'Governance of controversial Internet content in the European Union.'
In this chapter I will examine the emerging policy consensus at a European level on controversial Internet content. The impact of the policy making process on the final outcome will be examined along with the substantive policy recommendations. The latter will be critically evaluated as regards the feasibility of their enforcement within the infrastructural parameters of the Internet. I will conclude with a discussion as to whether these policies entail a novel pattern of governance.
All of the key policy statements under review are premised on the distinction between 'illegal' and 'harmful' content and go on to outline the types of content deigned to belong to these categories. This distinction has a large impact on the outcome in terms of the resulting emphases and recommendations. The Communication maintains that the illegal/harmful distinction is 'crucial' as:
These different categories of content (illegal and harmful) pose radically different issues of principle, and call for very different legal and technological responses. It would be dangerous to amalgamate separate issues such as children accessing pornographic content for adults, and adults accessing pornography about children.
The Communication later concludes that:
The main weapon for dealing with harmful content is in ensuring that practical means are available to limit access by the vulnerable to such content.
It would seem from this that what the Commission has in mind when speaking of harmful content is content harmful to the development of minors, for instance sexual content, which should not be universally restricted. However, this is at variance with the implied definition of harmful content elsewhere in the same document. Under the rubric of 'Harmful Content' a broader definition emerges:
Various types of material may offend the values and feelings of other persons: content expressing political opinions, religious beliefs or views on racial matters etc. What is considered to be harmful depends on cultural differences. Each country may reach its own conclusion in defining the borderline between what is permissible and not permissible. It is therefore indispensable that international initiatives take into account different ethical standards in different countries in order to explore appropriate rules to protect people against offensive material whilst ensuring freedom of expression.
This confusion in defining harmful content has the effect of projecting a broad category, yet one which in practise is reducible to children accessing adult material. Other possible forms of 'harmful' content are conveniently dropped early in the discussion, somehow devalued through the Commission's consigning of them to the domain of culturally relativist, subjective values somehow beyond the European Union's remit.
'Illegal content' is a similarly problematic category in European Union policy. This is so because of the diversity in the legal regimes of the member states. As the communication outlines:
The exact definition of offences varies from country to country. Within the EU, even child pornography, for example, where a high degree of consensus exists, is covered by specific legislation in some Member States and by more general rules relating to obscenity in others.
In view of this situation, the Commission stresses 'the fundamental importance of co-operation between the Member States in the field of Justice and Home Affairs, and that 'common definitions of what ought to treated as unlawful are needed' (Commission, 1996b). Since Justice and Home Affairs is subject to the intergovernmental regime of the Third Pillar, where unanimity is the order of the day, and given the sensitivities involved, such common definitions are likely to focus such areas of high degree consensus as child pornography. Illegal Content in any European regime is likely to be defined by the lowest common denominator.
If we turn to another area of controversial content where there is less consensus - racism related content - we see it fall between the two stools of 'illegal' and 'harmful' content. This is despite serious concerns about racism on the Internet, as exemplified by Michael Whine in his survey (1997) of its use by far right-wing extremists. He expresses concern that such extremists are using the new medium to avoid legal sanctions against racist propaganda in many countries, offering them 'undreamed-of possibilities'(1997: 209). Holocaust denial material now has its own newsgroups, he notes, and racist video games are being promoted (such as 'concentration camp manager'), some of which have fallen into the hands of children. (1997: 214, 220). Whine desires a more proactive approach by concerned authorities, citing the International Covenant on Civil and Political Rights and the Convention on the Elimination of all Forms of Discrimination as relevant instruments (1997:227). While the Communication's definition of harmful content seems to include such content, it is soon eclipsed by the reduction of the harmful category I have outlined above. As a potential area of illegal content, racism related content displays huge diversity in terms of the legal regimes of the member states. For example publication of Mein Kampf or 'revisionism' (ie holocaust denial) are forbidden in some member states, e.g. Germany, but not in others (Commission, 1996a: note 7). The Parliament, recalling in the Council's Resolution designating 1997 European Year Against Racism, lists racism, incitement to racial hatred or violence, and holocaust denial - among other categories - as problems which the European Union 'cannot stand aside from' (European Parliament, 1997: par J) and advocates the adoption of common legislation explicitly banning such content (European Parliament, 1997: par K). A long series of strong statements by the Parliament on racism related issues is invoked. The prospects for such common legislation are complicated not only by the exigencies of the Justice and Home Affairs apparatus, but also by the demand for homogenous regimes amongst diverse media.
