RE: [gnso-vi-feb10] So, what's best for consumers? Anyone? Bueller?

[Milton L Mueller <mueller@xxxxxxx>]

good contribution Anthony. My comments:

> -----Original Message-----
> 
> -- Single-registrant TLDs have been trotted out as a special case which
> get should get special favorable rules.  Leaving aside the fact that
> "single-registrant" is an egregious misnomer, since even within a
> "company-only" TLD there will be numerous registrants, the special
> treatment being considered is for the benefit of the applicant, not for
> the user.

Must disagree: to me, the interesting thing about self-provided TLDs is that 
the applicant _is_ the user. So I think this proposal does have a user/consumer 
rationale. It is like private networks in telecoms or provider-independent 
address space in ip addressing - it grants autonomy and a new option to end 
users. Granted, the users in this case are probably big organizations. 

I don't think the rules should "favor" self-provided TLDs and would not frame 
this option that way. Rather, I think users should not be penalized by the 
assumption that all TLDs are businesses predicated on selling registrations to 
third parties and taxed accordingly. Lots of enforcement and trademark issues 
go away with self-provided TLDs, so lower/eliminated fees are justified.

> -- Currently, the argument is being made that existing gTLD registries,
> with a ten-year head start over other applicants, should get a "level
> playing field" vis-a-vis those applicants who don't have ten years of
> revenues to spend on whatever they want -- for instance, buying up a
> bunch of registrars.  Maybe they should, but what has this got to do
> with the end user?

If you look at it _strictly_ from the standpoint of the user, whatever 
increases their options and intensifies competition is in their favor. So to 
eliminate entry by incumbents is to eliminate one source of perfectly viable 
entry. One could make a case for handicapping incumbents in the market entry 
race, but that argument could only be motivated by a desire to protect or 
"help" new suppliers. It is not a user-oriented argument. 

> The CRA report makes a case that no separation is best for the consumer.

Not my reading. It said that no separation would have lower costs. But there is 
the problem of lock-in, which could negate any advantages of lower supply side 
costs. Now if you allow end users to register a name for a reasonable flat fee 
that constitutes a permanent, non-expiring purchase of a domain name for as 
long as the registry exits, then I will stop worrying about lock-in. Until 
then, we need to think about registrars and, possibly, even price caps.

> If this group doesn't make at least a defensible attempt to develop
> policy from the perspective of the consumer, its work is highly suspect
> and, given the preponderance of registrars and registries participating,
> possibly anti-competitive.

Agreed. If we veer off in that direction I will make noise. We haven't yet, in 
my opinion. 

> - A cheap prices
> - B good choice of names
> - C good service
> - D portability
> - E perpetual right of renewal at a predictable price.

Yay. Especially on the last one. Good list.  I added letters for ease of 
reference.

> Maybe I'm missing something, but that seems like it to me.    So which
> vertical integration/separation policy gets us closest to that?

My read on this: 

1. Goal E (perpetual right of renewal) is a policy change that has little to do 
with structural separation. I think a truly competitive market would offer 
that, but ICANN prevents it now. 

2. Allowing 100% cross-ownership and allowing NEW cross-owned TLDs to sell 
their own name while RETAINING contractual and functional separation of Ry and 
Rr and equivalent and nondiscriminatory access is a no-brainer. You get the 
best of both worlds (VI and separation) in the first phases of new entry. You 
promote goals A, B, and probably C, you don't undermine D. You increase the 
amount of new entry and (gods willing) push toward E as well. If and when some 
of these new TLDs become successful, there may be enforcement issues regarding 
the nondiscriminatory access, but I think those can be dealt with. The real 
issue becomes merger and acquisition policy at that point. 

3. Creating a special category for self-provided TLDs and a fee structure for 
them that does not tax them for being in a business they are not in is another 
way of achieving goals A, B, C, D, and E. Organizations now have their own 
choice of a name, a self-provided (lower, non-taxed) price, service as good as 
they want to make it, no portability issues unless they outsource in which case 
they can change contractors, and in effect, E. 

4. Allowing an incumbent registry to act as a registrar in an existing TLD 
(while retaining separation) obviously does not help B. It may or may not help 
A, depending on whether registry acting as registrar eventually drives out all 
independent registrars and then jacks up its prices. If consumers are 
protected, however, by E, then these concerns diminish - however, I can see 
squabbles developing over the terms and conditions on which independent 
registrars and the registry offer perpetual registrations. 

5. Full vertical integration may lead to A, probably doesn't do more for B than 
the option #2 I described above, may improve C as long as the market is 
competitive, obviously defeats D, and may help E. 

A final note:  On the issue of portability there is a tradeoff. You might get a 
wider choice of names and lower prices by sacrificing portability. My view, on 
which I am flexible, is that you can let consumers decide for themselves on 
that score once the market moves beyond its current cartelized level and has a 
lot more options. That is, the fact that registry A uses independent registrars 
and registry B doesn't may be taken into account in a user's decision, with B 
being perceived as a riskier choice.