A policy framework is needed before a .com contract is approved
These comments are an expansion of my spoken comments during the board's public forum in Vancouver and in the dialogue between the Registrars constituency and the board. I am addressing these comments to the staff, the board and the GNSO council because I believe that all three play a critical role in successfully resolving the issues surrounding the proposed .com contract in a manner that is most beneficial to both ICANN and its community.
I recognize that these comments are quite lengthy and appreciate your willingness to read them.
NO ICANN POLICY EXISTS TO GUIDE REGISTRY CONTRACTS
The current registrar accreditation agreement memorializes the Registrar Accreditation Policy adopted on March 4, 1999 (http:// www.icann.org/registrars/policy_statement.html). In addition to qualifications for initial accreditation, the Accreditation Policy contains significant detail relating to the ongoing responsibilities and operations of all registrars, including: the process by which the accreditation is renewed, the process for resolving disputes under the agreement, and ownership of data.
On the other hand, there is no formal ICANN policy that broadly applies to the ongoing operations of a gTLD registry. There are a few consensus policies that have been developed that apply to registries, for example the Transfer Policy, but these describe a very narrow portion of the registry's obligations or the nature of its relationship with ICANN and the marketplace. Arguably, there are some policy statements, such as the green paper and the white paper, that precede ICANN's policy process and should apply to registries, but these are generally at a very high level and are not easily translatable into specific contract provisions. There is simply no registry analog to the Registrar Accreditation Policy.
ABSENCE OF A REGISTRY POLICY RESULTS IN BOTH INCONSISTENCY AND CONFLICTS WITH THE COMMUNITY
Because the staff cannot rely on a policy to develop registry contracts and to guide negotiations, the current gTLD registry contracts contain significant discrepancies on a variety of issues, many of which are issues of policy. For example, some contracts include presumptive rights of renewal and others do not. Some contain fixed prices, some contain price caps that can be adjusted upward periodically, and others contain no price controls at all. Some contain a specific process to approve registry services, while others merely incorporate the existing consensus policy on the matter. In all of these cases, the ICANN staff has been forced to make judgment calls about the appropriate approach to each of these policy issues. It is no surprise, then, that changes in staff have led over time to significantly different contractual terms. While this result is perfectly understandable, it has nonetheless led to significant inequities between registries in terms of the conditions under which ICANN has been willing to enter into contracts with them. This is especially problematic because the dominant market player, VeriSign, seems to be operating under the most favorable terms available to any gTLD registry. As a result, it is difficult for newer gTLD operators to become effective competitors.
Additionally, registry contractual terms have resulted in conflicts between the ICANN staff (and board) and the rest of the ICANN community. In many cases, the staff will make judgment calls about a specific contract that results in an arrangement that does not reflect the understanding or assumptions of the community about key policy issues that are reflected by registry contracts. Many of the largest controversies within ICANN over recent years have centered around these issues.
THE SOLUTION IS TO DEVELOP A POLICY FRAMEWORK FOR REGISTRIES
The obvious solution to this problem is for the GNSO to develop a set of policies that would provide a framework for the critical elements of the registry contract. The GNSO has a defined policy development process that can be used to create this framework in a timely manner. Examples of the the types of issues that might be addressed by this policy framework include:
1) Required services to be provided by registries
2) Mechanisms by which prices are set and adjusted, including consideration of relying on market forces
3) Standards of conduct to promote competition between registrars
4) Process by which registry contracts renew
The above list is not intended to be exhaustive, but it illustrates the range of issues that are currently not guided by policy, but rather rely on situational judgments by the staff.
Once the GNSO developed such a policy framework, the board would have an opportunity to consider whether or not the policies were in the best interests of the Internet community. If so, they could be approved as consensus policy to guide the staff in developing registry contracts. If not, the board would have the option to reject the proposed changes (along with reasons why) and request that the GSNO reconsider the policy framework.
A POLICY FRAMEWORK FOR REGISTRIES WOULD BENEFIT ICANN IN SEVERAL WAYS
A consensus policy framework for registries would benefit the ICANN staff, board, and community in several important respects. First, the framework would set expectations for registry operators about the rules they will operate under and should significantly streamline the contracting process by clearly delineating negotiable business terms from non-negotiable policy matters. This would improve ICANN's relationship with existing gTLD registries, as well as simplify the process for introducing new gTLDs to the point that they can be added in a scalable and predictable manner.
Second, this framework will reduce the number of "surprise" contracts that generate significant controversy within the ICANN community. The ICANN staff will no longer need to make a best guess as to what registry policies should be memorialized in registry contracts, a process which inevitably leads to dissatisfaction on behalf of a community that is not consulted in the process (due to the very nature of bilateral negotiations) as well as a considerable amount of additional work for staff. Instead, each contract would simply be aligned to the stated policy expectations of the community, as is the case with all other consensus policies.
