A study by Oliver Klug, Patrick Lutz, Anja Seidel, Alexander Svensson (University of Hamburg)
English summary by Alexander Svensson
May 2001

In the course of the year 2000, ICANN has developed from an organisation known only to few specialised journalists to an institution whose decisions are mentioned in the European mainstream media. Two main events contributed to this development: The election of five ICANN At Large directors in a process which began in February and ended in October 2000, and the decision to introduce new generic Top Level Domains (TLDs) at the ICANN meetings in July and November 2000.

In the beginning of 2001, an At Large Membership Study Committee (ALSC) was set up to study the At Large membership concept, structure and processes. So far, only a minute fraction of the 158,000 persons registered as At Large members have participated in the ALSC online forum. The importance of the study process should however not be underestimated, since the possible recommendations to the ICANN board range from another online vote for the four not yet elected At Large directors to an abolishment of the At Large directors' seats.

24.8% of the European At Large members have stated that they first heard about the elections directly via "Newspaper/Magazine" or "Print Media"; this does not even include the members which were reached indirectly by the media coverage -- 13.3% heard about ICANN's election from friends or acquaintances, 21.6% via e-mail (ICANN ALSC 2001a). The role of the media has been viewed critically. According to the German news magazine Der Spiegel, ICANN's Chief Policy Officer Andrew McLaughlin has called the media reporting "irresponsible" and claimed that thousands have been misled (31 July). ICANN press spokesperson Pamela Brewster explained the surprisingly high number of registrations by saying: "In Europe they think ICANN is the U.N. of the Internet" (ZDNet 2000).

Studying the European media coverage shall therefore contribute information about the perception of ICANN and the At Large elections in Europe. The study is based on the print media coverage: The special significance of the mass media lies in its reach and the mass impact of its perceptions and interpretations on the general public. The choice of print media as opposed to radio and television is due to the fact that print media coverage is more readily and completely available. The study comprises all articles from a range of high-circulation newspapers and magazines in the United Kingdom, Ireland, France, Belgium, the Netherlands, Germany, Austria, Switzerland, Denmark, Norway, Sweden, Spain and Portugal which contain the phrase "ICANN". The articles were obtained from Web archives and, in some cases, media databases; a total of 310 articles from January to December 2000 was analysed for the study.

In most cases, the first task of a journalist writing about ICANN was to explain what ICANN is and does. This is of special interest as it conveys implicit assumptions about ICANN's importance and future role. ICANN's self-description on its Web site is "a technical coordination body for the Internet". Former ICANN Chairman Esther Dyson has emphasized that ICANN in effect "governs the plumbing, not the people" (Dyson 1999). Critics like David Post have countered that control over the DNS is "quite literally a kind of life-or-death power" (Post 1999).

To gauge the perception of ICANN's role, the main description used in each article was classified into eight categories: 1. management/ administration/ authority, 2. government, 3. "government" (i.e. some form of relativisation), 4. government authority/office/agency, 5. Internet organisation, 6. standards-setting body, 7. company, 8. non-profit, 9. other. In a number of shorter articles, ICANN's role was not explained at all, instead the abbreviation was used or the organisation's name was simply written in full.

Of these classifications, "government" was the one most frequently used (11.6% of all cases). This classification was applied if the term government was used with some form of relativisation, e.g. "inoffiziell 'Internetregierung' genannt" ("inoffically dubbed 'Internet government'"; Der Standard, 12 October). In another 5.5% of the cases, the term was used without relativisation, e.g. "il governo della Rete" ("the government of the Net"; L'Espresso, 19 July). Interestingly, these descriptions have not been used mainly for brief reports; on the contrary. In almost all the cases, journalists using one of the two terms have written medium or long articles.

But also the terms implying a more subordinate role have been used quite often. Especially in the German language newspapers, the term government authority/office ("Behörde") has been used in 11.0% of the cases. The classification with slightly less emphasis on government, management/administration, is used in 9.4% of the articles. Even the very general descriptions as "Internet organisation" or "non-profit" are used in 10.3% and 11.3% respectively. ICANN is seldom described as standards-setting body (3 cases) or as company (7 cases).

