Testimony of Larry Irving
Assistant Secretary of Commerce for
Communications and Information
before the House Committee on Science
Subcommittee on Basic Research
September 25, 1997
Good morning. Thank you for inviting me to testify before the House Science Committee, on the issues of Internet domain names. No single force embodies the electronic transformation we are experiencing today more than the evolving medium known as the Internet. Once a tool reserved for scientific and academic exchange, the Internet has emerged as an appliance of every day life, accessible from almost every point on the planet. Students across the world are discovering vast treasures of data via the World Wide Web. Doctors are utilizing telemedicine to administer off-site diagnoses to patients in need. The Internet is being used to reinvent government and reshape our lives and communities. Finally, as the Internet empowers citizens and democratizes societies, it is also changing classic business and economic paradigms.
Given that over the next decade, advances in electronic communication will affect almost every aspect of our daily life, it is important that we look to preserving the integrity and stability of the Internet. And the way in which we address controversial issues surrounding Internet domain names will affect our ability to realize the full potential of the Internet in the years to come.
Background on the Domain Names Controversy
Domain names are the familiar and easy-to-remember names for Internet users (e.g. "www.thomas.loc.gov"). They correspond to unique Internet Protocol (IP) numbers (e.g. 18.104.22.168) that are used to address computers and route traffic on the Internet. Despite the growth and commercialization of the Internet, major components of the domain name system are still under the supervision of the U.S. Government:
The registration and propagation function for key top-level domain names (.com, .org, .net) is performed by Network Solutions, Inc. (NSI) under a five year cooperative agreement with the National Science Foundation (NSF). As of September 1995, NSF permitted NSI to institute user fees for registration and maintenance of domain names. Thirty percent of the fees must be deposited by NSI into an infrastructure fund, the disposition of which has not yet been determined. NSF's cooperative agreement with NSI expires on March 31, 1998, although the agreement provides for an optional six month "ramp-down period."
The policy oversight function for both the administration of domain names and allocation of network numbers is performed by the Internet Assigned Numbers Authority (IANA), operated by the Information Sciences Institute (ISI) at the University of Southern California (USC). IANA performs these services under a contract with Defense Advanced Research Projects Agency (DARPA) that runs through September 30, 1998. In March 1997, when DARPA funding ran out, additional funds were committed by NASA and the Department of Energy. We believe that less than $500,000 will be needed to fund IANA through FY '98.
The United States Government plays a direct or indirect role in the operation of half of the world's nine root servers. The root servers propagate the domain names and corresponding network numbers from the master database maintained by NSI.
U.S. research agencies have been trying to reduce and eventually eliminate their roles in these infrastructure functions, as they have done with respect to the provision of backbone and interconnection services. The decision to allow NSI to charge user fees privatized the allocation and management of domain names from a budgetary standpoint. NSI, however, remains in a monopoly position with respect to the most valuable name space -- ".com"-- creating a situation that is not acceptable to many in the Internet community. This situation could become especially problematic if NSI were to remain in sole control of the .com domain upon expiration of the NSF/NSI cooperative agreement.
In response to NSI's apparent monopoly, and NSF's announced intention to let the cooperative agreement with NSI expire in 1998, IANA and the Internet Society (ISOC) organized the International Ad Hoc Committee (IAHC or the Ad Hoc committee) in September 1996, in partnership with the World Intellectual Property Organization (WIPO) and the International Telecommunications Union (ITU). The Federal Networking Council (FNC) participated in the early deliberations of the Ad Hoc Committee.
The IAHC issued a draft plan in December 1996, which generated some 6000 comments. The final report proposed a memorandum of understanding (MOU) that would establish, initially, seven new generic top-level domains (gTLDs) to be allocated on a non-exclusive basis by new private registrars. Under the MOU, registrants for second-level domains would be required to submit to mediation and arbitration, facilitated by WIPO, in the event of conflict with trademark holders.
The MOU was opened for signature by companies and governments at a meeting sponsored by the ITU in April of 1997. To date, some 150 entities have signed the MOU, including a few major carriers, such as Deutsche Telekom and MCI. Only one government has signed thus far. Meanwhile, the Ad Hoc Committee itself has transformed into an interim Policy Oversight Council (iPOC) and has begun processing applications for new registrars. To date, 10 applicants have been qualified as registrars. This interim Policy Oversight Committee will continue to accept applications until October 16, 1997.
