While this may be better suited as a comment to the CENTR posting, I thought that its length might warrant a separate submission. Many thanks to CircleID for hosting this interesting discussion. Below is the full text of a comment I forwarded to CENTR earlier today in reply to its commentary on my recent study on national governments and ccTLDs.

Michael Geist

Date: Fri, 9 Jan 2004 10:13:48 -0500
To: centr.org
From: Michael Geist <>
Subject: CENTR Response - Geist Reply

Dear Mr. Mertens,

Thank you for forwarding CENTR's comments on my preliminary study on governments and ccTLDS. I think that greater discussion and debate are a good thing and I am grateful that CENTR has taken the time to read my work and provide some feedback.

Having carefully considered CENTR's comments I thought it would be appropriate to provide a few short responses. I should also preface my remarks by noting that I speak for myself -- not the ITU (see below), nor CIRA, (a CENTR member ) on which I serve on the board of directors, nor the Public Interest Registry, which manages the dot-org domain and on which I serve on the Global Advisory Council.

The CENTR response raises many issues that I believe can be distilled into five issues:

1. A misunderstanding about the relationship between my work and the ITU.
2. Concerns about the claims in the study.
3. Concerns about the survey response rate.
4. Concerns about one response from DENIC being ignored.
5. CENTR comments regarding conclusions in the study.

Let me address each in turn.

1. While the CENTR response suggests that the ITU is “awarding authority” for the study, this is simply not the case. I have not been paid anything by the ITU nor has the ITU sanctioned my work. As indicated clearly at the very outset of the study, the views are my own and do not reflect the views of the ITU.

This project was an extension of the ccTLD Governance Project that I launched months before the survey project began. The relationship with the ITU is properly characterized as one of cooperation for data collection purposes. The ITU agreed to use its large network of member states to distribute the survey and collect the data. We agreed to share the data to be used in our own respective ways. To that end, I understand that the ITU has published all of the responses online so that the survey is fully transparent. My approach was to issue a preliminary study with a fuller analysis to follow in 2004.

2. I appreciate the CENTR’s concerns that the study accurately reflect the results. Respectfully, I believe that the study does present an appropriate representation of the data. First, the title of the study, “Governments and Country Code Top-Level Domains: A Global Survey” is, in my view, an accurate description of a study that went out to 189 ITU member states (surely global) and includes nearly 60 responses from every region and every level of economic development. Further, as CENTR itself points out, the study includes a note of caution that non-response bias should not be ruled out.

Second, the survey consistently limits its conclusions by noting its application to those countries that responded. For example, in the conclusion cited by CENTR that “Governments are deeply involved in domain name administration at every level” the paragraph continues by stating that “Contrary to most expectations virtually every government that responded either manages, retains direct control, or is contemplating a formalized relationship with their national ccTLD” [emphasis added]. This limitation is included throughout the study as I consistently refer to “responding governments” rather than all governments.

3. With regard to the survey response rate, there are several points worth noting. First, I was very anxious to obtain as large a response rate as possible. I believe the ITU was as well. The survey was in fact distributed not once, but twice, to provide for a greater response. The ITU has continued to receive additional responses, including one from the UK, all of which will be included in my final study.

Second, I think the percentage described as a function of ITU members rather than total ccTLDs is the correct approach since the ITU distributed the survey to ITU members and not directly to the ccTLDs.

Third, given the cross-section of survey respondents from North and South America, Europe, Africa, and Asia, I believe the data does provide a very broad sampling of global approaches. The cross-section is certainly broader than CENTR’s own membership and also includes major ccTLD countries such as Australia, Hong Kong, South Korea, along with many African and South American countries that are not CENTR members.

Fourth, I find CENTR’s suggestion that the survey is unrepresentative particularly curious given the strong showing of the governments of CENTR’s own membership. In fact, with the recent addition of the UK, Iran, and Portugal (all posted on the ITU site), 54% of CENTR’s member and associate member governments actually did respond to the survey, suggesting that the data could potentially be even more representative of CENTR countries than of the rest of the world.

4. Contrary to some suggestions in the CENTR response and elsewhere, there was nothing sinister about the failure to include a response sent by DENIC in mid-June. Rather, its non-inclusion is the result of an oversight and a failed path of exploration. The survey was sent by the ITU to its member states in the spring. Given that I had previously contacted many ccTLDS themselves as part of my governance project, I was anxious to connect with them again with this survey. In fact, I thought it would be very interesting to compare the responses of national governments with those of their ccTLD administrators (where they are separate entities) to see whether they were consistent. Accordingly, an email attaching the survey was sent to all participants in the earlier ccTLD Governance Project as well as clarification to the ccTLD Discuss list. I also discussed potential participation with several ccTLD managers at the ICANN meeting in Montreal. Unfortunately, only one response - DENIC’s - was received to this separate call for responses and this line of investigation was regretfully abandoned.

When it came time to collect and assess the data, I must admit that I forgot about the DENIC response since my attempt to obtain a parallel set of data had failed. Had I not forgotten, I would have certainly referenced the DENIC response within the study and will do so in the final version. Unfortunately, the first time anyone ever mentioned the oversight was in the CENTR response—DENIC itself never followed up with any contact which would have allowed for the matter to have been quickly clarified.

5. The CENTR response suggests that I imply that the Internet functions particularly well as a result of strong government influence. I do not believe I say or imply any such thing—if readers believe that is the case, then perhaps the data is speaking for itself.

I ask that you please forward this email to the CENTR list so that the CENTR community can benefit from the discussion.

Sincerely,

MG