The Telecom Regulatory Authority of India (TRAI) expects to conclude public consultations on ‘net neutrality’ this month, culminating a process that began last year, with the regulator prohibiting “discriminatory tariffs for data services on the basis of content”. While considered a victory for neutrality evangelists, who want the internet to be free of monopolistic interests, it left several questions unanswered.
The central question is: under what circumstances can a network operator discriminate against applications or content available on the internet (a network of networks)? TRAI has the motivation to resolve this and other related questions, and the power to issue quality of service (QOS) regulations for telecom services (that run most Internet networks in India), as well as wider policy prescriptions that the Department of Telecommunications (DOT) may subsequently actualise.
In the latest round of consultations, it has emerged that TRAI is keen on contextualising its prescriptions, to fit India’s realities – which is a good sign. To this end, we propose that there are at least five fundamental realties that the regulator should consider (a) the prospects of consolidation in the telecom sector which may lead to long overdue rationalisation of revenue streams, (b) the absence of competition at the last mile of the distribution chain, (c) the imperative to connect the millions of unconnected Indians to the Internet, (d) the preponderance on wireless connectivity which requires greater traffic management (QOS) than wired connectivity, owing to limited spectrum and (e) technological ‘convergence’ of content delivery platforms. The following five ideas derive from these five realities, not necessarily in sequence.
First, innovations in network technology must be allowed to keep pace with consumer demand trends. Today, a large proportion of demand for data is linked to seamless access to video content, which accounted for 65% of total Internet traffic in Asia in 2015. And India’s inherent advantage in producing video content, with a vibrant media and entertainment industry, is clear. Technology has enabled entrepreneurs to develop content without investing heavily in production equipment. With low entry barriers, video content can become ubiquitous, and creative industries can potentially flourish with convergence of delivery platforms if requisite policy support is forthcoming. Therefore, even as the network will evolve over time, network regulations must be light touch and demand-led.
Second, a golden median between an ex-ante and ex-post approach is possible to achieve, if flexibility is at the core of regulatory ethos. Traditionally, internet service providers (ISPs) have favoured the ex-post approach as it makes strategic sense for large regulated companies: it gives them greater room to manoeuvre in terms of QOS, and large companies have the capacity to pursue resource-intensive litigation. Content providers on the other hand, have more to gain with a stronger ex-ante approach as it provides market certainty and predictability. Specifically, such an approach can help content providers avoid expropriation by ISPs (downstream monopolies beset the broadcasting supply chain for instance, a fact acknowledged in previous TRAI consultations), while ensuring some transparency in otherwise opaque traffic management operations.
Regulations need not reflect this binary. In fact, it is undesirable to emphasise one approach over the other. This would make little sense considering the pace of change in technology and economic incentives versus the inherent rigidities associated with regulatory precedents set in courts. TRAI must recognise that old norms may not fit next-generation innovations and that courts are not the best place to decide technological pathways. Any ex-post-facto approach should rely on technical expertise, extensive peer review and investigation. At the same time, a rigid ex-ante approach defies the very virtue sought to be protected by proponents of a free Internet – unfettered innovation.
India should adopt bright lines for ex-ante regulation, that should in turn be malleable and reflect the needs of a future population. That is, at the heart of the regime should be a recognition of the need for QOS, subject to regular review. This would ensure that broadband demand is not artificially suppressed owing to inadequate infrastructure; at the same time, norms should not encourage de-facto reliance on QOS, over improvements in network infrastructure. Net neutrality principles should instead help test whether a purported discriminatory practice is a deliberate attempt to reduce the quality or availability of a service, and responses to each principle should be verifiable.
Third, in lieu of verifying whether a practice is deliberate or not, TRAI must collaborate with independent actors for big data analysis. Statistical tests can be applied to samples of data received from ISPs as well as users, to ascertain the reasonability of traffic management practices. For instance, a practice may be reasonable, if data proves it to be exceptional and temporary. There is little doubt that such a measurement is not easy and will necessarily involve monitoring agencies and end users – but at least the scarcity of data is not a binding constraint, as operators maintain meticulous QOS logs. Although difficult to envision today, it is expedient to explore a co-regulatory model wherein end users – individuals or institutions – can empirically verify the claims of ISPs. Without such recourse, any transparency and disclosure requirements for checking adherence of networks to ex-ante principles will be unverifiable and, therefore, redundant.
Fourth, an effective network neutrality framework must centre on efficiency, which is where engineering and economics converge. For instance, large content providers are increasingly reliant on content distribution networks (CDNs) which are clusters of servers that bring content geographically closer to consumers, and concomitantly provide better QOS. There is a concern that CDNs can lead to de-facto prioritisation of data flows. There are two ways of looking at this: the first is that CDNs allow larger content providers which have deep pockets to invest in delivery infrastructure, to capture markets. The second, more palatable answer is, that while all data of a similar class should flow equally in a network, data that is closer to consumers, should naturally reach faster. Enhanced efficiency should not be penalised. Only economies that embrace Schumpeterian disruptions, that help deliver goods and services to consumers cheaper and faster, both online and offline, can remain globally competitive.
Fifth, following from the CDN example, it may also be inferred that a general principle could be developed wherein organic prioritisation of data flows is acceptable whereas forced prioritisation (based on modification of traffic protocols) is not. That is network engineers should not unreasonably discriminate between data flowing from point A to point B. If a third point C is physically closer to point B, the data from point C should be treated the same as data from point A, and it should organically reach the consumer faster. Such a framework may foster much needed last mile competition by lowering transit costs incurred by smaller ISPs on core networks.
The global discourse on net neutrality is by no means settled. Even as India makes its first holistic attempt to create a suitable regime, there are apprehensions that the US, under a new administration, may undo its extant framework which had set a benchmark. Nonetheless, perceived positions of other countries should not be a deterrent as our appetite for data is growing exponentially and a nuanced regime balanced by Indian realities can spur access and innovation.
Vivan Sharan is a Partner, and Prachi Arya is Head of Legal, at Koan Advisory Group, New Delhi.