Non-Commercial Users Constituency Letter - To the Governmental Advisory Committee
September 28, 2011
Chairwoman of the Governmental Advisory Committee (GAC)
The Non-Commercial Users Constituency (NCUC) would like to take the opportunity to express its views in relation to the Uniform Domain Name Dispute Resolution Policy (UDRP). NCUC understands that the GAC has sent a letter to the Generic Names Supporting Organization (GNSO), raising some concerns as to whether a Policy Development Process (PDP) should be initiated in relation to a review of the UDRP.
NCUC would like to welcome the GAC’s initiative in approaching the GNSO and submit its statement with its views on the UDRP. NCUC thinks it is very encouraging that the GAC is willing to speak with the GNSO and exchange ideas on the UDRP, one of ICANN’s oldest and most important policies.
It has lately come to our attention, however, that the views of the GAC on the UDRP were drafted with the solicitation and advise submitted not by the entire groups represented at the GNSO Council. NCUC is very concerned with this process and would like to submit for your consideration the statement and the comments it has submitted during the public comment period for the UDRP. We hope that you take into consideration the indisputable evidence that exists and our serious concerns on the current operation and practice of the UDRP.
NCUC would like to thank you and the GAC for your attention to the NCUC's comments and requests the opportunity to discuss the issues related to the UDRP with members of the GAC at the meeting in Dakar. We look forward to cooperating with you on this very important matter.
Dr. Konstantinos Komaitis,
Chair of the Non-Commercial Users Constituency
Comments of the Noncommercial Users Constituency (NCUC) on the Preliminary GNSO Issue Report on the Current State of the Uniform Dispute Resolution Policy (UDRP)
UDRP review should be the presumptive outcome of this process.
NCUC supports a review and believes that the burden of proof is on those who oppose it. We believe so for the following reasons:
All other ICANN policies must be regularly reviewed and improvements continuously sought.
The UDRP has been in place for a long time, was put together very quickly in the earliest stages of ICANN’s existence, and has not been reviewed or modified since.
Many country code TLDs have instituted similar Dispute Resolution Procedures (DRPs). Although many are based in part on the ICANN model, they often introduce slight procedural or substantive modifications. ICANN needs to assess whether it can learn from those differing experiences.
A significant amount of academic research and critical literature has developed around the UDRP which can be assessed for problem areas and mined for proposals to improve it.
No stakeholder or independent commentator now claims that the UDRP is perfect. Virtually all of the panelists on the workshop held at the ICANN 41 meeting in Singapore and in the earlier webinar noted specific improvements or changes they would like to see, even if they did not prefer to invest time in a PDP.
NCUC believes that resistance to the policy’s review comes from two sources. One is a fear of various parties that a formal Policy Development Process (PDP) might make the policy “worse than it is now”. In response to that, we reply that a PDP does not presume that the UDRP will be changed; it simply allows all the Stakeholder Groups within ICANN to systematically explore and answer questions about whether it should be changed and if so, whether there is agreement on ways to change it. If there is no consensus among GNSO stakeholders about any specific change, nothing should change.
Another argument heard against a review is that the timing is bad. Some believe that the timing is bad because we are headed into the new TLD program, which relies on the UDRP and the Uniform Rapid Suspension (URS) policy, and that consideration of change will “destabilize” these efforts. Others believe that other policy-making activities should be prioritized. The first claim seems entirely illogical to us. The new TLD program will expand the usage of the UDRP and the URS will rely on certain elements of the UDRP. Moreover, the URS is designed upon the same substantive and procedural ethos of the UDRP; if the UDRP is not working, we need to ensure that we fix it so that its existing flaws are not transferred to and will not affect the operation of the URS. It seems that the time between now and the actual coming to market of new TLDs is the ideal time to review the UDRP and fix any flaws. Further, we simply cannot agree with anyone who sees the UDRP and trademark – domain name conflicts in general as a low-priority area for ICANN. The UDRP touches on issues that are central to ICANN’s mission and fundamental to the interests of nearly all of its stakeholder groups. If ensuring that the UDRP is functioning properly is not a priority, what is?
