‘Monsoon Hungama ’ and the 2G Scam: Public interest and mobile spectrum policy in India,1999–2012

Introduction

India’s mobile phone industry is characterized by contrast, a term famously applied to many aspects of the country itself. Mobile operators here are allocated remarkably narrow ranges of airwave frequencies, known collectively as spectrum, but pay some of the highest per-unit costs (Mölleryd and Markendahl, 2014). Spectrum is a vital resource that makes wireless communication possible, and operators in other countries are typically given much larger swaths (Hazlett, 2011). Another anomaly is the extreme level of competition. Whereas most countries support a handful of mobile operators, at the time of this research (Spring 2014), more than 10 companies competed for market share within India (Curwen and Whalley, 2014: 69). This competition has yielded another stark contrast; India has some of the lowest mobile phone costs in the world, but because of its vast population, is the second largest market for the industry.

The goal of this study is to offer some insight as to how this peculiar situation developed. Why are so many mobile phone operators paying such enormous costs to access such narrow bands of spectrum in one of the most populous developing countries on the planet, where the technology might presumably have significant impact? In attempting to answer this question, particular attention is given to the concept of the ‘public interest’. In order to deal with the technical issue of spectrum scarcity, government regulators across the globe have invoked this ideal to justify a range of telecommunications policies, though nailing down a precise definition of public interest is notoriously elusive (Freedman, 2008: 73–70). India’s polices for mobile phones are likewise intended to maximize the public interest via spectrum allocations, though this goal has been pursued in complex, and sometimes puzzling, ways.

In revealing how Indian mobile phones have been regulated to serve the public interest, this analysis also highlights the general dynamics of spectrum allocation relevant to other contexts. Horst (2013) has urged mobile phone scholars to go beyond issues of usage and adoption to look at the fundamental infrastructure that governs the ecosystem in which the technology operates. This study is inspired by Horst’s appeal as spectrum is a significant, if not primary, determinant for mobile phone infrastructure. Decisions about how to utilize spectrum influence every aspect of mobile communication, including which companies may have access, what kind of services they may offer and, most importantly, the cost for users.

India’s spectrum policies have been critiqued by a number of scholars, although most of all the prior academic literature comes from economists or engineers (Hazlett, 2011; Kumar, 2011; Mölleryd and Markendahl, 2014; Prasad and Kathuria, 2014; Prasad and Sridhar, 2009; Sridhar and Prasad, 2011). Prior studies focus on such discipline-specific debates as the optimum design for spectrum auctions or methods for squeezing more data onto smaller ranges of frequencies. The findings are obtuse for readers not already familiar with a wealth of arcane details. Doron and Jeffrey’s (2013) The Great Indian Phone Book contains an overall history of mobile phones in the country, with some discussion of spectrum issues. This analysis is a supplement to, and clarification of, Doron and Jeffrey’s work, focusing specifically on issues of spectrum as influenced by the concept of public interest. In an attempt to demystify the issue of spectrum allocation and make it accessible to a broader range of scholars, acronyms (which are endemic to the mobile industry) have not been spelt out unless absolutely necessary.

In contrast to those nations that have enjoyed robust telecommunications networks for decades, mobile phones have added significance for developing nations such as India. In 1995, the country had approximately 1 phone for every 100 individuals (Sridhar, 2012: 35). As of February 2014, the Indian government estimated that phone lines outnumbered people in the urban areas, a development due almost entirely to the growth of mobile phones (Telecom Regulatory Authority of India (TRAI), 2014). The same 2014 estimate claimed one phone for every two individuals in rural areas. A rural-urban discrepancy certainly exists, but the overall statistics reveal a dramatic change in telecommunications access for the region as a whole. Beyond making voice calls, phones play music and videos, access information of every variety, and even serve as a flashlight (often necessary in a country with erratic electricity).

Scholars have studied the impact of mobile phones in India, including the effects upon business practices, agriculture, politics and gender relations (see Doron and Jeffrey, 2013, for summaries of numerous studies). A particularly influential study by Jensen (2007) found that fishermen in Kerala were able to increase their profits by accessing information via mobile phones. This study does not necessarily endorse all of these claims, which can stray towards determinism, but rather emphasizes the importance of spectrum allocations. Without access to spectrum, mobile phones cannot have any effect at all. An in-depth exploration of spectrum allocation policies is thus not merely a topic of academic interest, and such decisions can have widespread, dramatic effects upon developing nations.

