The counters to several submissions are grouped into the following broad points of conflict, where we differ:
1. There are other services where differential pricing is allowed
In several submissions1, one of the main arguments forwarded is how in case of delivery of services like electricity, water, hospitals and airlines, differential pricing is a blessed method for making use of consumer surplus.
It is acknowledged that this practice is prevalent in the provision of other services, however, the reason for resisting differential pricing for the internet is that the internet is distinct from the nature of the other services mentioned as examples.
While TRAI is a market regulator, the debate is hijacked wholly by market considerations. Connectivity to an open and free internet has been monumental in being an equaliser for the marginalised, by enabling access to knowledge and community that otherwise would not have been available due to spatial and geographical and cultural constraints. For those persons who face intersectional disadvantages of not only being poor and underserved, but also other barriers from identity politics, like the disabled, the sexual minorities etc., the internet is a portal of possibilities.
IPRG in its submission notes that
“…the banning of zero rating would result in the poor and developing nations blocking sure avenues of growth and prosperity”.2
In banning zero-rating, one avenue of growth of the market may be closed, but if allowed, it would threaten to foreclose not only foreseeable instances of permission-less innovation, but also use-cases that cannot be predicted. It would empower users to fully exercise their rights to freedom of expression, association and other human rights as the Internet currently allows.
The International Centre for Law and Economics cites Dropbox as an example of a service that differentially prices its services.3 One cloud-storage service doing so by offering free and paid versions of its service is by no stretch of imagination the same in effect as offering a limited set of websites for free. Dropbox does not shut off price sensitive consumers’ access to a whole range of other services and website which defy classification into categories.
In its statement that,
“There is nothing new or unique about internet companies that would justify breaking from this approach — in fact there is much to be gained in continuing to allow differentiation as internet platforms discover better ways to enhance consumer welfare”4
the submission reveals a refusal to acknowledge the internet’s open nature as specifically being different from voice or other services, and for precisely this reason requires that TRAI intervene.
Reliance claims that it is unproblematic to treat the internet and all its gifts the same as cable TV.
“This is similar to the DTH industry wherein TRAI has allowed packaging of various channels wherein consumers also have a choice to avail these channels in a-la-carte format, although the package is offering all these channels at a much cheaper rate compared to their a-la-carte value.”5
It is undesirable that there be a cableisation of the internet, and TRAI should recognise that the internet is distinct from a service like DTH!
2. Forbearance has been at the root of telecom growth
Much has been said about how the forbearance regime has been responsible for the explosive growth in telephony in India, and while that might be true, exceptions exist for reasons exemplified by the current situation. It is undoubtedly a great business opportunity for telecommunications service providers to expand the market through differential pricing, but in structuring the market in irreversible ways, it harms consumer welfare. TRAI has noted that exceptions have been made, and will be made if it thinks it fit. It is an important moment to invoke this exception by TRAI.
Somewhat counter-intuitively, the continued growth of the internet actually depends on TRAI stepping in ex-ante to strictly regulate against differential pricing, and allowing for accessibility to be free and open. Not regulating right now would skew the market in such a way that future possibilities will be closed off, because the free and open nature of the Internet cannot be taken for granted anymore. That not only harms consumer welfare, but also the orderly nature of growth of the sector.
Once the uptake is high and broadband penetration has reached a critical mass, TRAI could consider going back to the forbearance regime.
3. “Innovation” as used by those in support of the telecom sector versus “innovation” as understood by the user
Several submissions6 exalt the power of innovation in tariff pricing etc. The word “innovation” as used by the TSPs, is a proxy for “innovative ways to expand business and maximise revenues”. While this surely involves innovation, allowing for a free market of ideas by not offering an inferior connectivity to a chosen few websites is the kind of innovation that characterises the internet. Room for innovation, in the way that is truly valuable for the user (both creator and consumer of content), is when the user is allowed to choose his/her interactivity with the internet. Even in the TSPs’ best wisdom and good intentions, it is not possible to capture the benefits of permissionless innovation with gatekeeping. Simply because Idea Cellular chooses to call collusion and deal-based tie-ups between TSPs and content providers a “synergy” approach, does not make it so.
