Rechercher dans :

Lex Hebdo



Accueil / Lex Electronica

Lex Electronica

Revue électronique du Centre de recherche en droit public

The Form and Substance of Domain Name Arbitration

David LAMETTI (*)

Lex Electronica, vol. 7, n°2, Printemps / Spring 2002



Fort de son expérience d’arbitre, l’auteur présente ici une série d’observations sur l’arbitrage en ligne portant sur les noms de domaine.

Il note que grâce à sa simplicité, la tribune eResolution permettait non seulement à toutes les parties de faire état commodément et efficacement de leurs prétentions, mais contribuait aussi à relever la qualité des arguments puisque chaque partie y fournissait un aperçu préalable assez clair des faits qu’elle allait produire en preuve. Du point de vue d’un praticien, écrit-il, l’accès à toutes les prises de position des parties et aux documents les étayant était presque instantané. La disparition de eResolution révèle un certain nombre de problèmes sérieux relatifs à la gestion des noms de domaines telle qu’elle était abordée par ICANN et ses fournisseurs agréés de services. En particulier, elle souligne l’existence d’un a priori institutionnel en faveur des détenteurs de marques de commerce comme aussi la possibilité pour les demandeurs d’opter pour le forum où les positions qu’ils comptaient défendre avaient les meilleures chances d’être accueillies favorablement. Le rayon de soleil qui transperce ce nuage électronique est l’espoir que les fournisseurs de services en résolution de conflits qui restent en fonction et ceux de l’avenir — tant au sein de l’URDP que des systèmes nationaux de résolution de conflits en matière de noms de domaine — veilleront à incorporer à la fois la forme et l’esprit de l’approche eResolution. En ce sens, les observations que voici constituent vraiment un épilogue à ce qui était une recherche d’intégrité en ligne.


Drawing material from his experiences as an arbiter, the author presents a set of reflections on online domain name arbitration.

He observes that "eResolution’s platform, through its simplicity, allowed not only easy and effective pleadings from all parties to be entered into the record, but also helped to raise the quality of the arguments by giving a comparatively clear indication of what each party had to show in its arguments. From the perspective of a panelist" he writes, "access to all arguments and supporting documents was practically instantaneous. eResolutions’s demise underscores a number of critical problems with domain name management under ICANN and its sanctioned dispute resolution providers, and in particular institutional bias towards trademark holders and the possibility of forum shopping. The silver lining in this electronic cloud is that hopefully remaining and future dispute resolution providers under both the UDRP and national domain name dispute resolution systems might incorporate both the form and spirit of eResolution’s approach. As such, this set of reflections truly is an epilogue to what was an effort in “integrity on line”".

Table of Contents

I. Epilogue

II. The Form of Domain Name Arbitration

A. Quality of Arguments

B. Allegations of Bias in Favor of Complainants and Forum Shopping<

C. Should there be appeals?

D. Conclusions on Form

III. The Substance of Domain Name Arbitration

A. The question of “identical or confusingly similar”

B. The distinction between “legitimate connection” and “bad faith”

C. Awkward burdens of proof

D. The influence of (usually American) domestic law and the role of precedent

E. Conclusions on Substance

Conclusion (& Prologue)

Don’t it always seem to go, that you don’t know what you’ve got ‘till its gone.

Joni Mitchell

I. Epilogue

1. It may seem rather odd that this paper begins with an afterword instead of a prologue. Circumstances, however, dictate such a departure from accepted literary form. When I accepted to give this set of reflections on domain name arbitration from the point of view of an arbiter, eResolution was still a going concern. The project that was eResolution, from my experiences as an arbiter, turned out to be an excellent, efficient and just vehicle for domain name dispute resolution under the UDRP, and would have been an excellent on-line platform for other sorts of domain name and keyword disputes. eResolution’s platform, through its simplicity, allowed not only easy and effective pleadings from all parties to be entered into the record, but also helped to raise the quality of the arguments by giving a comparatively clear indication of what each party had to show in its arguments. From the perspective of a panelist, access to all arguments and supporting documents was practically instantaneous. eResolutions’s demise underscores a number of critical problems with domain name management under ICANN and its sanctioned dispute resolution providers, and in particular institutional bias towards trademark holders and the possibility of forum shopping. The silver lining in this electronic cloud is that hopefully remaining and future dispute resolution providers under both the UDRP and national domain name dispute resolution systems might incorporate both the form and spirit of eResolution’s approach. As such, this set of reflections truly is an epilogue to what was an effort in “integrity on line”.

