Commission on Intellectual Property Rights

Workshop 5: Copyright, Software and the Internet
21st January 2002

Participants: Alan Story (University of Kent), Richard Owens (BMR), Chris Zielinski (Informania Ltd.), David Vaver (University of Oxford), Michael Keller (Stanford University), Dervish Tayyip (Microsoft Corporation), Paul Goldstein (Stanford University), Eki Yemisi Omorogbe (University of Kent), Christopher May (University of the West of England), Georg C.F. Greve (Free Software Foundation Europe), Stuart Booth (Copyright Directorate, UK Patent Office)

Commissioners: John Barton (Chair), Ramesh Mashelkar, Carlos Correa, Daniel Alexander

Secretariat: Charles Clift, Tom Pengelly, Rob Fitter

Summary:  The first session of the workshop comprised a presentation by the author of the study report commissioned on this topic, followed by a response by a discussant and a general discussion of the paper. In the first session, Story focused on the barriers to the access of hard copy materials in developing countries, and prompted a discussion as to whether the creation of exemptions to copyright laws was necessary in order to provide assistance to the developing countries. The second session focused on copyright and software, in particular whether international copyright rules and practices are a significant constraint on access to computer software needed by developing countries, and whether proprietary software or free software models were
the best models to utilise in developing countries. The third session dealt with the relationship between copyright, and access to educational and scientific materials, and research over the internet and led to a discussion about encryption and fair use. The final session dealt with illegal copying, and protection for indigenous materials, and focused on the key issues of the workshop discussions as well as the areas in which the Commission could focus its recommendations.


Session 1: Overview of copyright and development issues

Presentation of Background Paper 
Alan Story referred in his Commission study paper to Robert Hunter Wade's book "Inequality of World Incomes: What Should Be Done", which presents evidence that global inequality has risen in the last 20 years.  It suggests a strategy for global income redistribution through the integration of LDCs into international markets by lowering tariffs and the privileging of foreign investment. The book argues that suggestions made by international organisations (World Bank and WTO), that it is in the national interests of developing countries to accept uncritically the IP models of rich countries (such as increasing the duration of copyright protections, further privileging of copyright holders, the spread of encryption processes, and other barriers to access) will not lead to fairer income distribution.

The overall theme of the paper is that developed countries can provide assistance to developing countries through exemptions to these Copyright laws.  Many of the recommendations could be undertaken at zero to minimal cost to the developing countries, with no loss envisaged.

There are barriers to access to information in Africa.  For example, in South Africa, nurses have to pay for HIV/AIDS information for their patients; the inability to distribute books compounds literacy problems; copyright on computer software has become a tax on developing countries. What is copyright for?  To protect electronic-book publishers or to ensure millions have a chance to read their first book?   

Discussant
Copyright and the system it represents bring the fruits of creativity to the public. There is conflict in the application of the international copyright standards (TRIPS) to the socio-economic contexts of the South. It is natural that problems should be encountered between rights and exceptions and it is right to seek to balance these. However the system not only benefits rights-holders, it also benefits users and the countries involved (through revenue generation), and without a stable framework of rules, the system cannot take hold.  Thus change should be made through the system not to the system; greater flexibility not legal change.

The report presents a view based on legitimate concerns, but does so without presenting a fair view of copyright holders and proprietors. There is an under-emphasis on the activities of UN organisations, the WTO, UNESCO, collecting societies, and RROs. 

The main issue in software is between market domination and piracy. And on Open Source and Licensing it is important to note that GPL is still a copyright licence based on notions of exclusivity.  

Indigenous knowledge and orally transmitted materials could begin to be covered by the TRIPS based IP system. Some of the case law, primarily from the Pacific region, relating to the use of copyright to protect indigenous knowledge could aid examination of sui generis protection systems.

Copyright has a positive effect on development, and more sustained licensing, tailored to the needs of developing countries, and further infrastructure and communications development is required.  The discussant did not believe that copyright is one of the main barriers to distribution of knowledge.  