Research for this study was conducted primarily in Spring 2014 in Delhi, India. Individuals currently or previously associated with the Indian mobile phone industry were interviewed with semi-structured, open-ended questions (see List of Interviews). These individuals included some who were directly responsible for major policy initiatives, operators of commercial mobile companies, representatives for trade organizations and consultants to the industry. The goal of these interviews was to gain clarity on the development of spectrum policy in India, a notoriously complex topic, and to uncover some of the behind-the-scenes logic that motivated particular developments.

These interviews have informed the entire paper, and direct quotes have been indicated as such. These first-hand recollections have been supplemented, and verified, with research from prior scholarship and original policy documents.

The study begins with a discussion of theoretical issues relating to spectrum allocation, especially the concept of spectrum as a public good. The research devotes significant emphasis to the New Telecommunication Policy of 1999 (NTP-99), a regulatory change that had profound implications for the evolution of the Indian mobile phone industry. A direct result of this policy was the introduction of code division multiple access (CDMA) mobile technology to India; this development directly inspired a dramatic boost for the industry, with details elaborated below. NTP-99 also mandated an auction for mobile spectrum in 2001, which paved the way for the infamous 2G Scam of 2008. The full significance of CDMA’s introduction and the 2G Scam become apparent when the broader scope of Indian mobile history is taken into account, and it is thus necessary to include some discussion of related government policies, tariff regulations and industry practices. The study concludes with some general remarks about spectrum allocation in India, and highlights the importance of infrastructure studies for communications scholars.

Conceptual issues of spectrum allocation

Theories of social construction recognize that technologies are not developed in isolation, but are integrally related to the cultural values of a given society and its institutions (Bijker et al., 1987; Bijker and Law, 1992). Spectrum allocation is an apt illustration of social construction because, unlike some other natural resources, it is not visible to the naked eye. Regulators first develop a mental conception as to how the resource could and should be utilized. These ideas then have direct influence upon how the technology of mobile phones may develop.

This analysis is further informed by Starr’s (2004) concept of ‘constitutive moments’. Starr developed this idea in his history of various mass media systems, focusing on the 17th to mid-20th centuries. His narrative stopped well short of the diffusion of mobile technology, though the idea of constitutive moments is an applicable framework for understanding this more recent technology. For Starr (2004), when critical decisions are made about the design of a technological system, ‘ideas and culture come into play, as do constellations of power, pre-existing institutional legacies, and models from other countries’ (p. 2). Once choices have been made, ‘mechanisms of entrenchment make it difficult to revoke such decisions’ (p. 4). The force of inertia propels a technology down a particular path of development, even after the initial factors that influenced the original decision have dissipated. Starr refers to such a dynamic of technological development as ‘path-dependent’; this specific term comes from the literature of economics and shares much with Hughes’ (1993) concept of ‘technological momentum’.

When applied to the current topic of study, initial spectrum allocations are these constitutive moments. Once these allocations have been made, they become entrenched and unlikely to change. Manufacturers create receivers and transmitters optimized for specific frequencies, and once they are deployed, a revised spectrum allocation would trigger unwieldy changes of infrastructure. The initial spectrum decisions could even be counterproductive to later development, and users are notoriously reluctant to abandon access to such a precious resource once they have commercialized specific bands.

The theory of diffusion of innovations also informs this analysis as it offers a framework for understanding the dramatic growth of mobile phones in India (Rogers, 1995). While users often view this technology as unprecedented and revolutionary, the pattern of its diffusion does conform to established theory. Successful innovations typically show a slow period of initial growth, followed by a dramatic boost in the adoption rate, followed by a subsequent slowing of growth. This classic model of growth, articulated by diffusion of innovations research, is referenced where applicable.

Regulating the airwaves in the public interest

Informing all official decisions about spectrum allocation in India is the declaration that the ‘airwaves constitute public property and must be used for advancing the public good’ (Narayan, 2014: 209). India’s Supreme Court established this principle in 1995 in a ruling that pertained to a televised cricket match, though the mandate has been applied to all uses of spectrum (PK Garg, 2014, interview). The rub, of course, is that there is no clear scenario as how to achieve the said goal. Balancing the needs of the public while also supporting a viable mobile industry is a complex process. The dispersed nature of spectrum policy contributes to inconsistencies in Indian policies. Policies originate with the Department of Telecommunications (DoT), which is part of the Ministry of Communications and Information Technology, though other ministries, departments and courts also influence the regulations and subsequent enforcement.