II. The Form of Domain Name Arbitration

2. Most readers will be familiar with the ICANN UDRP process. Registrants make certain affirmations when they register a domain name: they must warrant that they are not knowingly violating rights of third parties, are not using the domain name for an illegal purpose, and not using the domain name in violation of applicable laws and regulations2. In practice, these principles boil down to the substantive criteria for deciding applicable disputes in Paragraph 4: the domain name must not be identical or confusingly similar to a trademark or service mark (or image which is the subject of a right of publicity), the registrant must have a legitimate interest in the domain name, and must not be acting in bad faith by registering or using the domain name. Registrants further agree to be bound by at a minimum by a mandatory administrative proceeding before a sanctioned service-provider arbitration under the UDRP in cases where a dispute arises3. When a complaint is lodged and such proceedings are begun, registrants become respondents, and may choose to defend themselves in the proceeding. (They may also choose to defend themselves in a court proceeding in the jurisdiction in which the complainant has agreed to appear when lodging the complaint4).

3. There is now a large body of published arbitration decisions on various web sites5, as well as an increasingly large body of scholarly commentary and analysis6. Moreover, the literature contains not merely what one might call a first generation of expository texts, but also a great deal of criticism about the process from high profile commentators.

4. My set of reflections is on particular formal and substantive aspects of the domain name adjudication process that I have found troublesome as an arbiter or which I am aware are problematic. I have grouped the challenges faced by the parties under the UDRP into categories of form and substance; of course there is overlap between the categories — indeed one of my central points is that improving the complaint and response forms will improve substance — but the distinction is a useful one nevertheless. I shall deal in this part with the form of domain name dispute resolution, in the following part with issues of substance. In my view, the UDRP process is generally well-conceived; however, improving the form of arguments and deepening the understanding of the substantive components of the UDRP will lead to substantially better results in terms of justice and fairness. In particular, improving the written form of domain name disputes will lead to a marked improvement in the quality of argument, which in turn will lead to a better appreciation of substantive arguments. Unlike some commentators, I do not believe that three-person panels necessarily will lead to better decisions in all cases. Nor do I believe that there should be an appellate process in UDRP proceedings.

5. My formal reflections are grouped as follows: how to get the best written arguments from the parties; how to deal with allegations of bias and forum shopping, and whether an appellate procedure would be useful.

A. Quality of Arguments

6. It seems to me that the single most important challenge for proceedings under the UDRP — more important than unfairness associated with forum shopping — is getting the best written arguments pleadings from the parties such that arbiters have the complete substantive record in front of them. Coupled with greater substantive clarity on issues of burdens of proof and the distinction between legitimate connection and good faith7, this would, in my view, go a long way in rectifying any perceived problems of bias within the system. That is, with more meaningful participation and better pleadings it is my view that arbiters will get to better results more of the time, and in particular will find somewhat more often in favor of respondents.

7. The first issue is getting respondents to respond. Not surprisingly, early statistics on the outcomes of UDRP proceedings demonstrated that respondents who responded fared much better than those who did not. While my impression is anecdotal, it seems to me that an Internet-based platform, such as eResolution’s, is an ideal way to entice a largely technology-friendly population into participating in the arbitration process. As domain names are in effect international, the concept of Internet-based UDRP dispute resolution remains enlightened despite eResolution’s initial failure in that respondents could simply and efficiently choose to participate in the process. Moves in this direction by other providers is welcome8; indeed I would suggest that they consider purchasing the eResolution platform or licensing it from those who happen to own it now (if they have not already done so).

8. The second formal issue goes to improving the quality of argumentation. In my view the complaint and response forms of all service providers ought to clearly lead both complainants and respondents through the required argumentative steps. On such a form, one cannot simply repeat verbatim the provisions of paragraph 4 of the UDRP. Rather, the provisions must be broken down into a more simple format that allows even the lay person to follow the form and understand what is required to make his case, both in terms of argumentative formalities and in terms of supporting evidence. Specifically, the form should go so far as to say what sorts of evidence are necessary or useful to support claims. Such a form would have the effect of ensuring that all parties understand what is required of them. At present, in my experience, some parties do have a clear idea of what they need to argue and demonstrate, while other parties — even those represented by counsel — often do not. Both the complaint and response forms could be specifically tailored to account for different burdens of proof, etc.

9. In terms of forms, I would even go so far as suggesting a checklist format of direct, leading questions requiring fill-in-the-blank answers. Then, depending on the answers, longer prose answers would be required and the types of documentation necessary to prove the “answer” would be spelled out. So, for example, regarding “legitimate interest”, on the complaint form:

Q: Is the Respondent commonly known in business circles by the domain name?

If NO, please give evidence of why not (for example, attach a copy of a trademark or corporate registration, bills, receipts, business communications, other examples of third party documents, etc. which show that the Respondent does not use the domain name or do show the Respondent’s actual business name(s) in use).

Q: To the best of your knowledge, has the Respondent ever divulged in any way, directly or indirectly, plans to use the name in the future?

If YES, are the plans credible?

If NO, please give evidence of why not (for example, has the Respondent taken any steps to implement the plan, has a great deal of time elapsed without activity, is the Respondent’s plan linked to his present commercial activity, etc.).

10. The respondent’s forms would have similar questions framed from the respondent’s perspective, and arguments. As much detail as is practically possible ought to be incorporated into the form.

