Summary

  • As ALAC liaison to the GNSO Council, Alan Greenberg transmitted the committee’s request that ICANN staff prepare an Issues Report on domain tasting. ALAC, not Greenberg alone, made the request.
  • The report created a formal opening for investigation. Staff, the GNSO Council, public contributors and the ICANN Board then made separate decisions; the request was neither a policy mandate nor proof of a global user electorate.

A request in the liaison’s inbox

“The At-Large Advisory Committee is requesting that ICANN staff prepare an Issues Report on Domain Tasting.” Alan Greenberg’s 8 May 2007 email to the GNSO Council was brief. Its procedural point was unusually clear: the Council did not need to act yet. Once staff delivered the report, Council members would decide whether to begin a Policy Development Process.

That sentence is the useful unit of analysis. Greenberg was not announcing a policy, claiming a vote on behalf of Internet users or asking the GNSO to adopt ALAC’s preferred remedy. He was carrying a decision made by the At-Large Advisory Committee into the body that could examine the issue. ICANN’s announcement calls it the first Issues Report request made by any ICANN advisory committee. The bylaws supplied the route; ALAC used it; Greenberg served as its liaison.

The public record differs by one day. The GNSO announcement and archived email are dated 8 May; the June Issues Report says Greenberg notified the Council on 9 May. The email is timestamped 23:46 on 8 May at UTC−4, which is already 9 May in UTC. That is consistent with the difference, although the report does not spell out its date convention.

What the five-day window made possible

An Add Grace Period, or AGP, let a registrar reverse some new gTLD registrations and receive credit, typically during the first five calendar days. The mechanism had a legitimate purpose: correcting accidental registrations. Domain tasting used that same window differently. A registrar could register a large batch of names, test whether traffic or pay-per-click advertising might make them profitable, keep the winners and delete the rest before the fee applied.

The design problem was not simply that names were deleted. The zero-cost test could shift costs onto registries, users and the wider naming system, while the correction mechanism remained useful to ordinary registrants. Staff’s Issues Report framed the question for investigation; it did not establish that every AGP deletion was abusive or choose a single cure. That distinction matters because the tool being debated also protected legitimate error correction.

A report was a gate, not a verdict

Greenberg’s email spelled out what the request could—and could not—do. Staff had to prepare the report. The GNSO Council still had to decide whether a PDP was warranted. The Council formed an ad hoc group to gather facts, postponed its decision on 11 October 2007, and initiated the PDP on 31 October. A request had moved an issue into a documented channel, but it had not compelled the next step.

The later record is institutional rather than personal. Staff prepared reports and summarized comments. Constituencies and members of the public disputed both the problem’s scale and the appropriate remedy. The Council’s final report was dated 4 April 2008; on 17 April, the Council adopted a recommendation to restrict AGP refunds above defined thresholds and monitor implementation. On 26 June the ICANN Board adopted that recommendation and directed staff to implement it after further notice and comment.

There was also an interim lever. A separate FY2009 budget provision began charging a US$0.20 transaction fee on certain excess AGP deletions before the consensus policy was implemented. ICANN later reported that monthly AGP deletions fell from about 17.6 million in June 2008 to 2.8 million in July. That is a reported change after the budget provision, not evidence that Greenberg’s email alone caused it or that the later AGP Limits Policy had already taken effect.

The influence—and limit—of a liaison

A liaison can make an institution’s decision legible to another institution: who decided, what was requested, which rule allows the request and what decision remains with the recipient. That is meaningful procedural influence. It can place an issue on a formal work path, preserve a record and require an accountable answer. It is different from command, authorship or electoral authorization.

The domain-tasting episode therefore offers a more precise picture of advisory power than either “ALAC decided the policy” or “the liaison merely forwarded an email.” The committee used a bylaw-recognized request; the liaison transmitted it; the Council retained discretion; the policy process collected competing evidence; and the Board made the final institutional decision. A channel can affect what an organization examines without giving the channel’s participants authority over everyone affected.

That boundary is the point. Participation becomes consequential when it changes the route by which a question is considered. It does not, by itself, prove that participants represent all users, that their preferred answer should prevail or that one visible intermediary owns the result. The strongest account of Greenberg’s role is also the narrowest: he carried ALAC’s request into the GNSO record, and the record shows what happened next without turning that act into a mandate.

Sources