The European media are evidently varying in their perception and description of ICANN's role. The self-description -- "technical coordination body" -- is almost not present in the media coverage. Some of the articles turn to the discrepancy between the self-description and attributions by others and consequently emphasise the political, economical and social consequences of ICANN's decisions. Quotes like the above-mentioned by David Post (quoted e.g. in Der Spiegel, 31 July), but also statements by Center for Democracy and Technology (CDT) representatives such as Alan Davidson (Le Monde, 28 June) and Jerry Berman (Libération, 7 July) are used as a counterbalance. The technical self-description is even seen as veiling by some: "Wer so zungenfertig die eigene Macht kleinredet, der fürchtet fremde Ansprüche." („Someone who plays down his own power so eloquently is afraid of the demands of others"; Frankfurter Allgemeine, 17 June). In its review of the elections, The Economist presents both the view of ICANN as technical and administrative body and of ICANN as a virtual world government in the making and decides: "The truth, as so often, lies somewhere in between." (30 September).

The ICANN At Large elections were a long drawn out process. Voter registration began in February and ended on the 31st of July. In the meantime, on the ICANN public meeting in Cairo (7--10 March), the elections were fundamentally changed. The ICANN board agreed to hold direct instead of indirect elections, but only five of the nine board seats reserved for ICANN At Large directors were to be filled in 2000, and a post-election study should evaluate the results.

At the end of the voter registration phase, a Nominating Committee appointed by the ICANN board presented a list of candidates for each of the five electoral regions (Africa, Asia/Australia/Pacific, Europe, Latin America/Caribbean, North America). In a second stage, ICANN members could offer themselves as candidates and get on the ballot, if they could demonstrate sufficient support from their electoral region. ICANN had limited the number of candidates per region to seven, and as the committee had nominated five candidates for the European region, 78 persons were competing for only two free slots on the final ballot for Europe.

Andy Müller-Maguhn and Jeanette Hofmann, two German candidates, emerged as winners from the member-nomination phase. The final election with all seven candidates for Europe took place in the beginning of October. Müller-Maguhn won 52.6% of the votes from the European region, the other member-nominated candidate came in second.

During the days before voter registration ended, the server used by ICANN suffered from overload. Statistics published by ICANN show that over half of the registrations took place in the final ten days before the deadline (ICANN ALSC 2001b). Consequently, the point in time when the media started to cover the ICANN At Large elections had a considerable impact: Early coverage in one country could easily lead to a snowball effect, coverage shortly before the deadline to a higher number of people unable to register because of technical difficulties, while coverage starting after the 31st of July did not impact membership registrations at all.

There are two peaks in the coverage of the election over time: The first one in July, the month of the registration deadline, the second one in October, when the elections took place and the results were announced. Only 17 articles mention the At Large elections before July -- even though voter registration had started in February and the election procedure had been decided upon in March. The relative amount of election coverage compared to other topics increased over the year, until the introduction of new Top Level Domains became a more important issue in the media coverage in October/November.

To examine the European print media's assessment of the At Large elections, all articles were analysed for positive judgements, negative judgements, neutrality and mentions of both positive and negative judgements, all regardless of whether they were direct judgements or quoted judgements.

44 of the articles contained positive judgements about the election in general, 46 were neutral, 9 articles contained both positive and negative judgements. Only one article, a commentary in an Austrian newspaper, contained only negative comments.

The implementation of the election was however judged less positively. Of those articles mentioning the course of the registrations, there were 22 negative ones and 2 with both positive and negative judgements. Similarly, the total number of participants was judged positively in 6 articles, both positive and negative judgements were found in 5 articles, only negative ones in 29 articles. Many of the negative judgements appeared in July and October and involve a comparison between the number of membership registrations or voters and the millions of Internet users worldwide.

Even though the DNS management task is a global one and the electoral regions spanned whole continents, the media coverage had noticeably varying focuses in the different countries. There were however no nationalist tones in the articles reviewed.