11. Obviously these examples would need further refining. But the point is simple: the forms must be more “leading” and must help parties to understand their arguments, evidentiary burdens, etc. and must help the parties to frame them. While it is true that these forms will not be able to cover every conceivable instance of potential domain name conflicts and disagreements and that they will be somewhat longer than present complaint and response forms, at a minimum they must force the minds of the parties to examine the situation in some degree of rigour. A properly written and leading form, by forcing the parties to at least think by analogy, will also have the effect of organizing the substantive arguments. In turn, arbiters will have a better record before them. In short, better forms will lead to improved substantive decisions.

12. This is especially true on the response form, where in my experience there is less likely to be a lawyer involved. Clear questions will help explain the substantive issues, particularly regarding the difficult questions of “legitimate right or interest” and of “bad faith”, framed from the perspective of the respondent. Lawyers too would benefit, in particular from the substantive checklist such a form would provide. (I have, unfortunately, seen some poorly argued and organized complaints from lawyers.)

13. Moreover, the forms could be updated to take the results of major decisions into account. So for example rights of publicity and common law marks, now routinely accepted by arbiters as protected under the UDRP, could be incorporated into the questions on the form long before a formal amendment to the UDRP is made, allowing parties not represented by counsel and not necessarily experts in trademark law to make their own case.

14. It is true that complaint and response forms would be longer. However, they would not necessarily be more complex, and indeed, if done properly, might obviate the need for lawyers altogether, thereby decreasing the cost of the proceeding.

15. Let me conclude this section by noting that in terms of substance the eResolution electronic forms were superior to WIPO’s in this regard, in that the eResolution forms contained some measure of leading questions. From my admittedly small comparative sample, the quality of pleadings were significantly better in the eResolution cases that I have decided, perhaps helping to explain respondent success rates with that particular dispute resolution provider. In my view, even though eResolution is no more, all the other UDRP service providers would be well advised to trace this first step and go even further in articulating a clear and instructive set of forms.

B. Allegations of Bias in Favor of Complainants and Forum Shopping

16. Is the UDRP process skewed in favor of complainants? Statistics have shown that certain dispute resolution providers have extremely high rates of holding in favor of complainants9. A clear tendency for forum shopping by complainants — even the Canadian government — has led to the demise of eResolution, the service provider with the highest percentage of decisions in favor of respondents. The question boils down to whether a respondent can get a fair hearing under the UDRP at either NAF or WIPO with a process that is complainant-driven if not complainant-dominated .

17. In my view, there is nothing on the face of the UDRP and the UDRP process which automatically stacks the deck against a respondent. The UDRP places the burden of proving all of “identical or confusingly similar”, “no legitimate interest” and “bad faith” on the complainant. This is quite onerous on paper, and ought to mean that even in cases where there is no response, despite the reasonable inferences which might be drawn under the Rules10 by the lack of response, judgment should not automatically be entered for the complainant. Of course, panelists may not always appreciate what the burden of proof means, and as such any argument or information provided by the complainant is uncritically accepted as persuasive. Additionally, there are substantive problems understanding the nature of “no legitimate interest” and “bad faith”, but these are questions that can be elucidated in scholarly literature and incorporated into the various forms without changing the text of the UDRP11.

18. There is, unfortunately, great potential for skewing the process, if not potential for outright abuse, on the part of dispute resolution providers. The recent study by Michael Geist was alarming in that it appeared that two panelists for at least one major provider were deciding a very high number of cases both in terms of absolute numbers and in terms of percentage of decisions favoring the complainant12. Moreover, dispute resolution providers conceivably could stack the process in favor of complainants by choosing complainant-friendly single panelists, by not empanelling panelists with track-records favorable to respondents13, or informally questioning panelists about judgments in favor of respondents in an effort to influence the outcome. In a situation where a panelist needs the income from arbitration, the stress of not being picked from the rosters, or fear of being dropped from the roster altogether, is significant, and could lead to bias in favor of complainants. This would be a non-issue if one could ensure the integrity of dispute resolution providers, but the statistics cited by Geist are a serious cause for concern as they do indicate both an institutional bias in favor of trademark holders and a proclivity on the part of trademark holders to choose providers who are more statistically inclined to hold in their favor.

19. There are a number of potential solutions to this procedural problem. Of course, ensuring that there is a roster of well-trained, intelligent and ethical panelists is of the highest priority for any dispute resolution provider. But if the statistics are correct, that is clearly insufficient on its own. Geist has suggested mandatory three-member panels, for both the UDRP and the recently published CIRA dispute resolution procedure14. While there is obvious merit in a proposal that subjects any dispute to three persons instead of one, my own view is that three-member panels are cumbersome and not necessarily more substantively fair. Rather, I would prefer to see a situation wherein complainants choose the dispute resolution provider, but where respondents are allowed to help choose the arbiter, by providing the dispute resolution provider and complainant with a list of three possible panelists. If the complainant is not happy with the ultimate choice (if made by the provider), then the complainant can opt for a three-member panel, nominating in turn three potential panelists and allowing the dispute resolution provider to pick the third, lead panelist. At the very least, dispute resolution providers should be required to keep an even rotation of panelists as far as linguistic and national exigencies allow, and to publish statistics of how many cases each panelist has been asked to sit on, how many acceptances, decisions, outcomes, etc. This will at least ensure that the manner in which cases are delegated is open to scrutiny.

