To auction or allocate: The spectrum debate in the satellite era

In the case of 2G spectrum allocation, while the rest of the world was auctioning spectrum--with the United Kingdom realising a spectacular sum of £22.5 billion in 2001, albeit leading to several telecom firms going bankrupt—India was allotting spectrum in January 2008 at 2001 prices. This decision contributed significantly to the downfall of the UPA government  in the 2014 elections. The then Union Minister A Raja justified the allocation, claiming adherence to the recommendations of the Telecom Regulatory Authority of India (TRAI). However, after some equivocation, the then TRAI chairman admitted that despite unprecedented growth of the telecom sector from 2001 to 2008, TRAI  had recommended retaining the prices discovered in 2001 auction without revision for the sake of providing a level playing field. 

Interestingly, the recommendation came from the then TRAI chairman himself, rather than being the outcome of a formal meeting of the TRAI members. 

Post-facto, this procedure was widely condemned. The Comptroller and Auditor General (CAG), using simple arithmetic, estimated that imputable loss to the exchequer at Rs 1.76 lakh crore, and the Supreme Court of India subsequently cancelled the  allocations. Auction became the rallying cry of the time.  

Today, the debate has taken a new turn with the advent of Elon Musk’s Space-X and its Low Earth Orbit (LEO) satellites. The International Telecommunication Union (ITU) recommends the allocation of satellite spectrum, a position Musk supports, while Indian telecom giants Mukesh Ambani and Sunil Bharti Mittal are advocating for auctions. This raises a fundamental question: is auctioning the ideal mechanism for both terrestrial and satellite spectrum, or should the approach vary, with auctions suited to terrestrial spectrum and assignments better suited to satellite spectrum?

Satellite communication via Geostationary Orbits (GEO) has existed for decades.  But, as the Iridium project illustrated, technological success does not guarantee commercial viability. Space-X  recommercialised satellite communication by deploying about 6,000 relatively low-cost LEO satellites into space. 

The ITU distinguishes between satellite spectrum, a public good, and terrestrial spectrum, a private good. Public goods are defined by two characteristics: non-excludability and non-rivalry.  Non-excludability means it is impossible or non-optimal to exclude non-paying users, while non-rivalry means that use by one entity does not make it less for use by others. Satellite spectrum, with its global coverage and ability to transmit data across vast distances and geographical boundaries, allows multiple operators to use the same frequency bands, and meets the non-excludability criterion.  The coverage area of satellite spectrum is so large, the buyer operator will not be able to use it fully.   For satellites to function effectively, their allocated frequencies must be available across entire satellite footprints.  This requirement necessitates a shared approach, allowing multiple operators to utilise same frequency bands, thus maximizing the utility of the available spectrum.   

Additionally, within the bandwidth capacity, satellite spectrum is non-rival,  supporting use by multiple operators without degradation.  Thus, sharing is consonant with a public good and auction is consonant with a private good. Sharing and assignment will lower costs for each operator and lower entry barriers for newcomers and this will promote innovation and competition.  Of course, this may lead to interference, but that is taken care of by using advanced technology such as ‘spot beams’. By contrast, terrestrial spectrum is better suited to auctions because of its smaller coverage area, where the costs of preventing interference outweigh the benefits of sharing.  Even then, costs of infrastructure such as towers are
often shared.

Space-X’s Starlink and Amazon’s Project Kuiper are proposing to enter India.  Musk argues for spectrum assignment on the authority of ITU, which would grant him an early-mover advantage.  Space-X has also petitioned the US telecom regulator Federal Communications Commission (FCC) to remove power ceilings on LEO satellites, set 20 years ago to prevent interference with GEO satellites.  LEO satellites, capable of reaching remote areas without corresponding incremental costs, pose a formidable challenge to Indian telecom giants. Ambani and Mittal are now arguing for auctions, fearing that Starlink will do to them what Jio did to the then incumbent telecom majors, Bharti Airtel and Vodafone and the ailing BSNL.

Ironically, their demand for auctions echoes the rationale of ‘level playing field,’ used in 2008 to justify the assignment of 2G spectrum by the then Chairman of TRAI.  They had paid money to get terrestrial spectrum, and so Musk should also pay money arrived through the
same process of auctions to enter the Indian market. 

Ambani and Mittal contend that assignment would encourage inefficiency, enabling non-serious players to enter the field and hoard spectrum (and not use it)—a problem seen during the licence-permit era when Bombay Club oligarchs procured licences pre-emptively without using them.  In the light of the above analysis, these arguments have no merit. Under assignment, spectrum hoarding by non-serious players may be a possibility but this is avoided by having two part charge, viz. a fixed licence fee and a variable user fee based on aggregate revenue.  

LEO satellite communications is poised to disrupt the fast-evolving telecom sector. With prudent policies and regulation, this technology could herald a future of even greater benefits for consumers.

(The writer was member, TRAI and a professor at IIM Bangalore)