The IAHC process has been criticized for being exclusive, especially for lacking industry participation. iPOC has made concessions to these critics, however, and has removed the limit on the number of registrars for the new generic top level domains. More recently iPOC has invited further public input on several important aspects of the MOU, employing a formalized notice and comment procedure. Overall, iPOC has demonstrated a commendable degree of flexibility in the face of these criticisms.
Since the announcement by NSF that it would not renew the cooperative agreement with NSI, NSI has suggested that it may own a proprietary interest in .com, .net, and .org. NSI has not clarified the exact nature of this alleged proprietary interest, and has stated publicly that, not withstanding such an interest, it will act in the best interests of the Internet including sharing the valuable .com name space.
Request For Comments
The Clinton Administration supports the continued privatization and commercialization of the Internet and is committed to completing the transition to private sector governance. We realize, however, that this transition must be accomplished in a way that enhances the stability of the Internet and ensures its continued smooth operation.
An interagency working group has been studying domain name problems since March 1997. On July 1, 1997, President Clinton announced the Administration's Framework for Global Electronic Commere. As part of this initiative, the President directed Secretary Daley to support efforts to make the governance of the domain name system private and competitive and to create a contractually based self-regulatory regime that deals with potential conflicts between domain name usage and trademark laws on a global basis. This directive has the highest priority at the Department of Commerce and throughout the Executive Branch.
On July 2, 1997, the Department of Commerce, on behalf of the interagency group, issued a Request for Comments on the Registration and Administration of Internet Domain Names (RFC) in order to ascertain the views of the public regarding various domain name policy issues including the appropriate role of government in the operation of domain name system.
To date, we have received over 430 comments, amounting to some 1500 pages. The RFC elicited thoughtful comments from around the world, representing diverse perspectives.
In accordance with this Committee's desire to reduce the number of printed pages, I have not provided a hard copy of the comments received (although I would be happy to do so). These comments are readily available, however, at the NTIA site on the World Wide Web, the URL for which is "www.ntia.doc.gov". I have provided a summary of comments prepared by the interagency working group in Appendix A to my testimony.
Let me start by saying that the comments we received reflect overwhelming support for two things: private sector governance of, and increased competition in, domain name systems. Moreover, the principle of globalism received strong support. Most commenters agreed that the transformation of the Internet from its U.S. roots into a global medium should be reflected in the Internet's administrative bodies. Many warned that continued treatment of the Internet as a "U.S. asset" could provoke a negative reaction from foreign governments, businesses, and consumers.
Having said that, many commenters expressed a preference for making a smooth transition to the "right" system over rushing to adopt an ill-considered solution. We believe, and commenters agreed, that consensus must be achieved through an open and democratic process that is as expeditious as prudent consideration of these important issues permit.
The interagency group posited six principles that might be used to evaluate proposals for registration and administration of Internet domain names:
First, competition in and expansion of the domain name registration system should be encouraged, but interoperability of the Internet is essential. No addressing scheme should prevent any user from connecting to any other site.
The private sector, with input from governments, should develop stable, consensus-based self-governing mechanisms for domain name registration and management that adequately defines responsibilities and maintains accountability.
These self-governance mechanisms should recognize the inherently global nature of the Internet and be able to evolve as necessary over time.
The overall framework for accommodating competition should be open, robust, efficient, and fair.
The overall policy framework as well as name allocation and management mechanisms should promote fair and expeditions resolutions of conflicts over proprietary rights.
A framework should be adopted as quickly as prudent consideration of these issues permit.
In general, respondents supported these principles. Many proposed nuanced revisions or suggestions to emphasize different issues. A number of commenters suggested additional principles. Only a very small number disagreed with one or all the principles entirely. The interagency working group is studying the comments, and may refine the principles in light of the comments. At this point, however, comments did not reflect a view that the proposed principles should be modified in any comprehensive way.
According to respondents, the biggest advantages of the current system are its familiarity and the fact that it generally works. Disadvantages include the lack of competition, concerns about scaling capacity, and, notably, the potential for trademark conflicts. Not surprisingly, then, respondents thought the system could be improved by introducing competition and creating alternative dispute resolution processes to mediate conflicts of all sorts.
Commenters disagreed about the appropriate structure of domain name systems going forward. Some stressed that top-level domains are a global public resource and must be maintained as such. Others argued that top level domains are a private resource for individual businesses to develop exclusively. Nonetheless, commenters agreed that no single company should be allowed to monopolize domain name registrations.
Commenters that mentioned the IAHC proposal, more often than not, supported it (although sometimes for different reasons). Those who opposed the IAHC plan characterized it as an attempt to improperly assert control over the Internet. Even supporters felt that the size and makeup of any governing organization must ensure that policies promulgated represent a true consensus of the Internet community.