To serve the public interest with mobile phones, one might argue that a single company, funded by the state, could offer a universal, subsidized service. Such a system would fall in line with the older philosophy of the Indian government, before the liberalization of policies in the 1990s. Or, one could argue that a range of private companies best serve the public. This alternative scenario, which has been followed by the Indian government, is based on the assumption that competition will encourage the spread of telecommunication services and reduce prices for users. As part of NTP-99, India mandated that all phone operators pay 5 per cent of annual revenue into a Universal Service Obligation Fund (USOF), which in turn subsidizes phone service to under-served areas. Without getting into the long-standing debate about the relative effectiveness of the USOF and its numerous schemes, it is apparent that this particular method for serving the public interest was not inevitable or the only possible solution. Creating the USOF, and allowing for widespread competition within the mobile industry, was one of the constitutive moments for the Indian mobile sector.

Another significant choice made by Indian regulators is the decision to rely on the long-standing practice of allocating specific frequencies to individual groups. This method of discreetly segregating users to their own frequencies dates back to the earliest era of wireless telegraphy. The latest technologies of the 21st century, however, allow multiple users to share frequencies (Prasad and Sridhar, 2014). India has recently made changes to its spectrum policy in this direction, though for the time period under analysis, regulators relied upon the model of exclusive spectrum use.

To assign spectrum to mobile companies, India has primarily used auctions to assign rights, with the winning parties paying steep annual fees to the government. As users compete for spectrum, friction is magnified because different frequencies have different propagation characteristics (Song, 2011). Waves in the lower frequencies can, for example, travel more easily through physical obstacles, while higher frequencies can carry more data. Continued squabbles over the value of spectrum allocations, even when two companies receive an equal range of frequencies, is an inevitable outcome of the traditional method of spectrum division.

Heller’s (2008) concept of the ‘tragedy of the anticommons’ can be applied to this method of treating spectrum as private property. In contrast to the classic tragedy of the commons, in which multiple users over-use a commonly shared resource, ‘anticommons’ refers to a scenario in which a resource has been divided and parcelled out in a manner that similarly leads to less than optimum usage. In terms of spectrum, the anti-commons can be seen when companies are assigned a particular band of spectrum, but fail to fully utilize its capabilities. Rival companies, or entrepreneurs who may want to use the said spectrum for some other purpose, are barred from doing so because of the private-property framework that is employed. Innovation in wireless technology could thus be thwarted if a mobile operator refuses to allow a third-party service access to its particular spectrum allocation. Subho Ray, the president of a leading mobile trade organization in India, alluded to the tragedy of the anti-commons when he lamented that the largest mobile companies pay almost 30 per cent of their annual revenue in spectrum fees (Roy, 2015). According to this argument, all of this money devoted to securing spectrum rights reduces the amount that can be spent building new towers, or otherwise expanding the coverage area.

Alternative policies view the airwaves not as private property that can be assigned exclusively to one group but as spectrum commons. An administrative body might establish some broad parameters about the power and range of transmitters, but allow all users to operate in a given band of frequencies. The phenomenal success of the Wi-Fi standard is a testament to the innovation that such a system can encourage (Prasad and Sridhar, 2014: 38). In India, the bands of 2.4 GHz and 5.8 GHz are designated as unlicensed, following guidelines put forth by the International Telecommunications Union. In several underdeveloped areas, a project known as W4C (Wireless for Communities) uses this unlicensed spectrum to broadcast Internet to homes and schools (For general information, see http://wforc.in/). It is difficult to imagine a more succinct example of spectrum being used to benefit the public, although such programmes remain on the periphery in India.

Public interest and broadcast spectrum

Before delving into the particulars of mobile spectrum, it is worth taking a brief look at radio and television, two other communications technologies that are similarly dependent upon spectrum. For both of these mediums, the Indian government similarly seeks to maximize the public good. And, just as the spectrum allocation for mobile services is complex and often difficult to decipher, the policies for broadcasting are similarly byzantine.

