[ALSC-Forum] NAIS Response to ALSC

[Rob Courtney]

Attached is the response of the NGO and Academic ICANN Study to the 
ALSC's statements regarding the use of domain name ownership as a 
requisite to At-Large voting.

r

* * *

NAIS RESPONSE TO THE ALSC
WITH REGARD TO THE CONCEPT OF DOMAIN NAME HOLDERS

We welcome the opportunity, as members of the NAIS (NGO and Academic 
ICANN Study) team, to provide comments on the Interim Report of the 
At-Large Study Committee (ALSC), published shortly before ICANN's 
September meeting in Montevideo.

In this response, we address one key element of the ALSC draft 
recommendations: whether the membership of ICANN should be limited to 
Domain Name Holders.

Our own study, outlining recommendations for how ICANN should proceed 
to best incorporate the public voice into its decision-making 
processes, was published at the same time as the ALSC study and is 
available on the Web at http://www.naisproject.org/report/final/.

In our view, the public voice - and more importantly, the structural 
mechanisms that implement this voice - are essential prerequisites 
for ICANN's authority and legitimacy.  The need for a strong public 
voice is embedded in the historical process that led to ICANN's 
creation.  It is only through this established prerequisite that the 
U.S. government sanctioned the creation of ICANN - and the delegation 
to ICANN of the management of key Internet functions - in the first 
place.

The NAIS team and the ALSC are in agreement on one key overriding 
point: ICANN must provide some important voice for the Internet user 
in ICANN's processes of internal governance.

We are also in agreement on two important subsidiary points.  We 
agree that the At-Large Membership should be structured to facilitate 
participation in the ICANN decision making process.  And we agree 
that direct elections are the best mechanism for the At-Large 
Membership to select At-Large Directors.

We are pleased that there are these areas of consensus between the 
two reports.  We very much appreciate and support the recommendations 
of the ALSC regarding the need for a participatory membership and for 
direct elections.

But we also disagree on several key points.

First, we believe the number of At-Large directors should continue to 
be the same as the aggregate number of supporting organization 
directors, and should not be reduced from nine to six, as the ALSC 
recommends.

Second, we recommend that membership be open to any user of the 
Internet with an email address, and not, as the ALSC recommends, only 
to holders of domain names.

Third, we question whether a membership fee imposed in addition to 
the requirement of holding a domain name would impose such a 
financial barrier to participation in ICANN as to be an unreasonably 
exclusionary.

In our view, the ALSC approach on these issues fails to command the 
degree of "consensus" within ICANN that the ALSC draft report claims 
for it, particularly on the question of reducing the number of 
At-Large Directors from nine to six.  Serious questions have been 
raised as well with the idea of limiting At-Large Membership to 
domain name holders and with the reduction of At-Large board seats. 
We believe that these questions need to be resolved before any claim 
of consensus can credibly be made.

For this reason, any resolution presented to the Board at the Los 
Angeles meeting should not try to "lock in" any parts of the ALSC 
report which are not based on consensus - such as the recommendations 
for reducing the number of At-Large seats from nine to six, or 
limiting membership to domain name holders.  For the Board to adopt 
any such positions now would short circuit the need for much more 
discussion of these very controversial recommendations made by the 
ALSC.

THE ALSC'S DOMAIN NAME LIMITATION

The ALSC recommends that the universe of potential At-Large members 
be limited to "individual domain name holders", as opposed to some 
broader set of those who "use" the Internet.  While the criterion of 
domain name ownership may be attractive as an apparently simple means 
for limiting the number of potential voters, it is an inappropriate 
standard for a variety of political and practical reasons.

1.  The DNH approach excludes too many interested individuals from 
the At Large Membership.

The ALSC's main conceptual argument for limiting ICANN membership to 
domain name holders is that only these individuals "have a strong and 
tangible vested interest in ICANN activities, not just the Internet 
in general".  Indeed, every domain name owner has to sign a contract 
with his registrar that binds him directly to ICANN's UDRP.  He 
therefore does have an "interest" in ICANN, because ICANN defines his 
ownership of that one part of cyberspace.

But it is not clear why the ALSC chooses to limit the universe of 
interested parties to domain name holders only.   What about Internet 
users who have had their domain name taken away via the UDRP? What 
about users who cannot get the domain name they wish, because it is 
already taken? What about users who cannot afford to register a 
domain name (particularly in developing countries)?  What about users 
of the Internet for whom a domain name is impractical because 
bandwidth in Africa, for instance, makes only email-use feasible?

