Posted by Gary Herman, March-August 2000Disclaimer This paper was produced as a discussion document. It is based on a paper prepared for the European Commission's Information Society Forum Working Group on the Rights of Citizens and Consumers in 1999, but may be relevant to the current discussion about the Communications Bill. It represents only the views of its author and does not represent the policy of the Campaign for Press and Broadcasting Freedom or any other body.
Definitions In what follows, freedom means ‘freedom within the law', wherever a law may be held to apply. Since this is not a document on jurisprudence, we will further assume that the law is a set of agreed principles of social behaviour. This is a flexible definition, having the benefit of excluding the entirely private and personal. Of course, freedom is not an absolute - it is always qualified: freedom from something or freedom to do something, freedom within a given context. However, the Internet has seen a resurgence of naοve libertarian ideas which have gained some credence in the absence of appropriate international legal or regulatory frameworks to govern the medium. In this context, we should define the Internet as a medium of communication based on packet switching over computer networks. Packet switching means that communication works by splitting a message into pieces (the packets) each one of which is allowed to find its own way from source to destination, where they are all reassembled. To do this, they use a system for finding computers on a network based on so-called ‘Internet Protocol ’ (IP) addresses. It absolutely requires digital technology. It is a completely different technological paradigm from that offered by either so-called circuit switched telecommunications, in which a message is sent over a single fixed connection between source and destination (the circuit), or broadcasting, in which a message (the broadcast) is sent over an indefinite temporary connection between source and destination. Since each packet in packet switching can be thought of as a message in its own right, it is clear that packet switching can be implemented on a circuit switched or broadcast infrastructure. These two systems are often referred to as dial-up access and packet radio. The Internet The critical differences between the Internet and other mass media for our purposes are twofold: 1. Global reach 2. Relatively low cost of entry This has excited many commercial organizations who see the delivery of global markets at a negligible cost as an opportunity for them, but also a challenge because of the threatened emergence of new competitors. A recent ad for IBM in the UK shows an American couple visiting an old Italian olive grower. They reveal they have three grocery stores in Ohio, and the olive grower comments that she sells her olive oil in Ohio, and in California and Canada and... so on. The startled American woman realises even gnarled Italian peasants are on the Internet. ‘How can they afford it?' she asks. In fact, they probably can't - not at IBM's rates. And even if they can, it will not make a great deal of difference. When even Amazon.com can't make a profit, because they are crippled by marketing and shipping costs, how will one small olive grower fare? The economics of the mass market do not favour small companies supplying niche markets—they will have to charge premium prices or aggregate to form cartels or consortia. The Radio Model We can compare the Internet with radio in its early days. Here, too, we had something approaching global reach and a relatively low cost of entry. It's also the case that radio was a convergent medium, technologically and commercially. But the history of radio reveals political and economic dimensions which are often ignored in the rush to embrace the Internet: 1. Radio rapidly came under the effective control of national regulatory authorities who manage the right to transmit (that is, rules of access), through participation in international agencies for the allocation of radio spectrum and through the operations of national licensing authorities. They also often managed the right to receive—through licensing or the reservation of frequencies for special purposes. 2. Radio continues in a small way as a relatively open-access medium, but successful radio stations (that is, the ones with substantial audience figures or advertising support based on audience figures) are generally characterised by the commercial nature of their programming, reliance on comparatively costly marketing, and access to content provided by major networks. As part of the regulatory environment, most countries established frameworks for content control and imposed public service obligations on broadcasters or the medium as a whole, which meant that minority radio stations or minority content could still exist. (A phenomenon which is experiencing market-driven pressure.) But that was during a period when the idea of a ‘commonwealth' was not anathema. It was widely accepted that nations and citizens had rights which needed protection and, in a world where power was unevenly distributed, the political process could be used as a corrective. This is not happening with the Internet. One can argue that radio needed rationing, since analogue bandwidth is a scarce resource. It's a moot point whether digital bandwidth is, as some libertarian commentators argue, effectively