20. It may be naïve to think that improving the quality of the arguments (by improving the quality of the forms) will also have a significant impact on the substantive results of domain name decisions. Just as it is true that as a respondent if you do not respond, you cannot win, it is also true that the better your response, the better your chances of winning. If the platform, electronic or otherwise, increases your ability (a) to respond and (b) to do it effectively, you will be more likely to succeed. To this extent, dispute resolution providers who are accused of bias in favor of complainants might be able to remedy, at least in part, both the substance of their decisions and the appearance of bias by ameliorating their formal aspects. If I am correct, the phenomenon of forum shopping, no doubt prevalent under the present system, could largely be corrected.

C. Should there be appeals?

21. Mr. Scott Donahey, a leading commentator and arbiter, has suggested an appeals panels for UDRP decisions15. The point is to provide some means of ensuring that there is an adequate level of consistency as between decisions. An appellate panel could correct erroneous decisions, harmonize divergent views and give prescriptive guidance to participants and their counsel.

22. In my view this suggestion would not be a good idea in formal or substantive terms. In substantive terms, having an appellate body will change the nature of the arbitration process. Arbiters will begin to focus as much on making their decisions “judgment-proof” as they do on getting to correct and just results. Moreover, they will in all likelihood feel bound by previous decisions. Previously decided cases in the present system are not precedents in the way common law cases might be, but rather are at best persuasive arguments to decide a given case in a manner in which apparently similar cases have been decided. While I certainly have found previous judgments useful, I have never considered myself bound by any decisions as precedents. In my opinion, this situation would change if an appellate body were formalized within the UDRP process, with arbiters feeling in some way bound, especially by appellate decisions, to follow precedents.

23. In my experience, this tendency is already manifesting itself within the system. Parties represented by lawyers are constantly citing previous UDRP decisions as if they were binding. There have been suggestions to report arbitral decisions according to their holdings so that they would be rendered more accessible and could serve as precedents. Arbiters too have looked for “precedents”. I recall one arbitration with a three member panel where a co-panelist in considering an argument I was making on the wording of the UDRP asked me if I knew of any precedents that allowed me to make such a ruling. In my view there is no need to require precedents for motivating any arbitral decision; rather the focus should be on the UDRP and the actions of the parties. In this sense, if two decisions rendered under the UDRP on similar facts appear to be in conflict, it is not necessarily true that one of them is wrongly decided. Indeed, it may very well be that both of them are right. This apparently formal incoherence might be a sign of substantive justice.

24. In short, an appellate process is a step towards “legalizing” a system which was never meant to be formally legalistic. The theory of arbitration is predicated on flexible decision rendering. The whole point of arbitration is to provide a quick, efficient, less formal and less costly mean of adjudicating disputes than the formal court process. Arbitration is meant to be primarily party-focused and fact-driven, and not worried so much about the greater coherence of the whole system. As such, the idea of an appellate body is driven by legalistic assumptions of formal coherence and legal precedent which is emblematic of our Western and particularly common law court systems. It would be an error to graft these notions based on very different assumptions onto a more flexible arbitration system. Ultimately if one or both parties want a legal solution, they are always entitled to seek one, since the UDRP does not seek to preclude or pre-empt legal proceedings16.

25. The best solution for getting high quality decisions remains getting high quality arbiters and enhancing a formal system which produces high quality arguments.

D. Conclusions on Form

26. The form of UDRP arbitration is critical to its overall fairness. Form helps to create a dialogue in which parties advance their best arguments and arbiters can properly assess circumstances and make decisions. As such, the form of UDRP arbitration must strive to enhance this dialogue as far as possible. A user-friendly, electronic platform will encourage respondents to respond. Clear and detailed, leading forms will alert parties to the case they have to make or meet. These seemingly simple measures will have a substantial impact on the process. These changes, along with rules that allow the respondent some say in the choice of arbiter, might be enough to rid the UDRP of the appearance of bias towards the complainant, and should probably go a long way towards curbing the forum shopping phenomenon.

III. The Substance of Domain Name Arbitration

27. Obviously form and substance are intertwined. As I have argued, all of the issues of procedural fairness have a substantive impact, and separating issues into formal and substantive can lead to artificial distinctions. This is particularly true of questions of precedent and appeals, which could be found in either section (and to some extent, I discuss each in both the form and substance sections of this paper). Nevertheless, in my experience the following major issues of substance have been problematic for the parties and arbiters and need to be clarified.