(Only medium and long articles mentioning an issue as part, main or exclusive topic; multiple mentions possible)

a. Presentation of national candidates
In the member-nomination phase with its 78 mostly unknown candidates in Europe alone, several newspapers chose to present candidates from their own country: The Swiss SonntagsZeitung listed three of the Swiss candidates (27 August), La Repubblica's web site contained links to all the Italian candidates, the Frankfurter Allgemeine ran portraits of the three German candidates Müller-Maguhn, Hofmann and Zerdick. The portrait's author explained: "Ein deutscher Direktor ist also allein schon deshalb denkbar, weil die Kandidaten so namenlos sind, dass die Wähler im Zweifelsfall ihrem Landsmann die Stimme geben." ("A German director is thus already conceivable merely because the candidates are so nameless that, if in doubt, the voters will cast their vote for a fellow countryman."; 26 August).

b. United Kingdom: ICANN's role in dispute resolution
ICANN's domain name dispute resolution policy for trademark owners features prominently in the British media. 49% of the 51 articles covered this aspect at least marginally, in 16% of the cases it was even the main topic. The question of trademarks and domain names was exemplified by the case of the well-known writer Jeanette Winterson. She used ICANN's Uniform Domain Name Dispute Resolution Policy (UDRP) against a student who had registered the domain jeanettewinterson.com. The case, which was decided in favour of the author, was reported in the UK media, and Jeanette Winterson herself wrote two articles about "cybersquatting".

The At Large elections were discovered late by the British media. The Daily Telegraph regretted that "British web surfers appear apathetic about the future of the internet" (17 August) and pointed to the small number of At Large registrations from the UK. When this admonishment was printed, the deadline had already passed.

c. France: Focus on the election
The French media coverage was remarkably different from the British: Trademark domain name disputes hardly played a role (they remained unmentioned in 13 of 15 reviewed articles), but the elections were covered intensely, and the organisation was portrayed in several articles. Le Monde made the start before the registration deadline by calling the "cybercitoyens" to the ballot boxes to vote for "une instance de régulation qui, pour la première fois, pourrait casser le monopole américain" ("a regulatory authority which could break the American monopoly for the first time"; 28 June). Both Libération and Le Monde ran specials, covering different aspects of ICANN in several articles.

d. Germany: Extensive coverage
The number of in-depth articles about ICANN in Germany was considerably higher than in the rest of Europe. In mid-2000, all nationwide newspapers wrote at least once about ICANN; in July, the coverage reached its climax when five articles appeared in the Frankfurter Allgemeine within seven days. The beginning and the end of the final voting phase was even reported in the widely viewed evening news programme Tagesschau. Nevertheless, most articles also contained a short description of ICANN's responsibilities. For details on the Spiegel Online registration drive see below.

e. Italy: The own role within ICANN
The Italian media also managed to find national starting points for reporting about ICANN. Shortly before the deadline, L'Espresso called "Italiani, votate il governo della Rete" ("Italians, vote for the government of the Net"; 19 July). This call for participation contained a statement by a representative of an Italian Internet Service Provider (ISP) that Italy should show its presence. La Repubblica also reported about the elections before the registration was closed. It can be reasonably assumed that the activities of the Italian media -- partly initiated by an Italian member of ICANN's volunteer Membership Implementation Task Force (MITF) -- contributed to the large number of last minute registrations from Italy (from 160 on the 13th to 1670 on the 31st of July).

f. Scandinavia: Little media interest
Only a day before the registration deadline, the Scandinavia Online magazine of the Norwegian Aftenposten pointed to the fact that "verdens Internettregjering" ("the Internet government of the world"; 30 July) would close its registration on the following day. In the small number of other articles about ICANN, mainly short reports about new Top Level Domains, the elections remained unmentioned.

In Sweden, the high circulation dailies Aftonbladet and Expressen reported in November about ICANN, again because of the introduction of new Top Level Domains. In Denmark, it was curiously the Berlin correspondent of the newspaper Politiken who reported about the ICANN elections, mainly quoting the German candidate Jeanette Hofmann. For the Danish Jyllands-Posten it was again the foreign correspondent in Germany reporting about Andy Müller-Maguhn's victory. Despite the high number of Internet users in the Scandinavian countries and the fact that a Norwegian candidate, Alf Hansen, was on the ballot, the amount of media coverage was remarkably low.