Further, Idea goes as far as to state that
“Standard mobile data offerings do not generate interest in non-user/casual user who is still discovering the role internet in his life and society A standard offering of mobile data without any differentiation in terms of bundle on application, content or access will not be beneficial in driving adoption as straight jacket solutions will not work. For a billion Indians, “one size fits all” approach has never worked in the past and it will stifle growth. Every Internet user is different with different usage needs for e.g. younger consumer may be focused on messaging and communication, while an older consumer may have higher need for health or commerce (similarly rural consumer will differ from an urban consumer), the principle of “Different strokes for different folks” clearly needs to be applied.”7
In our opinion, Idea is right in concluding that a one-size-fits-all approach would not work for diverse users whose utility of the internet cannot be preemptively and prescriptively declared. It is precisely for these considerations that zero-rating or a limited number of websites would not offer the benefits that a free, open and secure internet would. Users should have full freedom to decide which parts of the internet works best for them as their lives, interests and circumstances evolve, which no TSP or platform provider can foretell.
4. Emphasis on the fact that there is cut-throat competition for the market to correct itself
Vodafone, like other TSPs claims that
“given the intense competition in the market with the presence of atleast 8-10 established operators in each telecom circle and the presence of strong regulatory and legal safeguards, the concerns around discriminatory or anti-competitive behavior can be addressed through existing mechanisms”.
While it is true 3-4 big telecom companies are competitive and seek to expand their market share, cartel behaviour is not new in the Indian telecom sector. One of the smaller telecommunications company, Citycom has noted its concerns about smaller TSPs being unable to compete meaningfully with the larger telcos should differential pricing be allowed.8
5. Same service, same rules
In this debate about what is a non-arbitrary way of classifying consumers to be able to differentially price them, many telecom companies have argued that ‘Same service, same rules’ should apply.9 A classification is sought to be made on the basis of whether an application allows users to communicate or not. But this distinction is simplistic, erroneous and framed in a way that speaks in binaries of what does and does not directly threaten TSP’s interests. In the consultation paper on OTT services, extensive arguments have been forwarded to TRAI about how this is a false distinction to make.10 This distinction is unnecessary in the current debate and should not be confused with the matter at hand.
6. Lack of data
Supporters of differential pricing often attack supporters of net neutrality for thinking theoretically about questions without empirical data, but there is barely any reliable and peer-reviewed data forwarded by companies like Facebook about rates of conversion to the wider internet etc. In the face of absence of data in the public domain, TRAI should ensure that characteristics of the internet which have been widely responsible for the Internet’s enormous positive impact should not be tampered with.
7. That CDNs are not pulled up for violation of net neutrality
Performance advantage is definitely a driver in the adoption of services on the internet, in a world where the user experience of being able to access something a second sooner leads to it being better liked and adopted rather than services which don’t have them. However, the ability of optimizing network traffic is different from using the limited spectrum resource to do certain things. As submitted in our comments to the consultation paper on OTT services, separate examination of methods and consequences of different traffic management is required.
8. Due to a double-sided market, users are best served through differential pricing
The internet is not as simple as a “double sided market”. Users are consuming and creating value and content for the internet at the same time, and blurring the ostensibly different sides.
“Multi-sided platforms — like Google’s search services, Facebook’s social network and many others — require critical mass on every side of their platform in order to maintain viability. Platforms must encourage participation from all parties — users and developers — to match supply and demand, and it is squarely within their interest to be as generally nondiscriminatory as possible. Thus, far from being a “gatekeeper” — that is to say a bottleneck in a process that uses its power to extract rents — these platforms typically operate using open API’s that encourage outside developers to design services for as many consumers as possible.”11
Contrary to what is expressed above, gatekeeping by platforms can be done in many ways, which are not always overt. Small players on the internet, whether it be small start-ups or users who double up as consumers and creators of content, but are not “creators” in the strict business sense would likely not be able to conform to the technical specifications of platforms and would have to jump through hoops to get onto them.