Policy on controversial Internet content takes place in the regulatory cluster of policy making in the Union, as opposed to constituent, distributive and redistributive clusters. The typical style of policy making in this regulatory cluster is characterised by 'disaggregated decisions, legal or quasi-legal processes and specialised interests' (Wallace and Wallace, 1996: 446). According to Wallace and Wallace (1996: 449) the structure of community policy making was
'designed from the outset to disaggregate issues wherever possible, to disguise broader political issues, [and] to push issues down from ministerial confrontation to official engrenage within the hierarchy of committees which formulated proposals for ministers to approve and the parallel hierarchy of committees which co-operated in their implementation' (Wallace and Wallace, 1996: citing Wessels 1990, 1992)
According to the argument, the rhetoric of technocracy and rational administration has reinforced this tendency to depoliticize issues. Members of Council of Ministers, preoccupied with their national dossiers, cannot focus on the technicalities of policy. Moreover, the Union's framework reinforces a tendency towards sectoral policy communities. Wallace and Wallace conclude that the massive growth of European Union regulation in recent years, placing it at the core of the Union's policy making, 'partly reflects a recurrent coalition between the Commission as policy entrepreneur and industrial interests preferring to move regulation to the European Union level'. This conclusion is one which mirrors the analysis drawn from Streeck and Schmitter in the preceding chapter, authors who emphasise that 'by fostering a transnational system of organised interest formation, the Community and its Commission in particular would thus contribute to its own growth as a policy arena and executive body' (Streeck and Schmitter, 1991: 134-5).
These elements in the policy process can be seen at play where Internet content is concerned. Moreover, it can be argued that these elements are reinforced by the nature of issues in this policy area. The technical complexity involved enhances Commission technocracy, and entails the use of its pattern of co-opting external, commercial, expertise. Preoccupied Ministers may be presented with a fait accompli. Moreover, the crucial role of intermediaries (Internet Service Providers) provides further rationale for their involvement in policy development. This pattern is borne out by the decision of the Council of Telecommunications Ministers of 27 September 1996 to expand the Working Party on illegal and harmful content (from a collection of national delegations) to include representation of the relevant industry interests, including content providers, telecom/cable operators, access/service providers, a hardware manufacturer and representatives of users (Commission, 1996d). Many of the resulting report's recommendations are reproduced word for word in the Council of Ministers Resolution. A Commission official has confirmed to me that non elected civil servants have played in the policy's development. Notwithstanding the crucial situation of service providers in the Internet's infrastructure, one recalls the makeup of the Working Party when the Report asserts that 'co-operation from the industry and a fully functioning system of self-regulation are essential to limiting the flow of illegal content on the Internet.' Besides the momentum toward European level interest organisations represented by the Working Party itself, any system of self regulation in the Commission's vision would involve European level organisations of service providers and self regulation bodies (Commission, 1996c).
Another relevant element in the policy-making process is the potential obstacle represented by the difference in outlooks among the member states. We have already seen a degree of diversity in the relevant constitutional and legal provisions of the member states, albeit with a certain amount of common ground. A representative of Ireland on the Working Party has outlined to me how disagreement between members on substantive common legal instruments has led to an increasing emphasis on a policy which has received unanimous assent, namely the promotion of the Platform for Internet Content Selection (PICS). This is a technical standard which aims to restrict the access of minors to adult content. The promotion of this standard represents a new form of governance in the context of content regulation, I will argue below, and presents unique problems as regards implementation. But it is worth pointing out at this juncture how much this policy option is supported by the nature of the policy-making process I have been describing. It eclipses nationally based discord. It involves close co-operation between the commercial industry and Commission technocrats on a protocol which is presented as a package and is thus unlikely to be questioned in its details by the Council or the Parliament. Moreover, in its presentation as a technically complex mechanism ethical nuances are obscured. Significantly, PICS has been developed by an institution charged with developing web standards which has been sponsored by the European Commission.
The emphasis in this section will be on the emerging policy on illegal content as expressed in various documents of the Commission, the Council of Ministers and the Parliament. In Chapter 5, I will examine policy on 'harmful content' as represented by PICS, the most concrete single policy measure on which there is large policy consensus. The European Commission's policies are enunciated in the Communication (1996a), the Green Paper (1996b) and the Report of the Working Party' (1996c). The Communication is aimed at presenting proposals for immediate action to deal with both illegal and harmful content. On illegal content it calls for reinforced co-operation between member states in the Justice and Home Affairs context, an assessment of the need for a 'common European framework to clarify the administrative rules and regulations which apply to access providers and host service providers' (liability) and the encouragement of self-regulation. The Green Paper is framed with a longer term outlook and is a consultative document. The policy options considered more or less mirror those in the Communication, however, including: 'systematic exchange of information, joint analysis of national legislative provisions, establishment of a common framework for self-regulation, recommendations for co-operation in the field of justice and home affairs, [and] common orientations for international co-operation'. However the Green Paper focuses on raising questions as much as on narrowing down the options. Questions such as liability and the appropriate fora for international co-operation are left open.
