ICANN: Defending Our Precious Bodily Fluids

Report and reflections on the November 2001 ICANN meeting in Marina Del Rey, California

It used to be that the machinations by which ICANN developed the agendas for its annual meetings was a bit of a mystery. Those agendas -- part by-product of backroom dealing, part concession to public pressures -- were usually handed down at the last minute by ICANN's staff, in routine violation of the scheduling requirements set forth in the organization's by-laws. But the 11 September massacres "changed everything," including ICANN: on 27 September, ICANN's staff dispensed with the smoke and mirrors when it announced that rather than "follow the usual ICANN format" the November 2001 meeting in Marina Del Rey, California, would "focus exclusively" on the pressing problems of "Security and Stability of the Internet Naming and Address Allocation Systems."

"Stability" is, of course, ICANN's favorite theme, maybe because it implies that skeptics of the organization are somehow mixed up with the forces of instability. That the term should figure prominently in this "special" meeting came as little surprise. However, the newer theme, "security," was the subject of some speculation. The terminally suspicious (of which this writer is a card-carrying member) viewed it, in Milton Mueller's cautiously crafted words, as "a conscious attempt to avoid the real issues facing the organization." Given ICANN's long history of manipulations, such cynicism was certainly justified. But that hardly precluded the possibility that other factors may not have driven ICANN's decision.

In messages sent to various ICANN-related mailing lists, former CEO and president Mike Roberts fueled speculations that the agenda change might have been a response to pressure from above. For example:

When civilization takes a step backward, as it did last week, it usually means a period in which the people with the guns make the decisions. Anyone watching the American President on tv in recent days knows that's where we're at. It may be a while before the luxury of debating what constitutes consensus in a terrorist-less society returns.

Eruptions like this were hardly new; in fact, ICANN critics had long, if quietly, treasured Roberts's talent for tone-deaf bombast -- so the question was whether to read anything into them at all. Was he responding to actual pressure or, rather, was he putting on an anticipatory CYA (Cover Your Ass) performance? After all, if the Domain Name System's rootservers collapsed under a distributed denial-of-service attack, the subsequent show trial (featuring Roberts in a starring role) would have been spectacular. It's fair to say -- without necessarily endorsing how they ended up in this awkward position -- that ICANN may not have had much of a choice about the agenda.


Normal ICANN meetings are bland bazaars teeming with policy wonks and telecom savants, but the labyrinthine schedule of overlapping constituency and council meetings at least allows for roving participation and haggling in the hallways of the hotel conference centers where it meets. This "special" meeting, on the other hand, looked really grim. The prospect of sitting passively through three days of PowerPointillism seemed bad enough to scare off even the most stalwart activists.

Worse: no sooner had ICANN grudgingly conceded that Monday, 12 November, could be devoted to the actual activities that, in theory, are ICANN, than it went out of its way to confuse matters again by announcing (on 26 October) that the meeting would begin not on the 12th but on the 13th, when the "security"-related presentations began. In response to my queries, ICANN staff publicist Mary Hewitt said that "the unofficial meetings occur on the 12th"; then, when pressed, she explained that the "unofficial meetings" are "still part of the ICANN official meetings."

It's no secret why ICANN would want to discourage attendance: the issue looming over this meeting centered on the At Large Membership (ALM), that is, the mechanism by which workaday internet users (as opposed to network operators, domain name registration services, intellectual property lobbies, and so on) are to be represented on ICANN's board of directors. More specifically, the At Large Study Committee (ALSC) -- a star-studded delaying tactic -- was scheduled to hand over its report on what to do about the ALM question.

The issue of the At Large has bedeviled ICANN since its birth in October 1998, when ICANN's "initial" board of directors in effect appointed itself. This arrangement, among other things, attracted enough attention in July 1999 to merit US Congressional hearings titled "Is ICANN Out of Control?", in which then-board Chairman Esther Dyson testified before the US Congress that:

it is ICANN's highest priority to complete the work necessary to implement a workable At-Large membership structure and to conduct elections for the nine At-Large Directors that must be chosen by the membership. ICANN has been working diligently to accomplish this objective as soon as possible. The Initial Board has received a comprehensive set of recommendations from ICANN's Membership Advisory Committee, and expects to begin the implementation process at its August [2000] meeting in Santiago. ICANN's goal is to replace each and every one of the current Initial Board members as soon as possible, consistent with creating a process that minimizes the risk of capture or election fraud, and that will lead to a truly representative Board.

What ensued is far indeed from what Dyson promised. After elections involving directors other than the self-appointees (who called themselves "At Large") the resulting "interim" board engaged in all manner of sophistry and procedural hijinks to defer and/or weaken the ALM. True, in November 2000, after a torturous game of bureaucratic maneuvers and countermaneuvers involving numerous civil society organizations, five At Large directors were elected (see my roving_reporter column passim for detailed coverage). But the remaining four At Large seats are occupied to this day by "initial" self-appointees who refuse to leave, or "boardsquatters" in the jargon: Frank Fitzsimmons, Hans Kraaijenbrink, Jun Murai, and Linda Wilson. (Of them, Fitzsimmons takes his refusal a step further than the others: he refuses even to participate in ICANN meetings.)