Further, to take the example of one of platforms that the ICLE submission defends, in on Free Basics, all traffic is routed through Facebook’s servers and decrypted and re-encrypted by Facebook and sent along the way to the destination. Even for sizeable competitors of Facebook who can meet the technical specifications, it is problematic that Facebook views and analyses all traffic on the platform. For a single company to have massive behavioural data of users across multiple applications and websites that the platform allows access to, is catastrophic.
Reliance has submitted that gatekeeping concerns are theoretical, but there is evidence to the contrary in the Indian telecom sector itself. According to Viren Popli, former head of Mobile at STAR,
“In the case of MVAS, in order to be allowed to deliver mobile content to those users who wanted it and were willing to pay for it, a provider would have to go and negotiate with the telecom operator, who would play gatekeeper, and try and squeeze as much revenue share out of content providers as possible. For smaller content owners, there was pressure to offer content exclusively to one of the largest mobile operators, on their terms. Once a deal was negotiated with one telecom operator, content provides would find that others were made aware of the terms with one: it was clear that there was collusion between telecom operators, thus ensuring that the content provider gets the worst possible deal. Many content providers were forced to partner with specific aggregators, who had their own terms. Because of this, the consumer got the worst possible deal, and the content provider had no freedom.”12
Further, International Centre for Law and Economics submits that
“As of 2012, for instance, Facebook alone had over nine million connected apps available to users of its platform, 17 and as of 2014 over 300,000 different developers integrated with Dropbox to provide the cloud storage service to their users. Thus, the economic incentives faced by application providers impel them not to restrict access, but to open it — frequently by charging nothing for either developer or end user access — in order to ensure that the platform is widely adopted. It simply makes no economic sense for an internet platform to engage in the sorts of discriminatory activity that would cause it to lose critical mass on any side of its platform.”13
This is a naive understanding that a global giant corporate acquiring big players in the respective service areas bodes well for application providers generally. For Facebook or other platform provider to have that kind of control further strengthens its market powers vis-a-vis any competitor, actual or potential. Even if other entities can initiate their own packages, platforms with popular big players will always be attractive, due to network effects. So zero rating by definition is likely to further the market power in favour of big players.
IPRG statements like
“The evidence from political movements in several countries (such as Arab Spring) is that zero rating of social media platforms greatly expands freedom of expression and political freedom”14
is false. In fact, if there were centralised points of control available within the country, the way zero-rating by TSPs would enable, these kinds of movements are very likely to have been stamped out.
In summary, for the consultation paper, the following is our response to the submissions made.
Question 1: Should the TSPs be allowed to have differential pricing for data usage for accessing different websites, applications or platforms?
At least until such a point is reached where the uptake of mobile data services is significant and a large number of now underserved users have adopted to the internet medium that is free, open and secure, TRAI should not allow differential pricing of data services.
Question 2: If differential pricing for data usage is permitted, what measures should be adopted to ensure that the principles of non-discrimination, transparency, affordable internet access, competition and market entry and innovation are addressed?
Telecom companies have argued that the FRAND principle provides sufficient checks and balances on a case-by-case basis to ensure consumer welfare and orderly growth is protected. However, the case-by-case basis of regulation that this would involve is a huge regulatory burden. Given that the auditing of data service offerings would not only have to ensure that platforms are sufficiently open, but also see how traffic to non-participating content compares with zero-rated or differentially priced services. Business arrangements between players and the technical compliance are opaque, and compliance monitoring burden would be enormous.
Question 3. Are there alternative methods/technologies/business models, other than differentiated tariff plans, available to achieve the objective of providing free internet access to the consumers? If yes, please suggest/describe these methods/technologies/business models. Also, describe the potential benefits and disadvantages associated with such methods/technologies/business models?
There have been several alternatives suggested in the submissions, like putting technical constraints like volume caps and sponsored data models instead of cableisation of the internet. These, along with community networks should be catalysed instead of compromising on a free, open and secure internet.