The Working Party Report comes out strongly in favour of self regulation, perhaps reflecting the Working Party's makeup, as we have seen above. Representative bodies for industry and users are envisaged as a key element in any self regulatory system. Such a system would include:
A Code of Conduct for Internet service providers (access providers, host service providers and anonymous remailers),
A hot-line for complaints from the public, with appropriate safeguards against misuse, (and)
An independent self-regulatory body, including representatives of industry and users, to advise on whether or not a breach of the Code of Conduct has occurred (without prejudice to the due process of law) (Commission, 1996c)
.
The self regulation system, it is asserted, would need to be in compliance with and supported by the legal system. Failure to comply with the system rules can result in legal sanctions. Moreover, the code of conduct could be made more binding by member states through a requirement that it be formally approved, or integrated into contracts between network operators and service providers. The Report advocates European co-ordination of representative and self regulation bodies and asserts that the Commission should collaborate in establishing this and should contribute towards the initial cost. Co-ordination should include common standards for national codes of conduct and the establishment of a European network of hot-lines.
The areas of consensus in the above documents are reflected in the Council of Ministers Resolution politically agreed at the Telecommunications Council of the 28 November 1996 and formally adopted at the Council of 17 February 1997. The resolution welcomes the Working Party Report, and repeats many of its recommendations. It requests the Commission to 'ensure the follow-up and coherence of work on the measures suggested in the report, taking into account other relevant work in this field (the Communication and Green Paper) and to reconvene the Working Party as necessary to monitor progress and take further initiatives if appropriate'. Furthermore, it requests the Commission to 'foster co-ordination at Community level of self-regulatory and representative bodies and to promote and facilitate the exchange of information on best practice in this area.' Finally, it requests the Commission to foster research into technical issues, in particular filtering, rating, tracing and privacy-enhancing technologies, taking into account Europe's cultural and linguistic diversity and to consider further the question of legal liability for Internet content'. (Filtering and rating will be dealt with below in Chapter 5.) Liability remains an outstanding unresolved issue. The mention of privacy-enhancing and tracing technologies (originating in a recommendation in the working party report) may reflect an oblique acknowledgement of the thorny issue of encryption. Interestingly encryption provides not only for the evasion of regulation but also for secure business transactions, which impinge greatly on the European Union's project of developing the Internet as an engine of economic growth. The aspiration towards tracing technologies is thus aimed at reconciling material interests as well as those of combating crime, all in a pre-given context of Convention protected privacy rights for those who have legitimate needs for anonymity (e.g. sexual abuse survivors, those recovering from addictions, etc.) However, given the nature of packet switching and the availability of strong encryption, it is likely that tracing technologies will remain merely an aspiration. The Resolution also invites the Member States to take the following actions:
- To encourage and facilitate self-regulatory systems including representative bodies for Internet service providers and users, effective codes of conduct and possibly hot line reporting mechanisms available to the public.
Encourage the provision to users of filtering mechanisms and the setting up of rating systems (for instance the PICS standard launched by the international World-Wide-Web consortium with Commission support should be promoted).
To participate actively in the International Ministerial Conference to be hosted by Germany and encourage attendance by representatives of the actors concerned. (The Bonn Conference).
(Council of Ministers, 1997)
The Resolution further recommends that the Commission and the member states take all necessary steps to enhance the effectiveness of the above measures 'through international co-operation building on the results of the International Ministerial Conference and in discussions in other international fora'.
In response to the above policy statements, a Resolution of the European Parliament has been passed on the subject of illegal and harmful content. In addition to the parts addressed to the issue of racism, as outlined above, the resolution echoes the emphasis of the Commission and the Council on self regulation and filtering. A 'purely punitive approach', it is argued, would 'substantially undermine' the positive contribution of the new technology. The Commission is called upon to 'draw up a European quality rating system for providers of Internet services and to support international co-ordination of such ratings'. The Parliament's concern with the public interest is apparent when it states that Justice and Home Affairs co-operation under the TEU, being purely intergovernmental in character 'fails to provide for the democratic control and impetus which it is up to Parliament to provide in this area'. (European Parliament, 1997: par m.) As noted above, the Parliament calls for common European legislation banning racist content (1997: par K).
