An appropriate form of Net neutrality for a digital India


An appropriate form of Net neutrality for a digital India

  • facebook
  • linkedin
  • twitter
  • whatsapp

Prof. Arvind Sahay


Net neutrality appears to be in fashion. But what does it mean exactly, and what kind of its application would enhance the probability of reaching the goal of Digital India faster? 

Intenet access in India is predominantly wireless. As of May 2015, an estimated 270 million Indians have some Intenet access and approximately 80% of them are using mobile devices. As a policy goal, I am going to assume that by 2019, we would want at least 750 million Indians with Intenet access at a reasonable price and that 95% of these Intenet users will access the Net using a mobile device. In effect, therefore, Intenet access in India for the aam aadmi (common man) will be wireless unlike the US, where the on-ramp is still substantially fixed line. Wireless access requires spectrum.

The total spectrum that is available to all mobile operators in India, as of May, is across 800MHz, 900MHz, 1,800MHz, 2,100MHz and 2,300MHz bands. Bharti Airtel Ltd just renewed or bought new rights to about 120MHz block of spectrum in the March 2015 auction. As per news reports, in October 2014, the Airtel network carried an average of 5,200 terabytes (TB) of data daily. Airtel has about 20-28% market share in India. Extrapolating to all networks on a pro-rata basis at 20% market share, the total mobile data traffic in India would be about 26,000 TB daily, assuming the current number of base stations and 3G data usage. 

The first question, therefore, is that since universal Intenet access requires capacity and the capacity bottleneck in India, given the nature of the on-ramp to the Intenet, will be spectrum, even if a fibre optic backhaul exists from a base station, what is the capacity that we have and need? 

Let us remember at this juncture that the goal is a Digital India that has Intenet access on demand at a reasonable price for the end consumer. What is a reasonable price and consumption of data for the end consumer? 

Let us make another assumption that a reasonable price is Rs.100 per month for 300 megabytes (MB) of data (this is slightly cheaper than most plans out there today on a prepaid basis but actual consumption can vary hugely). That would mean 3 MB of data per person per day. As many as 750 million people doing 3 MB per day on average would be 2,250 million MB (2,250 TB per day). Which appears to be well within the data-carrying capacity of the networks. Of course, data requirements around the six large metros are much higher. 

The second question is, how will Net neutrality, if at all, contribute to the goal of universal digital access, given these limitations of the on-ramp to the Net? To answer this question, we will first delineate Net neutrality. 

Take the following four features of a “strong form” of Net neutrality, which proponents want to preserve because they promote competition and consumer choice (a) It should be possible to reach between any two end points of the Intenet without any restrictions by the end user (b) Each end consumer should be able to get content without any throttling or discriminatory treatment, regardless of the type or content of data by the telecom operator (c) Information on traffic management practices on the Net should be available to the public and Intenet service providers (ISPs) must be transparent about these practices, and (d) Network service providers (ISPs or telecom operators) should not prevent or reduce access to any website through their actions. This definition, by definition, if implemented, should provide access.

Notice that in the above definition, nowhere is there a requirement that any person that accesses the Net should not pay to use. Also, consumer welfare requires that there should be competition and that providers should be able to practise price discrimination. The third question, therefore, is that what level should these four dimensions be operationalized in India—each country has followed what works best for it. 

The idea is that more end consumers (the aam aadmi) would use the Intenet through mobile devices more, if they saw more access to websites and apps. Simultaneously, more service providers (the Facebooks, the WhatsApps, the Hikes, etc.,) and apps (the Flipkart app) would use mobile Intenet and use the last mile provided by the telecom operator if they saw more end consumers (in fact, service providers do not have a choice—they must use the mobile platform to access consumers). Volume growth in one market reinforces volume growth in the other.

The premise of Net neutrality activists is that only a strong form of Net neutrality will lead to this virtuous circle. The second assumption is that a “reasonable price” for access to the Intenet will provide sufficient monetary heft to telecom providers to build and operate the wireless infrastructure and the long-distance fibre optic backhaul. Are these correct?

Putting it all together, therefore, if there are capacity constraints, one needs to either create more capacity by asking the telecom operators to pay more to the govenment or the over-the-top (OTT) service providers to pay the telecom operators—when there is sufficient capacity, meeting all four conditions strictly is not difficult. 

In India, arguably, capacity will be a constraint, though the preliminary calculations suggest otherwise. So, someone has to pay for creating the capacity, if that is needed. Or there has to be rationing. Rationing in the offline world has not really worked. And in any case, why should Indian citizens suffer from data rationing? 

So who pays? Arguably, OTT providers and e-commerce firms that depend on data as a critical “raw material” for their business should pay—at least, that appears to be the argument of the telecom operators. For example, if an e-retailer depends on 20 TB/day of data traffic to its website to fuel its business, why should the cost of all that data traffic be paid by the end-consumer? Why not the e-commerce retailer? And if they do pay the telecom provider, would it violate Net neutrality?

As of now, only end users in the mobile and Intenet ecosystem pay the telecom operator for access for the movement of traffic on the Intenet. Intermediaries such as WhatsApp, Facebook, etc., do not pay. Charging intermediaries does not mean that free movement of data between end points, transparency of charges and of prohibition of throttling go away.

In a market like India, where the capacity of the citizen to pay is limited and the goal is to increase digital access, it may make marketing, economic and political sense to charge according to what the market will bear. Individual end users are able to pay less and use less data so they should be charged less. OTT service providers use the Intenet more and are able to pay more, so they should pay more.

The key is that there should be no paid prioritization: telecom operators should not favour some lawful Intenet traffic over other lawful traffic in exchange for consideration of any kind—in other words, no “fast lanes”. This rule also bans ISPs from prioritizing content and services of their affiliates. In other words, on the Intenet, while one can charge for more volume of traffic, one cannot change the pre-sold level of speed of traffic flow. Here again, capacity is an issue. With sufficient capacity, there would less of a Net neutrality debate.

The (currently hypothetical) extra payments that OTT service providers make should then be used to build more capacity, increase the efficiency of the network and provide more access.


Arvind Sahay is dean (alumni and extenal relationships) at the Indian Institute of Management, Ahmedabad (IIMA). Prior to this, he had been teaching at the London Business School. Sahay is a B. Tech from IIT Kanpur and holds a PhD from the University of Texas, Austin. He is also an alumnus of IIMA, from where he received his PGDM.

This article presents the author’s personal views and should not be construed to represent the institute’s position on the subject.


First published in 'View from IIMA' column in Mint.