It isn't mere speculation that the replacement agenda for the November 2001 meeting was intended in part to preempt the At Large issue. In each of his militant messages sent on 16 and 17 September, former CEO Roberts -- a former military man -- explicitly discouraged any support whatsoever for the At Large. He derided "populist notions about any old terrorist around the globe getting to vote on how to run the DNS," declared that "there is going to be much less interest in who is represented by whom on the Board," and even sought to undercut funding in advance:

"If you were thinking about contributing to an ICANN ALSO [At Large Supporting Organization], send it to the Red Cross instead."

Just over a month later, ICANN announced that Roberts, who had left the organization less than a year earlier, was to chair the "informal" Program Committee that would plan the detailed agenda for the meeting, which was just three weeks away.


In a way, these procedural maneuvers were overkill. The At Large Study Committee itself had been formed to answer a question whose very existence had been conjured up only by means of fantastically scholastic readings of ICANN's own by-laws (which have been revised no less than eleven times in the organization's three-year existence). From its firm formulation at ICANN's March 2000 Cairo meeting, to the June 2000 Yokohama by-law revisions that initiated it, to the January 2001 charter that christened it, countervailing -- indeed, paradoxical -- currents defined the ALSC's mission. On the one hand, it was to "study" whether an At Large should even exist; on the other, it was to "forge consensus" for the conclusions it reached. But, realistically and politically, it would be impossible to forge a consensus under which there would be no At Large at all. As a result, the ALSC was, more than anything else, an exercise in diplomatic finesse -- which, perhaps, is why former Swedish Prime Minister Carl Bildt was chosen to head it up.

ICANN's priorities in this regard are perhaps best summarized in the way it allocated budgets to different aspects of the At Large question. When it first undertook the actual global election of the five At Large directors, it asked the Markle Foundation for a measly $100,000 (Markle prudently doubled this amount); but in ratifying the ALSC, ICANN gave it a budget of $450,000. Granted, the learning curve separating these two budget allocations was steep, involving, as it did, the actual execution of a worldwide election of unknown scope. (Initially, ICANN had guesstimated that between 10,000 and 20,000 people would sign up for the At Large; more than 158,000 signed up, of which 34,000 actually cast votes.) Neverthless, if conducting a study of the At Large costs a half a million dollars, one can only wonder how much it would cost to execute another round of global At Large elections.

The ALSC can hardly have failed to notice such a big-ticket item. Nor, given the participation of former ICANN directors Esther Dyson and Pindar Wong, can they have failed to consider the fact that ICANN spent its first few years in debt and on the brink of financial chaos. ICANN's financial woes were resolved only when it began the process of soliciting proposals for new top-level domains (TLDs), for which it charged (through some creative accounting); but Roberts's replacement as CEO, Stuart Lynn, recently told CNN that the new TLD process may well come to an end. Thus, to the ALSC's untenable diplomatic mission we might well add completely untenable financial constraints.

In fact, there are many more problems the ALSC faced, notably logistics. In executing the first At Large election, ICANN was very reasonably concerned that the potentially high stakes, which bear strong similiarities to spectrum auctions, would create a strong incentive for "capture" -- that is, concerted efforts to subvert the election process. One obvious way to accomplish this would have been enrolling fictitious people in the At Large electorate, which in a purely virtual environment could have been trivial. As a result, in the first round of elections, ICANN insisted on including a "snail mail" authentication component: sending the voter's PIN via a physical letter. Unfortunately, ICANN's method of doing this (an envelope letter rather than a postcard) was needlessly expensive, and added yet another source of difficulty for voters in less-developed and/or countries (like China) where Latin scripts are unfamiliar.

Obviously, this litany of problems -- and these are only a sample -- is staggering enough to invite some seriously "out of the box" thinking about how to effect broad, public participation in ICANN. Unfortunately, the ALSC report's lines of inquiry and analysis were thoroughly and neatly boxed up: unimaginative, predictable, and above all sloppy. Moreover, the report relies much more extensively on ICANN's staff than on "outreach", let alone on independent reporting. Three of the ALSC's four "meetings" took place at ICANN meetings, and the ALSC describes the remaining twelve chances for face-to-face input in other contexts thus:

"You may catch Committee Members for a chat at..."

The result? A report that offered up, among other thing, the following "compromise" recommendation:

The ALSC carefully considered, but ultimately rejected as having an unsound logical basis, the notion that users' interests are best served by giving At-Large half of the Board seats. A functional approach makes more sense considering ICANN's evolution and responsibilities, and it has the greatest likelihood of achieving a consensus. Reserving six seats for At-Large Directors enables users to have a proportional role but also prevents any one of the constituency groups from exercising undue influence within ICANN.