infinite (see George Gilder ’s ‘The Telecosm ’ for this argument). In fact, the argument is technicist, because it ignores the practicalities of bandwidth utilisation (particularly so-called network loading which tends to increase as fast as or faster than bandwidth availability). In fact, digital bandwidth - the capacity of a network to carry signals—is not actually infinite; it is more accurate to describe it as indefinitely expandable, but the expansion is subject to often overwhelming technological, economic and political constraints. More to the point, even digital bandwidth can be rationed and, when it comes to digital TV, is rationed. Current Developments on the Internet Current developments on the Internet are characterised by the rapid growth of high profile commercial sites offering frankly dubious value propositions on the basis of three questionable assumptions: 1. The Internet is a global medium 2. The Internet will continue to grow exponentially 3. E-commerce is set to become the dominant form of trade There are a number of counter-arguments to these propositions, most of which suggest that the Internet is at risk of becoming a very skewed and inegalitarian medium. Some recent academic research suggests that it already is: 1. The Internet is a predominantly anglophone, predominantly US-based medium with (currently) around 300 million potential users (that is people with Internet access—usually through an office computer - who may or may not actually use the network). It is only a global medium in the sense that short-wave radio is a global medium. There is an important distinction in broadcasting between reach and share, and while the Internet may eventually have global reach, it makes little sense to talk of a medium by itself having global market share. Some Internet content providers may become global players, but linguistic and cultural differences remain significant obstacles to monopolisation, as they do in many other industries. 2. Internet growth appears to be exponential because of a number of contingent factors - IP addresses are never withdrawn, single users may have many IP addresses or Internet accounts, Internet connectivity is bundled with most new computers, and so on. In fact, real Internet growth is probably flattening out. Digital television is likely to make the Internet - in a restricted and pre-packaged form - just another channel option. 3. E-commerce is a very small proportion of the total retail (never mind trade) market. In some areas, it will undoubtedly become significant (it already has in, for example, retail share trading in the US), but it is by no means a universal model. Issues of security and liability are important (particularly in the business-to-business area), and have generally not been sorted out, while the history of most business-to-consumer e-commerce suggests that profits are extremely elusive. 4. The web is increasingly dominated by a few very popular sites, such as AOL/Netscape, Microsoft Network and Yahoo! According to Nielsen, around two-thirds of the web ‘audience ’ is accounted for by the top ten sites. The dominance of portal sites like most of these may effectively restrict access to the medium for both users and producers on the Internet. Deregulation and the Public Sphere However, enthusiasm for the Internet is unstoppable in some quarters. The Internet has developed against a background of growing market liberalisation and deregulation. Its internationalist aspect has ensured that it has become a trailblazer for neo-liberal ideas. Internet ideology promotes free-market access and open competition. It is against regulation in any form except self-regulation. It is against any authority except its own. In practice, it promotes American values and American business, while Europe - disturbed at the prospect of losing this new trade war - has responded by accepting the challenge of a fight under terms set by its global competitors. Let's not forget that the myth of the Internet brushes over the fact that it was funded, developed and supported by the US state machine. If now it seems occasionally to challenge a political settlement with middle-American moralism, the White House is not allowed to forget that IT and telecoms businesses are fuelling the growth of the US economy to an unprecedented degree. Some 30-35 percent of economic growth in the US is due to digital electronics—not, it should be said, through the use of such technology, but through its sale by the new industrial giants such as Microsoft, Intel, Cisco and Sun. The result is that the public sphere has been dangerously backgrounded wherever the Internet and other new communications technologies have taken hold. We have seen a subtle movement of emphasis in political discourse away from citizens' rights towards consumer rights. In other words, the paradigm used when discussing the Internet is increasingly a market model, not a public service model - at least in part for the very obvious reason that if the Internet really is a global medium, or is at least believed to be one, where is the authority that can enforce public service obligations? Consumer Rights Versus Citizen Rights? Consumer rights do not address all the issues that free market propagandists claim. For example, you cannot transform a citizen's right to information into a consumer's right to buy information services—even if you can substitute the one