To conclude, regular review and updating of policies is often presented as an “ICANN value” or way of policymaking, and it would be terribly odd for that value to not apply to one of the only substantive policies that ICANN has created -one that impacts every domain name registrant. The presumption of any ICANN policy should be constant review and improvement and the UDRP is no exception to the presumption. After more than a decade of practice and evidence, this policy is beyond ripe for a review.
The Preliminary Issues Report
NCUC is not satisfied with the quality of the published GNSO Issue Report. We expect Issue Reports to involve the collection, compilation and analysis of relevant factual evidence, and a neutral and balanced assessment of relevant literature. The Report should be enhanced to take account of the factual evidence and analysis generated in the 12 years since the UDRP has been in effect, to provide the GNSO Council with informative background materials that can prepare it for policy discussions and debates.
Over the past twelve years, the UDRP has generated a significant amount of independent academic research and critical examination. There are also statistical sources available, both in the research literature and from the UDRP providers. If one uses focused search techniques and devotes two or three days to examining the most important materials, one finds about a dozen useful empirical studies and analytical law review pieces on the UDRP. These reports could have and should have been summarized and referenced in the Issues Report. While it does take time to sift through this material and glean its findings, the amount of time consumed would almost certainly be less than that consumed by organizing a “webinar” and a survey. Instead of webinars and surveys, that merely elicit opinions from the handful of people already involved in ICANN who happen to be available, a literature review considers the evidence amassed over time, and subject to peer and critical review. There is an important distinction between an internal opinion poll among insiders with an agenda and a fact-based Issues Report.
ICANN Issues Reports, and GNSO policy development processes generally, should not rely exclusively on casual internal soundings, but rather apply a higher intellectual standard and depth of understanding. If professionals, social scientists and researchers outside of ICANN have devoted resources and expertise to the analysis of ICANN and its policies, our processes can only benefit from drawing upon these materials. We note that the Security and Stability Advisory Committee (SSAC) reports routinely involve reviews of the relevant research literature and often do a good job of compiling and summarizing such literature. The GNSO needs to do the same. As the primary policy making arm for generic domains on the Internet, a higher level of analysis should be sought, as befitting its responsibility to good governance and the global public interest.
Summary of research literature
To aid the compilation of a more complete report, we list and provide brief summaries of the relevant research in the appendix to this comment. The sources are in chronological order. We ask that this information be incorporated into the revised Issues Report. The overview exposes a huge disparity between the Preliminary Issue Report conclusions and the conclusions of most of the research literature on UDRP. The preliminary report portrays the UDRP as completely consistent, fair and universally supported. Nearly all scientific reports, on the other hand, while noting the UDRP’s success at reducing the costs of resolving domain name disputes and rectifying gross forms of cybersquatting, tend to be critical of both substantive and procedural aspects of the UDRP. In other words, independent, neutral research does not support the preliminary report's assertions that UDRP is fair, consistent and in no need of review.
Helfer, Laurence R. and Dinwoodie, Graeme B., Designing Non-National Systems: The Case of the Uniform Domain Name Dispute Resolution Policy. William & Mary Law Review, Vol. 43, p. 141, 2001; Stanford/Yale Jr. Faculty Forum Paper No. 01-05. Available at SSRN: http://ssrn.com/abstract=275468
The article critically assesses the Uniform Domain Name Dispute Resolution Policy (UDRP) as a potential model for solving the legal challenges presented by transborder activity. It describes the conditions that led to the UDRP's formation and considers whether the UDRP can and should be replicated elsewhere. While authors believe that UDRP succeeded in bypassing cumbersome mechanisms of national and international law making and in fulfilling demand for effective dispute settlement, they do not believe that it should be uncritically extended to other contexts. This article contains no empirical or statistical analysis of UDRP results.
