It should reflect the economics of expanding networks aggressively to connect the unconnected
Raising issues of interconnection in any telecom regime is like opening the Pandora’s Box, from which fly out many explosive issues on policy, competition, costing methodologies, etc. Regrettably, in telecom, the Ecclesiastes’ lament, that “there is no remembrance of former generations”, is true. On March 25, 1997, the then new entrant cellular operators (today’s incumbents) took on the monopoly government operator, DoT, with similar demands of multiple points of interconnection and opposed a horrendous hike (about 24 times) of the (interconnection) tariff for landline-to-mobile calls—the incumbent could have crushed the fledgling new operators. However, the newly set up Trai ruled otherwise, and the rest is history.
The telecom sector is amazingly unique because each player, for its basic sustainability and customer benefit, it depends on being connected with rivals/competitors. A Taj Hotel does not need an ITC for its daily sustenance. Nor does a Fortis an Apollo, or a Maruti Udyog, a Toyota Motor Co. In telecom, however, the value accruing to consumers is not derived by using the service in isolation, but rather from having access to many other customers who also subscribe to the same set of interconnected networks. Hence, understandably, the avalanche of issues in telecom relating to interconnection. The ongoing dispute between Jio and the incumbents needs to be seen from this perspective. It may be couched in technical terms and principles, but viewed dispassionately, it is all about commerce and increased competition and the impact waves from a powerful new entrant. A replay of the 1997 struggle, with some role reversals!
Apart from the competition aspects, there are a couple of vital basic issues dogging interconnection policy and derivation of mobile termination charges (MTC) that merit serious consideration. Over the years, the analytical framework that is applied to MTCs has reached some level of consistency amongst most regulators across the world. This analysis is typically couched in terms of ‘bottleneck’ access: Put simply, the terminating operator provides the only network that can terminate calls to its customers—there are no adequate substitutes—and therefore, in the absence of regulation, this operator would have an incentive to raise the cost of termination. These are powerful forces and, given the strong incentives on operators to raise termination charges and the need to protect customers, MTCs are almost always subject to regulation. Obviously, MTCs should not be set in a vacuum and changes are not a ‘zero-sum game’ merely shifting value from one operator to another; their level and their path, over time, impact the attainment (or otherwise) of critical policy objectives.
The most important policy aspect for India is rural connectivity and bridging the urban-rural digital divide. The divide remaining almost as large as it was years ago—as the accompanying graphic shows—is distressing.
In rural India, where income levels are much lower, the customer’s incoming calls are 65-70% of the total. Hence, if the MTC doesn’t cover the cost of terminating the call, there is no incentive for the operator to roll out the network at great cost and with much difficulty in the villages. In Malaysia, some years ago, the regulator fixed MTC above cost in order to bridge the urban-rural divide existing at that time. Trai must consider this if Digital India is to be realised.
Mobile networks benefit from the economics of density. The costs per unit of output are lower when sites, distribution outlets and customers are clustered together. Thus, it costs more to serve rural customers. Given the proportion of rural customers has increased from 28% in FY08 to about 40% now, the average cost of serving customers has increased. Termination charges, thus, should reflect the economics of expanding networks aggressively to connect the unconnected. It is a double-whammy for the operator if it extends the network to rural areas at high costs and then bleeds due to MTC not covering the termination cost.
Moreover, the the urban-rural divide shown in the graphic is largely a voice connectivity divide. This is rather secondary for today’s India. Universal data connectivity and ubiquitous broadband are what Digital India demands, and, if one plots the divide on data, the picture is far sorrier.
In the two years preceding the reduction of MTC to 20 paise in April 2009, the industry registered a profit before interest and tax (PBIT) of R17,000 crore. After the 20-paise MTC, PBIT plummeted steeply, to a loss of R17,499 crore in FY12!.
Curiously, the Trai itself, in a July 2013 consultation paper, indicated that in FY12, the loss/under-recovery of cost was as much as R15 per subscriber per month—16% of the average revenue per user (ARPU). This means that the industry is losing about 4 paise/minute of airtime! Updating the ARPU to today’s actuals, and applying the inflationary cost escalations, the position would be probably worse, with about a R20 loss per user. Such a situation is harmful for connecting rural India. Our interconnection policy needs to address this aspect urgently.
Another view doing the rounds is that MTCs have to be reduced to zero, and we should have a BAK—bill and keep—regime, like there is in the internet world. This prima facie seems very attractive and is justified by claiming that telecom networks are becoming packet-based networks (PSNs), based on internet protocol. This is incorrect. While it is true that there is a conversion in progress from circuit switched networks (CSN) to PSNs, in a country as large as India—where CSNs are already in place and operators still have a huge voice demand (70%) from customers—it would be naïve to cost on the basis of a 100% PSN. Trai data show that almost all the mature and advanced networks of developed regimes follow MTC, not BAK.
To ensure investments on expanding network, the actual cost of call termination must be paid to operators. No wonder, most countries require MTCs to be cost-based and compensatory. This means the terminating network must be compensated for the value of the resources it uses to provide the service—including the capital cost of those resources. While the method of calculating the value of these resources may be debated, compensating the company for recources used must be the touchstone.
Incidentally, the Trai has tabled, in page 38 of its current consultation paper, MTC data of 34 advanced countries and the average MTC is $0.0197, i.e. about 2 cents, or about R1.33 per minute—10 times the 14 paise per minute currently charged in India. Little wonder that operators are not investing enough to expand networks and quality of service and that the urban-rural voice and data divide is unacceptably high.
A CY16 set of PPP-adjusted MTCs of some countries is given in the accompanying graphic. Many complex and theoretical discussions about whether MTC derivation should be by LRIC (Long Range Incremental Cost) or LRIC+ or pure LRIC or FAC, etc, still rage. The reality, however, is that most advanced economies have adopted LRIC or pure LRIC only a couple years ago, after attaining almost 95% coverage of both population and geography—generally, with both voice and data. India is far away from that. We have a billion connections, but these represent only about 600 million or less unique users and, that too, largely for voice. Thus, a LRIC—whether pure or hybrid—makes little sense. We must first connect the unconnected.
India needs to urgently revamp its interconnection policy if we are to bridge the urban-rural connectivity divide and move expeditiously toward Digital India.
The author is honorary Fellow, IET (London), and president, Broadband India Forum. Views are personal