A. The question of “identical or confusingly similar”

28. With the exception of parody and critical web sites, for example, the “-sucks” web sites, the question of “identical or confusingly similar” appears to me to be dealt with reasonably. No doubt because most arbiters are well versed in trademark law and passing off where confusion is a staple subject area. Nevertheless, the degree to which “initial interest confusion” is to be tolerated is contested. Some arbiters would hold that as long as the web site made it clear that the respondent’s web site was not the complainant’s, it would be legitimate to use a domain name which was initially confusing. The average internet user is sophisticated enough to know that the site is that of a competitor, is unofficial, etc. Others would hold that the mere use of the initially similar or confusing domain name is enough to divert web traffic to the respondent’s site.

29. Both these solutions are too rigid. The latter position eliminates at the outset any sort of criticism/ parody or product information sites, while the former allows competitors to trade for profit on the well-known signs or symbols of a competitor by using the competitor’s name as a magnet to attract both the inadvertent navigator and the sophisticated search engine.

30. The solution, then, lies in the middle. Some initial diversions using a confusing domain name might be legitimate. On my view, fan sites, information sites, etc. while initially confusing, might very well represent a legitimate right or interest in the domain name since they do not use the domain name to compete with the trademark holder and are bona fide, legitimate activities. On the other hand, cases where the initial interest confusion leads to direct competition are examples of what the UDRP is meant to prevent, and should not be tolerated under the UDRP. Thus competition is the litmus test for determining what sorts of initial interest confusion will be tolerated17. If the initial interest confusion leads to the web site of a business competitor, then no degree of initial interest confusion should be tolerated no matter how clearly demarcated the competing web site is.

31. However, some latitude on domain names can be given to non-competing sites. This position also leaves room for parody and criticism web sites, at least to the extent that they do not infringe on other intellectual property or civil laws, such as those aimed at trademark dilution or defamation. These other intellectual property rules would form a useful boundary for determining when parody sites may have gone too far; as such, the UDRP perhaps ought to allow such sites where the legitimate interest of the respondent is present.

B. The distinction between “legitimate connection” and “bad faith”

32. The lack of understanding of the distinction between “legitimate connection” and “bad faith” is the most serious substantive problem that I have encountered in pleadings and decisions to date. There seems to be little or no idea on the part of parties, lawyers or advocates of the distinction between these two concepts, both of which have to be shown by the complainant. Oftentimes, the issues are presented in argument in a confused way, with the same evidence repeated for both criteria. Nevertheless, I believe that there is a valid and important distinction to be drawn between these two criteria, and indeed I have tried to draw it in various cases.

33. Paragraph 4(c) of the Policy gives three examples of circumstances where the respondent has a legitimate interest:

i. before notice to the respondent of the dispute, the respondent used or prepared to use the domain name in connection with a bona fide offering of goods or services;

ii. the respondent has been commonly known by the domain name even if it has not acquired trademark or service mark rights; or

iii. the respondent is making a legitimate non-commercial or fair use of the domain name, without intent for commercial gain to misleadingly divert customers or to tarnish the trademark or service mark at issue.

34. The critical point is that there exist in these situations some connection, determinable on the facts of a case, between the respondent and a domain name. Thus the underlying criterion of illegitimacy — or framed in terms of the respondent, of the determination of the respondent’s legitimate right or interest in a domain name — focuses in my view on the objective or relatively objective links that the respondent has to the domain name in question. These connections might arise either at the outset through a natural or visible connection to the respondent’s business name, trademark, bona fide offering of goods and services, or with a bona fide non-commercial use to the domain name. These connections might also develop over time through the nature of the respondent’s bona fide activities as they relate to the domain name. These connections might be obvious given current uses or practices in the instances just mentioned, or they might be demonstrable, as in the case of planned connections to a domain name for these various bona fide activities in the future.

35. The question then becomes whether the respondent, on the evidence, can be shown by the complainant to have no such connection to — and hence no such legitimate interest or right in — the domain name. In effect, then, the respondent has gotten to the domain name first, but given the objective and determinable connection, had every right or could have been expected to get there first, register and use the domain name for a commercial or non-commercial purpose. This connection and use might be independent of the complainant in some cases, perhaps even in most cases, but could even be dependent on the knowledge of the right-holder’s interest (a non-profit product information site for example, or a fan web site).

36. Unlike illegitimacy, the criterion of bad faith goes to the circumstances surrounding the registration or use of the domain name by the respondent, and focuses on the respondent’s intentions, either as they are stated explicitly or as can be inferred from the respondent’s actions and the facts of the matter. A list of partial criteria is articulated at paragraph 4(b) of the Policy, and the burden of proving bad faith falls on the complainant:

i. registering or acquiring the domain name primarily for the purpose of selling, renting, or otherwise transferring the domain name registration to the owner of the trademark or service mark or to a competitor of that owner for valuable consideration in excess of documented out-of-pocket costs;

ii. registering the domain name in order to prevent the owner of the trademark or service mark from reflecting the mark in a corresponding domain name (and engaging in such a pattern conduct);

iii. registering the domain name primarily for the purpose of disrupting the business of a competitor; or

iv. by using the domain name, intentionally attempting to attract for commercial gain Internet users to the respondent’s web site, by creating a likelihood of confusion with the owner’s mark as to the source, sponsorship, affiliation, or endorsement of the owner’s web site or location, or of a product or service on the respondent’s web site or location.