Mueller, M. (2001). Rough Justice: A Statistical Assessment of ICANN's Uniform Dispute Resolution Policy. The Information Society Volume 17, Issue 3, 2001, 151 - 163
A statistical analysis of UDRP case outcomes based on the concern that ICANN allows the complainant to select the dispute resolution service provider. The statistical tests indicate that complainant selection of dispute resolution service providers does lead to forum shopping that biases the results.
Geist, Michael (2001), Fair.com? An Examination of the Allegations of Systemic Unfairness in the ICANN UDRP
This study finds that influence over panel composition is likely the most important controlling factor in determining case outcomes. Providers steer a majority of the cases toward complainant-friendly panelists. Fifty-three percent of all NAF single panel cases were decided by only six people, and the complainant winning percentage in those cases was 94%. Provider influence over panelists diminishes in three-member panel cases, 4
since both complainant and respondent choose one of the panelists and exercise some influence over the choice of the third; predictably, complainant win rates go down in three-member panels. The study concludes by proposing changes to the UDRP to instill greater fairness and confidence in the process.
Froomkin, A. Michael (2002), Uniform Dispute Resolution Policy—Causes and (Partial) Cures. 67 Brooklyn Law Review 605.
This paper describes the main features of the UDRP and ICANN’s accompanying rules of procedure and compares them to the recommendations of the WIPO proceeding and Report called for in the 1998 White Paper. The article documents four "unfair aspects of the UDRP rules:" (1) the selection and composition of the arbitral panel; (2) the failure to provide a registrant with adequate time to reply to a complaint; (3) the limited opportunity for complainants who lose a UDRP action to get their cases into court; and (4) the absence of any meaningful check on the providers’ creation of supplemental rules that effectively tilt the playing field. Notes that in practice, UDRP’s requirement that names be registered and used in bad faith “appears to have been completely lost on numerous arbitrators, who have read 'and' as if it meant 'or.'” Notes how the noncommercial use defense imported a tarnishment concept and that this "undermines a substantial part of the free-speech value of the non-commercial and fair use defenses" and "could be used to deny protection to legitimate criticism sites.”
Mueller, M. (2002). Success by Default: A New Profile of Domain Name Trademark Disputes under ICANN’s UDRP. http://dcc.syr.edu/PDF/markle-report-final.pdf
Study supported by the Markle Foundation based on a comprehensive review and classification of the first 4,000 UDRP cases. The UDRP has been an effective remedy for cybersquatting primarily because it makes it economically inefficient for abusive registrants to defend their names. Known cybersquatters default 70–100% of the time. The study’s compilation of case statistics provides many insights into the UDRP’s results. E.g., eighteen percent (18%) of UDRP claims are based upon unregistered trademarks; the UDRP has protected personal names as strongly as registered marks; of the 20 cases UDRP panelists cite as precedents most often, all were won by Complainants and all but 4 were Respondent defaults. The study further provides proof that decisions regarding gripe and/or criticism sites are inconsistent, and calls for clarifying standards for a finding of “confusing similarity” to a trademark such that criticism, parody and commentary are not suppressed.
Selby, John (2004). Competitive Justice?: The Role of Dispute Resolution Providers under ICANN’s UDRP, 1 Macquarie Journal of Business Law 23
Having examined the quantitative and qualitative means by which Dispute Resolution Providers can compete for market share in UDRP disputes, the main thesis of the article is that the requirement that "competition" between dispute resolution providers be a design principle in the UDRP was (and continues to be) flawed. Competition between 5
such providers results in flawed incentives and likely injustices, most notably it can bias those providers towards making it easier for complainants to achieve victory over respondents (particularly through panelist selection processes and differences between provider supplemental rules). Competition between providers of justice in a situation where they are competing for selection by only party is antithetical to the principle of equal treatment of parties in a dispute.