In reality, even the basic individual Internet user - i.e., one who 
uses only e-mail or simply surfs the web - is no less directly 
impacted by ICANN's decisions than is a domain name holder.  If ICANN 
makes policies that negatively impact the Internet's stability, or 
that fail to address the monopolistic practices of registries, or 
that retard the development of new name spaces in the issuance of new 
gTLD's, the simple e-mail or web user is also directly affected by 
those bad decisions, and should therefore have a voice in the 
processes by which these decisions are made.

There is no doubt, as well, that interested individuals will be 
excluded from membership in ICANN simply by the financial barrier the 
ALSC proposal would erect.  As we note above, the ALSC proposes not 
only to require domain name ownership, but to impose a membership fee 
in addition to that.  In Africa, for instance, the retail purchase 
price of a new domain name is at least $15 per year.  But a 
university professor in parts of Africa may earn as little as $120 
per month for a permanent post.  The costs of participation in the 
ALSC model would, as a practical matter, be an insurmountable barrier 
in Africa, and in many other parts of the world.

For these reasons, domain name ownership as a criterion for 
membership in ICANN is a radically under-inclusive standard, in that 
it excludes many who should be included in ICANN membership because 
they have a direct and legitimate interest in the work of ICANN, and 
ICANN decisions have a direct impact on them, even if they do not own 
or cannot afford a domain name.[ 1 ]  ICANN's own Membership Advisory 
Committee recommended in its final Report "If membership fees should 
be assessed in the future, they shall reflect the economic 
differences of the various geographic regions." The ALSC Draft Report 
does not address this critical point at all.

2. The DNH approach includes too many business interests in the At 
Large Membership.

But the domain name standard is over-inclusive as well. Current 
statistics indicate that 80 percent of domain name owners world-wide 
are commercial enterprises. Individuals, of course, have the right to 
register dot-com domains, but they are vastly outnumbered by 
corporate entities which do so.  Thus, under the ALSC's membership 
proposal, the vast majority of eligible members will functionally be 
corporations, not individuals.

Even if corporations are "represented" through designated individuals 
listed as their "administrative contact," those individuals will be 
functioning essentially as corporate agents in their capacity as 
ICANN members.

This approach will turn the At-Large Membership into yet another 
forum for corporate interests, which are already well represented - 
indeed, over-represented --in the supporting organization structure. 
Corporate and business domain name holders are already eligible to 
participate in ICANN through the business constituency, for instance, 
in the DNSO.  It is a clear distortion of what the At-Large 
Membership is supposed to be - a forum for the expression of the 
views of individual users of the Internet- to allow the At Large to 
become another forum dominated by commercial interests.  The MAC 
Final report clearly stated in its introduction that "the At-Large 
Membership should primarily represent those individuals and 
organizations that are not represented by the Supporting 
Organizations (SOs)."[ 2 ] By limiting the At-Large membership to 
Domain Name Holders, the distinction between members of the At-Large 
and of the Supporting Organizations becomes far less clear and 
unbalanced towards the interests of Domain Name owners, not the 
entire scope of the Internet users.

While corporate perspectives may not be monolithic, they do tend to 
embrace certain consistent values.  If the At Large is comprised 
largely of corporate interests, that view would be positioned to 
capture every At-Large seat, thereby depressing the diversity of 
views that should ideally be represented through the At-Large.


3.  The DNH approach does not account for ccTLD registration anomalies.

A further problem with a membership based on domain name registration 
is that the registration practices for country code domains varies 
widely.  In Germany, for instance, an individual may register any 
name, and as many names, as he or she wishes in .de, but in France 
and the Netherlands individuals' registrations are significantly 
curtailed.  Until recently, individuals were not permitted to 
register for .jp names in Japan.  And in other regions, these types 
of limitations are even more profound.

Creating a domain name standard for membership will thus filter 
participation in ICANN elections based on national policies governing 
registration.[ 3 ] In the example above, Germany would be advantaged 
over France, solely on the basis of the ccTLD registration policy. 
There is simply no principled reason to do so, and national 
manipulation of these policies can create unfair advantages for some 
countries in the ability to control ICANN elections.

4. The DNH approach does not solve the authentication problems.

The ALSC further argues that using an existing structure - the DNS 
and the registrar system - is more practical than reliance on a 
hybrid e-mail/postal system for defining and authenticating potential 
members as real persons, because the number of potential voters would 
thus be limited.

