[arin-ppml] ARIN-prop-176 Increase Needs-Based Justification to 60 months on 8.3 Specified Transfers

[Tom Vest]

Hi Jeff, 

Thanks for the response. Reactions inline...


On Jun 29, 2012, at 9:00 AM, <jeffmehlenbacher at ipv4marketgroup.com> <jeffmehlenbacher at ipv4marketgroup.com> wrote:

> Hello Tom,
> 
> Thank you for weighing in on this proposal.  Please permit me to preface
> my statistical suggestions with a couple of points:
> 
> 1.  I neither suggested it was valid to wait for statistical evidence
> before considering the merit of extending the window for justification
> nor did I suggest the current statistics available through ARIN were
> sufficient to evaluate the appropriateness of the current justification
> window.  Two other AC Members suggested that would be a sensible
> approach.  I do not believe published statistical tables currently
> exists to support such an assessment and it's perhaps a deferral
> strategy on behalf of those individuals opposed to the policy proposal.

Although it reads that way, I'll assume that you're *not* implying here that proposed policies could or should be evaluated based on their supporters' and/or detractors' "true" motivations (?). Please correct me if I'm mistaken. 

However, if that was not the implication, then your criticism of a proposed policy evaluation mechanism solely because it results in "deferral" -- i.e., taking (more) time (than some alternative) -- must be based on some underlying theory of policy process "timeliness." Could you share the details of that theory with us? At what point does the time required to evaluate specific proposal (x) go from "reasonable" to unreasonable"? Does the time vary depending on the specifics of (x), and if so what factors justify more vs. less evaluation time? 

> 2.  My proposal is founded on the very simple principle: reallocating an
> organization's unused IPv4 blocks to a legitimate ISP or End User via an
> ARIN-approved 8.3 Specified Transfer is a good thing for the community
> and should be made as painless as possible.

While I can understand why your self-declared commercial interests might lead you to assert this, the assertion itself is literally indefensible unless one believes that (a) the act of making such transfers "as painless as possible" does not and *cannot* give rise to any net adverse collateral effects, perverse incentives, unintended consequences, etc. under any circumstances (i.e., relative to not acting to absolutely reduce all transfer-related pain), or alternately that (b) such "externalities" cannot or should not be have any bearing on the decision(s) to accept or reject specific transfer policy-related proposals. Do you subscribe to either or both of those beliefs? 

>  60 months justification
> promotes a much, much higher degree of certainty to enter into an 8.3
> transfer and thus all parties (sellers and buyers) will be more willing
> to participate.  

Please correct me if I'm mistaken, but you seem to be using the word "certainty" as a substitute for "private incentive." Granted, under certain circumstances, some kinds of "certainty" can incentivize private actors to participate in a market, but it's not clear to me how a 60-month justification period would contribute to net *market* certainty -- i.e., the kind of certainty that is widely advocated as a goal for public policy. One the first day after a 60-month justification period were approved, aspiring IPv4 buyers (and, to a lesser degree, aspiring IPv4 dealers) might have greater private "certainty" that a one-time purchase of IPv4 *on that day* would serve their own private interests -- i.e., they would have a greater private incentive to participate in the market. However, that one-time gain in private "certainty" would be purchased at the expense of their own (and everyone else's) *diminished* certainty that a nonzero supply of transferable IPv4 would exist every day thereafter. How is overall "certainty" improved in any way that might merit policy intervention?

> Unused IPv4 blocks are revitalized,

If auditing IPv4 utilization was a costless, non-disruptive, perfectly accurate, and universally embraced practice, this would unquestionably be true.
In the world we actually live in, the "IPv4 revitalization" effects of the proposed change would likely be much more ambiguous.  

> the registry is more accurate, etc.

Practically speaking, this would be the expected result ONLY to the extent that the current registry is populated with many entries consisting of (1) institutions that do not exist and (2) address resources that are completely idle. If the current registry entries for institutions that have resources in production are 100% accurate today, then bringing formerly idled resources would grow the registry, but have no effect on registry accuracy. Conversely, if current registry entries for institutions that have resources in production are somewhat less than 100% accurate today, then why would you expect incumbent operators and/or new entrants to voluntarily subject their newly purchased IPv4 to a higher accuracy standard than they currently maintain? 

