Last month, the Telecom Regulatory Authority of India (Trai) recommended to the government that telecom operators be allowed to share unused spectrum. This is a logical and a necessary first step, but hardly a leap. The government will have to take more substantive measures than spectrum sharing to deliver the amount or type of spectrum that companies need urgently.
India has a bigger stake in spectrum than any other developing country including China. Wireless devices like mobile phones need spectrum to work and over 95% of Indians have no other means to make telephone calls or access the internet.
Spectrum sharing makes sense for consumers as well as government. Trai points out that the nature of spectrum sharing is such that the resulting capacity to carry wireless traffic exceeds what players had prior to sharing it. So, a company with a large subscriber base can reduce call drops in busy commercial areas if it shares spectrum with a later entrant who needs less spectrum to support its new services. Additional traffic means more revenues for operators and more for government, since it gets a significant share as licence fees and spectrum usage charges (SUC).
Trai?s current recommendations on spectrum sharing are less restrictive than those it made two years ago. If the government accepts the new proposals, any operator can share any spectrum allocated to it and not just the part it acquired in auctions, as Trai had recommended earlier. It will need no permission from the government to share spectrum. They need to only inform the WPC (Wireless Planning and Coordination Wing) and pay a fee of R50,000.
Trai was right to set up a committee of operators? representatives to advice on spectrum sharing. This would have helped produce a framework that meets most industry concerns and anticipates several, if not all, of the reservations that have caused the government to delay a decision on spectrum sharing. These recommendations will assuage some, if not all, of those fears.
The proposals for sharing will not allow companies to escape future auctions. The framework allows a company to only augment the spectrum it holds, but not to acquire spectrum in a different band. For instance, it cannot share, say, 900 MHz spectrum belonging to another player if it never obtained any spectrum of its own, either administratively or through bidding in an earlier auction.
The proposals create no new rights. A player will not be able to use sharing as a backdoor entry to obtain fully liberalised spectrum against current rules which require that administratively acquired spectrum be used only for its originally stated purpose (for example, 2G services). New proposals will not allow a company sharing spectrum to deploy any technology that it was not permitted to earlier. To be used for all approved technologies, both companies need to have acquired it through an auction.
Companies sharing spectrum cannot escape any obligations. Either company will still be bound by the terms of its licence to roll out services in the prescribed manner which includes covering 10% of all district headquarters in 1 year and 50% in 3 years of allocation of its spectrum.
Existing liabilities relating to fees will continue. As will market caps on spectrum holdings for purposes of mergers or acquisition. Half the spectrum shared by one player will count towards the other players? spectrum holding and vice-versa. The players will also pay an additional 0.5% of their adjusted gross revenues (AGR) as SUC.
All this will soften the resistance in the Telecom Commission (TC) and the Department of Telecommunications (DoT) who arguably see protection of government revenues as a bigger responsibility than any other.
However, it would be naive to overstate the impact of spectrum sharing or see its role as somehow transformational. For a start, sharing is not a technically trivial task. Equipment already deployed by a player can make it tougher to combine networks through sharing. But the bigger challenge is that India?s operators have some of smallest allocation of spectrum in the world. They simply do not have much to share.
Private 3G players will have little to celebrate about sharing. There are major gaps in their networks since none won spectrum in all the circles. They have little to gain since sharing cannot help a player who does not already have spectrum in the same circle. The 3G operators have implemented intra-circle roaming to plug the gaps. However, the government wants the Supreme Court to rule on its legality. Not surprisingly, 3G players are urging the government to release spectrum from other sources; for example, defence, which has significant amounts of unused spectrum.
Further, 4G players like Reliance Jio, with 2300 MHz, too, might be disappointed. They could have used the 1800 MHz spectrum for LTE voice to augment their existing data services in 2300 MHz (earlier this year, 4G player Reliance Jio bid for this spectrum in some circles). However, since most of the 1800 MHz spectrum was allocated without auctions, an operator sharing it will be restricted to its original 2G technology.
So, while spectrum sharing will ease some pressures, it cannot tame the rising demand for spectrum from increasing use of data services, especially multimedia and video.
Clearly, it is spectrum trading or leasing that would enable telecom operators to make robust and flexible choices. It can help create a much-needed secondary market for spectrum where players can acquire or sell some or all of their spectrum based on commercial considerations.
However, spectrum trading is easier said than done. A competitive secondary market in spectrum is bound to drive spectrum prices down and hurt the proceeds of auctions. A commercially driven transaction could pose new challenges to computation of operator revenues. Will inter-operator payments for spectrum be excluded from adjusted gross revenues, as interconnection charges are currently? If not, will Trai determine a base price for traded or leased spectrum or will operators take that call? These are complex questions.
Spectrum trading and related reforms like introducing mobile virtual network operators are central to efficient use of spectrum. But they could undermine the revenue sharing regime for government levies. They would make it tougher to accurately estimate or attribute operator revenues for computing government levies.
However, the answer is not to abandon spectrum reform but for the government to move to a method of determining levies that is simpler to operate and protects the government?s targeted revenues. There are several, including basing fees on the size of an operator?s spectrum holdings. But continuing with current revenue sharing approach will unfortunately doom the sector to ?weak? regulatory solutions like spectrum sharing or the earlier move to unified licences. They are right in principle but will probably change little.
Mahesh Uppal
The author is a telecom consultant.