T. Lee, D. Hunter and D. Orr, "Cohesion and Coherence in the UDRP," (2008)
A study that tests the hypothesis that UDRP panelists are more likely to side with complainants or respondents who are of the same nationality as they are. The study used automated techniques to process UDRP decision documents to sample 2944 disputed domains where the nationalities of the panelist and both litigants were known. Based on this data, the hypothesis of a nationality bias appears to fail. However, the overwhelming number of Americans in both the respondent and complainant sample may limit the results’ generality.
Komaitis, Konstantinos. The Current State of Domain Name Regulation: Domain Names as Second-Class Citizens. Routledge, 2010.
Book-length analysis that argues that domain names are a form of property, and the property rights held by domain name registrants need to be recognized in law – independently of, and carefully distinguished from, the limited rights associated with trademark protection. The book discusses the history, legal basis, procedural aspects and performance of ICANN’s Uniform Domain Name Dispute Resolution Policy (UDRP). Analyzes the differences between arbitration and the UDRP, which is sometimes characterized as an arbitral process; most of the differences are shown to weaken the rights of registrants, binding them to a procedure and rules while allowing the complainant more choice and options. The book concludes with some procedural and substantive recommendations that, if applied, will help the UDRP to become a more fair and balanced system.
Substantive and/or Procedural Review of UDRP?
For NCUC, the key question is not whether to conduct a review of the UDRP, but rather, how to conduct a review of the UDRP. NCUC members agree that at the very least, a review of the noted procedural flaws of the UDRP should be thoroughly examined by the impacted community. A number of NCUC members further believe there should be an additional review of the underlying substantive policies contained within the UDRP, including their ability to protect freedom of expression guarantees and the fair use or other noncommercial rights of domain name registrants. Thus while the precise scope and framing of the UDRP review is open for discussion and input from all impacted stakeholders, NCUC contends that at the very least, a procedural review of the UDRP is necessary at this time.
Use standard ICANN processes, not a hand-picked group
Existing GNSO policy development processes are precisely what this type of examination was set up for and allows for balanced input from all impacted stakeholders. NCUC does not support having this process carried out by an arbitrarily selected group of “experts”. The GNSO was established for exactly this purpose and has put in place processes that provide room for multi-stakeholder participation and can reach results that reflect the views of all stakeholder groups within the ICANN structure. The community learned a valuable lesson with the failure of the single-constituency-driven Implementation Review Team (IRT), and created the subsequent Special Trademark Issues Team (STI), which reached unanimous consensus within the GNSO and drew much less criticism concerning issues of bias or misrepresentation, because it had incorporated members from across the community. We, therefore, suggest that ICANN make use of the existing GNSO processes that were established to evaluate policies and make recommendations as the community deems necessary.
Ever since the issue of a potential review of the Uniform Domain Name Dispute Resolution Policy (UDRP) was placed on the table, two camps were created: those advocating against a review – the Intellectual Property Constituency (IPC), some big brand owners and their lawyers and the UDRP accredited centers, who have been basing their arguments on issues of timing. Their main (and only) argument has been that a potential review of the UDRP would upset the current status in the adjudication of domain names and would create instability in the new gTLD environment. The other camp, the non-commercial users, have been advocating that the timing for a review is right, given the fact that the UDRP is one of the oldest ICANN policies and has been operating for more than ten years now. In the comments submitted during the public comment period, the Non Commercial Users Constituency (NCUC) outlined the academic research that has been conducted in relation to the UDRP, its process and its substance. For the past ten years, academic scholars, including myself, have indicated that the UDRP is in serious need for a review and that the system is flawed to the extent that it does not have any checks and balances, it is occasionally biased and it is procedurally flawed. These are not arbitrary accusations, but are based on extensive scholarly research and study.
