has been in the interests of NTIA (for political reasons) and Network Solutions (for competitive reasons) to keep the costs and other particulars of this arrangement unspecified.
In comments on .us, VeriSign has argued that running a registry is capital
intensive. (8) Adding policy services to
registry services demands a scope of operation that guarantees there will
be few bidders. This is confirmed by the evaluation criteria in the solicitation:(9)
VeriSign/Network Solutions is only U.S. company that has managed a large public domain name registry. The model for the expanded .us registry is nearly the same model that Network Solutions evolved for .com, .net, and .org. This is the model that turned Network Solutions into a $21 billion company (based on its sale to VeriSign in March 2000) less than five years after the National Science Foundation allowed it to charge for registering names.
VeriSign/Network Solutions has parlayed a five-year cost-plur cooperative agreement with National Science Foundation into a monopoly on the .com registry that will run until 2007 and probably beyond -- without facing a recompetition. While one may disagree with its vigorous efforts to maintain control of .com, its expertise and experience cannot be questioned. Indeed, the company's role in domain name litigation, lobbying, and protracted negotiation with Commerce and ICANN is legendary. It alone has grappled in depth with the full spectrum of technical, managerial, legal, and policy issues within the broad scope of the solicitation. It alone can prepare a fully informed and detailed response and point meaningfully to past performance.
The Value of Domain Names
Just as earlier attempts to address policy for domain names sought to frame the issues as "technical management," (11) the issuance of the .us solicitation as a "request for quotations" directed at registry operations downplays unaddressed policy questions. The real value of domain names (as well as the value of a contractual interest in allocating the names) is ignored. Like other property interests, domain names are exclusive, transferable, and enforceable. Like trademarks, their value can be expressive or denotative, but unlike trademarks, domain names can be generic. It is the generic names that are uniquely valuable and may occasionally be worth millions. "Business.com" reportedly sold for a record $7.5 million, and Great Domains, a VeriSign company, claims to have sold "loans.com" for $3 million. (12) Like spectrum, some names are much more valuable than others. But domain names are used, bought, and sold as property - no less than the classic forms of intellectual property.
Unlike other national domains where the most valuable generic terms have been registered long ago, .us remains undeveloped beyond the two-character state codes and three special domains. (13) While the requirement of a U.S. nexus may limit the universe of applicants, most electronic commerce companies and users are still U.S.-based. In terms of consumer confidence and business credibility, .us should be prove more valuable than an artificial generic like .biz.
The inherent value of domain names derives from years of U.S. taxpayer investment in the development and deployment of the Internet. This value can be offered to the operator of the .us registry to extract in high-volume registrations. Or it can be returned to the Treasury for the benefit of taxpayers or new investments. (14) Years of experience in spectrum management show that the best approximation of value - as well as the economically efficient method of allocation -- is achieved through properly designed auctions. True, lotteries were favored for a time before the opportunities for abuse became apparent, but no one suggested that they be outsourced.
NTIA has been absorbed in the many complex issues of process and substance around ICANN and generic top-level domains. While it has solicited input on .us from time to time; it has done little to develop and articulate policy. The jury is still out on ICANN as an experiment in private-sector lead international policy development -- but at least ICANN does not run registries itself.
Except for NTIA's failure to develop reasoned policies for .us, there is no justification for privatizing policy development for the U.S. national domain. If solicitation of policy development services can be justified and safeguards provided, the award should not be tied to registry services. Combining the two functions prejudices the policy outcomes and severely limits the universe of potential bidders.
The Department of Commerce should develop a meaningful vision and high-level strategy for the .us domain. Thanks to the early use of .com, .net, and .org, there has been little pressure to develop .us. It should not suddenly be offered up as another me-too generic top-level domain following the models developed to remedy the .com monopoly. The plan for .us should acknowledge and address the property-like characteristics and value of domain names - and the reality that .us is valuable asset that belongs to the citizens of the United States.
The current "Request for Quotations" should be withdrawn. Any future solicitation should be offered by NTIA directly and not by NIST, which has no experience with domain names. If Commerce lacks confidence in NTIA's legal capacity to do so, it should seek authorization from Congress.
NTIA should reassign responsibility for .us from its Office of International Affairs to its Office of Policy Analysis and Development. The administration of .us is a domestic issue and should not be addressed in the shadow of ICANN and policies for generic top-level domains.
Congress should recognize .us a national asset that merits policy development on its own terms. It should help provide long-term guidance by holding hearings on the future of .us.
The current National Research Council study of domain names (15 mandated by Public Law 105-305 should be asked to consider the special circumstances of .us, including alternatives for expansion and allocation of new second-level domains. NTIA may wish to defer new action on .us pending the outcome of this independent analysis.
NTIA should develop a proper policy analysis and course of action for .us, as was done for generic top-level domains in the 1998 green and white papers. (16) NTIA should explicitly evaluate options for .us and come to a reasoned and articulated conclusion on principles, policies, and a course of action.