For several decades, broadcasting was entirely under the control of the Indian government: All India Radio established in 1930 and Doordarshan (DD) starting in 1959, for television. In 1999, as part of the same liberalization efforts that gave spectrum to mobile services, private companies were able to bid for frequency modulation (FM) radio frequencies. Non-profit, community radio stations were also permitted in India, although the licensing process is notoriously complex and lengthy and only 150 exist across the nation, a modest number given India’s expanse. Community radio stations are often denied licences in urban areas because of spectrum crowding, and are prevented from broadcasting news, a heavy-handed attempt to preserve the dominance of the All India Radio service.

For television, India similarly tries to maintain the authority of DD, which was technically freed from total government control in 1997. DD remains, however, the only network licensed for terrestrial, television broadcasting, and operates a number of channels in various regional languages. In this regard, the public good is served by a single entity. Television viewers in India, especially in wealthy urban areas, do have a growing number of options, thanks to the rapid growth of satellite and cable services that began in the 1990s. In a comprehensive analysis of television in India, Narayan (2014) wrote that while the state ‘appeared to be caught unprepared for this onslaught’ of new competition, it ‘did not discourage it by imposing an outright ban on satellite television’ (pp. 251–253). By 2009, Narayan (2014) wrote, over 450 different television channels were available in India, a seismic shift from the previous era in which DD was the sole viewing option (p. 89). One could say that the great many television options available to the viewing public are not because of specific government action, but rather because of inaction. Even while maintaining sole control over terrestrial broadcast frequencies, India turned a blind eye and allowed satellite technology to circumvent the rules and take hold.

‘Monsoon Hungama’, introduction of CDMA into India

The story of CDMA’s entry into India illustrates the shrewd ways entrepreneurs have circumvented official spectrum allocation. This particular scenario is not unique to India, as entrepreneurs in the United States and China also launched successful mobile services on frequencies not officially designated for such use (Hazlett, 2012: 73–74; Sridhar, 2012: 108).

Commercial mobile services were launched in India in the mid-1990s as part of an overall wave of reform. In the early 1990s, the government released its tight control over spectrum and allowed access to private companies. The DoT previously had sole control over the Indian phone system, though it realized that it did not have the financial resources to install enough physical telephone lines to satisfy the public. Wireless was seen as a cost-effective way to expand telecommunication services (PK Garg, 2014, interview).

A defining constitutive moment of the time was India’s selection of a technical standard for its mobile system: global system for mobile communications (GSM), a platform developed in Europe in the 1980s, was chosen over CDMA. Invented by the American company Qualcomm, CDMA is capable of transmitting more data on a narrower band of frequencies, although it was less developed at the time. Another pivotal decision was to carve up India into more than 20 geographic circles, with spectrum allocated separately in each. A total of 900 MHz was chosen, as this was the frequency already used by the European GSM systems. After negotiating with private oil companies and its own defence forces, the Indian government was able to acquire just enough spectrum in the 900 MHz band to allow two mobile companies to operate in each circle. It was believed, at the time, that two companies would allow at least a modicum of competition and more spectrum would eventually be vacated (PK Garg, 2014, interview). To use Starr’s terminology, the original users of the 900 MHz band were well entrenched, which greatly limited the frequencies available to the mobile companies.

For most portions of India, auctions were used to dole out spectrum rights. Mobile phones were seen as a luxury to supplement traditional telecommunications service, and the government established steep licence fees and complicated tariff regulations. There were different rates for a call between two landline phones versus a landline and mobile phone though ‘whichever direction the call went, the cellular subscriber paid for it’ (Desai, 1996: 48). A popular perspective on Indian mobile history has been that private companies originally over-estimated demand and paid far too much for spectrum, an auction scenario known as ‘the winner’s curse’. Desai (1996) challenged this notion and emphasized ‘the conditions of their license ensured that their revenue would fall short of the costs’ (p. 49).