The ICANN Staff Report that came out in the beginning of August suggested against reviewing the UDRP, but the decision ultimately sits with ICANN’s Generic Names Supporting Organization. One thing that we need to understand, which might assist in also understanding why there is such a great upheaval when it comes to the UDRP, is that there are many interests at stake. The accredited centres do not want to see the UDRP reviewed because for them the adjudication of domain names has become a business. WIPO and the National Arbitration Forum have managed to generate a great deal of business from domain name disputes and many of their panellists have become quite comfortable in earning their living out of the UDRP. For the trademark community, the UDRP has provided the means through which they are able to exert their power over the domain name registration culture and, to a certain extent, control the ownership of names and words used in our everyday vocabularies. At the same time, through the registration of domain names, they have managed to receive international protection for their marks, something that they could not have achieved by following the traditional trademark registration channels. So, I do understand (although I don’t agree nor do I accept it) why these bodies are pushing ICANN not to initiate a PDP on the UDRP.
But, there is something that I do not understand. Only yesterday, a letter sent by the GAC surfaced in the email list of the GNSO. According to the letter, the GAC is asking the GNSO not to proceed to a UDRP review at this stage. “The GAC is aware that the GNSO is considering initiating a Policy Development Process (PDP) for the purpose of reviewing the Uniform Domain Name Dispute Resolution Policy (UDRP). The GAC wishes to convey to the GNSO its view that a decision in favour of such a PDP at the time of launching the new gTLD program would have public policy implications of great concern to the GAC”. I have too many problems with this (initial) statement. First of all, I do not recall another time (and I might be mistaken) when the GAC has sent a letter to the GNSO asking for anything; doing it for the UDRP is quite alarming. But, more importantly, why the sudden interest? Considering that the whole issue of the UDRP started in May, why sent this letter now? The GAC did not participate in the UDRP webinar, did not participate in the Singapore UDRP panel session and did not submit any comments on the issue during the public comment period. Generally, I have never (and mean never) heard any of the GAC members talk or being concerned about the UDRP. So, why now? How did the UDRP become a priority of the GAC?
From what the letter states, it is because the GAC is concerned about the impact on 'public policy'. Yes, you have heard right - again the never-ending public policy argument. But, the idea that the UDRP should not be reviewed due to ‘public policy implications’ is as ridiculous as the idea of the UDRP being a fair system. First of all, this does not make any sense. What ‘public policy implications’ could a review of the UDRP create? Isn’t it a public policy issue the protection of free speech through clear provisions? Is it not an issue of public policy the need to provide to an international system of adjudication with clear checks and balances? These are quintessential issues that any democratic government should seek to protect. So, why doesn’t the GAC strive to secure these fundamental issues on the basis of public policy concerns? Well, the answer appears to be, because the letter did not come straight from the GAC. According to the meta info and the information retrieved from the properties of the word document, the letter was written by a trademark law firm – specifically the US based law firm of Covington & Burling. I am wondering whether the rest of the GAC is aware that their official position on a very important issue concerning the dispensation of rights between trademark owners and domain name holders is coming from a US law firm.
The rest of the letter follows more or less the same arguments we have so far heard from the trademark community: “The new gTLD program will see the introduction of several new rights protection mechanisms. The inevitable uncertainty associated with the introduction of these untested RPMs would be compounded if simultaneously the future of the primary, pre-existing and proven PRM – the UDRP – were also subject to uncertainty as a result of a long-running PDP”. I, personally, never understood this argument, no matter how much I have tried. But, let’s say that I am just dumb - in any event, the first step of a the PDP will not be to change the UDRP – it will be to ask the community to determine whether the UDRP needs to be reviewed. So, there is really no reason to panic or to make such extreme statements, seeking to spread unnecessary fear to the community.