for the other. The excuse for ignoring citizens ’ rights when it comes to new media in general and the Internet in particular is that media and Internet companies operate in a global environment in which the idea of citizenship cannot be interpreted. That is what is meant by encouraging competition in a global market. Of course, we could be promoting the idea of global citizenship—and there is some evidence that this will happen. Yet models and institutions are already available for safeguarding citizen rights in an international context: Internet content, for example, is produced somewhere and consumed somewhere and is therefore governed, in theory at least, by national jurisdictions and international agreements. For many multinational and transnational companies, this is actually a problem. A company operating in 17 countries, would rather not have to negotiate 17 rights regimes. Hence the proposed Multilateral Agreement on Investment (MAI), new developments in copyright law, and the argument by many US companies that—despite operating globally—they should only have to obey the legal regime of their home country (or even the state they are registered in). William Gibson's nice definition of cyberspace as ‘the place where telephone conversations happen' should not be taken to mean that cyberspace is actually a place where people live. They live in Lisbon, or Manchester, or Copenhagen, or Akron. Cyberspace is a medium not a location. To think it has the characteristics of a place is to take the post-Newtonian weltanschauung a little too far. The critical thing that has happened as a result of the victory of neo-liberal economic and political theory, from our point-of-view, is this: in the generally positive urge to free ourselves from the officious constraints of authoritarian or paternalistic government, we have begun to forget what government was for in the first place. The rights of citizens are still vital. Regulation versus Self-regulation? Even if we accept the need for citizens ’ rights, the concept is under threat from attempts to redefine the very idea of a citizen for the ‘information age ’. We seem to have accepted wholesale two dubious propositions framed by the ‘organic intellectuals' of the new ruling class. The first is that there can be such a thing as an Information Society based entirely or even predominantly on non-material production, and the second is that globalisation is an equally distributed phenomenon affecting everybody to the same degree and in which everybody can participate to the same degree. Cyberspace may not be a place, they seem to say, but it can be a community and a polity. In April 1999, Vint Cerf (the so-called ‘father of the Internet') made a speech entitled ‘The Internet is for Everybody' at a conference on Computers, Freedom, and Privacy in which he said: ‘The Internet is becoming the repository of all we have accomplished as a society. It is becoming a kind of disorganized Boswell of the human spirit.' He made some pretty startling claims about the ultimate universality of the Internet, which it may be fair to say amount to this: the Internet will become the dominant (if not the only) communications medium in the world and within the local solar system, as long as we all behave responsibly, and as long as there is no regulation - except for the ‘the protection of intellectual property' and voluntary regulation by ‘parents and teachers' in order to protect children. The enabler of this ‘noble vision' will be technology, which can solve all problems, eventually. These are widely held views among the techno-elite. In late 1999, in an interview in the Guardian newspaper, the operations director of the company that makes the Tomb Raider video games said this about the Internet: ‘I hope that it might almost regulate itself, with the result, for instance, if there are people writing horrible things or putting disgusting images of Lara Croft [the cartoon heroine of Tomb Raider] up then it ’s up to the community itself to drive that out. ’ Yet, in 1995, Thabo Mbeki reminded the assembled dignitaries at the Global Information Infrastructure (GII) summit in Brussels that, ‘Half the world has not made a telephone call'. Despite satellites and cybercafes, this figure has not greatly changed four years later, although last year (* 1999) the market capitalisation of Microsoft out-stripped the gross domestic product of the Netherlands. Bill Gates, needless to say, sees the Internet as the future of his company. Accordingly, Microsoft now controls a large proportion of the cable TV industry in the US and the UK and is in on-going dialogue with AT&T and BT. Rupert Murdoch, meanwhile has had a Damascene conversion and declared News Corporation an Internet company, when he realised that Internet companies could double or treble their share price in weeks. And, in case you thought this was all smoke and mirrors, the US company, Cisco, has a market capitalisation greater than General Motors, and is evidently destined to become the world ’s first trillion dollar company, all because it provides 80 percent of the highly specialised computers called routers that make the Internet work. (* In fact, a collapse in share prices meant that Cisco's coup de bourse has so far failed to materialise.) What is happening on the Internet is the result of a largely unregulated market favouring the dominance of US technology and US companies. That ’s not necessarily a bad thing, despite the disputed sustainability of the economic model which informs the World Trade Organisation. The problem is that globalisation rests on free access to markets which may have very good reasons not to be free or which may, according to the relevant democratically elected government, choose to be regulated. So, regulation is increasingly depicted as a barrier to market entry, rather than a means of supporting pluralism, local economies, cultural diversity, and public service. According to the EC Green Paper on Convergence: ‘The global nature of the platform [for converged media] and the difficulty of exercising control within a given Member State are leading to solutions which draw on self-regulatory practices by industry rather than on formal regulation ’. This is a counsel of despair, and a recipe for failure. Globalisation and the new media challenge our conventional wisdom about values, rights, quality and content, and they allow transnational corporations to use market power to wipe the floor with their competitors. The hope of an information superhighway leading to freedom will be illusory, as long as the issue of regulation is relegated to an apologetic footnote (where it ’s classified as ‘light touch ’) in the attempt to liberalise markets and investment opportunities. And this applies to content as much as infrastructure. When companies like IBM begin to develop a global strategy for selling education on the Internet, and governments like the UK ’s show them an open door, it is time to think seriously about controlling the information superhighway before it becomes little more than an information supermarket offering up bland and processed product. Access and Regulation Under these conditions, how do we approach the question of communications freedom? We need some sort of framework within which issues of communications freedom can be usefully discussed. It may seem as though there are only two issues: access and regulation. The received wisdom is that access is desirable but not always possible, while regulation is undesirable but may be necessary. The libertarian tendency is to interpret communications freedom as meaning complete freedom of access and complete freedom from regulation, but there are other options that make more sense when balancing freedoms against responsibilities within the context of civil society. Many people treat access and regulation separately, following the fashionable separation between ‘conduit' (or ‘carrier') and ‘content'. But the separation breaks down in too many places to be convincing. In practice, many content companies may have given up their control of networks, but others, not subject to government strictures—like News Corp—retain it, while network companies (notably, telecommunications operators) increasingly sell content (as premium phone services, Internet portals, or cable TV programmes). Electronic programme guides are likely to make this confusion even greater. Internet service providers are increasingly funding their operations through advertising and, in general, Internet content is so wound up with Internet technology that it is difficult to separate the two. Content shades into software which shades into hardware with no easily detectable divisions. It is possible to separate content from hardware, but where does that leave software? In European directives, the software issue is often disregarded. In practice, the arguments about access and regulation resolve into a debate about details: how much access?, how much regulation?, where should the regulation be?, who should regulate?, how will access be paid for?, and so on. These may seem entirely pragmatic matters, but the models underpinning the debates are important. For example, the content/conduit distinction is felt to be necessary because infrastructure industries are increasingly ‘horizontalised' while content provision is still largely vertically structured. There seems to be a tendency on the part of some observers to equate horizontal structures with free markets and vertical structures with regulated ones. In this model, horizontal structures are already subject to competition (but may need competition law to be enforced), while vertical structures need to be deregulated to stimulate competition. Consumers are ruled by the great god ‘choice ’. Actually, to be more accurate, content provision is increasingly the subject of attempts to vertically integrate on the part of companies like BT, Deutsche Telekom, or Microsoft. In any case, this model does not address the issues of public service. Accordingly, there is a somewhat more elaborate model, in which producer access should be open (and regulated for such, using competition law wherever possible), while user access should be protected (and regulated for such using copyright, privacy, and defamation law). ( ‘User ’ carries fewer precise connotations than ‘consumer ’.) In this model, we divide access into: 1. User access (that is, access to content and services) 2. Producer access (that is, access to networks and the market) Universal service, public service broadcasting, censorship and so on would then be matters of user access, while open network interconnection, asymmetry and so on would be matters of producer access. But this analysis does not address the issue of user choice. A better framework might involve the points-of-view of ‘the