[gnso-vi-feb10] Re: cross-ownership definition

[Avri Doria <avri@xxxxxxx>]

Hi,

>> 
>> While I agree with what the end result would be, i am somewhat troubled by
>> bundling Recommendation 19 into the definition of CO.
> 
> ?? I don't see any mention of Rec 19 there. Can you explain?

Recommendation 19 "
Registries must use only ICANN accredited registrars in registering domain 
names and may not discriminate among such accredited registrars."

I.e i assumed that this was well know shorthand, in a group such as this, for:

>> while maintaining the
>> contractual and functional separation and equal access arrangements
>> required by ICANN policies and contracts

Apologies for my error.


>> 
>> And I think
>> there has to be some notion of verifying that in those cases were cross-
>> ownership is allowed, the fair and equal access can be verified.
> 
> Here I think you are getting away from definitions and into specifying policy.


I don't think so.  I think trying to set what those protections should be would 
be policy and defining the methods for verification implementation.  I think 
defining something as protected and verifiable is common sense and is inherent 
in the bylaws requirement of ICANN and the in compliance responsibilities of 
ICANN.

But I admit if maybe borderline (one person's common sense is often another 
person's error) and if others think that it is policy, then I would recommend 
another objective such as:

Determine whether the full and equal access by all registrars requirement on 
Registries needs to be protected and if so, what those protections should be. 
Also determine whether this needs to be verified by ICANN compliance measures. 
(note, i admit ignorance about the extent of protections and verification that 
may already be in place)

cheers,

a.



On 7 Feb 2010, at 14:19, Milton L Mueller wrote:

>>> Don't we also need a definition of cross-ownership, just to avoid
>> further confusion about that?
>>> 
>>> Here's a stab at that:
>>> "Cross ownership" is defined as the ownership of a controlling share of
>> a registry by a registrar, or vice-versa, while maintaining the
>> contractual and functional separation and equal access arrangements
>> required by ICANN policies and contracts. As long as equal access
>> arrangements are in place, cross-ownership that permits the registrar to
>> sell the names of the cross-owned  TLD registry shall not be considered a
>> form of vertical integration.
>> 
>> 
>> While I agree with what the end result would be, i am somewhat troubled by
>> bundling Recommendation 19 into the definition of CO.
> 
> ?? I don't see any mention of Rec 19 there. Can you explain?
> 
>> I also do not know
>> if controlling share is a necessary part of the definition.
> 
> Ok, it doesn't have to be. My reason for putting it in was that I was under 
> the impression that minority ownership is already allowed by current policy. 
> So the purpose of the definition is to make it clear that the WG will be 
> concerned only with CO issues that involve acquisition of a controlling 
> share. With a controlling share, CO moves somewhat closer on the spectrum 
> towards VI, which makes it a concern of this PDP.
> 
>> And I think
>> there has to be some notion of verifying that in those cases were cross-
>> ownership is allowed, the fair and equal access can be verified.
> 
> Here I think you are getting away from definitions and into specifying 
> policy. One of the chief policy debates about cross ownership has to do with 
> the verifiability of equal access/nondiscrimination. The registries have 
> argued that unless there is a ban on selling a cross-owned TLD, we cannot 
> enforce nondiscrimination. The registrars have argued that such a ban is not 
> needed. By including the word "verifiable" in there you may be viewed as 
> pre-judging that issue.
> 
> The definition of CO I proposed above was meant to define the type of CO this 
> WG will consider as in scope, and to distinguish it from pure VI. It was not 
> intended to specify the appropriate policy for allowing or not allowing CO, 
> nor is it appropriate for a DT charter to do that. 
> 
> I think the issue of verifiability should be left to the WG as a policy 
> issue, and left out of the definition. If you like, we can explicitly mention 
> it in the description of the WG scope or objectives. 
> 
>> What would we call a relation where a registrar owns a strong minority
>> share of a registry and they have concluded a marketing deal that gives
>> priority to that registrar in selling the names of that Registry's new
>> gTLD.  It seems that we would have a cross-ownership situation with a
>> question of VI and degree.
> 
> Sure, there is a gradation here but again I am under the impression that 
> current contractual arrangements already allow minority shares. Perhaps the 
> registrars and registries on this list can clear up that factual issue. 
> 
> If we decide to remove the "controlling share" language from the CO 
> definition, then the easy solution is to specify a threshold (say, 30%); 
> above that we consider the share large enough to be considered "cross 
> ownership" below that we don't. Advantage: the definition is clearer. 
> Disadvantage: the threshold is somewhat arbitrary. 
> 
> --MM
>