As we gear up for another round of telecom spectrum auctions in March, it is useful to recall the lessons from previous auctions and to internalise these in subsequent auction design. Like all historical narratives, the story of spectrum in India has many strands—in this column, our focus is only on the method of assignment. The watershed was undoubtedly 2008, when the 2G scandal broke. To start with, until 2001, spectrum was assigned through auctions and then via a contrived administrative procedure. In the beginning (circa 1994), a single-stage ‘sealed bid’ method was used that produced very high licence revenues for the government, but simultaneously created onerous financial burden on the operators. In the absence of coordination among bidders, sealed bid type auctions produce high prices since you do not know how others are going to bid. The ‘irrational exuberance’ of the nascent yet promising market may have contributed to the exaggerated bids. Within a few years of their operation, in 1999, the mobile operators collectively and successfully petitioned the government to reduce the licence fee burden, which was converted from the fixed bid amount to a more risk-neutral revenue share payment. In 2001, auction design was changed to multi-stage auctions in which information revealed during the bidding process reduced operator risk compared to the single stage, sealed bid routine.
The inherent contradiction during this time, however, was that vital spectrum to provide mobile services came bundled with the licence to operate. Assignment of additional spectrum was ill-defined and conferred discretionary power on the department of telecommunications (DoT). The subscriber-linked criteria (SLC) and the notorious first-come first-serve (FCFS) methods employed by DoT during the period 2003-2008 created perverse incentives to game and undermined the system. This was easily the darkest of times for Indian telecom. Certain operators inflated the subscriber base to gain additional spectrum, others successfully lobbied to literally jump the queue when FCFS was selected as the assignment method. The underlying motivations of doing that are now as clear as crystal. Spectrum—rightly referred to as the lifeblood of the wireless industry—had become a ‘priceless’ asset. For incumbents, it meant greater market power, for new entrants it provided a huge arbitrage opportunity.
The Supreme Court came down heavily on the abuse of power by DoT and the impropriety in the assignment of spectrum to favour a small number of select operators in February, 2012. It justified its judicial overreach in the larger public interest and pronounced that the assignment of spectrum, and indeed all natural resources in India, be henceforth done through auctions. Few objected to the Court’s assessment or their unstated position that institutional failures stretched well beyond the telecom sector but there was a push-back on their recommended method for allocating all natural resources. Article 39(b) of the Constitution asserts that natural resources should be distributed so as to best serve the common good. But it is not altogether clear that the common good can only be served through auctions. Auctions are known to often conflict with the inclusion agenda. For telecom spectrum, however, auctions have to be the preferred assignment method. What’s not obvious is whether maximising government revenue should be the overriding aim of the auction design. In any case, a successful revenue maximisation strategy for the government will depend not so much on a high reserve price, but on the prevailing conditions of the mobile ecosystem.
The accompanying table shows the outcome of auctions since 2010. If success is judged based on the premium over reserve price, the 2010 auctions were by far the most triumphant. Indeed, the reserve price for the 3G auction scheduled for next month (March) is 15.53% higher than winning bid of 2010, revealing government’s expectation from the auction. Outcome of the approaching auction will depend upon, inter alia, the amount of spectrum being auctioned, number of bidders, expected demand for spectrum enabled services (demand for spectrum is a derived demand) expectations and timing. In a depressed market, as was the case in 2012 and 2013, auction proceeds equalled the reserve price. There is no fear of that happening this time if the auction is held as scheduled. India is poised to become the second-largest smartphone
market, replacing the US and just behind China with an expected user base of 382 million by 2016. More spectrum will enable faster and reliable wireless service and will be crucial to handle the increasing amount of data traffic that will flow over smartphones.
It is also essential to recognise the merits of increasing access and the role that digitisation can play in overcoming other beleaguering infrastructure constraints. Mobiles currently provide more than 900 million points of connectivity enabling information and data flows. The latter has transformative potential for it does improve efficiency, trigger new business models and boost economic growth. This is well documented in research around the world on the economic and social impacts of greater voice and data access. India’s biggest challenge in telecom is increasing and improving access to underserved areas and also extending that access to the internet. 3G spectrum will play a crucial role in that since most data in India will be accessed through mobile devices.
A primary objective of the auction should thus be to maximise the future benefits of access to consumers and, by implication, to the economy.
This is done by awarding spectrum to those operators who are able to make the most effective use of it and not necessarily by maximising return to the exchequer. All this, while maintaining integrity of the process. In the words of renowned physical chemist CP Snow this means combining the two cultures—the science and the art—to design an auction that meets several objectives all at once. Applying this principle for the March auction for example would have meant (as Trai had recommended) auctioning even those 15 MHz of airwaves in possession of defence, but likely to be released for commercial use by telecom. The ‘art’ lies in increasing the supply of spectrum that would reduce scarcity and temper unit price, yet generate healthy bids since data traffic is set to grow. The volume effect may even increase government revenue and finesse operator arguments of sluggish roll out of data services. Even if it is too late now, future auction design should focus on transparent assignment while also recognising that along with government, bidders too could be subject to budget constraints. And there is always the possibility that sector growth could contribute to the exchequer through future tax receipts. But for now, a bird-in-hand appears to be the preferred canon for the government; for the future, it might be worth opting for two in the bush.
By Rajat Kathuria and Nilesh Basu Roy
Kathuria is director and chief executive & Basu Roy is research intern, ICRIER. Views are personal