But this approach would require that an "At-Large Membership 
Contact," which the ALSC defines as "a new term to be incorporated 
into the domain name registration procedures," be designated for each 
domain name. Via the registrar, the registrant who then chooses to 
become an "At-Large Member" would be required to pay an annual fee.

In practice, this would mean that ICANN would need to amend all its 
contracts with registrars to include such a field, that the registrar 
would be required to handle the administration for collecting the 
fee, and that the registrar would be responsible for verifying that 
the identity of the individual domain name holder is correct.

All this depends on large-scale changes in the registrar structure, 
as well as a reliance on an untested technical protocol for 
"registration information and fee collection." Collecting this 
information with any degree of accuracy or security is a daunting 
task, as anomalies in the existing Whois records show.   As we found 
in our report, "to provide false administrative contact information 
is trivial and hard to detect." [p. 126].

In fact, for this reason, the DNS system is a poor choice of means to 
serve the purpose of voter authentication.  The only usual 
prerequisite for registration is a functional email address and a 
payment.  There is nothing in the system as it currently functions 
which authenticates the registrant.

Further, we question the wisdom of placing a key element of the board 
electoral process - the registration of voters - under the control of 
any single group in ICANN that is highly affected by the decisions 
and policies of the board.  Great care would have to be taken in 
order to ensure that the requirement that all voters be filtered 
through the domain name registrars does not give registrars an 
opportunity to exercise undue influence over the actions of voters or 
their access to vote.

The "registrar partnership" element of the ALSC plan does have merit, 
if adapted to a more inclusive approach to membership eligibility. 
Instead of contracting with registrars to undertake authentication 
based upon domain name ownership, ICANN could instead contract with 
geographically based "At-Large associations or chapters" that would 
distribute Personal Identification Numbers (PINs) by local snail 
mail. This would be less complex and more secure than a risky 
experiment with the DNS, which is neither a practical improvement 
over the snail-mail system nor acceptable on legitimacy terms.

CONCLUSION

We hope the ALSC will address the concerns stated above in continuing 
the work on its recommendations.  We are similarly addressing 
concerns that have been raised about the proposals set forth in the 
final NAIS report, and are examining proposals to "cap" membership as 
a way of limiting the pressure on ICANN's budget in supporting an 
At-Large Membership, proposals to supplement the postal/PIN 
authentication system with less expensive alternatives, and 
alternative funding sources as a means of supplementing ICANN's 
current revenues in order to support the costs of the At-Large 
Membership in more sustainable ways. We plan to deliver these 
proposals along with its implementation plan in the next ICANN 
Meeting in Marina Del Ray.

We look forward to a continuing exchange with the members of the ALSC 
and the ICANN community on all of these issues.


The NGO and Academic ICANN Study


FOOTNOTES

[ 1 ] In its final report, the ICANN Membership Advisory Committee, 
chaired by then-ICANN Directors Greg Crew and George Conrades, began 
its report with the recommendation  that the At-Large "include any 
Internet user with access and verifiable identity in order to reflect 
the global diversity of users (membership should not be limited to IP 
address or domain name holders)," which was further followed by 
commentary that: "Limiting membership to IP address or domain holders 
would unreasonably ignore the interests of the majority of the users 
of the Internet who make significant financial payments for their 
access. An open, bottom-up decision-making structure, however, must 
be balanced by reasonable steps to assure that the membership is 
authentic and representational. Limiting membership to IP address or 
domain holders would unreasonably ignore the interests of the 
majority of the users of the Internet who make significant financial 
payments for their access. An open, bottom-up decision-making 
structure, however, must be balanced by reasonable steps to assure 
that the membership is authentic and representational." See 
http://www.icann.org/berlin/membership_commentary.htm#1.

[ 2 ] See http://www.icann.org/berlin/membership_rec.htm.

[ 3 ] There will be severe limitations in certain countries where 
bundling the sales of Domain name under monopoly situation (most 
ccTLDs are exclusive channel of selling Domain names in its 
jurisdiction) with ICANN At Large membership may be regarded as 
anti-trust practice. Before DNHS approach is to be implemented, 
significant study in its legal implication should be carried out. 
ALSC report has not addressed this critical issue yet.

-- 

Rob Courtney
Policy Analyst
Center for Democracy & Technology
1634 Eye Street NW, Suite 1100
Washington, DC 20006
202 637 9800
fax 202 637 0968
rob@cdt.org
pgp id: 0xAD7123FB
http://www.cdt.org/