What happened in Brussels? A wreck.

After 13 years of ICANN and the putative “privatization” of
DNS governance, we are still having fundamental debates about the role of territorial
governments in ICANN. The significance of this debate goes well beyond the
governance of domain names, and implicates governments’ role in Internet
governance more generally. The February 28-March 2 showdown between the ICANN Board
and its not-so-advisory Governmental Advisory Committee (GAC) in Brussels
resolved very little, except to demonstrate that conflict over this issue is
hardwired into the structure of ICANN. It is a big problem and it is going to continue. This article provides our take on what happened.

(Another summary of the meeting was
written by William Drake, who attended the meeting as an observer on behalf of
the Noncommercial Users Constituency (NCUC). You can see my response
to that summary here
, and Drake’s response to my response here.)

The Brussels meeting produced a lot of nice words about “mutual understanding,”
“improved communication,” and converging positions from the participants. I have no doubt that these effusions were made sincerely, and
that they accurately reflected the feelings of the participants at the time. But
from a more objective perspective informed by knowledge of international
institutions and governance processes, the whole process – both the meeting and
the activities leading up to it – was more like an automobile wreck. Yes, the
colliding partners emerged alive. Yes, they politely, even cordially exchanged information
about their insurance companies. Yes, they managed to agree on some facts
regarding what caused the accident (while disagreeing about others). And yes,
they learned a lot about the perspective of the other driver. But this doesn’t
change the fact that it was a collision, that damage was done, and that it never should have happened.

From a superficial perspective confined to that of the
bargaining parties (GAC and Board), the meeting could be considered a partial
success. The ICANN Board wants to satisfy the GAC (which means, primarily, the
US Government) so that it can proceed with the implementation of its policy for
adding new TLDs. This meeting allowed it to respond to specific demands made by
the GAC and to clarify which concessions it could/would make and which ones it
would not. From the GAC’s perspective, they finally got what they crave, which
is the full and undivided attention of the ICANN Board. And for a while, GAC’s time
in the spotlight made many of these mid-level bureaucrats feel as if they were in
the correct “role;” that is, one of exercising their sovereign “right” to make something
called “public policy” (despite the fact that none of them are accountable to
or elected by the global polity represented by the Internet and its users).

But that perspective, as I said, is superficial. Look beyond
bargaining among the powerful and take a broader, public interest-oriented take
on the meeting. ICANN went through 5 years of policy making and two paroxysms of
last-minute changes demanded by trademark holders at the end, and another 11th-hour
attempt to revise its objection/censorship process demanded by the GAC. The very
fact that the Brussels meeting took place at all meant that the ICANN Board,
which tells the world that it makes policy in a multi-stakeholder, open, fair, bottom-up
process, has conceded that it really has no defined process. We can all
participate in a process and come to a decision that satisfices us, but if anyone
powerful demands a “re-do” then ICANN can and will
oblige. Remember, the GAC scorecard was not a meeting about one or two
especially bad or objectionable parts of the new TLD policy; it was a demand to
revise and resubmit practically the entire thing. This approach to policy
making has terrible consequences for the long term. It means that it makes
little sense for anyone to invest time and effort in ICANN’s process. Practical
people will concentrate their efforts on lobbying their national governments
and directly lobbying the Board members.That makes the policy development process restricted to a privileged few.

Worse, we need to ask why the governments acted the way they
did. What drove them to make this extraordinary attack on the outcome of a five
year process?  The answers are
depressing. It is not too outrageous to characterize this meeting as a street mugging.
Governments and brand owners conspired to hold up and threaten the very
existence of a new TLD program at the last minute, in order to extract
concessions that were clearly and explicitly rejected by other stakeholders
during the normal process. Governments acted primarily as agents of a few powerful
economic interests and secondarily as protective of the narrow institutional
self-interests of national governments themselves. With one notable exception,
they did not act as public interest advocates seeking to improve ICANN’s
accountability.* It was abundantly clear that the only members of the public that
the German and UK governments listen to and care about were the big brand
owners; they practically said as much during the meeting. The US government,
too, cares mostly about trademark interests and its own geopolitical games. Other
governments care about whether they have to pay fees to object or whether they
control geographic names which no existing international law or treaty gives
them control over.

Even if you restrict your perspective to that of the Board
and GAC, the meeting failed in a very important way. The Board, which is
committed by its bylaws to listen to GAC “advice” and make a decision about
whether it agrees or not, thought that this meeting was part of a “bylaws
consultation.” In other words, it thought that this meeting would make them
ready to tell the GAC at the end of the San Francisco meeting (March 17) which
aspects of the “GAC Scorecard” it would agree to and which it wouldn’t. According
to the bylaws, that would be the end of the process and they could proceed to
implementation. The GAC, however, would not allow any such straightforward
resolution. It announced, unilaterally, that a “bylaws consultation” could only
occur when the GAC agrees that it is in a “bylaws consultation.” The GAC
complained that every member of the 60-government committee had to go back to
their national capitals and find out whether the Board’s positions were
acceptable or not. And of course, that could not be done in the two weeks
remaining before the San Francisco meeting. As far as GAC was concerned, it was
not quite ready for such a formal consultation. By this logic, the GAC can
delay indefinitely simply by refusing to admit that it is in a consultation
until it gets whatever concessions it wants.

The difference between cumbersome
intergovernmental procedures and the decision making of a private corporation’s
board does exist, and is important. But what boggles my mind is that no one on
the government side seems to remember that this slow, bureaucratic 200-sided
multilateral process is precisely why the U.S. Commerce Department and practically everyone else involved agreed to put DNS governance in the hands of a
private corporation to begin with! We wanted to get DNS out of a
situation in which 200 different jurisdictions would try to impose their own
law and spend ages trying to work out their differences.

The biggest issue here is what constitutes a “public policy
issue” and “public policy advice”? GAC has had many, many opportunities to make
its advice known during the policy making process and to participate in that
process. GAC issued its basic policy advice on new gTLDs in 2007, and has never
had the capacity or the inclination to make use of opportunities for ongoing
consensus development in the GNSO, making it a bad faith partner in the policy making
process. Instead, it has repeatedly intervened ex post after the policy making process is finished.
Over time, GAC has gravitated more and more toward demanding an operational
role, i.e., giving itself direct control over outcomes, rather than providing
advice on general policies that should be adopted. The TLD veto – or even the
watered down version of giving “advice” on individual applications through a GAC
Communiqué – illustrates the difference. When you look at individual domain applications
and ask to be able to object to them “for any reason” you are not providing
“policy advice”: you are asking to be the one who makes the final decisions
about what TLDs can and cannot exist. And even though the US government was
forced to back down from its earlier position that a GAC veto should kill an
application, US representatives still reiterate their strong belief that it
would be politically unacceptable for ICANN to accept any TLD that the GAC
objects to. This is typical of the train wreck that constitutes ICANN-GAC
relations. We are told that the GAC’s power is advisory, but if you don’t take
the advice, it is implied that you will get into serious trouble.  And yet the GAC’s authority and standards for
making these objections is completely extra-legal: no laws, no treaties, no appeals
process no legislative ratification are required for it to make these life-or-death decisions for internet

GAC is starting to look less and less like an
Advisory Committee and more and more like an emergent intergovernmental

* The one exception is when GAC asked ICANN to reconsider its decision to make itself completely immune from lawsuits for mistreatment of TLD applicants. This is indeed a major public interest issue that was not adequately dealt with by the ICANN process, an instance of the Board putting its interests as a corporation above the public interest. But of course, the GAC doesn't press very hard on that issue.

One comment