Discussion
The report recommendations were criticised for the proposed overhaul of Copyright Laws and positive discrimination to Open Source software. The report's premise that there are "no credible economic models that project economic development for the developing countries" was not balanced.

There is growing evidence of direct correlation between strong IP protection and economic development for developing countries and reference was made to Edwin Mansfield's work on the willingness of multinationals to make direct investment. The OECD Report supports this view.

Multinational software companies understand that there should be balance between the public domain and copyrights protection.  Bad decisions made now will cause all to suffer, particularly, the poor and developing countries. 

The US focused perception of copyright protection, particularly of the software industry, ignores the global dimension despite the US companies involved being multinationals.

Large companies are better able, in the short term, to weather IP piracy.  An Annual Report by Price Waterhouse Coopers addresses this issue, and finds that these industries create local jobs upstream, and downstream thorough their distribution and training services. They generate huge sales taxes, and increase productivity. The focus of the multinationals is the development of local industries to make a difference to the economic well being of these countries. The latest figures (1999) show 41,000 jobs, US$900 million tax revenues generated in these countries. The projection for 2004 is 72,000 jobs and US$1.7 billion in tax revenue, even though piracy occurs in those countries. Even a small reduction in piracy would make a huge difference.

Differential pricing has been recommended in the past. However, discounted price products immediately leak back to markets in the West. If companies could be assured that this would not be the case, then the paradigm would be acceptable.

Pricing is not the only issue. The main difference between proprietary software and freesource is the freedom to use, study, modify and redistribute. There is a common misunderstanding that free software means non-commercial software, but the terms of GPL (the Licence for the Free Software Foundation) protects the rights of the author. It does not allow use without having recourse to the terms the author wants. The software industry does help developing countries, and does create revenues. But these revenues are totally independent of whether the software is proprietary or free.  It was suggested that proprietary software restricts access and lowers revenues.  

Multinational companies do what the law allows them to do. Their actions are a natural consequence of the proprietary system.  Local software systems based on the proprietary system, will create a local system dependant on it and the bundling of hardware and software. 

Protection of software and other types of literary work should be separated.  The books and software markets are different as the products have different features.  Software could fit more neatly, into the patent model, as it is a functional tool.  Although softer patents are being discussed in Europe, studies suggest that they are harmful and that the bigger companies have them because only they profit. Notice should be taken of the US situation where they are struggling to readjust from a reliance on softer patents.

India and Brazil are examples of the failure of PPP model to develop computing facilities, as private funding could be obtained.  Parallels with the pharmaceutical industry suggest global funding should be used.

Fair use policies and open licensing for educational purposes are potential options. There is support for increase of fair use in academic works.  The Appendix to the Stockholm Protocol refers to compulsory licenses, other than fair use, given under tight, stringent conditions. Discussion about fair use for educational materials reverts back to the discussion earlier, that if there is no promise that companies can recoup their investment, if there is no spill over from a particular market, then, these works will not be produced. In other words, there will be under investment in the creation of goods if there is no IP investment. 

It was asserted that WIPO and UNESCO are unaware of any disgruntlement with the copyright treaties, from any of the developing country delegates invited to attend WIPO functions and training programmes. 

Various open access providers are increasing the scope to provide materials for free to developing countries, but the chance of repealing Berne is small. The way forward is to work on the commercial psyche of shame and guilt, or to try to do something with the legislative system, perhaps to include "essential information" in the original language. Such "essential information" would include "information necessary for human development". The problem is that information is incorrectly treated as a monolith within the same legislative framework. 

IFFRO has functional difficulties. There are four different agreements operating on a flexible basis between IFFRO member societies regarding remuneration and licensing agreements. One problem with compulsory licenses is that only two countries, India and Thailand (and perhaps Egypt) signed, however, their memberships have not been reactivated. The framework for the design and administration of licence is Article 20 of the Berne Convention 1971. Whilst it is conceivable that the Appendix could be reopened and renegotiated, at an international level for variation, the political likelihood of this occurring is unknown. The political muscle of developing countries was shown in the enactment of the Appendix to the Berne Convention, but it has not been utilised.

