[CWG-Stewardship] Update on the Integrated model.

[Suzanne Woolf]

On Feb 22, 2015, at 1:22 PM, Andrew Sullivan <ajs at anvilwalrusden.com> wrote:

> The IANA function is really extremely small.  It's a critical but
> basically boring book-keeping function.  As near as I can tell, there
> have been practically no cases where there was any accusation that
> IANA did not do exactly what it was supposed to do.  There were
> historically some complaints that IANA didn't act expeditiously, and
> there were _lots_ of historic complaints that an IANA function was
> being used by ICANN to try to impose ICANN policies.  The former is an
> SLA issue; the latter is actually a policy matter with enforcement
> attempts in the policy side of the organization, and is not actually
> an issue with IANA at all.  So in my opinion, accountability _for
> IANA_ would help with the SLA stuff (and also with the case that IANA
> "goes rogue") but would not help with the policy issues.
> 
> Does that match the accountability concerns others have?

Possibly as a help in thinking about this question….

There's a question here of what people are being held accountable *for*.

In addition to separability as a final backstop, the other feature of the model the IETF has, and the RIRs are seeking, is that their contractor for the IANA functions that affect them is held accountable for two things: 
	1. Doing what they're told, in a timely and correct fashion
	2. Not doing anything else

If the contractor doesn't do things they *are* supposed to do, or does things they are *not* supposed to do, they're accountable for their poor performance-- but the directions come from the policy body, as implementation of its policies, and the contractor's accountability is to the policy body. In ICANN's case, this is where the Board's accountability touches IANA as an operational function, and is separate from any role of the Board or accountability of the Board regarding ICANN as a policy body.

The contractor is not accountable for following improper directions from the policy body, as long as the agreed-upon processes between the policy body and the contractor are followed. It doesn't matter whether those directions are improper because process wasn't followed internally to the policy body, or because properly-executed policy nonetheless leads to unacceptable outcomes. In both of those cases, there's no wrongdoing by the contractor, and the only accountability that matters is that of the policy body to its users. 

In the IETF model, accountability for making bad decisions about what to tell IANA to do is maintained in a number of ways, but they're outside of the IANA SLAs and I don't see any of them changing in the event that a new contractor is selected for the protocol parameters registries. 

Bluntly, isn't the major accountability concern for root zone management that ICANN will give inappropriate directions to IANA staff in the course of implementing policy? If so, in the case we're concerned about, what's broken is the policy body. Changing the status of IANA staff from employees to outside contractors, or changing contractors, wouldn't help at all; the problem is internal to the policy body, and any IANA contractor (internal or external) would be obligated just the same to follow its directions.  The place to prevent or stop or fix the results of bad directions to IANA is in ICANN, not IANA.

I think this argues that the group needs to refine the definition of what kind of accountability belongs in the IANA transition discussion. It also argues for some care in applying the IETF/RIRs' model for names, or expecting them to participate in a model that may fit names better than protocol parameters or addresses.


Suzanne


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mm.icann.org/pipermail/cwg-stewardship/attachments/20150222/995e7ec9/attachment.html>