Multiple disputes within the industry, including objections to the lopsided interconnection charges, led the government to establish the Telecom Regulatory Authority of India (TRAI) in 1997 (Jain, 2001). Continued frustration motivated the NTP-99. Under this dramatic regulatory change, mobile phones were no longer characterized as a luxury, but, instead, a necessity (Thomas, 2007). As part of the attempt to better serve the public, NTP-99 authorized a new spectrum auction in 2001. The immediate goal was to increase competition in the mobile market, which in turn might lead to lower costs. At the same time as the spectrum auction, a separate auction allowed private companies to finally enter the landline phone market. A provision of the new landline licences allowed CDMA, and other wireless technologies, to be used to solve the famed ‘last mile’ problem of connectivity (Sridhar, 2012: 41). A landline phone provider could lay physical lines up to the edge of a neighbourhood and then use CDMA to complete the final connection to a user’s home or office.

Two of the companies that acquired the new landline licences, Reliance and Tata, became the principle users of CDMA in India. It is critical to note that both companies had failed to acquire any premium circles during the mobile spectrum auction of the previous decade. TV Ramachandran, a long-time participant in the Indian mobile industry, recounted the loophole that allowed CDMA (2014, Interview):

As part of the new license, these companies were given 8 MHz of spectrum. This was a lot, in which to use some wireless technology. The point was for ‘non-feasible’ areas, such as Chandni Chowk [one of the main areas of Old Delhi] or Nariman Point in Mumbai or remote hill stations. You can’t really dig there, so you need wireless to provide phone service. This was supposed to be a temporary solution, and was known as Wireless in Local Loop with Limited Mobility (WLL). Nobody had ever heard of this before, it was an unheard of license. But those companies cleverly took that license and used it to offer full mobile services.

These new entrants into the phone market had been ostensibly licensed to provide landline service, and were not viewed as competition for the officially licensed mobile operators. To this end, the DoT mandated that Wireless in Local Loop (WLL) remain within the short-distance calling area (SDCA), a region that determines the price of phone calls (Sridhar, 2012: 41). WLL phones were treated, for the purpose of tariff regulation, as fixed landlines and their service thus was necessarily cheaper than that of mobile companies. Some regulators championed WLL as the way to finally provide phones to the poor and to those in remote areas (O’Neill, 2003).

In 1999, the founder of Reliance Industries, Dhirubhai Ambani, made a prophetic claim about mobile technology: ‘Make a phone call cheaper than a postcard and you will usher in a revolutionary transformation in the lives of millions of Indians’ (Keskar, 2009: 199). The new WLL licences turned this vision into reality. Reliance interpreted the regulations, and even re-engineered transmitting equipment, to offer low-cost mobile services. The service was possible thanks to the technical option of call forwarding. When a Reliance subscriber travelled outside their home circle, the network automatically assigned a number corresponding to the new region. Calls were automatically forwarded, without any effort from the subscriber (Anil Kripalani, 2014, interview). The end result was a full-fledged mobile service, with users able to travel anywhere in the country.

Reliance promoted its disruptive service in July 2003 with a lavish campaign known as ‘Monsoon Hungama’, which loosely translates to ‘rain of excitement’. In a figurative and literal sense, Reliance rained cheap mobile phones upon India. For Rs. 500 (approximately US$8), a Reliance subscriber received a CDMA-capable phone plus credit for free phone calls. The cost of service was a fraction of what others were offering, and incoming calls were not charged. Samir Saran, Senior Fellow, Observer Research Foundation, explained (2014, Interview):

Until then this (mobile) service was designed to cater to only the elites, this was a wrong estimation of the market potential. Reliance was able to acquire a new class and generation of customers very quickly with a bold initiative. In retrospect one can say that it also made good business sense. The amount foregone in this scheme more than offset the customer acquisition costs of its competitor.

Reliance paid a sizeable sum, in one flamboyant burst, to subsidize the necessary CDMA phones, and amassed a strong user base. If Indian regulators had enforced the letter of the law, the WLL services would never have been allowed to spread since the calls were clearly extending beyond the SDCAs. But in a larger sense, the public interest was indeed served by the diffusion of this low-cost, non-sanctioned service.

Rival companies protested and appealed to the government. To combat the growth of Reliance, the official mobile operators slashed their own costs, and their own numbers similarly rose. The phone statistics for India reveal a sharp increase in this time period. To use the terminology of diffusion of innovations, this was the take-off period for the technology. In 2001, there were less than 4 million mobile phone subscriptions in India; by 2005, more than 50 million (Figure 1).