The RRO model dramatically increases transaction costs.  To provide course-packs for students, an author who has assigned copyright to the publishers, then has to buy back, and sell to the students, his/her own work. It would seem unlikely that one could get royalties for photocopying African writers work in the USA, and that very little photocopying takes place in Africa. However, African music is being used quite widely so countries could earn royalties.

The Report was criticised for having a single viewpoint on the way publishers deal with the public. Highway Publishers offer 300,000 articles free to the world and include toll-free subscription. In Scientific and Medical Publishing publish over 60,000 of the most used articles. Not for profit publishers have started distributing at very low cost to the Third World.  Reporting such evidence could show publishers that little or no costs are involved, thus persuading them to distribute freely to developing countries. 

It was suggested that the flexibilities in Berne make the Appendix unnecessary, and no developing countries had brought copyright problems to the attention of the TRIPS Council. Both developing and developed countries were eager to come on board, such as Malawi, which is trying to bring its laws into line with TRIPS by 2006.  The recommendation that copyright should be abolished for 20 years in developing countries would remove any incentive to produce their own indigenous work. 

The international copyright system operates through Berne and WIPO, and this will not change, because these agreements operate by consensus. There is very little possibility for radical changes, but the Commission can work with the system, for gradual change, and one opportunity is the imminent European Copyright Directive.

The report highlights that even in developed countries, copyright has problems, thus raises concerns on its applicability of the TRIPS 'one-size-fits-all' regime to developing countries.  However it was argued that the Berne Convention is ambiguously worded to allow consensus and flexibility, to enable governments to draft laws according to its own system and needs. 

Authors have creative incentives other than payment, and Jessica Litman in "The Copyright Myth" states, most do not understand copyright law and do not necessarily benefit from copyright. 


Session 2: Copyright and software

Are current international copyright rules and practices a significant constraint on access to computer software, and related Information Technologies, needed by poor countries? What is the evidence?

Is the length of copyright protection for software too long, given its sometimes short shelf life? Should there be differential software pricing between rich and poor country markets and/or encouragement of cheaper non-proprietary types of software?

Discussion
Is copyright protection relevant to relieving poverty in developing countries? Firstly, basic needs; clean drinking water, health infrastructure etc. have to be satisfied.  But it was argued that access to technology and information was an essential element of sustainable development. The software industry has developed in some parts of Asia from piecemeal work out-sourced by the US to flourishing national industries.

In the absence of price discrimination, illegal copying is necessary, but a system which depends on an illegal action is not stable.  The enforcement of software copyrights in developing countries will increase costs and thus reduce access to information.  Price discrimination is a possible solution, though it may not concern IP.  Multinationals are not fundamentally against differential pricing, a structure based on self-help rather than charity, but it was not thought to be the answer to the access problem.  

Software development, like science, increases incrementally, but IP means that the first company to file will drive other companies out of the field, and stifle development. Distinctions must be made between operating systems, applications and the different models of research and development.  Although developing countries may not prioritise access to computer programmes now, the new transition communication technology will be central to future development, and it is important to retain access right. 

Recently, Open Source companies are having difficulties in generating profits, suggesting this is not a sustainable business model. It was contended that total cost of ownership for Open Source is higher than proprietary software, and the specific economics of the software industry make the dominance of the proprietary model inevitable. The proprietary software model has produced the most significant innovations, with businesses often building on university ideas in successful PPP arrangements, which would not work with Open Source and GPL. The GPL model will not help developing countries. 

Proponents of Open Source argued that it is the proprietary companies that have not survived. In Europe, free software does have successes because they work with established business models. Studies on the Total Cost of Ownership (TCOs) suggest that TCOs are higher for free software, but these are reports that are sponsored by proprietary software competitors. Open Source software rates are lower. And if most innovations start at universities using public funds, the products should be used for the benefit of the public. The proprietary model has failed to produce technology transfer, it binds consumers to one company's technology, which is effectively legal subsidisation of multinational corporations.  Government should favour a system which allows fair competition among companies and access to products by all. 

