Are we putting the cart before the horse?

Trai has taken a technical view of net neutrality to the neglect of India's circumstances and policy imperatives

There is no universally accepted view on what constitutes net neutrality (NN). This is why regulators evolved certain basic principles: no selective blocking or throttling and transparency in traffic management. In 2015, I said that these ought to be accepted as inviolate (The Wire, May 13). And, this is precisely what Trai has recommended. The surprise is that when its own consultation paper (March 2015) argued for these principles why has it taken the authority so long to reach this conclusion. Most disappointing was that it failed to situate the matter in the larger public policy context.

Trai has taken a technical view of the entire matter to the neglect of India’s circumstances and public policy imperatives.

India’s circumstances: Rural teledensity 40 per cent; 50,000 villages with not even 2G connectivity; internet access in rural areas not worth mentioning; spectrum costs 10 to 20 times that in the rest of the world; lowest tariffs in the world; highly indebted telecom industry; financial distress in most telecom companies. Public policy goals: ensure affordable access to all; rural teledensity of 70 per cent by 2017 and 100 per cent by 2020; 600 million broadband users by 2020; promote investment in infrastructure and development.

Now, congestion lies at the heart of the debate on speed and quality of service. First, if there are constraints of bandwidth (and speeds), network overload can become a serious problem. Second, when bandwidth is constrained and data-hungry applications compete for the same bandwidth with others that are relatively data-intensive, there is a clear crowding-out problem (This happens even if the second application is not data-intensive). One consumer hogging bandwidth excludes another user’s availability; hardly “neutral”.

Economists agree that the only way to deal with this is that costs be borne by the person creating the negative externality. This is why Nobel laureate economists in the US argue against the purist version of NN. It is also the rationale for enabling broadband suppliers to treat data-hungry applications somewhat differently (in pricing) from others. The counterargument is simple: if networks are overprovisioned (as they are in developed countries), then the case for congestion is significantly weakened.

An even more difficult public policy problem pertains to investment in infrastructure, spectrum etc. to deliver broadband and access to all. In India, investment was largely driven by the private sector through licence conditions. True, there is the Universal Service Obligation Fund and other public investment. But we are nowhere near access to all, forget broadband delivery.

The public policy problem is this: the sector needs an annual investment of $15 to 20 billion. Where are the resources going to come from? The private sector’s profits are measly and the recent tariff war has not helped. The fisc is under strain, so massive public investment is a pipe dream. And, for all the hoopla, BharatNet is some time away. So, are we unconsciously expanding the Digital Divide?

On congestion, even the haves carp about quality of service. The availability of spectrum (40 per cent of what is provided in China) and extent of “overprovisioning” in India’s networks is severely limited. Once again, this is a problem of investible resources. India, therefore, needs to view NN in its larger public policy perspective. Even with high tariffs and profits, universal access has not been achieved in many developed countries. Hence, India’s situation on access and congestion is truly grave. The question that arises is this: how do the telcos raise more internal resources?

India will need to consider some forms of departure from pure NN if additional revenues are to accrue to the telcos. (TRAI has conceded this reality — specialised services). That is the only way resources for investment will materialise; and, only investment can tackle problems of congestion and expanding access. Some practices have a strong economic rationale, for instance treating data-hungry applications differently. The trick is to consider those practices that do least damage to the NN framework.

There should be no misunderstanding. Core principles on NN such as selective blocking and throttling and traffic management are not negotiable. This article is a plea for taking a pragmatic approach to NN. We need to customise our approach to NN to fit in with our circumstances and policy imperatives as also to bridge the widening Digital Divide. In the US (and elsewhere), the NN debate is about how large profits are distributed across telcos/ISPs and the OTT service providers, viz. business versus business. In India, it is about how to achieve public access to broadband quickly in a tight resource situation.

The author is ex-chairman, Trai