ARIN: Questions & Comments

[Stephen Satchell]

At 9:11 AM -0800 2/21/97, William Herrin wrote:
>  ARIN needs to build and maintain a well defined cash reserve. Daniel
>Karrenberg mentioned this in passing, but I want to emphasize it. The
>Internet is unpredictable, and when (not if) budgeting mistakes are made
>there needs to be a buffer that allows ARIN to continue operating
>effectively until changes in the fee structure can be made. Ideally the NSF
>or Network Solutions should contribute the initial reserve, but if not then
>I think most of us ISPs will understand and support the higher initial fees
>necessary to build it.

     The problem with keeping a cash reserve is that existing tax law and
IRS regulations frown heavily on a not-for-profit maintaining such a cash
reserve.  As I understand it, the thinking of the federal policy-makers is
that a charity should dole out all it gets during the fiscal year, so there
is no "profit" at the end.  If money is needed, you go beg for it.

     This isn't the best thing.  I ran into the same problem with my
not-for-profit, and ended up *not* getting 501(c)(3) status because of the
cash reserve portion of the corporate policy.  I *could* operate as a
private foundation, but I would pay tax on any cash I had at the end of the
year.  <sigh>


>  Speaking of the fee structure, creating fixed dollar amounts strikes me as
>a mistake. The fees collected from each activity should be expressed as a
>percentage of ARIN's budget, and on a particular day each year, an accountet
>should sit down and combine that percentage with the projected activities
>and budget in order to determine the next year's fees. Accountants are
>roughly as intelligible as lawyers, so tying the fees to the costs is really
>the only way that we can be sure ARIN is really acting like a non-profit
>organization (i.e. isn't gouging us.)

     Frankly, when I did my back-of-the-envelope alternative budget for
ARIN, I discovered that the expenses of the Registry are pretty much fixed,
or easily tied to specific activity.  The cost of supporting a Member was
roughly $117, the cost of making (or unmaking) an allocation was around
$300, and that was that.  The rest of the costs were pretty much
independent of activitiy.  If you dump ARIN having to maintain
IN-ADDR.ARPA, the need to get heavy-duty computers and connections goes
away -- the only significant net activity would be applications and
database look-ups.

     (By the way, that $117 was assuming that *all* members participated
using traditional paper systems; the cost goes down further if you operate
ARIN the same way as much of the Internet Society operates.)

     So what you have is a large fixed budget and a small task-based
budget.  Sad but true.

     I might add that I'm still waiting to see a strawman budget.  Kim has
been a bit gun-shy about presenting it because of what happened when the
first draft of the proposal was posted.  I find that interesting, though,
because I don't remember getting a lot of heat when I presented my
alternate budget proposal.  Maybe because I'm not "offical", eh?  :)

---
Stephen Satchell, {Motorola ISG, Satchell Evaluations}
<http://www.accutek.com/~satchell> for contact and other info
Opinions stated here are my PERSONAL opinions.