Piracy is not always linked to organised crime. Piracy has been there since the beginning of copyright, but for a long time, because of the difficulties of copying, it has been a de facto problem only. But now, the development of copying technology requires a reassessment of copyright law.

Government may have a responsibility to provide equal time and access to information on both proprietary and open source software.  But there are examples of open source and proprietary source providers supporting each other and the two systems working together in a way that maximises the benefit for the end user.  It would be to the benefit of the developing countries to have the two systems working together.


Session 3: Copyright, education and research and the Internet

Will the use of "rights management systems", sui generis protection of databases, and techniques such as encryption deny developing countries the prospect of improved access via the Internet - including access for "fair use" - to material necessary for their development, (e.g. scientific journals, genomic information, meteorological and geophysical data, other educational material)?

Given that access to and use of both computers and the Internet is still relatively limited in a number of poorer countries and use of traditional "hard copy" materials remains critical for their educational programmes, what barriers do current copyright rules and practices with regard to "hard copy" materials create for poor countries in the attainment of their educational goals?

Discussion
Encryption may be outside the Berne and TRIPS Conventions. It has nothing to do with copyrights, only with company secrecy laws. There is no standard international law. The control of access at local and international levels is made on a political basis, and similar to digital rights management. The rhetoric about encryption has its roots in privacy, the right to strong encryption.  But to support the private right of encryption, and public dissemination of information is a contradiction.
 
In most cases, there is no extra cost for student who uses it in a developing country. There is a new copyright rule which gives a blanket licence to cover all things, but this kind of scheme is outside any regulatory oversight, unlike hardcopy licensing fees which are regulated. There is a need to change the copyright designs act.

Will the market or state intervention lead to tiered pricing? Is the market model working? It was suggested that more pricing programmes are being initiated because it improves the public profiles of the companies concerned. It has been recognised that passwords and encryption technology are blocking potential paying customers, and there now are many legal journals from Humanities which allow rights of access to developing countries free or at reduced rates. Copyright Tribunals have made decisions that supported first, blanket licence, then course-packs and blanket licence, and have now moved back to blanket licence.  UK Licensing Schemes are very complex, and are framed to extract as much as possible from the market place, and to reflect market conditions. The Copyright Tribunal rule on prices, and so the market does prevail. But does the market works for poor countries?  

Essential information has some form of legitimacy, but the line between essential information and privacy is blurred. Regulation may not be entirely satisfactory, but some regulation is needed.  The Soros Foundation, MIT, and BOXM, are working on this issue. Should regulation be set in law or would it best be left to a voluntary mechanism, and who should decide?

The term "essential information" determines that there is "non-essential information" which is impossible to define and a dangerous concept (the term "essential information" is derived from the notion of "essential drugs"). But information is to be made available under the Stockholm Act, e.g. as it relates to educational needs. Also Human Rights Laws and Conventions help build a picture of what is essential information, and proffer an adequate definition. Essential information could mean 'essential for human development'.  Libraries should provide free access to users of information they hold, and to other libraries for other users. However they are part of the market model, and are consumers of copyright, and should respect copyright.

There was a need to draw a distinction between essential information and essential uses. Essential uses would be easier to define, for example locations such as universities are a natural site for educational access to the internet. A possible compromise could be the adoption of a site licence, one flat fee on an annual basis.  Essential use is a modification of traditional fair use doctrine. Government funded research should be made available on the internet as essential information.

Attempts to regulate ECD and DMCA have been problematic because they do not define fair use. It is suggested that a global restriction should apply, but to which information?  This is not possible under DMCA. Traditional fair use, such as passing on a book to a friend, becomes unfair use when it is done over the internet. Local laws should thoroughly explore what it will deem to be fair use.  

However, others argued that much of the DMCA works well for the library system, and there are many inbuilt fair use provisions in the DMCA. As there is no general definition of fair dealing, cases are being dealt with on an individual basis, and countries should find their own solutions. Within the EU, it has been decided not to take developing countries to court. 