Perhaps you are assuming that registry accuracy will be improved by (someone) first identifying resources that are currently registered under org (x) but actually used by org (y), and then either compelling those unauthorized users to correct the registry record, or alternately forcing them to cease using the resources so that they can be conveyed to someone else who is both able to put the recovered resources to productive use (as defined by policy) and also willing to establish and maintain a more accurate registry entry linking themselves to the resources. If this is your assumption, and further you believe that the "someone" who will do that private detective/bounty hunter job is the newly emerging commercial IPv4 brokerages, then I would have *lots* of questions for you that I'll defer pending your response. For now, I'll just point out that no matter who that hypothetical "someone" might be, the same questions asked above would apply, i.e., how would the 60-month justification period that you advocate change circumstances ( e.g., imbue facilitators with greater investigative/enforcement powers *and* commercial incentives that actually favor a higher level of accuracy over higher profits) in any way that might lead to improved registry accuracy? 

> Having said that Tom, I believe statistics that might lead the community
> to a better understanding of what has transpired the last 17 months
> under 8.3 are readily available.  They simply are not published by ARIN.
> What we know with certainty is how many 8.3 transfers have been
> requested, and how many have been completed.  We do not understand the
> one to one relationship of request and approval/rejection because it's a
> rolling monthly total.  I believe that before we attempt to use any
> statistics we should understand the data elements that are captured, the
> accuracy, and the utility to evaluate such things as whether the 24
> month justification window is appropriate.  I would suggest the
> following data elements (not a comprehensive list) are captured but
> currently not published...if Mutual NDA's prohibit said, it's really a
> matter of modifying said Mutual NDA to ensure that generic data may be
> collected and published:
> 
> 1.  Date of Request for 8.3 Specified Transfer
> 2.  Status of Transferring Block (Legacy, LRSA, RSA)
> 3.  Size of Block (/16, /17, etc.)
> 4.  Source SIC Code
> 5.  Recipient NRPM Definition: ISP or End User
> 6.  Recipient Customer Status with ARIN: Customer or Non-Customer
> 7.  If ARIN Customer, date and size of last allocation from the free
> pool
> 8.  Current percent utilization of existing blocks
> 9.  Number of Justification re-submissions for transfer
> 10. Approval or Rejection decision
> 11. Approval or Rejection rationale (codes 1-?)
> 11. Date of Approval or Rejection
> 12. Facilitator Involvement (yes/no)
> 13. Price paid (US$)

I agree that more information is almost always better than less -- but I suspect that you've already overshot the level of information that would be deemed acceptable to most potential market participants by a very wide margin. Moreover, even this is not enough. For example, why exclude the identity of the aspiring buyer and the aspiring seller? With the possible exception of smallish entities that are both privately held AND have zero customers (i.e., that have no PA allocations and at most a very small number of ASNs), every institution that has ever received resources from ARIN or one of the other RIRs has critically important "external" stakeholders -- including either or both current and potential future stock buyers/holders/shorters, and (imo, more important still), a potentially large number of downstream network service "customers" -- i.e., entities whose own IPv4 addressing needs often provided much of the original "justification" that enabled the aspiring IPv4 buyer/seller to obtain whatever resources are already in their possession. Surely the network customers whose future service expectations, and the investors whose future earnings expectations may be substantially impacted by a network service provider's evolving IP addressing strategy also have a right to know? Promoting a larger/more active IPv4 market while simultaneously shielding the identities of aspiring IPv4 market participants would dramatically reduce whatever level of market "certainty" that those important stakeholders currently enjoy -- how can that be justified?

Also, assuming that "facilitators" are free to differentiate themselves by adopting different commercial strategies that could in turn materially affect the outcome and details of the transactions that they broker in different ways, why should their identities be exempted from public disclosure?

> I'm surely missing a half dozen other data elements that are captured
> during an 8.3 but my point is, there is considerable data...will it
> support the notion that 24 months is the correct window?  Very loosely
> at best, but it would assist all interested parties in better
> understanding the 8.3 transfer market.  Publishing price however is
> going to be an entirely separate debate isn't it?  I quietly muse at all
> that would suggest 60 months will permit speculation or hoarding. 
> Really?  What are you speculating on in the absence of known prices and
> continuous updates?  If however, prices are published, now you have the
> basis to speculate as is the case in every open market system (real
> estate, stock market, etc.).
> 
> As for other well-established markets, I have thoughts but would suggest
> there are individuals who participate on the PPML better qualified to
> suggest objective models for assessment purposes. 