people ’ (which would include companies) and ‘society ’ - or if, along with Margaret Thatcher, you believe there is no such thing as society, the people and governments or public authorities. Such a model seems to fit a ‘consumer-citizen' model. Access could be considered a ‘people issue ’ focusing on the relationship between users and producers mediated by the market, while regulation is a ‘society issue ’ focusing on the relationship between the market and society mediated by the political process. The first is an essentially individualistic and competitive arena - we all want as much as we can get and as good as we can get it—and the second is a collective and cooperative arena - we must protect the weak and make sure that everybody else behaves themselves. This model seems to cover all the areas of concern: choice, open networks, public service, privacy, censorship and so on. (As an aside, this raises an interesting question. If we have consumer rights [beyond the simple right to consume], then high-school sociology tells us we should also have consumer duties [beyond the simple duty to pay for what we consume]. But what would a consumer duty look like? Be nice to Carrefour? Take care of Ahold? Other than to protect the earth's resources, and that's hardly a duty on consumers as consumers, it is difficult to think of one. Citizens have both duties and rights, in that they must uphold the law and - in return? - are entitled to the protection of the state. In a world where consumer rights erode or encroach upon citizen rights, and the protection of the state becomes less and less significant, we risk creating populations with a profound sense of duty but with no reward. Is it any wonder that crime is rampant in consumption-obsessed societies? ) That said, there are some clear issues in the debate, even if there are no clear frameworks yet or clear solutions. Issues Access to the technologies, devices and services allowing content to be received. Access to technologies and devices is generally considered a consumer issue, typically negotiated by consumer markets - although not necessarily (for example, minitel in France). The continued existence of capitalism means that it is likely that software and hardware will continue to be the products of commercial companies, yet there are a number of issues which bring such companies into potential confrontation with public authorities. These issues do not have a direct bearing on consumer interest, but they are significant to any discussion of the wider aspects of communications freedom. At the level of services, citizen rights are increasingly foregrounded in discussions of USP and social priorities. USP is seen as a financial burden by telecoms operators and there is pressure for it to be abandoned. This can only be resisted if USP is seen as a citizen right. Related issues a) Standards Technical standards make life easier for consumers, but may be a barrier to market entry. Standards are themselves products with commercial value and may be the property of governments, public sector bodies, private companies or even (improbably) individuals. Consumer interest is not necessarily served by the existence of a single set of de facto or de jure standards (Betamax was a better system than VHS/ almost any operating system is better than Windows), but nor is it served by a multiplicity of contending standards. The standards making process, as everybody has discovered, is generally too slow for consumer technology markets, and governments have sought to modify it or withdrew from it (at least in part). On balance, standards making should probably be left to the market, but standards should be approved and formalised by international bodies governed by majoritarian principles. Thus the US government will not be allowed to act as a proxy for Microsoft or for the American telecoms industry and enforce international acceptance of Windows or CDMA. Ultimately, standards approval bodies should be answerable to the UN. This is more-or-less what happens now. b) Pricing This is a complicated issue and acts, in part, as a mechanism within the standards making process in the form of premium pricing for new technologies producing effective market testing. In general, pricing is a function of markets and may or may not be subjected to regulation. Governments frequently use the tax system to control prices for social purposes - at least, to a degree (for example, increasing the tax on tobacco and alcohol to discourage use or decreasing the tax on capital gains to encourage investment). Consumers may be able to push down prices in a competitive market, but it is equally likely that ostensible competitors reach explicit or implicit cartel agreements on appropriate price levels. This should be a justification for regulatory measures. Communications technologies increasingly provide essential social functions (for example, access to emergency services). If access to basic services is considered a social priority, then it should be available free or at least at affordable prices - a principle of universal service provision (USP). Regulatory authorities do have a range of instruments with which to force commercial suppliers in communications to adjust their pricing strategies, and I believe that this is an essential function of government. Price regulation should not be used to create barriers to market entry, however, and may