Already in the planning for the At Large elections, ICANN's Membership Advisory Committee feared capture by special interests (ICANN MAC 1999). The committee's report from March 1999 did not mention the possibility that stronger media interest and activity in one country may lead to a distortion of the relative participation within an electoral region. The following section covers the Spiegel Online initiative "I can! -- eLection 2000" that started with the headline "Aufruf zur Wahl der 'Regierung des Internets'" ("Call for the election of the 'government of the Internet'"; 2 May).

a. Initial situation
On the day the Spiegel Online initiative started, only 14,167 persons had signed up for the At Large elections. The first statistics showing the distribution of registrations among countries were published twenty days later, on the 22nd of May. At this point, 16,919 persons were registered worldwide, including 4,107 from Germany and 6,915 from the United States. The remaining 5,897 registrations were distributed over 141 countries, largely from the OECD. The number of registrations from Germany were thus proportionally high already in the early stages of the election; the relative share of worldwide registrations dropped from 24.3% (22 May) to 12.9% (31 July).

b. Candidate selection and reporting
The Spiegel Online initiative consisted of two elements: a procedure, according to which readers should propose candidates, and a special page within the Netzwelt section of the Spiegel Online web site with a call for participation. The original plan to collect and present candidate proposals, to organise an preliminary online election and to submit them as a petition to ICANN, was dropped later. But the search for suitable candidates had begun and drew additional attention to the At Large elections. Shortly before the registration deadline, the German Federal Education and Research Ministry suggested to organisations and associations by e-mail to nominate their own candidates.

The special ICANN coverage was at least equally important: Continuously, articles about the organisation were added to the special sections of Spiegel Online and Heise Online. Until the registration deadline, 31 articles about ICANN appeared in Spiegel Online, 20 were published in Heise Online.

c. Reactions outside Germany
The Spiegel Online initiative was not only judged positively. The Economist wrote: "What critics fear most is that the membership […] will be captured by special interests. To some extent this is already happening: Because German magazines and newspapers such as Der Spiegel and Die Zeit are running a voter-registration campaign, a fourth of the membership is currently German" (10 June). Except for the Economist article, the Spiegel Online initiative was seldom mentioned in print media outside Germany (four marginal mentions).

In fact, the print media Der Spiegel and Die Zeit did not participate in the initiative, only their online equivalents. Not all the online media that supported the initiative put up their own web pages about ICANN. In the end, the ICANN coverage in Spiegel Online led to an increasing interest, but the pages about ICANN fell considerably short of the usual access figures for one of the largest German online news sites.

Outside Germany, the online magazines of Le Monde (France) and Der Standard (Austria) participated in the initiative. Many other European online media, among them sites from the United Kingdom, Italy and Spain, were invited before the start, but did not choose to join. Responsibles at Spiegel Online assume that it has been partly a problem to assign ICANN issues to an editorial department, since it is a mixture of political, economical and computer issues.

d. Direct and indirect consequences
The high participation of German Internet users cannot be directly attributed to the Spiegel Online initiative: After the critical Economist article, the extent of the coverage was reduced, whereas the highest increase of registrations took place in the fourteen days before the registration deadline. This was the time period when registrations greatly increased in a number of countries, even though the ICANN registration server problems did not allow everyone to register.

The involvement of Spiegel Online and the printed Der Spiegel has nevertheless been far from ineffective. In March, an article about ICANN in the print edition of Der Spiegel was among the first to make the organisation know to the German general public. Additionally, Der Spiegel has a considerable readership among other journalists, so the article titled "Herrscher des Netzes" ("Ruler of the Net"; 13 March) set the agenda for subsequent media coverage. The difficult task of the volunteers of ICANN's Membership Implementation Task Force -- making ICANN and the elections known -- was in Germany taken on early by the media.

The ICANN At Large elections have generally received a positive media reaction in Europe, while their implementation and the number of participants have been criticised. In Europe, the amount and focus of the print media coverage varied, but more importantly, not all the reports about the At Large election appeared before the registration deadline. Judging the direct media impact in the various European countries would require data from ICANN's membership database to a degree that the amount of registrations from a specific country can be more accurately matched to registration drives or media attention.

The study of the media coverage also shows that ICANN is not viewed as solely technical body; instead, many articles focus on the non-technical consequences. The fact that the bold term "Internet government" is used even in in-depth articles highlights this discrepancy. However, in the reviewed articles there were no calls for an enlargement of ICANN's role to other areas of Internet governance. ICANN is associated in the European media mainly with two developments: the introduction of new Top Level Domains and the global online elections. Both have led to an increased media attention and close observation of its decisions. From ICANN's point of view, the perceived importance of its tasks may make a reduction, let alone an abolishment of the elections a dangerous venture unless it can convince the public that it corresponds to a reduction of the organisation's importance. If, on the other hand, the organisation chooses to hold another election, it would be well advised to inform the media directly in order to avoid considerable imbalances due to more or less successful volunteer efforts to draw media attention to the elections.