In Econ 101, we are repeatedly reminded that incumbent firms expend considerable amount of resources in attempting to raise the height of entry barriers in order to create conditions that make doing business costly for rivals. After all, competition, while good for the consumer, is not very helpful for producer bottom lines. Thus, attempts to limit it are as prevalent in India today as they were during the licence permit raj, although methods have changed. The question is, whether outcomes are different? Clearly, in terms of access, price, reliability and reach, the outcomes in Indian telecom are far better now than in the wretched days of an inefficient public sector monopoly; although we argue in this piece, keeping the recently concluded spectrum auctions in the backdrop, that there’s a lot of unnecessary gaming of the system in play that thwarts full benefits of the telecom sector from being realised.
The transformation of India’s telecom landscape from the days of an insidious monopoly to the frenetic competition of today is nothing short of spectacular. Aided in part by rapid technological progress and in part by an increasingly liberal policy environment, the makeover, however, did not happen overnight. DoT to start with and then its successor BSNL were not exactly welcoming of private participation and sought to enfeeble new entrants through both fair and foul means. The latter included what can crudely be characterised as rent-seeking—a term coined by economists to reflect expending resources on activity that increases individual profit without boosting overall welfare.
The effects of rent-seeking activity have manifested in delay in the emergence of competition, excessive and avoidable litigation and, of course, loss of the famous and fabled growth dividend from telecom. According to latest available estimates, every 10% increase in the number of internet subscribers in India leads to a 1.08% increase in the growth rate of state-level GDP. This calculation is vastly underestimated compared to the potential since the positive spillovers of pervasive broadband penetration have not even begun to be realised. Conservatively, the number could be double the earlier estimate. Besides, mobile networks spawn new businesses routinely and inject efficiency in a range of market-based transactions. Over the last decade, and particularly over the last five years, India has registered impressive growth in telecom, soon to cross a billion telephone connections! Over 95% of these are going to be wireless, making spectrum an invaluable resource. An overwhelming share of internet access is and will continue to be on wireless devices.
Now, cut to the 2015 auctions. Several incumbent licences were expiring and for business continuity reasons they needed to win back airwaves that were assigned to the common pool. Else they risked losing assets worth several thousand crores of rupees built over the last two decades. The incumbents whose licences were about to expire were challenged by new entrants and by other incumbents whose licences will expire a few years later. With different rules governing these bidder types and with different amounts of spectrum available across bands, the auction produced a bonanza for the government, huge cash commitments for incumbent operators and a significant amount of unsold spectrum especially in the 2100 MHz band valued by the government at R7,935 crore.
The simultaneous and counter-intuitive occurrence of very high bid prices especially in the 900 MHz band and unsold spectrum in the same band occurred due to auction design (minimum bid amounts for both new entrants as well as operators with expiring licences), desperate demand for business continuity reasons and perhaps naughty bidding by parties whose principal interest was to jack prices for incumbent providers (table 1). Although specific details of round-wise bidding are yet not available on the DoT website, this conjecture may have some truth. Raising rival costs of doing business is a time-honoured and legitimate business practice and in this instance has resulted in the government successfully extracting huge returns. Earlier, in 2010, rent-seeking in spectrum that nearly paralysed the government led to the ethereal rents being captured illegally by private and individual interests. In that sense, we have a better outcome this time.
But do we really have an outcome that is good for overall welfare and in the long term? This is a more difficult question and will depend on how complementary policy addresses the burden created by expensive spectrum. For one, the current bidding has helped consolidate the positions of the incumbents (table 2). Bharti Airtel, Vodafone and Idea Cellular together control over 55% of the market share, own 58% of the total spectrum made available so far and spent 79% of the total amount spent in the 2015 auctions on airwave acquisition. This suggests that the smaller players will face increased difficulty in doing business and, therefore, a workable M&A policy that encourages rather than obstructs consolidation should be put in place. Since spectrum value has already been extracted and will be in future auctions by the government, further taxing it while service providers merge appears unwarranted. Simultaneously, spectrum trading, sharing and leasing are the policies that address the current problems of fragmentation and lack of contiguity of spectrum and ought to be encouraged. The telecom market and institutions that govern it are now mature enough to handle such progressive policies.
A dominant theme of our research over time in telecom has been the growth dividend. Our hypothesis has been that the growth dividend from broadband will be higher than for voice for all sorts of reasons, including the new app economy that internet enables. In this auction, the focus has been on the 900 MHz and to a lesser extent on the 800 MHz band, which is data friendly. The interest in 1800 MHz and 2100 MHz bands has been considerably lower (table 3). In terms of the government’s expectations of relative revenue contribution per band, realisation from all bands except 900 MHz has been lower than expected. Does this mean the much-needed mobile broadband ecosystem will be further pushed into the future? If so, this would be another case of lost opportunity in telecom. For operators too, the business model is shifting slowly but surely to data. What would help catalyse it is an early deployment of the much-delayed National Optical Fibre Network project of the government. But that, as they say, is another story.
Rajat Kathuria is director and chief executive; Mansi Kedia is a research associate; and Parnil Urdhwareshe is a research assistant at ICRIER