OPEN IN VIEWER

Figure 1.

Growth of phones in India, numbers of subscribers, 1998–2014 (numbers in millions).

The conflict between the incumbents and WLL providers was resolved with two major policy revisions in 2003. These changes were directly motivated by the way WLL systems had circumvented the official spectrum allocations. The first change was the billing system known as Calling Party Pays. Virat Bhatia spearheaded the industry’s campaign to change the tariff regulations that made regular mobile calls so expensive (2014, interview). WLL providers had not paid the same entrance fees, nor were they subject to the same interconnection charges, and Bhatia argued before regulators that it was therefore only natural that the new, unofficially licensed mobile service was cheaper. Bhatia recalled that when a regulator claimed that WLL phones ‘served the public interest’, he responded that all mobile companies could serve the public more effectively if the uneven interconnection fees were abandoned. This argument was successful and the skewed regulations were changed. The revised scheme shifted the burden of cost to the caller. This policy change made phones more accessible to low-income users, who may use them only to receive calls (or play music and videos). The second policy change was the creation of the Universal Access Service License, which legalized the aggressive behaviour of the WLL operators. Under the new licence, landline and mobile services were lumped together, with operators free to use any technology they chose, including CDMA.

The CDMA systems were allocated to the 800 MHz band (to prevent interference with GSM). GSM proponents openly acknowledge that CDMA requires less spectrum to operate, but in a twist of logic, this efficiency was then used to limit its growth. When the government officially allocated spectrum to CDMA mobile systems, they received only half the amount given to GSM. Anil Kripalani, an engineer with extensive experience in the Indian market, mocked this decision: ‘It’s like telling the government that you need a road that is twice as wide because you have an old, out-of-date buggy’ (2014, interview).

The constitutive choice to limit the allocation for CDMA was more than just a symbolic slight. When mobile systems have more spectrum, the increase in call volume is not purely linear. Having an additional 3 or 4 MHz of spectrum would allow for a dramatic increase in quality. If the Indian regulators truly sought to serve the public in the most technically optimum manner, why not fully endorse the CDMA systems? On this matter, Kripalani also offered his perspective. India has historically modelled its policies on those of Europe, not the United States, and the perceived link between CDMA and the American company Qualcomm may have discouraged India’s support of the technology. Admittedly, this observation is more speculation than established fact, but the decision to handicap CDMA cannot be defended on purely technical grounds. As the social construction framework suggests, we must look for factors beyond technology itself to account for decisions about its regulation and usage.

Once the CDMA mobile systems were officially allowed, competition within the industry increased. CDMA never challenged the dominance of GSM in terms of market share, though its introduction changed the economics permanently. By allowing the WLL services to grow so broadly that dismantling them would have been detrimental, the Indian public was served by a policy of inaction rather than action. To this day, India remains one of the cheapest places to make a mobile call.

The 2G Scam

The initial allocation of mobile spectrum in India was marred by controversy (Doron and Jeffrey, 2013: 46), although that sequence of events pales in comparison to the flagrant corruption of 2008. The so-called 2G Scam was not limited to the pages of the trade press and is known to most Indian citizens. The sordid tale warrants a retelling here not only because the events proved so influential upon later spectrum policies, but because the events remain shrouded in confusion. Many observers and critics misunderstand the exact nature of the illegality, and this analysis is one attempt to clarify the events. The scam began with a strange attempt to serve the public by preserving the low cost of mobile service, thought this attempt inadvertently laid the foundation for the one of biggest acts of corruption in modern Indian history.

To understand the 2G Scam, it is necessary to go back to the 2001 auctions authorized by NTP-99. For the next several years, India avoided further mobile spectrum auctions and maintained the 2001 prices. This decision was rooted in the desire to deliver telecommunication services at affordable rates. Nripendra Misra, former head of the TRAI, and PK Garg, former head of the Wireless Planning and Coordination Wing within the DoT, both stated that low-cost service was the primary motive for spectrum decisions of the time (2014, interviews). Misra stressed that in the 2001 auction, there were no successful bids in some of the less-profitable circles. Wary of a repeat situation, the DoT avoided further auctions.

