[address-policy-wg] 2013-03 New Draft Document and Impact Analysis Published (No Need - Post-Depletion Reality Adjustment and Clean up)

I don't support the approval of 2013-03 "No need" at this time.

Needs-based allocation is more than an operational procedure for the
RIPE NCC to follow. It encapsulates a core policy objective for the
management of the IPv4 address space, not only by RIPE NCC but also by
LIRs.

I am not convinced by the the principle argument being made in favour of
abolishing the concept of "need", and concerned that we do not fully
appreciate the potential consequences. The proposal adopts one
particular type of "reality adjustment" in response to IPv4 address pool
depletion, abandoning the core objective. There might be other types of
responses, but if we abandon the core objective, as 2013-03 would do,
we're basically closing the door on any discussion of what those
alternatives might be.

Finally, as noted by the RIPE NCC, this takes away a key message the
RIRs have used to justify their independence from government. That may
not be sufficient reason to oppose the policy if it is otherwise
necessary, but it does seem to me to indicate caution and counsel
against acting precipitously.

In short, the potential downsides seem to me to be more compelling than
the claimed upside, and certainly sufficient to justify an extended
delay (by which I mean, broadly indefinite, or until the community feels
the implications of depletion are clearly understood and well established).


Let me elaborate.

1. The main argument in favour of 2013-03, as I understand it, is in the
title "Post-Depletion Reality Adjustment and Clean up". It has been
expressed in the summary

/'The goal that precipitated the requirement to demonstrate need for
delegations is explicitly stated in section 3.0.3 of the policy, which
reads:

Conservation: Public IPv4 address space must be fairly distributed to
the End Users operating networks. To maximise the lifetime of the public
IPv4 address space, addresses must be distributed according to need, and
stockpiling must be prevented.

Following the depletion of the IANA free pool on the 3rd of February
2011, and the subsequent depletion of the RIPE NCC free pool on the 14th
of September 2012, the "lifetime of the public IPv4 address space" in
the RIPE NCC region has reached zero, making the stated goal
unattainable and therefore obsolete. This proposal attempts to recognise
this fact by removing all policy requirements that was working
exclusively towards attaining this "Conservation" goal.'/

However this argument is fallacious. The Conservation policy, even as
stated, expresses *two separate* policy objectives: 'fair distribution',
and maximising the lifetime of the public address pool. Depletion means
that reality has superseded the second objective, but not necessarily
the first.

So my first question to Tore is: "Why should 'fair distribution' of
addresses between users no longer be considered an overarching objective
of IPv4 address management?"

2. "Fair distribution" establishes a basic goal for IPv4 address
management policy. Other policies exist to pursue that goal.
If we abolish the goal, we not only abolish those other, secondary
policies (which may indeed be out of date) but we also deny ourselves
the right to introduce new, updated policies to pursue that goal.

2A.  Incidentally, I've seen comments in this thread that I read as
claiming that we effectively abolished the goal of fair distribution
back when we set a precise allocation size for the final /8. I can see
this argument, but disagree: limiting the maximum allocation size
directly relies upon the notion of fairness (we're only giving you X,
even though you want Y, so that the next guy can have some at all). By
limiting the minimum allocation size we made a pragmatic choice to trade
off some precision for efficiency in the allocation process, rather than
fundamentally abolishing the policy goal.

3. The proponents of 2013-03 have argued on this list that the existing
policy is unsupportable because, in a post-depletion world, RIPE NCC has
lost the means of enforcing it. It is certainly true that post-depletion
it will have lost the means by which it has traditionally enforced it.
But that doesn't necessarily mean that the policy should be abolished:
it might alternatively suggest that the required "reality adjustment" is
that the community develop new policies that would give the RIPE NCC
alternative means of enforcing it in the post-depletion world, where
IPv4 addresses continue to be an important resource.

a) I anticipate being asked "What might such policies be?". My answer
is, without shame, that I don't know. But perhaps the community should
retain the ability to set such policies in the future, if they seem
necessary. 2013-03 removes the basis on which such policies would be
introduced.

b) I anticipate being asked "Are such policies needed? Can you prove
this?". My answer is again that I don't know. More than that, it may
very well be that such policies are *not* needed, yet. But proof that
may change in the future. I would prefer that at that time new policies
could be introduced if they were proven to be needed. I would prefer
that such proof were not met with the response "When we abolished needs
based allocation back in 2013, we accepted that RIPE has neither the
responsibility nor the legitimacy to address these issues.". I fear that
the effect of passing 2013-03 will be to invite that response.

c) Actually, contrary to what I seem to say in the paragraph above, I'm
open to be persuaded that we should say "RIPE is not responsible for the
efficient and fair management of the address space, that's a purely
market process". But that's so far away from what we've had in the past,
I strongly believe we should set an unusually high bar for community
support for this, in determining a rough consensus, even at the cost of
taking a very long time to achieve it. Extraordinary measures require
extraordinary support.

d) Suppose that we do end up saying "It's nothing to do with us, let the
market rule". Who do we think should be responsible for clearing up the
mess if this does result in a clear, unequivocal disaster? Will we take
that as a failure of the new address policy, and change the policy
again? Would that even be legally possible, once we've commoditised
addresses? Or would we consider that a failure of the market, and leave
it to government regulators, competition authorities etc?