The international system works reasonably well (no developing countries have yet complained to WIPO/EU), TRIPS Article 13 provides plenty of flexibility, and the debate about encryption is a red herring, because it is not a problem at the moment. It will only become one when all information is in electronic form. Hard copy will remain, but encrypted e-info will become a medium of choice in the future. The fear about encryption is that people are shut off from access to information. Encryption is purely technical, and has no morality. It is unclear whether one can input fair use into encryption, but as fair use is user specific not information specific, this is not possible. It is hard to imagine a digital owner refusing to make their product available at some price, unless they have a monopoly on the market, and this is something that competition laws should address. 

Toll-gating systems control access into a library. One possibility for redressing the balance would be through amending Unfair Contracts Terms Act as a way of reducing the power of multinational publishers. These publishers require that the authors give them the digitalisation rights over the work. 

Encouraging fair use results in knowledge transfer and thus knowledge creation. Few other mechanisms seem possible within an IP regime. Millions of students have no possibility of acquiring books. The problem is how to ensure access to students in poor countries, in order to create more equal societies through education. The problem is access to the materials, not copying. 


Final Session: Conclusions and key issues

Are changes needed to current copyright rules and practices? Are international rules on copyright unduly weighted towards the proprietors of copyright? Is the length of copyright protection provided too long? Might lower and/or more targeted levels of copyright protection help to address the extent of illegal copying, recognising that this practice has some benefit for consumers and local producers at the expense of copyright holders?

If copyright is potentially an "engine of development", to what extent are poor people and poor countries reaping the economic, social and cultural benefits from copyright protection of their own indigenous materials? What are the main obstacles and how important is the level of copyright protection compared to other factors? Is there more that could be done to help? 

Are there any other copyright issues that the Commission should consider?

What are the key issues and where should the Commission focus its recommendations?

Discussion:
In 1960, many newly independent countries of Africa had to deal with the Copyright Convention. As they were part of the British Empire, they were part of the Berne Convention. African countries have come to realise European copyright models not appropriate for them. In 1963, at the Brazzaville-Congo meeting, changes were proposed, but not accepted, and the UK's attitude to the proposed changes was hostile. The Stockholm Protocol term 'educational purposes' and the Paris wording have reduced the flexibilities in Berne.  Why more countries have not signed up to the Appendix to Berne is not known.  Action is required as the situation 29 years ago was very different from now both in terms of technology and rights-holders. 

The proposal that the Commission suggest changes to Berne, TRIPS and other international agreements was considered, and it was suggested that the system should not be too strictly enforced, and the good and bad practice of companies be made public. Perhaps could be broadened to include work being done for information of a higher or lower quality. It would be a type of kite-marking. For a government to measure business according to these criteria would be controversial. But there would be a market for that kind of information within organisations such as OXFAM. An example of a successful scheme would be OXFAM's green rating, or http://www.scidev.net/ which makes information available on the net to LDCs.

Recommendations and projects (eg Senegal) regarding copyright and TK were questioned on grounds of funding, economic reform and whether a system-oriented approach could help in this context.  It was stated that holders of TK viewed knowledge in custodial transmission as to whether there was interest in keeping knowledge secret or in divulging knowledge. Is IP the best method for protecting TK?  Most developing countries believe there is a need for sui generis protection in traditional knowledge terms, including protection for unauthorised use, access and appropriation. However, it is principally a non-economic interest that traditional folk have at present in their TK. Copyright is not the appropriate form to protect all forms of TK.  

Folklore is periferal due to the eventual failure of provisions developed by WIPO. No African nation has implemented the provision, but it is believed that the provision would not really work. Folklore does not really fit into the formal legislative models which exist.  WIPO has stated that documentation is essential, but lack of documentation should not destroy the ability of indigenous community to seek these rights.

Clarification of the criteria for protection of TK was provided; one must prove that it is traditional, distinctive and preserved. The problem with sui generis rights is there is no guarantee that it will be accorded protection outside the country, and it is difficult to define traditional.  Also there is the problem of loss of perpetual protection due to loss of distinctiveness, for example, the Keinte Cloth of Ghana has been copied so much over time that it has lost its distinctiveness.