The allocation process during this time was simply ‘first come, first served’ (Nripendra Misra, 2014, interview). In contrast to earlier practice, obtaining a licence to offer mobile service no longer came with any spectrum. Companies paid an entrance fee, based on the 2001 auction prices, and then waited for frequencies to become available. These years involved ongoing negotiations with the Department of Defence to free up new bands of spectrum.

The DoT micromanaged the available spectrum to an extreme degree. Companies could obtain additional spectrum by proving a certain number of subscribers (Sridhar, 2012: 112). This is a relatively uncommon technique for managing spectrum and encouraged companies to distribute subscriber identity module (SIM) cards at a low (or even no) cost (Mahesh Uppal, 2014, interview). A company could then use these numbers to represent subscribers, even though many SIM cards were not in active use. As a spectrum usage fee, mobile operators paid a percentage of annual revenue to the government.

Using the diffusion of innovations framework, an initial glance at the number of mobile phone users in India might suggest that the plateau effect occurred in 2011, with growth slowing dramatically at this time. This perception is not entirely accurate however, as the statistics for earlier years may have been artificially inflated by the SIM card issue; additionally, India now uses stricter criteria for determining active phone users. This observation is not meant to downplay the explosive growth of phones in India, as their usage has undeniably grown exponentially over time, but is rather meant to urge caution when relying upon official estimates.

A significant constitutive choice at this time was to tightly limit the amount of spectrum given to mobile companies. As previously noted, the quality increase from increased spectrum is exponential, not incremental. One way to serve the Indian public would have been to award significant swaths of spectrum to a handful of companies, who might then dominate the market with superior service. The decision instead was to serve the public via maximum competition, so small slivers of spectrum were doled out to several companies in each region. Under the subscriber-based criteria of the time, GSM operators could acquire a maximum of 20 MHz of spectrum, only half of the global average at the time. CDMA was even more even restricted: a maximum of 10 MHz was possible, compared to the global average of 30 MHz (Sridhar, 2012: 113).

In 2007, faced with a severe backlog of spectrum requests, the TRAI issued recommendations that set the stage for the scam. The policy document stated that the government should not limit the number of mobile operators in each circle and should instead allow the market to settle the issue (TRAI, 2007). Standards for subscriber-based spectrum allocation were also tightened, making it more challenging for incumbent operators to obtain additional spectrum. The result of these policies was to introduce a slew of new mobile operators into a market that some felt was already overcrowded. Most significantly, however, was the recommendation that spectrum not be auctioned. The TRAI guidelines admitted that the 2001 price was perhaps no longer applicable, but auctioning the latest round of spectrum would put new entrants at a competitive disadvantage. Misra defended this policy recommendation, saying that auctioning off new bands of spectrum would have led to higher prices for consumers (Nripendra Misra, 2014, interview).

Under the guidance of Minister of Communications A. Raja, the ensuing allocation process triggered a wave of lawsuits, prison sentences and debates that continue to this date. On 24 September 2007, the DoT announced that it would award spectrum through the standard ‘first come, first served’ process. In Doron and Jeffrey’s (2013) account, the process was labelled ‘spurious’ (p. 58), but as misguided as it may have been, it was in fact the long-standing policy. The date of 1 October was put forth as the application deadline.

On 10 January of the following year, the DoT announced via press release that only spectrum applications that had been submitted before 25 September would be considered. In the most questionable move of all, the press release further announced that preference would be given to those who submitted the requisite entrance fee of Rs. 16.5 billion; this figure was based on the price of a nationwide licence using the 2001 auction prices. The window to deliver cheques to the DoT was a mere 45 minutes after the press release was issued (Doron and Jeffrey, 2013: 58). Despite the absurdity of such abbreviated notice, 122 separate spectrum licences were awarded. It was apparent that some companies had been advised in advance, as they arrived with checks dated prior to the ‘surprise’ press release.

The fortunate companies were able to reap immediate rewards. Share prices of the new spectrum licence-holders increased exponentially, and critics insisted that Raja had cheated the government out of much-needed funds. There were also allegations that Raja and other bureaucrats had personally benefitted through secret bank transfers. A subsequent investigation found that the questionable process robbed the government of approximately US$32 billion, as the spectrum licences could have generated much higher prices (Open Society Foundations, 2012: 85). This estimate was based on the substantial increase in spectrum prices in two auctions conducted in 2010.