therefore be tied to profitability. New market entrants would be able to set their own prices, while established and profitable players would be subject to regulation. It is also possible for governments to use their purchasing power to establish low market prices (as the French did with, once again, minitel). How they recoup the costs (through general taxation, means tested taxation, leasing or whatever) is a matter for individual governments. We should also like to see an obligation on equipment suppliers to provide public access terminals tied to given services. c) E-commerce and universal service We cannot leave a discussion on pricing without mentioning the current trend towards ‘free' Internet services in the UK (and Germany), or the idea of using advertising to finance communications services. This is not the place to discuss business models in depth, but it is important not to take the arguments for ‘free' or advertising funded services at face value. ‘Free' Internet services are not free. They are funded out of the costs of telecommunications services and are therefore unlikely to take root in the US where telephone access to the Internet is unmetered. In Europe, it has been suggested that ‘free ’ ISPs are effectively fulfilling a public service obligation and that therefore there needs to be no regulatory intervention to ensure USP applies to the Internet. But to leave USP to the whims of the market is plainly wrong; the principle must be established in law. In any case, the cost of Internet access is not just the cost of an ISP and a phone line. The (* somewhat hesitant) movement towards unmetered calls in the UK suggests that ‘free ’ ISPs may suffer financially. In some models, Internet service provision is a kind of loss-leader. In this view (the ‘e-commerce paradigm ’), the Internet is seen as a vehicle for making money. This paradigm is already losing support in the US, and although a few people will emerge from the e-commerce boom as very rich, it is inevitable that most of the companies currently involved in Internet service provision and e-commerce on the Internet will collapse. Advertising on the Internet is a difficult phenomenon, largely because all the evidence is that it doesn't work. The idea that advertising can be used to fund USP in Internet and other services is not objectionable, but it may not be viable. Consumers should be allowed to choose advertising funded services, if they wish, but not at the cost of USP. The implication is that the USP obligation should be shared out, according to a formula that takes into account the financial arrangements of the providers involved. It may be possible to define USP activities in such a way that they can produce tax benefits for the companies concerned, thus allowing companies to choose to balance the cost of these activities against their tax obligations. d) Technology awareness Purchasing a complex piece of communications technology is only the first step. You must also know how to use it. We believe that the public education system should ensure that basic skills training is available to everyone. There are a number of mechanisms for delivering this, but ultimately technology awareness should be included in basic education. Since technologies change with such rapidity, we also propose that funds be made available for ongoing technology training through each country's adult and continuing education provision. Where philanthropic contributions are not available (eg. from the Bill and Melinda Gates Trust), technology companies could be given ‘tax breaks' for contributions to the funding of such programmes. Alternatively, they could be funded out of windfall taxes on the excessive profits of technology companies. A good case could also be made (and has already been proposed by a number of US companies) for a tax tied to share price or dividends. Luc Soete's suggested ‘bit-tax' is worth further investigation, and this may be used to fund training. Access to networks and content This is partly a consumer issue and partly a producer issue. Access in this context is often identified with consumer choice, but to the extent that governments are concerned with economic development and the stimulation of a pluralistic market, access should also be seen as a citizen right. Access to networks underpins diversity of content, and this in turn is fundamental to the continuation of public service and quality. Related issues a) Censorship In general, we reject censorship by dictat, although it is clear that there will be certain cases where content is unacceptable. These cases will typically involve illegality, defamation, or the invasion of privacy. The issue is complicated because the law changes even within a country and differs from one country to another - a particular problem for satellite TV and the Internet. Some form of censorship may be necessary in certain cases - for example, web sites used to organise racial violence (rather than merely express racist opinions) or targeting attacks on medical personnel. In most cases, editorial responsibility and professional and ethical standards are the right vehicles to ensure acceptable content, but there should be a system of accountability and explicit sets of principles which ‘content providers' sign up to. This must be a voluntary arrangement, although one which could be monitored by national agencies who are able to award ‘seals of