In general, respondents identified the root domain database and the root domain servers as an essential core that needed to be preserved as a separate, independent entity, free from government and commercial pressures. To the extent that IANA performs these functions today, most respondents were complimentary of the job it was doing.
As I said, respondents overwhelmingly favored private sector governance of the domain name system, and urged government to take a back seat in the registration and administration of Internet domain names. Jon Postel, Director of the Computer Networks Division at USC said: "The role of government should be to foster a fair system of self-governance for the Internet that embraces open competition where possible on an international scale."
Commenters had more varied and even contradictory attitudes toward the role of international, inter-governmental organizations (IGOs), especially the ITU. Some commenters expressed deep misgivings about IGOs taking a leading role in Internet governance. Many commenters viewed organizations like the ITU as unaccountable, unelected, and unlikely to consult with the Internet community. Others criticized it as moving too slowly to address rapidly developing Internet issues. Still, a sizeable minority of commenters argued that international organizations provide ready-made fora for reconciling competing national and commercial interests. International organizations were also seen as exerting a useful check on domination of Internet governance by the United States.
Creation of New gTLDs
The comments reflected extensive support from the technical community for the addition of new gTLDs. On the other hand, trademark owners for the most part weighed in against the creation of new gTLDs out of concerns that they would increase policing burdens and give bad actors more opportunities to infringe.
A substantial number of commenters addressed the type of new gTLDs that might be created. Several suggested that any new gTLDs that are created should be distinct enough in their intended purpose to minimize the chance of confusion created by the desire of name holders to register their name in all domains. Though many commenters found the concept of creating special-purpose gTLDs attractive, they expressed concerned that the specific top level domains proposed by IAHC could "raise questions about global transparency and potential duplication and user confusion." Several commenters urged that gTLDs correspond to the categories in an industrial classification system. Others suggested that we address the need for complementary business gTLDs by encouraging the use of ".com.us" and perhaps by "cloning" .com (that is, by allowing the use of ".com1", ".com2", etc.) to mitigate the demands for the same name space by companies with similar names.
Policies For Registrars
Most respondents favored a system in which domain name registrars generally exercised non-exclusive control over gTLDs. These commentators believe that shared registries will protect consumers by promoting competition and enhancing choice, avoiding problems of "lock in" and high switching costs. Even in this group, however, commentators noted that there may be circumstances under which exclusive TLDs may be desirable. Several cited ".gov" as an example of an appropriate exclusive TLD.
Another sizeable group of commentators favor a mixed shared/exclusive system, noting that these distinctions can serve as a useful dimension of competitive offerings, creating more choice for consumers. This group favored allowing the market to determine the optimum mix of exclusive and non-exclusive top level domains.
In general, respondents did not believe that technology would ultimately limit creation of shared or mixed shared/exclusive registry systems.
Many, however, noted that as the technology for shared registries does not yet exist, any transition plan should assess technical obstacles realistically, and plan accordingly. (Several commenters did refer to the shared registry now operating in the United Kingdom. Our understanding is that the underlying technology used in this system is not sufficiently robust for a global DNS system.)
A majority of the respondents favored establishing threshold requirements for registrars, citing the need for stability and consumer protection. Suggested qualifications involved: technical skills, operations skills and experience, and financial resources. A number of respondents cited the IAHC requirements approvingly, although one commenter described them as "too U.S.-centric." The iPOC itself cautioned that threshold qualifications should be kept to a minimum in order to promote diversity and participation in the DNS by developing countries.
A sizeable minority of respondents asserted that the marketplace should determine whether and what requirements and responsibilities a domain name registrar should have. Registrants, in the view of this group, should be free, after full disclosure, to deal with any registrar they choose.
The comments indicated general agreement that trademark rights (registered and common law trademarks, trade names, business names, etc.) should be protected. Commenters focused, however, on how trademark rights should be protected, i.e., by national courts or some other types of dispute resolution mechanism.
All commenters agreed that the national courts should remain an option for trademark protection, and indeed the technical community expressed a preference that this should be the only forum for trademarks disputes. However, trademark owners and attorneys indicated that they would prefer an additional alternative dispute resolution mechanism for some kind of first level clearance to deal quickly with cyber-pirates.
There was general consensus among the commenters that there should be no preliminary review process. First, such a review would add unwanted delay to the registration process. Second, most commenters did not want the registrars to review applications on any substantive grounds. Finally a standard of review would be too difficult to establish and enforce.
