A new instrument for digital trade?
When E. M. Forster began writing his novel Howards End in 1909, could he have imagined that the simple phrase “only connect” he uses in the work could have resonance for the burgeoning development in digital trade seen today? Naturally, Forster was writing of something entirely different; the need to connect man’s internal and external desires. And yet, we see also that this phrase is apt for digital trade: the need for a nation state to ensure an effective communications network not only within its borders, but also to the wider Net. Also, the transactions and content that flow over such networks are directly correlated with Moore’s law, which talks directly of the density of transistors on an integrated circuit board. But perhaps Moore also indirectly suggested the benefit of network externalities and the power of reaching millions of users through the effective interconnection of networks.
The major trading nations have been busy with FTAs and bilateral trade agreements incorporating TRIPS-Plus and WIPO Internet Treaty Provisions into respective schedules impacting the trade in digital intangible products. Many agreements talk of the interconnection of networks at both the physical and logical level. As communication networks have evolved from “real time,” the value and speed of communication has expanded exponentially. Indeed, the move from point-to-point dedicated circuit switched connections over copper cables to complex multi-point, packet switched connections over distributed fibre optic networks has fundamentally changed the global communication landscape. Now, complex audio visual media services run over a series of layers in the packet-switched chain. Dominance is not just about control of the physical infrastructure layer but all layers of the Internet Protocol Stack, from application and messaging to content layers, where most of the intellectual property resides. But the question is whether such trade agreements take account of these nuanced changes and will be stepping stones or stumbling blocks to any future multilateral provision on digital trade?
Could we make do with what we have to regulate trade in intangibles? This includes the GATT, GATS, and TRIPS as any intangible product could inevitably fall under one, or possibly all three of the framing agreements (a complex classification conundrum that is yet to be resolved). The Annex on Telecoms talks of using “any protocol of choice” in the delivery of services over public telecommunication networks and scheduled in a member’s schedule of specific commitments. This is an incredibly powerful and far-sighted provision dropped by the early drafters into the Annex and reflected now in some telecommunications schedules of the new trade agreements, such as KORUS (Korea-US agreement). The early GATS drafters could not have envisaged the wide array of applications such as Java, HTTP, FTP, etc., that we see today. And yet potentially all of these applications are covered by this simple wording. The Annex on Telecoms is an insurance policy for providers of services requiring access to a basic telecommunications network to deliver enhanced or value-added services, whether financial, distribution, legal, or other.
By contrast, the Basic Agreement on Telecommunications covers basic telephony services and incorporates further commitments on telecommunications services post the Uruguay Round. Also, there is the additional commitment in the form of the regulatory reference paper, which entails a combination of competition safeguards, including provisions on abuse by a major supplier and cross-subsidisation, interconnection of networks, licensing, access to spectrum and universal service amongst others. Is the Reference Paper (RP) fit for purpose for digital trade? Can it capture the sophistication of mergers and joint ventures that results in the most complex combination of services operating at multiple points of the internet protocol stack as provided by Google, Facebook, Microsoft, or Sky? Not really; it is meant to cover basic services. The RP talks of super-dominance and access to and control of an essential facility. This is more about control of infrastructure and less about control over the access gateways of the internet, both physical and logical. We need something more nuanced and able to cover complex packet-switched networks. Otherwise, we may see the issue left to dispute resolution to resolve interpretation conflicts in much the same way as the panel in US-Telmex passed judgement on the meaning of cost-orientation for telecommunications carrier traffic. The decision also simultaneously blew away the existing gentleman’s agreement on accounting rates for cross-border interconnection of public telecommunication networks between member state national carriers.
A number of WTO member states, including the US, have inscribed packet-switched data communication services as a basic service in their schedules of specific commitments. Basic Communication Services fall under the capture of the Regulatory Reference paper inscribed as an additional commitment by such states. The Reference Paper contains clear requirements on interconnection for all operators who are major suppliers of public communications services. All it takes is a further case brought to the WTO’s Dispute Settlement Body (DSB) on challenging the meaning of a commitment to packet-switched data services. If such services are classed as a basic service under the regulatory capture of the Reference Paper, would this blow apart the confidential transit agreements of international operators who control the international backbone networks for internet traffic? These services can generally be classed as packet-switched services. Would these predominantly US-based operators have to now negotiate transparent cost-orientated rates for interconnection of packet-switched networks as required by the Reference Paper rather than charge (confidential) higher commercial rates for transit interconnection? This would be game changer for smaller cash strapped internet operators in the developing world.
Multinational companies keen on cross-border data transfers and anti-localisation provisions would also balk at blocks to data transfers by member states keen to keep data local. In a range of FTAs and bilaterals, a number of the more powerful states, including the US, have restricted the use of performance requirements by developing countries associated with processing data and to keep data local. Often reliance by those states wishing to promote cross-border data transfers is placed on the annexes on telecommunications and financial services agreements that protect cross-border data flow in the ordinary course of business. But these annexes talk of data transfers being in “real time.” With complex packet-switched networks, data transfer is not real time; packets are disassembled and transmitted through a series of routers either by way of private Virtual Private Networks or over the public internet, mostly through a store and forward mechanism, and reassembled at their points of delivery. This is not exactly real-time. Is this a problem? Perhaps not. For example, the annex on telecoms talks of communication via any protocol of choice?
We do have a collection of far-sighted provisions within existing WTO agreements. The question is whether and how we make them work more efficiently for digital trade at the multilateral level, or continue to work piecemeal regionally, with the hope of folding new commitments into existing trade agreements.
Rohan Kariyawasam is a member of the E15 Expert Group on the Digital Economy. He trained as a telecommunications engineer and is currently Professor of Law and Director of Research at the Faculty of Arts, Law & Social Sciences at Anglia Ruskin University, Cambridge (UK).