approval'. This can be tied to an international ratings system which can be used, if so required, to filter out unwanted material. (Filtering and rating systems are currently being widely discussed in relation to the Internet). For voluntary agreements to work, they must be well publicised and subject to complaints procedures for those people who feel that the agreements have been broken. National procedures are well-established in many countries, although the privacy, data protection, right-of-reply, and freedom of information continue to be problematic within the UK and elsewhere. An international complaints procedure would be unmanageable and would run counter to the principle of subsidiarity, but complaints generated in one country could be channelled by an international monitoring agency to the relevant authorities in the country or countries from where the offending item or items originated. Such a system has been proposed for on-line news media by the International Federation of Journalists, and a similar system for Internet content in general was outlined last year (* 1999) by the Bertelsmann Foundation. Much of the debate about offensive material on the Internet is a ‘moral panic'. The material is certainly there and should be managed, but for the most part the law and parental control are adequate means of doing so. The threat of prosecution has proved to be enough to stimulate self-regulation. Governments clearly believe they have a right to monitor communications traffic, even where it is private. This right should extend only so far as the threat or occasion of illegal acts can be demonstrated. No blanket surveillance mechanisms or activities are acceptable. Current proposals by the UK government regarding the interception of communications and freedom of information are not acceptable (* the Regulation of Investigatory Powers Bill). These would allow governments to maintain their own secrecy and yet control private communications on the basis of unproved assumptions. Meanwhile, the legal position and responsibilities of transmitters and receivers of information must be clarified. We believe that the need to develop liability mechanisms for e-commerce will ensure that this happens. b) ‘Jamming' This is difficult to classify, but means that technological measures should not be used to prevent or hinder access to a network or programme, all other things being equal. In other words, governments should not interfere with transmission or reception and network operators or content providers should not hinder users who wish to access alternative networks or programmes. A classic case of jamming in the second sense would be Microsoft's attempts to force Windows users to accept its Internet browser (which has a built-in - though not permanent - ‘home page portal') rather than competitive browsers with different home page portals. c) Market power/concentration of ownership Individual companies owning infrastructure should not be allowed to prevent consumers accessing whatever networks are available to them. Network access should be subject to market mechanisms. Consumers should be allowed to buy into any network they wish, subject to its legality, and producers should be able submit content to any network. Backbone networks should be open to any network operator, and producers should not be excluded from networks by ‘gatekeepers' in order to distort the market. Concentration of ownership is to be deprecated because it limits choice and diversity. It is potentially profoundly anti-democratic. Regulatory instruments, typically based on licensing and invoking quality thresholds and market share, are available to governments but are generally irrelevant in the case of international media such as the Internet. Pro-active measures might be used to encourage new market entrants (tax breaks, government backed low interest start-up loans) and to nurture appropriate technical and creative skills in individual countries, but it seems clear that international agreements should be formulated to ensure that global media networks do not exercise disproportionate market power within local or regional economies. Mergers and acquisitions should be subjected to rigorous scrutiny and agreement tied to quality thresholds. In the longer term, a global anti-trust mechanism should be established to avoid effective monopolies within the media. Conclusion This paper has attempted no more than a preliminary survey of ideas and issues. It is highly opinionated and probably naive in parts, but its basic hypothesis is this: the Internet has created some new concerns for advocates of communications freedom such as the CPBF. These focus on access and regulation within an increasingly globalised media environment. Communications freedom is a complex of issues demanding the attention of governments, international organisations, advisory bodies, campaign groups and, of course, companies and individuals, acting as citizens and consumers. The neo-liberal position taken by Vint Cerf and many Internet enthusiasts is not adequate to the issues, because it is technicist and fails to acknowledge the influence of the cultures, opinions, beliefs, and institutional frameworks that shape the world. Freedom must always be negotiated and that is a process that involves a continuing dialogue between all the ‘stakeholders' in society which must take place in the real world, not in cyberspace.