The emphasis on self regulation in the emerging policy reflects, in my view a realistic appraisal of the limits of a traditional government role in this area and an acknowledgement of the key place of 'intermediaries' in the governance of Internet content, given the unique infrastructure of the Internet as a communicative medium. Self regulation has other positive dimensions, as argued for instance by Clive Walker:
Self-regulation in this field has a number of advantages. Rules devised by the media are more likely to be internalised and accepted. In addition, it may avoid heavy-handed legal intervention which carries with it the spectre of government censorship (1996: 537-38).
As the Commission states, 'the issue of self-regulation and liability are... closely connected and need to be examined together' (1996c). As we have seen above, however, liability remains an unresolved issue which the Commission has been asked by the Council to consider further.
At times, the Working Party report seems unrealistic, for instance in seeking to apply a code of conduct to anonymous remailers. The whole point of anonymous remailers is that they are usually located outside of the relevant jurisdiction, thus allowing the user to benefit from Froomkin's 'regulatory arbitrage'. Thus I believe that European self regulation is likely to be bypassed in this respect, in the couple of seconds it takes to type a web address on a keyboard. Maybe it is thought that anonymous remailers operating within European jurisdictions may be restricted from facilitating the actions of their overseas clients. Such clients have the whole globe at their disposal, however. This is before we take into account the implications of the availability of encryption technologies which receive little attention in the documents under consideration here (Froomkin, 1997). Astonishingly, the word encryption is not actually mentioned in any of the key policy statements on controversial content. This may reflect the disadvantages of the highly sectoralised nature of policy-making in the union. Other policy sectors have been highly active in promoting the legal recognition of strong encryption as a means of guaranteeing European pre-eminence in the sphere of electronic commerce (Campbell, 1997c; Commission, 1997).
The Working Party report also seems to be unrealistic elsewhere, again treating the Internet as a sub-global infrastructure:
Where the (illegal) content emanates from abroad, they should pass information to the host country's self-regulator. They should also if requested transmit appropriate information to the police to allow them to fulfil their tasks, or to exchange information with another police force.
This is fine in a European context, and even in this context self regulation and police regulation is embryonic. What of illegal content originating in, say, Thailand? The obvious question of global structures for governance in this area is raised, one to which I will turn later.
The conclusion to the Commission's Green Paper contends that:
.. faced with the limits to purely national solutions and the difficulty and of devising and applying world-wide solutions, the European Union has a fundamental role to play. But the potential for transnational development of the decentralised services is such that common or at least compatible solutions are worth seeking in the European Union. (1996b)
This encapsulates an important issue to be faced. First and foremost, the obvious question that arises is whether actions at a European level do not suffer from the same limits as those on a national level. How can the Union's role be fundamental? Obviously, where offending material originates in a member state common European Union rules can facilitate action. Where material originates from outside the community it may be argued that any European action will have greater influence at international level than isolated national action, an argument made to me by the secretary of the Working Group. However any actions suffer the same limitations imposed by the decentralised nature of the Internet. The CompuServe case in Germany, among others, illustrates the process of regulatory arbitrage which would affect European enforcement as much as that in any member state. In the CompuServe case the public prosecutors considered that certain items available on newsgroups were illegal, and requested CompuServe to block access to these newsgroups. Since CompuServe's software did not initially make it possible to differentiate between German subscribers and others for access to newsgroups, CompuServe suspended access to a number of newsgroups to all of its subscribers world-wide. This action spawned wide-spread protests that German standards of morality were being exported. Subsequently, CompuServe restored access to most of these newsgroups except to its German subscribers. No action was apparently taken against other access providers based in Germany, so their subscribers could continue to consult this content, if the access provider chose to carry the newsgroup in question. At the latest count, the material is mirrored on 43 web sites and 2 newsgroups and is available from an e-mail listserver (Commission, 1996a). This process of mirroring illustrates the way in which the Internet interprets censorship as damage and routes around it. In another case, a court injunction ordered the removal of a leaked county council report from web sites in England. The result was to 'produce a rash of mirror sites on the Internet, outside the jurisdiction of British courts'. The Council dropped its legal action against the offending site, appearing to concede that it had been counterproductive (Campbell, 1997a). The conclusion that must be drawn is that safe havens, regulatory arbitrage and exit from rule-sets can only be prevented in a situation of globally harmonised regulation.
Daniel
Dunne.com