...based on an entirely innovative -- and imaginary -- new division of the ICANN eighteen-member board into three groups of six representatives one for the At Large, one for "developers," and one for "providers." The problem? The ALSC implicitly acknowledged that giving normal internet users nine seats on the board would hobble ICANN's "consensus"-development process; in other words, that users probably wouldn't support ICANN's typical "consensus" positions. And further, while advocating diminishing the At Large directors from nine to six, the ALSC didn't also suggest changing the existing mechanisms for electing other directors. Clever, that.

Why ICANN would go so far out if its way to minimize the possibility for open discussion of such a banal report -- by substituting a "security"-related agenda rather than augmenting the existing agenda with parallel sessions, by diddling the dates of the Marina Del Rey meeting in announcements, and so on -- is anyone's guess. But they did.


It's one thing to hijack the agenda of a public meeting, but quite another to actually deliver the goods, and on the latter front ICANN failed miserably.

Some of the speakers, notably AT&T's Steve Bellovin and security guru Bruce Schneier -- were variously a little entertaining and a little informative. There were moments of unplanned comic relief, for example, when the keynote speaker much-touted by ICANN, Kenji Kosaka (Senior Vice Minister for Public Management, Home Affairs, Posts and Telecommunications of Japan) closed his vacuous speech with the naive declaration, "Long live internet! Long live ICANN!" There were some riveting moments of drama: for example, a series of not entirely collegial exchanges between weighty attendees -- "DNS inventor" Paul Mockapetris, BIND author Paul Vixie, and venerable net-hand Randy Bush -- on the one hand, and ICANN board chairman Vint Cerf, on the other. There were some dodgy moments, given the ostensible context: for example, when DNS security honcho David Conrad basically explained how to bring the DNS system to its knees. And there were even some moments of cruel irony: for example, despite Randy Bush's exemplary warnings about using unsecure protocols over the wireless network (he read numerous passwords aloud during public exchanges), echt boardsquatter Hans Kraaijenbrink -- by far the most outspoken opponent of At Large "lusers" -- repeatedly demonstrated his 31337 technocrat status by checking several email accounts using plaintext POP.

Ultimately, though, these isolated "moments" were sideshows to what remains ICANN's core business, if not its core competence: "technical coordination" of domain names and IP numbers (and protocol parameter and port numbers, not that ICANN has even mentioned it). And if ICANN had to scramble to address the issue of "security," it's only because they have neglected their mandate in favor of the political opportunism of catering to rabid intellectual property lobbies.

Remarkably, the board's discussion at the tail end of the meeting turned out -- despite the thinning crowd -- to be something of a salvation.

At Large director Andy Mueller-Maguhn led a series of aggressive challenges to pivotal points in the summary resolutions handed to the board, which for the most part functions as a passive rubberstamp for ICANN staff's initiatives. In a particular surprise, Mueller-Maguhn challenged the presence of WIPO honcho Francis Gurry on an ICANN committee ostensibly devoted to "internationalized" -- that is, non-Latin -- domain names. Why, he pressed, should ICANN both nominate and subsidize intellectual property interests in this area? Or, alternatively, why not do the same for free-speech advocates? After much back-and-forth, an upset vote on a move to depose Gurry from the list (6 for, 6 abstentions, and 5 against) drew surprised gasps then wild cheers from the audience. In a way, it was a pyrrhic victory: board chairman Vint Cerf hastened to add that anyone -- Gurry of course included -- was welcome to participate. Nevertheless, the vote demonstrated (and the crowd's visceral reactions confirmed) what a vital role At Large representation on ICANN's board can play in representing broad social interests in the face of organized industry pressure.

But if the irritable response of ICANN's engineer and external counsel, Joe Sims, was any indication, it was the vigorous debate about the ALSC report that struck hardest. Again, it was Mueller-Maguhn who spearheaded the challenge, this time supported by two other At Large directors, Ivan Moura Campos and Karl Auerbach. The deceptively banal wording of the board resolution on the report (very likely crafted by Sims) was a model of ICANN's "pro-active" sophistry: both "accepting" the report and directing CEO Lynn to continue working with the ALSC to begin "continency planning" for the "possibility" of an "option" to act on the ALSC's recommendations, the resolution was a have-your-cake-and-eat-it-too ambiguous invitation to the kinds of abuses of that led to the ALSC process in the first place. The board's discussion of the resolution translated into only a handful of changes to the wording, but the sharp discussion made clear that the board -- concurring with almost everyone but Kraaijenbrink and Lynn, who both threw tantrums over the specter of board involvement in further At Large developments -- viewed the ALSC report as deeply unsatisfactory and, at best, a provisional basis for further work.

Before night had fallen, the ALSC website was pushing a press release headlined "ICANN BOARD AGREES WITH AT-LARGE REPORT PRINCIPLES, MOVES TOWARD FINAL ACTION."