Tour de Table on Key Themes and Issues for the Commission

* Examine the terms of IP license agreements - revisiting the UNCTAD model on licensing agreements for technology transfer may be a good starting point.

* Explore Berne Appendix 1 in terms of rights/exception to access to essential information for human development.

* Take note of the number of "open access" initiatives of all kinds which have arisen recently amongst publishers and copyright holders in favour of developing countries 

* Explore potential for, and constraints, to differential pricing for proprietary software and consider whether proprietary software would be less expensive for developing countries

* But bear in mind the other non-IP factors in the digital divide in developing countries.

* Consider merits of public education campaigns on open-source (non-proprietary software) for developing countries.

* Emphasize changes in practice rather than national or international IP law and treaties (e.g. good practice on access initiatives for developing countries or on fair-use in encryption technologies for digital rights management). "Burning Berne" is not an option. One means of doing this would be a "best practice and naming and shaming" type permanent international body or NGO to oversee practice by copyright holders (companies and countries).

* Build coalition/common cause for more public debate on copyright and development related issues. Consideration of how to achieve greater democratisation of copyright regimes and rule making: how to overcome "regulatory capture" of policy makers and policy institutions on copyright policy and get them listening more to needs of poor people. 

* Look at MPEG21- development of a standard for identification and description of content rights and rights owners. Consider whether if MPEG-21 is successful, it could facilitate the documentation of traditional knowledge formations in digital form that could be extremely helpful to indigenous groups and other custodians of TK in both exercising IPRs, and preventing others from doing so where appropriate.

* Put a strong emphasis on capacity building in terms of public education and public information in developing countries as a counter-balance to the decrease in access which stronger copyright protection leads to in poor countries (eg more resources for public libraries in LDCs).

* Think creatively about developing exceptions to copyright restrictions in international rules for LDCs (over the longer term).

* Bear in mind that, despite the growth of the Internet, hard copy materials are likely to remain the most important modality of accessing information in developing countries and LDCs.

* Regarding software and copyright rules (e.g. DMCA in the US), these increasingly reflect the interests of (proprietary) software producers and copyright holders. But within the system of rules, Commission should insist that policy/law makers are careful to ensure there is enough room for initiatives like open source software.

* Free software should be recommended for developing countries, it has a number of benefits as outlined in Story paper (e.g. access to programming code, encourages development of local IT programming skills). On the other hand, its not an "either or equation" in terms of open source versus proprietary software, and indeed it could be counter-productive to encourage developing countries to see it in this way. The two models can productively co-exist.

* Primacy should be given to private initiatives and market based solutions/reforms. Equally, in cases where there is market failure, priority should be given to tackling the root causes of market failure.

* We should constantly bear in mind that, from a welfare perspective, information is a public good, strongly linked to issues of public interest. While copyright protection is necessary, it should extend only as far as necessary and be very carefully measured against the nature of the information content as a public good.

* A specific example of the above, in the field of software, would be the regime towards reproduction and use of software programming code: too strict a regime here could inhibit local IT industries in developing countries and may not be necessary to protect the core business interests of existing software producers/copyright owners.

* Implementation of TRIPS Article 66.2 and Article 67 should be approached by developed countries within a paradigm of generosity rather than a narrow legalistic interpretation of what is the minimum required.

* Regarding copyright and software, Commission should concentrate on maximizing fair-use and areas where LDCs need exceptions in order to access and integrate with the global economy, rather than unpicking the exiting copyright regime wholesale in LDCs.

* Public policy makers in all countries should realize the fact that the "knowledge gap" between rich and poor countries is increasing and needs to be checked by policy instruments and funding initiatives (e.g. more resources for public libraries and internet access in poor countries would be one specific intervention).

* At the same time, the Commission should recognize and encourage the potential role of private philanthropy (e.g. the Bill and Melinda Gates Foundation).




1