Enforcement of TLD Registration Restrictions: Why Restrictions Matter - Part II
[draft -- not for redistribution or citation]

Last month, I wrote to describe the state of registration restrictions in .BIZ, .US, and .NAME. I noted trends among nonconforming registrations in these TLDs, and I suggested that certain automated enforcement systems might serve to improve compliance. But an important larger question remained unanswered: Why care about registration restrictions in the first place? Much as registries might like to ignore the restrictions, I submit that the Internet community nonetheless ought to hold them to their contracts.

One rationale for enforcing registration restrictions focuses on the community consensus arguably embodied in the restrictions. New TLDs resulted from a lengthy process in which constituencies reviewed the proposals, the interested public submitted written comments, and the community gathered for a public forum. Some continue to dispute the legitimacy of the procedure and the merits of its results. For a registry to impose unilateral changes to the resulting specifications might therefore be taken to undermine ICANN's entire process of policy development.

Furthermore, ignoring consensus results reaches beyond procedure to substance. Community discussions produced registry policies entailing a bundle of compromises that, taken as a whole, generally satisfied the diverse interests involved. But changing the rules risks upsetting that delicate balance. For example, although Neulevel may consider unimportant the prohibition on offers of sale of .BIZ names, this provision was of significant interest to the Intellectual Property Constituency, for this rule (if enforced) would much reduce certain high-volume registrations considered costly and undesirable by IPC members. But the IPC and major commercial registrants are not the only parties who suffer when enforcement is lax; end users also have an interest in requiring compliance. After all, if a non-American impermissibly registers a desired .US name thanks to Neustar's lack of nexus enforcement, a legitimate American registrant cannot cheaply register that name. To most would-be .US registrants, it's little consolation to learn that dispute procedure offers a possible recovery of the disputed name -- with a delay of several months and a cost of thousands of dollars paid to lawyers and arbitrators. Instead, registrants would prefer that Neustar take further efforts to prevent non-Americans from impermissibly registering .US names in the first place.

Letting a registry operator ignore its promised registration restrictions also risks unfairness to other registry applicants. Other prospective registry operators may have, to their detriment, forthrightly proposed precisely the unpopular policies the selected registries now implement -- less enforcement of IP rights or other restrictions, or increased fees to fund such enforcement activities. But it would be an odd allocation procedure that grants TLD registry contracts to those applicants with the most optimistic plans for service -- then lets them change such plans to suit their subsequent convenience.

Lurking in these examples is a concern that a registry's business interest -- in selling more domains -- may dominate its contractual obligation to enforce the rules on the books. No business wants to turn away interested customers, but enforcement of registration restrictions necessarily entails refusing those transactions that break the rules. By conveniently disregarding registration restrictions, registries well serve the interests of their owners and shareholders -- cutting enforcement costs and selling additional units beyond expectations. Such self-serving behavior should be viewed with suspicion and with a clear understanding of the indirect costs imposed on consumers who follow the rules.

High school civics class offers two further reasons to seek enforcement of rules on the books. For one, if violations are allowed to mount in the background, there's cause to fear arbitrary enforcement in the future. It would be an unhappy world in which most registrants stood in violation of one rule or another -- and risked losing their domains on the whim of a registry, arbitrator, or well-heeled challenger. In addition, when numerous background registration restrictions are ignored by many registrants, there comes a danger -- to new or inexperienced registrants in particular -- of confusion as to which rules are to be taken seriously (the UDRP, for example) and which may typically be ignored without concern. It's preferable, civics class suggests, to strike the unimportant rules rather than preserve rules that are consistently flouted.

Counterintuitively, increased enforcement need not entail stricter limits on behavior. Instead, enforcement of the rules might be combined with a careful reconsideration of which rules ought to continue in effect. Such changes are not unprecedented: Later this year, the Netherlands' .NL will remove its longstanding restrictions on geographic location of its registrants. The Netherlands' change tacitly admits the ineffectiveness of country-based targeting; any registrant who wants a .NL domain can already get one with minimal risk of detection or punishment, notwithstanding the official rules requiring a Dutch affiliation. The change also admits the rule's relative lack of importance -- that the distinguishing characteristics of .NL result not from its formal prohibitions but rather from trends in use of its space. Neulevel's .US could follow this change with equally minimal ill effects but with a laudable reduction in dead-letter law and, most importantly, with an end to perceptions and allegations that some foreign registrants have taken advantage of the system.

Accordingly, a portion of changes to TLD registration restrictions might sensibly follow .NL, phasing out restrictions that cannot be effectively enforced and that matter little to a TLD's true purpose. But for registration restrictions important to a TLD's focus or necessary in gaining widespread support for the TLD's creation, enforcement is not only appropriate but, I suggest, obligatory. With this in mind, Neulevel's .BIZ ought to develop new programs to ferret out domains offering themselves for sale in violation of registry policies. Other chartered TLDs must also develop procedures -- presumably automated systems, to keep costs manageable -- to enforce their respective restrictions. The necessary code isn't hard, but its implication is significant: That when the Internet community develops a procedure for addressing a difficult problem, the businesses charged with implementing the procedure must in fact do so, even when contrary to their direct financial interests.