The following excerpts from the solicitation ("Instructions For Submitting Quotations") illustrate the absence of policy framework, the very broad scope, and conflict of interest problems in the solicitation:
C. In light of the "Statement of Purpose" in the SOW, present a detailed narrative describing the Offeror's overall vision for future management of the usTLD, including how the Offeror proposes to make the usTLD more attractive and useful to United States Internet users and the Offeror's expectations for the number of potential usTLD registrants.
E. Demonstrate clearly, concisely and accurately, in written narrative form, the Offeror's understanding of the current state of the usTLD domain space.
H. Include a proposed draft of any contract(s) that the Offeror proposes to use between itself, as Contractor, and expanded usTLD registrars to ensure the stable operation of the expanded usTLD and implement the necessary policies (which may include shared registration system license agreements, registrar accreditation agreements, and registrant agreements)….
I. Include written policies (including implementation details) that the Offeror proposes to follow, as Contractor, during the start-up phase of registrations in the usTLD. Such description must include the following considerations:
1) How the Offeror would design and implement the "Sunrise Policy" that permits qualified trademark owners to pre-register their trademarks as domain names in the expanded usTLD space prior to the opening of the expanded usTLD space to wider registration.
How the Offeror proposes to implement and enforce the United States nexus requirement intended to preserve the usTLD for use by the community of United States Internet users.
Describe any proposed additional, alternative, or supplemental policies or programs the Offeror considers relevant and essential towards the locality-based usTLD space or the expanded usTLD space.
J. Address the following considerations in the description of the registration process:
How the Offeror proposes to address the potential initial "rush" for registrations at the opening of the expanded usTLD space.
Describe the proposed application process for potential registrants;
Describe the proposed mechanisms for ensuring that registrants meet registration requirements;
K. Describe in detail … the mechanism by which the public can suggest or recommend additional policys or procedures for the usTLD.
L. Describe, in detail, how the Contractor would fund the requirements of this acquisition at no cost to the United States Government.
M. Project/estimate and explain annual Contractor costs for this acquisition in such a way to permit the Government to match those costs to specific SOW Contractor Requirements.
N. Include detailed proposed financial plans, including, if appropriate, the manner in which fees levied for services rendered by the Contractor would be derived, considering cost plus a fair and reasonable profit.
1. http://www1.eps.gov/EPSData/DOC/Synopses/252/SB1335-01-Q-0740/SB1335-01-Q-0740.doc NTIA has no in-house capability, no history of network administration, and its authority to contract for such services is unclear. Accordingly, the solicitation is being handled by the National Institute of Standards and Technology on NTIA's behalf. Like other domain issues, the .us domain is being handled out of the NTIA's international division even though a U.S. "nexus" will be required for all registrants in .us.
No changes are proposed for the legacy structure of .us, including the thousands of exclusive delegations of localities to private registrars; see http://www.nic.us/domain-delegated.txt.
2 Statement of Work, page 2 of the solicitation, footnote 1. above. A similar statement appears in the draft Statement of Work issued in August 2000; see http://www.ntia.doc.gov/ntiahome/domainname/usrfc2/dotusrfc2.htm, D. Expansion of the .us Space.
3 There are actually three separate instruments between the Department of Commerce and Network Solutions. See the GAO report on the relationship between ICANN and the Department of Commerce, GAO/OGC-00-33R at http://www.gao.gov/new.items/og00033r.pdf
4 PL 106-113
5 IANA was the Internet Assigned Numbers Authority a service performed by Information Sciences Institute at the University of Southern California for DARPA until 1998. The IANA functions including oversight and coordination of domain names, IP addresses, and protocol parameters such as TCP port numbers.
6. Until two years ago, the registry was subcontracted to the Information Sciences Institute at the University of Southern California (ISI also had the contract for the IANA), but VeriSign has since managed .us directly.
7. See Amendment 19 to Cooperative Agreement # NCR 92-18742, Section I.B.11, http://www.icann.org/nsi/coopagmt-amend19-04nov99.htm
8.Letter from Roger Cochetti, VeriSign, to NTIA, October 2, 2000, http://www.ntia.doc.gov/ntiahome/domainname/usrfc2/latecomments.html
9. Last page (34) of the solicitation (1. above).
10. Ibid. "Technical excellence and comprehensiveness of the overall service for usTLD operation is significantly more important than proposed price(s) to .us registrants."
11. See the January 1998 "Green Paper," a proposed rulemaking entitled, "A Proposal to Improve Technical Management of Internet Names and Addresses," http://www.ntia.doc.gov/ntiahome/domainname/dnsdrft.htm
12. http://www.greatdomains.com/; see listings on homepage.
13. The ISC domain survey shows 89 second-level domains, http://www.isc.org/ds/WWW-200101/dist-byname.html, which suggests that there are some additional names that remain grandfathered in from earlier policies.
14. The author has proposed a Digital Opportunity Trust as a possible in broadening participation in the digital economy; see http://cip.umd.edu/projects/dot.htm. See also http://www.benton.org/Policy/US/
16. See footnote 11. above ("green paper"); the "white paper" is a statement of policy, Management of Internet Names and Addresses, issued June 5, 1998, http://www.ntia.doc.gov/ntiahome/domainname/6_5_98dns.htm