Only a few respondents supported the institution of a waiting period after the filing of an application but prior to registration so that disputes could be filled. However because there are currently a fairly low number of actual disputes, it appears that establishing a short period after registration, during which a domain name could be suspended in the event of a conflict, would be a less intrusive procedure. For the majority of domain names there has been no trademark conflict.
There is general agreement that domain name/trademark conflicts cannot be prevented in an international arena - but they can be minimized with certain technological solutions. The commenters agreed that a searchable domain name database with up-to-date contact information would certainly be helpful for clearance purposes. There were also a few comments suggesting that a worldwide trademark registration database would be helpful to deter conflict, however it is generally agreed that such a database would be too difficult to maintain.
Overall, we found the comments to be extremely thoughtful and quite helpful. Obviously, there is much that is not settled under the proposed plans for domain name system governance, and the public is justifiably concerned about the stability and reliability of the Internet environment in the wake of any new system of governance and new technologies for registries.
As I said, the comments supported transition to a private sector driven, competitive system of domain name allocation and management. But a number of issues must be resolved in the course of that transition.
For the sake of simplicity, I believe we should divide these issues into three categories.
First of all, it does not appear that the technology to manage a fully competitive registry system involving shared top level domains currently exists. Although almost everyone concedes that such technology can be readily developed, estimates about development time and cost vary widely. Some assert that existing technology can be easily modified and ready for testing quite soon. Others predict a six to eight month development path, or longer. Cost estimates range from virtually nothing, on the theory that this technology will be developed cooperatively, to several million dollars.
Second, it seems clear that the governance structures proposed to date are not adequately democratic or representative of the ever-expanding Internet community. Additional discussion is needed to reach consensus on a private sector governance structure for DNS, with attention to important questions such as: "Who gets to decide? How are the decision-makers chosen? And what recourse is available in the event that the Internet community becomes dissatisfied with the performance of the selected decision-makers?" Moreover, we must transfer the policy oversight functions now performed by IANA to a permanent setting with stable funding, deep expertise, mechanisms to guard against anticompetitive behavior. As I said, the Internet community by and large expressed satisfaction with the current arrangement, so that the need for change here seems less pressing. Nonetheless, the inter-agency working group, DARPA, and USC all agree that the current arrangement is not viable on a long term basis.
Finally, additional consensus is needed with respect to important intellectual property problems, especially, but not exclusively, in the area of trademark. The IAHC has set out the contours of a dispute resolution process, the details of which require further attention. Trademark owners must be comfortable that their rights will be protected adequately, if not perfectly. The number and type of new generic top level domain names must thus be carefully considered.
Resolving the question surrounding technology, governance and trademark will, unfortunately take some time. We are quite confident, however, that adequate will and the basis for agreement can be mustered to complete the transition to a robust, stable, competitive, global, and private-sector driven system over the course of the next year.
First of all, we believe the U.S. Government must act promptly to assure the Internet community that it will preserve the stability and integrity of the Internet during a transition period. We believe that IANA should be funded during the transition period. We also believe that the government should, if necessary, utilize the NSI "ramp down" option in the cooperative agreement, providing the option to rely on the current system through fiscal year 1998. Finally, we also believe that the government should, if necessary and in accordance with the provisions of the cooperative agreement, receive a copy and documentation of any and all software and data generated by NSI under the cooperative agreement in such form and sufficient detail as to permit shared registration in the .com space.
Second, the interagency working group must complete and make available a detailed analysis of comments submitted in response to the Department of Commerce request for comments. Based on this analysis, we would expect to issue policy recommendations with respect to each of the questions
Although I do not want to pre-judge the final recommendations, I believe that they are likely to contain the following elements:
The memorandum of understanding proposed by the Ad Hoc Committee should be amended, or another vehicle should be created, to reflect the concerns about governance, dispute resolution, and trademark voiced in the comments received by the Department of Commerce.
Although I have outlined a transition roadmap in very broad terms, the Administration does not underestimate either the importance or the complexity of this task. I look forward to working with all interested parties, including the Congress, throughout this important transition period.
I would be happy to answer any questions that the Members may have at this time.
The Government should support private sector development, testing, and deployment, based on the procedures set forth in the Ad Hoc Committee's memorandum of understanding and the Council of Registrars Memorandum of Understanding or another vehicle, of technology needed to administer shared top level domains.
The United States should work with the global Internet community to establish an independent, self-sufficient policy oversight body with sufficient authority, global credibility, as well as a structure and process that will minimize the risk of anticompetitive conduct to carry out the core aspects of Internet governance that must be undertaken on a centralized basis.