4. For the reasons above, I don't think it's accurate to claim that
2013-03 is a mere "reality adjustment"; I think it's a fundamental
change in policy. I believe the following represents a reasonable
alternative "reality adjustment", that more accurately states the
current reality:

"It has always been the policy of RIPE to seek to ensure that IPv4
addresses are allocated fairly and efficiently. While the RIPE NCC had
free IPv4 addresses to assign, this policy was implemented by requiring
End Users to provide documentation to the LIR (and/or the RIPE NCC) that
proved their need for address space, and the LIR need to provide similar
documentation to the RIPE NCC. In all cases, the documentation had to be
thoroughly vetted and archived for later auditing. After depletion of
the IPv4 address pool this process is no longer an effective control.
With the introduction of a transfer mechanism, it is possible that
market pressures may prove adequate to ensure efficient and fair
allocation of IPv4 addresses; in this transition period, the
consequences of run-out are uncertain. RIPE has chosen not to act
pre-emptively, to give market mechanisms a chance to succeed. However if
market mechanisms prove inadequate to safeguard the fair and efficient
use of IPv4 address space, RIPE stands ready to introduce new measures
as necessary."

5. We have lived with "needs based allocation" for a long time, and it
has served us well. "Delaying run-out" may previously have been a
sufficient justification for the policy, but that doesn't mean it's the
only justification for the policy.

I do accept that the onus should be on those that wish to retain the
policy on this basis to produce and convince the community of an
alternative justification. However, given that we're talking about
abolishing the core rule at the heart of allocation policy for the past
many (nearly twenty?) years, I believe it would be prudent to allow
quite an extended time to conduct that discussion.


6. My sixth reason for speaking out against the adoption 2013-03 at this
time is the potential political fall-out.

Certain governments are keen to end the independence of the RIRs and
move them from being run by and on behalf the community to being made
directly answerable to governments. In practice, this would mean the
abolition of RIPE (or at least, its policy making function) and perhaps
having the RIPE NCC board appointed by an intergovernmental body, likely
the ITU.

Some nation's governments remain very keen on this. Others, notably
developing countries, used to be very keen on this but have subsequently
been persuaded that the RIRs do a good job and that "bottom-up" /
community-based policy-making just works. This change of heart has been
hard won, and a key part of the message of RIRs and their defenders
(myself included) has been that these governments should support the
management of address space on a purely technical basis, with allocation
according to need, rather than risk politicising the issue by placing it
under inter-governmental control.

As someone who has worked in that space for the past eight years (some
of you may remember I proposed and chaired the RIPE Enhanced Cooperation
Taskforce, which recommended measures to address that debate) I can say
from experience that 2013-03 will be politically uncomfortable for the
RIPE NCC, and will give ammunition to those governments that still want
to gain intergovernmental control. The RIPE NCC impact assessment also
alludes to this risk.

Now, if there were a strong technical necessity for a policy to go
through, I would agree that the politics would have to take second
place. The NCC, and its supporters, like me, would have to justify the
change as best we can. However 2013-03 doesn't seem as important as
that, and certainly not urgent: it is just a "clean up" after all. So
this political downside should also be taken into account.

Moreover I would ask that the WG Chairs take this sixth issue into
account as evidence that the community hasn't fully examined the
potential consequences, and should be wary of proceeding too
precipitously. I would suggest that the NCC (or perhaps another NRO
member) be invited to give a presentation on this topic at the next
Working Group meeting.


Sorry for the long message on a long thread, and thank you for
considering it.

Regards,

Malcolm.

P.S. Direct response to specific questions for objectors from the WG
Chair follow.


On 25/07/2013 21:49, Sander Steffann wrote:
> 
> But I have seen lots of cases where Tore asks people to point out
> which part of his policy proposal causes concern, and I haven't seen
> any answers to that. 

For me, it's not a question of a particular element of the proposal that
needs improvement, it's a concern that its essence is misguided as a whole.

> So for the sake of an open, transparent and honest discussion about
> 2013-03 please include references to existing policy that is removed
> by 2013-03 or to specific sections of 2013-03 itself. 

My comments relate mostly to, and at the minimum to "Removes
"Conservation" as a stated goal of the policy".

Since this seems to be the core of the proposal, I would think that most
of the rest of the proposal would fall away if this were deleted.
However it's possible that I'm wrong: perhaps there may be some
compromise whereby we continue to assert that that the objective of
policy is fair allocation, but when considering the small assignments
from the final /8 "assume" that there is need (without documentation)
unless there is reason to believe otherwise.

This would allow for future policies that established reasons deemed to
suggest otherwise.
For example, in the unlikely case where someone registers 5,000
companies just to get 5,000 allocations (as has been suggested on this
thread), it would also allow the RIPE NCC to step in and say "No, that's
just gaming the system and we're not fooled".

It would also allow other policies to continue to require the
demonstration of need for the transfer of large blocks and of
intraregional transfers, if such policies were thought necessary and
someone could come up with a proposal to make them enforceable.

-- 
            Malcolm Hutty | tel: +44 20 7645 3523
   Head of Public Affairs | Read the LINX Public Affairs blog
 London Internet Exchange | http://publicaffairs.linx.net/

                 London Internet Exchange Ltd
           21-27 St Thomas Street, London SE1 9RY

         Company Registered in England No. 3137929
       Trinity Court, Trinity Street, Peterborough PE1 1DA