37. In all of these examples, non-exhaustive according to the UDRP, the inquiry is centered around the respondent’s behavior and motives — admitted or attributable on the facts — for registering or using the domain name. It is, in my view, a much more subjective analysis than that undertaken for legitimate interest, and is a serious burden for the complainant to prove.

38. Where it is the case that the domain names are identical or virtually identical to a very popular, well-known trademark, service mark or other protectable right or interest, it seems to me that the complainant’s burden is somewhat easier to bear, given that it is easier to show that the respondent must have had the complainant’s intellectual property right in mind at the time of its registration of the offending domain name. Where the mark has accumulated a great degree of good will, and absent some clear justifying factor (such as a clear legitimate interest), bad faith is more easily attributable on the facts. Where it is a popular site, it can be inferred that the respondent must have had it in mind when it registered its domain name for the purpose of drawing traffic to a competing web site. It knew or should have known that it at least might have been infringing on legitimate rights protected under the UDRP. Under the UDRP, such registration or use ought to constitute bad faith.

39. This last point also illustrates the link between legitimate right or interest and bad faith. The presence of a legitimate right or interest — relatively objective — on the part of the registrant might be evidence of — subjective — good faith. Where the two criteria effectively intersect is in those situations where the mark is well known and where the proof of the respondent’s legitimate interest relies on facts not objectively present in the record — unsubstantiated claims of planned future use of the domain name. In these cases, the evidence for both good faith and legitimate interest will likely be the same. Moreover, these will be the cases in which the complainant will have the most serious challenge of showing neither criterion existed. In effect, the best that the complainant can infer is that upon registration the registrant must have known of the owner’s strong right or interest in the mark, and could not reasonably have thought that his future interest in objective terms was strong enough to override the mark. Without changing the burden of proof, arbiters ought to give some latitude to complainants in this situation. In other words, arbiters ought to be slightly sceptical where planned use of the domain name is advanced by the respondent as both evidence of legitimate interest and evidence of good faith. (I elaborate more fully on this point in the next part.)

40. In keeping these two inquiries, one primarily into objective connections and the other primarily on subjective intent, arbiters might well provide less confusing decisions. Parties, especially those represented by counsel, would advance far better arguments if the distinction is understood and applied. And, as argued above, the complaint and response forms ought to make the nature of each inquiry clear.

C. Awkward burdens of proof

41. A related difficulty on both legitimate use and good faith is the awkwardness of the burdens of proof as framed in the UDRP. The UDRP places the burden of showing all the component parts of a complaint on the complainant. This burden is relatively simple with respect to “identical or confusingly similar”, in that the complainant is usually the trademark holder who is in the best position to show that he owns the mark, that it has a great deal of goodwill embodied in it, and that the domain name is similar.

42. Regarding both “legitimate interest or right” and especially for “bad faith”, the practical problem, which is mirrored in the unsatisfactory wording of the UDRP, is simply that the respondent in many if not most circumstances is the best party to adduce evidence of both legitimate interest or right and good faith. How can one encourage the obtaining or even require the furnishing of this critical information without overtly or subtly shifting the burden of proof onto the respondent? Moreover, if the respondent does enter evidence on these two points, the evidence may be novel to the complainant, who at least formally does not have a right to reply under the UDRP.

43. As stated above, in the more objective inquiry into legitimate interest, the situation is usually manageable: there hopefully is enough on the record for the complainant to make an argument (for no legitimate right or interest on the part of the respondent). In the case of determining legitimate interest based on proposed uses or plans to use the domain name, this problem is especially acute, and this inquiry is much more subjective. It often happens that the proposed uses were not divulged during the course of pre-arbitration communications. Hence there may be nothing on the record when the complainant files the complaint: the most he can say is that there have been no plans for future use disclosed by the respondent. If the respondent does disclose such plans in the response, the arbiter will have to assess the plausibility of the planned use. In such situations, it should be routine for the arbiter to accept or request the complainant’s submissions on the proposed use, and he should make certain that the respondent’s reasons for non-disclosure prior to the commencement of the arbitration proceeding are present in the dossier.

44. Similarly, with respect to good faith, the inquiry turns on the subjective behavior of the respondent. As with planned use of a domain name, the respondent’s intentions are not always clear on the record, and may not become clear until the response is filed. In both of these situations, we are in effect asking the complainant to make an assertion, often in the absence of any specific proof, of certain component parts of the case to be borne. Without any facts on the record, these assertions necessarily will be weak.