In that case I guess we'll have to wait for someone who is both (a) knowledgeable of a potentially illuminating present-day exemplar market and (b) actually believes that it would be beneficial for IPv4 to be traded more like a "normal" commodity to speak up.

For my own part, I am quite familiar with some other markets that are, IMO, quite relevant and illuminating to the problems under discussion, but at present those markets provide only a wealth of examples of what not to do (i.e., if we value stable markets, a stable, self-governing industry, continued/future growth and prosperity, etc.), and what to expect if we proceed anyway. 
 
Regards, 

TV

> 
> Jeff Mehlenbacher


> -------- Original Message --------
> Subject: Re: [arin-ppml] ARIN-prop-176 Increase Needs-Based
> Justification to 60 months on 8.3 Specified Transfers
> From: Tom Vest <tvest at eyeconomics.com>
> Date: Thu, June 28, 2012 11:59 am
> To: <jeffmehlenbacher at ipv4marketgroup.com>
> <jeffmehlenbacher at ipv4marketgroup.com>
> Cc: Daniel_Alexander at Cable.Comcast.com, arin-ppml at arin.net
> 
> Hi Jeff, 
> 
> Out of curiosity, could you suggest ( at min.) one set of data that
> would be sufficient, in your view, to determine the appropriateness of
> a(ny) IP number resource transfer market, under current or any other
> imaginable policy criteria? How would you define "appropriateness" and
> what sort of data would you need to make such a judgment? 
> 
> If this question seems too abstract, perhaps you could point us to some
> other, well-established market that demonstrably satisfies its own
> "appropriateness" criteria based on "objective" data that is, if not
> publicly available, at least available to some disinterested
> "appropriateness evaluators"? 
> 
> Thanks, 
> 
> Tom Vest
> 
> On Jun 28, 2012, at 6:25 AM, <jeffmehlenbacher at ipv4marketgroup.com>
> <jeffmehlenbacher at ipv4marketgroup.com> wrote:
> 
>> <snip> I have a real concern that the body of data available for assessment is neither comprehensive
>> nor focused on dictating success or failure of the current 24 month
>> justification period. I would be interested in understanding what data
>> will be monitored to determine the appropriateness of current policy for
>> 8.3s.
>> 
>> Thanks!
>> 
>> Jeff Mehlenbacher
>> 
>> 
>> Date: Thu, 28 Jun 2012 08:11:24 +0000
>> From: "Alexander, Daniel" <Daniel_Alexander at Cable.Comcast.com>
>> To: "arin-ppml at arin.net" <arin-ppml at arin.net>
>> Subject: Re: [arin-ppml] ARIN-prop-176 Increase Needs-Based
>> Justification to 60 months on 8.3 Specified Transfers
>> Message-ID:
>> <B64177493F39BA4A81233AA84B50049E5A2BA30E at PACDCEXMB12.cable.comcast.com>
>> 
>> Content-Type: text/plain; charset="Windows-1252"
>> 
>> Jeff,
>> 
>> One of the primary justifications used during the debate of 8.3
>> transfers
>> claimed that transfers would put underutilized resources to use. By
>> stretching the period to five years, we start trading one underutilized
>> resource holder for another. This is a contradiction to the claimed
>> benefits that everyone was supposed to accept who objected to these
>> transactions.
>> 
>> No sooner than section 8.3 was created the policies came in to expand
>> the
>> timeframes. The timeframes have already been expanded before having any
>> data as to the benefits or consequences of the changes that have been
>> made. While I don't claim to know what the magic number should be, I
>> think
>> this change would be irresponsible at this time, based only on the
>> speculations made on this mailing list.
>> 
>> I'm opposed to this proposal.
>> 
>> Dan Alexander
>> Speaking as myself
>> 
>> 
>> _______________________________________________
>> PPML
>> You are receiving this message because you are subscribed to
>> the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
>> Unsubscribe or manage your mailing list subscription at:
>> http://lists.arin.net/mailman/listinfo/arin-ppml
>> Please contact info at arin.net if you experience any issues.
>