Another interesting point in the letter is the following: “Further, the GAC notes that almost 50% of the country code top level domain (ccTLD) operators have either adopted the UDRP itself or a variant of it, such that any potential changes to the UDRP that could result from a PDP would potentially have broad implication beyond the remit of the GNSO”. Rubbish – even if we accept this figure, why make a review of the UDRP dependant on the stability of these processes? Needless to say that many of these ccTLD operators have taken the UDRP and have created much fairer and more balanced versions of it. Take, for example, NOMINET’s equivalent policy. It started as a copy of the UDRP, but it is currently a policy document that has managed to surpass the UDRP in many ways: it has incorporated clear safe harbour provisions and it has adopted an internal appeals process, providing this way a system with more checks and balances. (The STI actually suggested its safe harbour provisions based on NOMINET's wording). Why was NOMINET allowed to amend and create new provisions within its own body of rules and the UDRP is not allowed to do the same? Wouldn’t the GAC like to see a more balanced, better and fairer UDRP, especially since there is evidence that this can be achieved?
Therefore, this GAC letter to the GNSO only tells us one thing: the same way the trademark community used the GAC to channel its concerns on the STI recommendations, the same way it is using it now to prevent a potential review of the UDRP. But, whereas in the STI recommendations, the GAC had sort of made its own research and sort of knew what it was talking about, in this case the GAC appears clueless and it is embarassing. If the GAC knew the history, process and the various issues of the UDRP, it would realise that not reviewing it has public policy implications (to use their own language) that governments should be concerned with.
So, I would recommend the GAC to go back and read all the academic writing that exists and dates back to 2000. It is really an eye-opener and, in any case, it would help the GAC form its own conclusions and possibly write its own letter.
Comments submitted on the draft policy promulgated by NOMINET, UK: “Domain Names used in connection with criminal activity”
I would like to welcome NOMINET’s initiative in recommending and seeking to draft policy related to ‘Domain Names used in connection with criminal activity’. In the current Internet environment such policies are necessary to protect consumers and the associated public from any form of criminal activities that are conducted through the registration of domain names. In particular, criminal activities relating to financial crimes, such as phising, or those targeting children constitute a plague to the Internet’s ecosystem.
To this end, such an initiative is a very important step towards a more secure Domain Name System (DNS) and the Internet. However, it is also important that any such policies adhere to principles of proportionality and due process. The fact that NOMINET remains committed to the protection of free speech and political expression is a very positive step; but at the same time, more work is necessary in ensuring that appropriate rules are in place to safeguard those principles.
One thing that should be made very clear is that NOMINET should not engage in determination of criminal conduct, but should act as the agent assisting in enforcing decisions that are taken in other fora. It is important that an independent panel is established that will be responsible for making determinations regarding the criminal conduct that domain name registrants engage in and is promoted through domain name registrations. NOMINET should not act solely on advice received from public law enforcement agencies. The latter can operate as the agents flagging the potential criminal activity, but any final determination should ultimately rest on the expertise, impartiality and judicious authority of an independent panel.
Moreover, an additional value deriving from having an independent panel relates to the complexity of domain names. Domain names are multifaceted articles and, in many instances, one cannot make final determinations on the basis of their content. For example, experience has shown that domain names may host content, which is both legal and illegal. It is important, therefore, that any policy that is in place is in the position to distinguish and differentiate between illegal and legal material hosted under one domain name. In such instances, the role of the independent panel will be to advise the registrant of the illegal content hosted on his/her domain name and request for immediate amendments. In cases like this, suspension of the entire domain name should be the last resort.
Finally, I would strongly encourage the implementation of an appellate mechanism within this policy. It is pertinent that for any such disputes, an appeals process is available to registrants. Considering the international nature of domain names and the jurisdictional issues that the Internet has generated, registrants are not prima facie aware that their domain name registrations engage and/or encourage criminal activities. Subsequently, an appeal process will provide registrants with the opportunity to explain themselves and amend the domain name or the criminal part thereof. Needless to say that an appellate machine contributes to a more robust, balanced and fair process and ensures that justice is appropriately served.
I would like to thank you for the opportunity to submit.
Dr. Konstantinos Komaitis,
University of Strathclyde Law School,
Author of the book: “The Current State of Domain Name Regulation: domain names as second-class citizens in a mark-dominated world”.
Views are my own and my own only!