This would be a race to the bottom, due to jurisdictional arbitrage, especially given the existence of countries like Tuvalu and the like who’ve essentially allowed for-profit companies to take full control of their TLD.

Every TLD needs to work within a legal and regulatory framework.  By a single ICANN policy framework that is restricted to the US legal framework and the ICANN regulatory framework.  There is already one country with ‘disproportionate rule-making power over all TLDs’ - the US - and that’s only going to get stronger unless ICANN legally moves to another country and then whatever country it moves to will be in the same position.

Setting the standards that a local policy framework must meet is nowhere near the geo-political issue that you describe.  And it would be a community driven process, not a GAC process.  If some are preferred among those that make the cut then that’s competition not gaming, a key principle that ICANN should enshrine.  I see no reason to let a TLD move countries as you suggest, nor would it simply being a case of ‘picking a jurisidiction’.  Legal and physical residency is not a trivial requirement.

Finally, yes they most certainly should be paying whatever fees to the local regulator and not ICANN.  Again that’s competition not gaming.

Competition as a principle is clear to me - if you don’t allow it then there needs to be a very good reason why not.  There isn’t a good reason to disallow competition of policy frameworks for new TLDs.  Many of us fought for new TLDs to enable competition even though it had the potential to reduce our own sales because we knew it was the right thing.  The argument doesn’t suddenly change when it comes to policy frameworks.

Shows us the way, then, using dot-nz as the testbed. By your “logic”, gTLD registries (who are delegated a domain to the left of the “root”) should be able to choose their own policies, essentially.

Continuing with that “logic”, registrants within .nz (who are delegated a domain to the left of “.nz”) should be able to choose which country’s policies their registration is governed by, and pay fees accordingly to that country. For example, a Canadian company registering example.co.nz would pay fees to CIRA (and follow Canadian rules). Similarly, if a UK company registered example.co.nz, it would pay fees to Nominet, and so on.

Let us know what happens when such “policy competition”, as you call it, is brought to the .nz namespace. Get all the wrinkles out, before you try to assert it will work a level higher, in the root.

One trouble with ICANN is that it could become a policy monopoly.  It’s by no means perfect, and I among many other criticize it on a regular basis.

But the ccTLD world, although it has well-run TLDs, also has really abysmal failures. All of the qualities you claim for ccTLDs are shown on their dark side by rogue ccTLDs. As you well know, some are run by barely-accountable ministries, some by near madmen, and some by out-and-out scoundrels. There’s nothing about the ccTLD model that makes it better—rather, there just happen to be some enlightened administrators. 

Although I agree with you about ICANN “growing a backbone” to help it deal with overweening IP and government intrusion, ccTLDs with excessive government intrusion are a dime a dozen. And the ccTLD grouping within ICANN maintains a polite silence on these issues - wasn’t Bahrain recently welcomed as a ccNSO member despite their recent well-publicized jailing of bloggers?

In theory, I agree with a lot of what you say.  But in theory, theory and practice are the same thing…

Antony

Jay,

I have long said publicly that ICANN should stop try re-inventing the wheel and to take the best practices from well-run ccTLDs as a great starting point for policy development and as a sanity check for already-developed policies.  But that’s based on the actual policies, not the idea that ccTLDs are an inherently better policy development arena. 

The political interference in ccTLDs is legion and longstanding.  I spent the late 90s and early 2000s at ICANN fighting a largely losing battle against the notion of “sovereignty” over ccTLDs and to prevent nonsensical re-delegations.  Those have occurred and continue to occur, and ccTLDs can be “disappeared” as well: .PF, .MQ, .WF, .BV, .PM, .UM—there are many examples.  Many of these “barely-accountable scoundrels” responsible for making ccTLDs disappear are so-called respectable national governments—the U.S., France, and Norway in the examples above. 

In other ccTLDs, the pricing, the rules, the corruption or other barriers prevent any noticeable use at all—and there is no appeal.  This to me is the essential point—for all of its craven subservience to the GAC (and they do stand up to the GAC from time to time by the way) ICANN conducts its business largely in public and when something untoward happens, there are lot of people like you and me who will scream foul.  That is not the case in the ccTLD world, where Bad Things can happen in the middle of the night. 

In a country like New Zealand, where civil society is strong and there are norms of transparency, the ccTLD can easily be better than anything ICANN can come up with.  In a country like Zimbabwe or Turkmenistan, which are run by vainglorious madmen (and mind you, there are more than a few), the opposite is likely to be true.  In countries like France and the US, where there is great lip-service paid to good governance, but a strong tendency to do the opposite while no-one is looking, ICANN is useful because it’s pretty hard to slip one through unnoticed.

Having watched with alarm as more than a few ccTLD operators I used to correspond with completely disappear, and the ccTLDs they ran revert to silence, and having watched the very unseemly strong-arming of local ccTLD admins by governments, I cannot honestly say that the ccTLD set-up is necessarily a better one.  It can be better, certainly, but often it is worse.

Antony