962 words

[scratch]

the approach is nonetheless worthwhile for the more productive and efficient use of the domain name system. Through deference to previous decisions, registries better serve the consensus processes that regulate them, and as registration rules come to better comport with intuition and [morals], registrants will come to take registration rules more seriously. In the long run, such rationalization will likely result in a more productive and efficient use of the domain name system.

• The rules are the rules. Whatever one thinks of certain regulations -- the [controversial] [UDRP], for example -- there's no denying that registrants ought to conform their behavior to comply with at least certain such rules. A savvy registrant may in time come to understand that some rules are, effectively, more important than others. But to ask this of everyone is unduly
• Encouraging sensible uses of the domains that have been registered. Consider the problem of a registrant who picked up some impermissible domain -- say, "the.best.name" -- with the initial intent to develop it into an active site. Now that the registrant realizes the domain is impermissible, he might understandably hesitate to develop it; after all, under .NAME's ERDRP, the name could be taken from him at any time, and he'd then lose the promotional efforts invested in the site. In practice neither .NAME nor any third party is likely to pay the $1000+ required to challenge the domain, but with the prospect of enforcement lurking in the background, the registrant may fail to develop the site and may ultimately let the registration lapse. Little harm results -- the registrant can always find another domain for the content of interest -- but it's hard to see how the .NAME rule makes the world a better place.
• Avoid encouraging rule-breaking by non-bad actors. Certain poorly-thought out rules can lead even good-faith actors to break the rules. Federal Reserve and J&J in .NAME. Their respective registrations of ?? and ?? .NAMEs -- none of them the names of any actual people -- certainly don't comply with .NAME policies. Their registrations may have reflected that they didn't understand the rules -- thought it was acceptable to register their defensive marks as ordinary names, as it is in .COM, .NET, and .ORG. But could also result from unwillingness to pay .NAME "defensive registration" fees as high as $?,000. If .NAME pricing (charging a lot for def regs) is truly a consensus policy, reflecting the deliberation of the many constituents that come together in ICANN, it's wrong to let these entities take advantage of the rules and fail to pay their fair share. If .NAME pricing is a sham -- if the community thinks it's wrong or otherwise unacceptable to charge def registrants ??10?? times as much to prevent use of a name as a registrant would have to pay to register it -- then the rules on the books should enforce as much. But it's wrong to put J&J in the position of having to choose between breaking the rules and spending lots of money.
• Respect for registration rules generally. High school civics class teaches that legal prohibitions ought properly be limited to behavior truly of concern. When reasonable behavior is prohibited by law, theory and experience suggest, even law-abiding individuals lose faith in the law -- and may subsequently ignore more sensible laws they wouldn't have "Dead-letter" law without enforcement won't be taken seriously, theory and experience suggest, and a legal system without enforcement risks losing its moral sway in encouraging the behavior it seeks to require. In the context of domain names, the moral importance of compliance is arguably less compelling. But to require registrants to take seriously certain rules (the UDRP, for example) while simultaneously making a mockery of others (like registration restrictions) risks tainting the [system/process] in a way that reduces its overall credibility.
• Fairness to other firms who applied to run registries. Others may have planned .BIZ etc. with registration restrictions that they intended to enforce, that they thought they could actually enforce, and that they would spend money to enforce. These plans no doubt made their applications less attractive relative to the accepted applications -- made them some commbination of more costly (due to enforcement expenses) or seemingly less favorable to IP interests (for lack of enforcement rules on the books). Setting policies for a TLD isn't easy -- lots of interests to please, and a difficult balancing act. But we ought not penalize those who were in the first instance forthright about the challenges of doing the task; it seems odd to grant the TLD to those who made the wildest and most unrealistic promises, only to fail to follow through later.

What to do? I suggest that registries ought to rethink their registration restrictions to cut the "dead letter law" that is currently ignored both by registrants and registries. Every surviving registration restriction ought to be rigorously and promptly enforced; restrictions without enforcement provisions ought to be simplified to facilitate enforcement, or removed if effective enforcement truly is not possible. Such changes are not unprecedented: The Netherlands' .NL recently removed its dead-letter registration restrictions, tacitly admitting both the ineffectiveness of the rules and their lack of importance. Neulevel's .US could do the same with equally [minimal] consequences. [Little harm comes to the Netherlands, and their resulting policies yield less fear or perception of gaming the system relative to current practice in .US. ]

TLD registries -- and the consensus policies that guide them -- ought to focus their registration restrictions on rules of true importance to Internet users. Doing so will encourage registrants to take the rules more seriously ... and will make for more productive, efficient, and sensible use of the domain name system

) Better for Loraine and Andrew Mitchell, registrants of LoraineAndAndrew.Mitchell.NAME, might someday [Just as cities may sensibly fear that everyone breaks loitering rules sometimes and therefore refuse to enact such restrictions. Lest domain names reach a similar situation, enforce rules now, before we reach an "everyone breaks" point.]