45. As bad as this situation seems for the complainant, it is even worse for the respondent. If the respondent does not respond, as it is his right to do, the arbiter is in the awkward position of having to decide without the benefit of the best information in front of him. Moreover, the power of an arbiter under the UDRP to draw potentially negative inferences from the failure to respond might lead to a presumption of bad faith or no legitimate interest against the respondent. If the respondent does enter such proof (and particularly where the complainant furnishes counter-arguments) the process may have unwittingly shifted the burden of proof onto the respondent to prove a legitimate interest or good faith. Some arbiters have responded to this situation by explicitly or implicitly placing a prima facie burden of proof on the complainant which then shifts (rather quickly) to the respondent once the allegations have been made. The burden is then de facto on the respondent. In a sense, the procedure has created a potential Catch-22 for respondents.

46. Is there a way out of this conundrum? It is undoubtedly true that clearer complaint and response forms would help, instructing complainants to do more than merely assert, and outlining the potential benefits of responding and the drawbacks of not responding to respondents. In the end, the complainant must not be absolved from bearing the burden of proof.

47. As an arbiter, my suggestions focus on the decision-maker maintaining a sense of balance. First, it would be helpful if a simple lack of response was not sufficient to draw negative inferences against the respondent. As an arbiter, I have hesitated to draw such conclusions. Rather, such negative inferences should only be drawn when the non-response is coupled with some other indicia on the record of bad faith. This alone, would assist in keeping the burden on the complainant to prove his case. However, when there is no response after allegations of bad faith and no legitimate interest, one can be slightly more lenient on the burden of proof to be met by the complainant. While the complainant’s assertions cannot be simply accepted uncritically, they can be accepted if they are plausible given the rest of the record. And certainly, if evidence of good faith or legitimate interest is adduced by the respondent for the first time in the formal response, then the complainant must be given the opportunity to reply.

48. In all events, arbiters must keep the record in front of them and the burdens of proof straight. While these suggestions do not resolve the formal difficulties with the structure of the UDRP identified above, they at least keep arbiters focused on the task at hand.

D. The influence of (usually American) domestic law and the role of precedent

49. As an arbiter, in argument forms I have often seen cited American and occasionally Canadian, English and even Italian law. What should be the weight given to domestic statutes in such cases? In my view, cases under the UDRP should be decided on the provisions of the UDRP, as clarified by principles of common sense, even though the UDRP does open the door to such applications18. The arbitration proceedings should not be discussed with reference to domestic law. If for example, one has to decide whether a breach of contract is an indicium of bad faith on the part of a respondent, then one should use one’s common sense examining the breach of contract in the context of all of the other facts in the proceeding, and then decide. The decision should not rest on the whether such behavior would constitute a legal definition of bad faith under the laws of a domestic legal system. If the parties wish to seek redress under a domestic legal system for the breach of contract, they are always free to do so; arbiters however should not use a domestic legal system to either decide the case (as a domestic court would) or decide not to decide where UDRP has empowered them to decide. The domestic legal remedy should neither influence nor preclude the outcome of the UDRP process. This is not only sound in substance, in my view, but it is also desirable in form. If the UDRP procedure is to have any ongoing coherence as a process, the UDRP document must be elevated to the status that a civil code has in civil law jurisdictions: as the primary articulation of norms guiding a set of legal relationships.

50. Conversely, and equally important, focusing on the UDRP, and rendering decisions according to the UDRP, should have the additional benefit of restricting the scope of arbitration to the UDRP and not domestic laws where arbiters may have no competence, a problem identified by Ian Stewart19.

51. This does not mean that individual decisions themselves must always cohere with previous UDRP decisions. As I have stated above in the section on formal procedure, arbiters should place just results above adherence to previous UDRP and domestic court decisions. In my view, arbitration ought to be focused on flexible standards of justice meted out to the individual parties in individual cases. Put simply, sometimes analogous cases should not be decided analogously, since very different motivations might be present as between parties in different cases. Blind adherence to precedent will not lead to just results. For example, an offer of $5000 by a prospective respondent to turn over a domain name to a prospective complainant might be evidence of bad faith in one set of circumstances but not another.

52. Moreover, formalizing a doctrine of precedent, in my view, will lead to arbiters adopting a common lawyers’ or judges’ mentality of needing or requiring a precedent in order to decide an issue. So, to reapply the example above, if one has to decide whether a breach of contract is an indicium of bad faith on the part of a respondent, then one should use one’s common sense, and so decide. The decision should not rest on the whether there is a precedent for such an interpretation already handed down by a previous panel (or an appellate panel). Such rather timid behavior is, I am afraid, already present to too large a degree in the UDRP system.

E. Conclusions on Substance

53. There is still a long way to go to sorting out the conceptual status of at least two of the three main substantive prongs of disputes under the UDRP. Here commentators (and arbiters) have a role to play in articulating the distinctions between these concepts while remaining faithful to the burdens of proof imposed by the UDRP, not to mention imperatives of fairness, particularly in a system where the cards already appear stacked against respondents.