In public perception and numerous press reports, this estimated financial loss has gained great notoriety. It should be clarified, though, that if the DoT had followed the announced process, and allocated spectrum on a legitimate ‘first come, first served’ basis using the old 2001 price, most of the controversies would have been avoided. The government only had enough free spectrum to allocate 122 licences, so the amount of money received would have been the same. This observation is not meant to endorse ‘first come, first served’ as a viable policy of spectrum allocation, but rather to clarify the illegal activity of the scam. The illegality was not undercharging for spectrum, but rather manipulating the allocation process to favour certain entities. The decision to maintain 2001 prices, however, remains one of the most well-known aspects of the scam.

In 2012, the Supreme Court invalidated the licences that had been allocated in 2008 (Doron and Jeffrey, 2013: 59). Companies participated in an auction for the same spectrum. Some of the new entrants disappeared from the scene, though some remained. The result has been a dramatic increase in competition. Some circles, for example, have spectrum allocated to more than 12 companies. According to the chief executive officer (CEO) of Vodafone, one of the dominant mobile companies in India, this level of competition makes profits difficult, and many smaller operators are sustained only through continually refinanced bank loans (The Hindu, 2014).

A critical outcome of the scam is a government ruling that all spectrum must now be auctioned, including the old 2G bands as well as frequencies not yet allocated. As a positive development, one could say that this creates a more transparent process, and limits the influence that administrators can bestow upon favoured parties. As a negative, the mandate establishes price as the prevailing criteria. Advocates for unlicensed spectrum highlight the great social benefits that may come from offering free or low-cost wireless services to the public, a scenario that is unlikely if price remains the defining standard for spectrum allocation (Ponapa, 2011; Song, 2011).

Conclusion

It is undeniable that utilizing the airwaves in the ‘public interest’, as the Indian Supreme Court has mandated, is a worthy goal, lest mobile phones become a luxury reserved for the wealthier segments of society. And while this study did not begin as a critique of India’s telecommunication policies, but rather an attempt to explain them, it is impossible to ruminate on the findings without offering some thoughts on the matter.

The evidence presented here offers ample evidence for both defence and criticism of India’s attempt to serve the public with mobile services. Supporters of official policy could point to the rapid diffusion of phones, the extreme low cost of service, and the exponential increase in phone penetration, even in rural areas. Indeed, it is difficult to go anywhere in India without seeing some evidence of mobile phones, as the technology has worked its way into almost every level of society.

Indian regulators certainly made some constitutive choices with the goal of expanding phone service, though critics could point out that the increased competition, which ultimately reduced phone costs, came through decidedly unintended means. Recall that the original rules of WLL were explicitly designed to prevent, not encourage, competition with the GSM incumbents. Further competition was introduced via the fraud of 2008, which brought new companies into the mobile sector. With the growth of satellite and cable services in India in the 1990s, the government displayed a similar level of tolerance, and only took steps after the fact to bring in some level of regulation. But tolerating rule breakers is hardly a prescription for future success when regulating public resources. It would be a perverse form of optimism, for example, for policy makers to ignore problems with current communications regulations in the hope that some future, disruptive technology will evolve to serve the public interest.

Critics of Indian spectrum policy could also claim that, as rapidly as mobile phones have diffused, a different sequence of events could have produced an even more rapid rate of growth. Invoking Heller’s concept of the ‘tragedy of the anticommons’, critics could, for example, emphasize that the significant revenue devoted to spectrum fees thwarted the growth of mobile networks. An even more significant choice was the decision to restrict CDMA systems, a technology that is more efficient in terms of spectrum usage. Postulating alternative histories is a risky endeavour, though one cannot help but wonder how the situation would have differed if CDMA had been given equal spectrum access as the GSM systems.

Debates about spectrum allocation are not likely to disappear, even though engineers continue to produce more efficient systems for utilizing and sharing spectrum. In this manner, the supply of available spectrum is increasing. However, demand is also increasing, as companies are rolling out more and more products, from automobiles to kitchen appliances, which access wireless networks. As indicated by the Indian scenario, regulations about how this spectrum is allocated have direct and immediate effects upon the diffusion of wireless technologies and, by extension, their impact. Infrastructure issues are notoriously opaque to users, though scholars of modern of communications technologies should keep such fundamental, underlying issues in mind as they evaluate their effects.