[ARIN-consult] Fee restructuring

[David Farmer]

On 10/27/12 18:11 , Bill Woodcock wrote:
>
> Two things I'd like to point out:
>
> First, this doesn't disincentivize IPv6 deployment, it only disincentives people getting allocations that are way, way larger than they need.  Same as in IPv4.

I disagree at least partially.  But first, I really like what has been 
done with regards to the one-time fees for end users and IPv6.

The Policy experience report last week, slide #7, gives use good data 
about IPv6 assignments since March 2011;
https://www.arin.net/participate/meetings/reports/ARIN_XXX/PDF/thursday/nobile_policy.pdf

Combined that with the current and proposed fees, and you get;

53% of end users are getting /48, current fee $1250, proposed fee $500
25% of end users are getting /44, current fee $1250, proposed fee $1000
15% of end users are getting /40, current fee $2250, proposed fee $1000
5% of end users are getting /36, current fee $2250, proposed fee $1000
2% of end users are getting /32, current fee $2250, proposed fee $2000

If these percentages hold for the long term, then 53% of end users will 
pay $500 (a 60% reduction) and 45% will pay $1000 for their IPv6 
assignment.  I think that is quite reasonable, and represents a good 
reduction for most end users, and everyone is reduced by at least a 
little.

However, the problem and disincentive isn't with the one-time fees, it 
is the annual maintenance fee.  So, lets look at a small organization 
that was an early adopter and has been listening to us about IPv6;  They 
paid $1250 (there never was a IPv6 discount for end users) for their /48 
of IPv6 and now they have to pay $300 a year (assuming they have one 
assignment each type of resource; ASN, IPv4, IPv6), $100 of which is 
directly attributable to their IPv6 assignment.  They don't get any 
advantage from the reduction in one-time fees and are seeing a 300% 
increase in their annual fees.

Next, lets look at a small organization that hasn't deployed IPv6; They 
will now pay $200 a year (assuming they have an ASN and IPv4 
assignment).  When they add IPv6, they pay a $500 one-time fee for their 
/48 and an extra $100 annual maintenance fee, again directly 
attributable to their IPv6 assignment.  Its the extra $100 a year that 
is directly attributable to IPv6 that I contend is the disincentive.

My recommendation is to include one resource assignment of each type 
with the base annual maintenance fee for each organization, be that 
$100, $150, $200, etc...  The first organization, that was an early 
adopter of IPv6, no longer sees any increase if the base fee is at $100, 
and in any case no increase that is directly attributable to their IPv6 
assignment.  The second organization, only pays the one-time $500 fee 
for their /48 when they receive it, no change in their annual fee when 
they get IPv6, even if their annual fee gets increased overall.

With the recommendation, as long as an organization has only one IPv6 
assignment, which under current policy should be the norm for the 
foreseeable future, there is no increase in their annual maintenance fee 
that can be attributed to IPv6.  However, as currently propose there is 
a $100 increase in the annual maintenance fee for every organization 
that is directly attributable to IPv6 if they have it, I believe that to 
be a disincentive for IPv6.

> Second, arguments that some other hypothetical person might suffer an increased burden, that are not borne out by the statistics, don't carry a heck of a lot of weight.  If _you_ are unhappy with how _your_ fees are changing, I want to hear about it.  Arguments about what someone else, who might or might not exist, might or might not think, are feckless.
>
>                                  -Bill

I believe Owen and a few others have said that they are directly 
effected and have taken issue.  I'm merely proposing a solution and 
discussing the advantages of such solution.  John has said ARIN is 
generating numbers related to such a solution, only ARIN can actually 
generate the numbers.

I generally like the idea that annual maintenance fee for end users will 
scale up with the number of registrations that they have, that even 
remains true with what I'm recommending.  I just think that having a 
$100 increase in the annual maintenance fee that is directly 
attributable to IPv6 is a bad idea.

If you want all end user organizations to pay more than $100 for their 
annual maintenance fee, then make the annual maintenance fee $150, $200, 
$250, or even $300 per organization, but include one resource 
registration of each type and then charge $100 per additional resource 
registration beyond the first one of each type per organization.