Conclusion (& Prologue)

54. The title of these reflections is an obvious allusion to Duncan Kennedy’s well-known article, “The Form and Substance of Private Law Adjudication20. That piece focused on the concepts of “individualism” and “altruism” and their analogues in the adjudicative process, rules and standards. Kennedy argued that the substance of private law adjudication, focusing on law as a particular ideal, was more caught up in rules and individualism, while the form of private law adjudication was more inclined to both the flexible normative concept of standards and the less-identifiable concept of altruism. Standards were more holistic than rules, requiring a knowledge of the whole system, and providing for a greater degree of (desirable) discretion than did rules. Ultimately, Kennedy believes that form helps to condition substance.

55. Like Kennedy, my experience thus far as an arbiter leaves me with a sense that we ought to be giving a greater role to what he calls the concept of standards and the related principle of discretion ­ as well as their many analogues and proxies ­ with a focus on improving the form of the domain name arbitration system. While domain name arbitration is not private law adjudication, this insight is still quite (and perhaps even more) appropriate. By increasing the degree to which parties communicate with each other and the ease and effectiveness with which they can put their best arguments forward, we improve the quality of proceedings without changing any formal rules. By maintaining the discretion and the flexibility that arbiters have, we keep the system anchored on individualized justice, a foundational point in systems of alternative dispute resolution. My reflections have centered around improving the form of the UDRP process, viewing at all times the UDRP process as one of communication between parties. Enhancements in form result in improvements in substance.

56. In that sense, the enhanced dialogue advocated in this paper might be seen as a type of prologue to a more substantively fair and flexible system of domain name arbitration, all of which is easily within our grasp. At the end of the day, nothing is more important than a competent roster of arbiters genuinely trying to fairly decide cases, trusting their own instincts and intuitions, not worried so much about precedent and what results would be produced in adjacent, and more formal, legal systems. This case-by-case integrity, in my view, aided by the substantive prodding of well-written forms, will solve a great deal of what is perceived to be wrong with the UDRP system.


*Assistant Professor of Law and Member, Institute of Comparative Law, McGill University; B.A. (Toronto); LL.B., B.C.L. (McGill); LL.M. (Yale); D.Phil. (Oxford). Panelist for WIPO Domain Name Panel and former panelist for eResolution. I wish to thank Ryan Rabinovitch (B.C.L./LL.B. 2002), for his able research. Research support was generously provided by the McGill Faculty of Law’s Wainwright Trust. This article is dedicated to the project that was eResolution, and the fine people behind it: Karim, Robert, Fabien, Joëlle, Aubert. Email :

2UDRP, para. 2.

3UDRP, para. 4.

4UDRP Rules, para. 3(b)(xiii).

5For WIPO decisions, see For National Arbitration Forum decisions, see For decisions rendered by the CPR Institute for Dispute Resolution, see The Asian Domain Name Dispute Resolution Center,, did not have any published decisions as of 4 March 2002.

6See e.g. M. Geist, “ An Examination of the Allegations of Systemic Unfairness in the ICANN UDRP” (2001) [available at]; B.J. Freedman & R.J.C. Deane, “Trade-marks and the Internet: A Canadian Perspective”, (2001) 34 U.B.C. L. Rev. 345; and I.L. Stewart, “The Best Laid Plans: How Unrestricted Arbitration Decisions Have Corrupted the Uniform Domain Name Dispute Resolution Policy” (2001) 53 Federal Communications L.J. 509.

7Discussed in the next part.

8WIPO now allows electronic submissions.

9M. Geist, “”, supra note 6.

10UDRP Rules, para. 14(b).

11I deal with these two issues in the next part of this essay.

12M. Geist, “”, supra note 6.

13I may very well be in this situation as I write. After having rendered my first two decisions for one dispute resolution provider in favor of respondents in September and October of 2001, I have yet to be empanelled in a new proceeding. For the time being, I am content to accept this as mere coincidence.

14For the CIRA Policy, see

15M.S. Donahey, “A Proposal for an Appellate panel for the Uniform Domain Name Dispute Resolution Policy”, (2001) 18 J. Int’l Arbitration 131.

16On the complementary nature of UDRP proceedings and legal proceedings in the US, see J.M. Osborn, “Effective and Complementary Solutions to Domain Name Disputes: ICANN’s Uniform Domain Name Dispute Resolution Policy and the Federal Anticybersquatting Consumer Protection Act of 1999", (2000) 76 N.D. Law Rev. 209.

17For support and elaboration on this point, see Freedman & Deane, “Trade-marks and the Internet: A Canadian Perspective”, supra note 6.

18See e.g. UDRP, para. 2, where the registrant must state that he is not in violation of “any applicable laws or regulations”.

19See Stewart, “The Best Laid Plans”, supra note 6.

20(1976) 89 Harv. Law Rev. 1685.




© copyright 1995-2008 Lex Electronica Tous droits réservés / All Rights Reserved ISSN 1480-1787



© copyright 2009-2014 Lex Electronica
Tous droits réservés
ISSN 1480-1787
Regroupement droit et changement Faculté de droit de l'Université de Montréal Centre de recherche en droit public