Third, a policy framework will help frame the discussions of the proposed .com contract and determine which elements, if any, need to be adjusted in order to conform with the community's expectations. Once these policies have been developed, they can be compared to the proposed contract. If the contract is consistent with the policies, the board should simply approve it. If there are areas in which the contact and policies are inconsistent, the policy can guide the staff in renegotiating a contract that does conform. This approach would also allow other registries to reasonably expect that they could negotiate their contracts to bring them into alignment with the policy as well.
Finally, the creation of a policy to guide gTLD contracts would send a clear signal that ICANN is able to successfully use its bottom-up process to make critical policy decisions, and to do so in a manner that can resolve pressing real-world concerns. This approach helps ICANN establish greater credibility both with its own community and to the rest of the world by demonstrating that it is a functional and successful model for Internet governance.
THE .COM CONTRACT SHOULD NOT BE MODIFIED WITHOUT FIRST ESTABLISHING THIS POLICY FRAMEWORK
ICANN's bylaws provide that contracts with significant policy changes should be posted for public comment, and this requirement has been fulfilled for the proposed .com contract. Posting contracts serves a critical function in achieving ICANN's core value of transparency, and allows the board to be fully informed when determining whether or not a specific action is in the best interests of the Internet community. However, public comments should not be taken as a substitute for the bottom-up policy development process. As chair of two GNSO policy development processes that have produced a number of consensus policies, I have learned that developing policies that benefit the entire community requires dialogue and compromise by a variety of interested parties. Groups that have policy preferences which initially seem to be fundamentally opposed can, through good- faith discussion, eventually agree on compromise solutions that many times do not resemble any of the initial positions. If the board and staff simply examine public comments and attempt to use them to guide further contract negotiations, they will quickly discover that a large number of "specific recommendations" have been presented, but many will conflict with one another with no obvious mechanism to resolve those conflicts. More importantly, such an approach robs the community of the opportunity to develop creative compromise solutions to many of the issues that may seem intractable based on a reading of a collection of public comments advocating for each individual or organization's wish list of specific changes to this particular contract.
Using public input to modify a specific contract also encourages similar controversies around other gTLD contracts. Because the fundamental problem--lack of a policy framework--will not be resolved, ICANN staff will continue to have to make arbitrary policy decisions based on their best judgments at the time. As we have seen time and again, this approach is nearly certain to produce controversy. Now is the time to direct the considerable energies that have been evoked by the proposed .com agreement to create a policy framework, not to allow a patchwork solution that will assure us of similar problems in the future.
Some may argue that the GNSO will be unable to produce a policy framework in a timely manner. I believe that the community understands that it needs to act with considerable (perhaps even unprecedented) speed to produce the required consensus policy and will succeed in doing so. Certainly, the GNSO needs the minimum time demanded by ICANN's bylaws to complete the Policy Development Process, and the board should ensure that no agreement is approved until the GNSO has been given the opportunity to comply with the bylaws and produce a policy. If there are specific issues that require resolution prior to the completion of the policy process, the staff should endeavor to work with VeriSign to resolve those specific issues, rather than allowing those issues to rush the approval of a contract as important as the .com registry agreement. Failure to do this simply because the policy development process might not be timely in some circumstances is to deny the GNSO the opportunity to fulfill its most fundamental purpose within ICANN--developing policy on matters of interest to the community.
The proposed .com contract contains a provision that allows it to perpetually renew, and identifies a number of areas that are not subject to the consensus policy process. Approving this contract without first resolving the underlying policy issues means that we have lost the opportunity to align our most important gTLD registry with the community's policy expectations (and, by extension, other gTLD registries) not just now, but forever.
A FINAL NOTE
I ask you to step back for a moment and to consider the roles of the various parts of ICANN imagined by those who helped create the organization, as codified in the bylaws. The job of the community is to work through the bottom-up process to develop policies. The job of the board is to consider whether those policies are consistent with the best interests of the Internet community and to provide strategic guidance to the staff. The staff works in tandem with the board and community to implement those policies, including by developing contracts that reflect approved consensus policies. Today, we seem to be asking the community and board to do the job of the staff by resolving issues around specific contract language. This approach is doomed to fail because the competing interests in the discussion are not engaged, and because the board lacks any objective means to resolve the conflicts between the many views that have been presented.
There is a better approach that will produce better results. Let's all do our jobs, work together, and develop a policy framework that can simultaneously help resolve the current .com crisis, resolve inequities that restrain competition between gTLD registries, ease the introduction of new gTLDs, and announce the success of ICANN as place where the bottom-up ideal can result in the best possible stewardship of vital naming and numbering resources.
---- Jordyn A. Buchanan Director of Policy Register.com