ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Spectrum Auctioning and Licensing in Telecom Industry

With the growing mobile subscriber base and the introduction of newer technologies like 3G, it is essential to carefully regulate spectrum amongst various telecom operators. In India, where the net worth of the telecom industry is about $30 billion, the revenue earned through spectrum auctioning becomes quite significant. This article focuses on the issues involved in the allocation and auctioning of spectrum in existing scenario for 3G technology in India.


had a few shortcomings: first, though

Spectrum Auctioning and

there was considerable privatisation of the telecom industry, the service roll-out still

Licensing in Telecom Industry

remained slow. This was apparently due to the fact that the winners had shelled out huge sums to win a circle and hence, sivasankari S V, Mahim Sagar, D P Agrawal had financial crunches to provide faster

With the growing mobile subscriber base and the introduction of newer technologies like 3G, it is essential to carefully regulate spectrum amongst various telecom operators. In India, where the net worth of the telecom industry is about $30 billion, the revenue earned through spectrum auctioning becomes quite significant. This article focuses on the issues involved in the allocation and auctioning of spectrum in existing scenario for 3G technology in India.

Sivasankari S V ( and Mahim Sagar ( are at the Indian Institute of Technology, Delhi. D P Agrawal ( is at the Indian Institute of Information Technology and Management, Gwalior.

Economic & Political Weekly

january 17, 2009

he operations and functions of the Indian telecom industry were monopolised by department of telecom (DoT) until 1994. The DoT set-up in 1984, acted as the policymaker, regulator and operator for all those services that existed in the industry then. The allocation of spectrum for different purposes was and is managed by the Wireless Planning and Coordination (WPC) of ministry of communications. Eventually, after almost 10 years since the establishment of the DoT, there arose a realisation that telecom services on demand at affordable and reasonable prices combined with a world class standard for their quality of services must be provided as charted out in National Telecom Policy 1994 (NTP-’94). NTP 94 recognised that the required resources for achieving these targets must also come through private investments and the involvement of the private sector is mandatory to bridge the resource gap.

1 Spectrum Auction in India

In the year of 1995, spectral auctioning began for global system for mobile (GSM) cellular operations alone. According to the National Frequency Allocation Plan (NFAP), frequencies 890-915 MHz and 935-960 MHz were earmarked for GSM operations. Similarly frequencies from 824-844 MHz and 869-889 MHz were allocated for code division multiple access (CDMA) operations by the WPC. As the first step towards GSM spectral auctioning of 2G services, India was divided into 21 circles – six states forming six “A” circles, eight states forming eight “B” circles and 12 states forming seven “C” circles. The four metros – Delhi, Mumbai, Chennai and Kolkata – were considered as separate areas of operations. This stage of auctioning of 2G spectrum was called “direct auctioning” and was conducted in two stages. Two qualified bidders who quoted the highest prices for the circle emerged as the winners in the second stage. This stage of auctioning services. Second, various non-winners of the bidding process had complained of a lack of transparency in the auctioning process, while the winners complained of high inter-circle interconnection costs fixed by the DoT. These were believed to be the reasons for the slow service roll-outs.

In order to resolve the above issues and to create a regulatory structure, the Telecom Regulatory Authority of India (TRAI) was set-up in 1997 and the New Telecom Policy ’99 (NTP99) was announced. The focus of the NTP99 was telecommunication for all and telecommunication within the reach of all. It offered a new scheme to the bidders called the “migration package”, according to which the existing service operators could make a one-time payment called “entry fee” and subsequently pay a licence fee, on a quarterly basis which is a part of their adjusted gross revenue (AGR).

Initially after the bidding process, the winners were allocated a spectrum bandwidth, which was decided to be 4.4 MHz for GSM operations (2x, i e, inclusive of uplink and downlink) and 2.5 MHz (2x) for CDMA operations. Today, GSM operators hold a maximum spectral width of 11.2 MHz in the metros, 9.2 MHz in circle A, 8.2 MHz in circle B and 6.2 MHz in circle C. The maximum in any circle for a GSM operator is 15 MHz. However, some GSM operators are demanding spectral widths as large as 22.5 MHz for their 2G network expansion besides their separate spectral claims for 3G services.

In this scenario, there arise questions regarding the efficiency with which these operators are utilising their allocated frequency bands. GSM operators need to use the spectrum they already have much more efficiently, noting that they serve 3.9 million subscribers per MHz, compared to China, where mobile carriers serve 8.5 million subscribers per MHz. There are also criticism concerning GSM operators for cartelising and hoarding spectrum. As a learning outcome from 2G spectrum auctioning, we believe for 3G spectrum auctioning and licensing the following key


aspects have to be thoroughly debated and detailed: (i) deciding the frequency band for spectrum allocation; and (ii) the minimum reserve price and auction design.

2 Spectrum Allocation for 3G

Before a country may start the licensing for 3G services, it is important to identify available spectrum. This requires earmarking newer frequency bands and the migration of incumbent users to other bands. According to India’s NFAP (2002), bandwidths ranging from 1920 to 1980 MHz will be used for International Mobile Telecommunication (IMT) 2000 services (which include 3G) for uplink and 21102170 MHz for downlink. In addition to this, it will be required to allocate more bands in the wake of newer technology and services. In the recently held World Radio Communication, under the aegis of International Telecommunication Union (ITU), more frequency bands were identified for IMT applications, viz, 450-470 MHz, 698-806 MHz, 2.3-2.4 GHz and 3.4-3.6 GHz. However in India, the stakeholders such as defence, Doordarshan and police expressed their concern about the use of the frequency bands 450-470 MHz, 698806 MHz and 3.4-3.6 GHz for IMT. TRAI also recommended the use of 450 MHz for 3G operations as the band has excellent propagation characteristics well suited for coverage in rural areas. However, the DoT has not approved of this idea yet and hence the spectrum allocation for 3G services is getting delayed in India. A similar scenario existed in the US when the US commerce department and US Congress had postponed their 3G spectrum auction by two years – from September 2002 to September 2004. The delay occurred so that the F ederal Communications Commission (FCC) could take the time it needed to identify the available spectrum and arrange for the incumbent users to migrate to other slots.

In a recent development on July 1, 2008, the DoT has informed TRAI that due to non-availability of spectrum in 450 MHz and 800 MHz bands, it is proposed that spectrum will be allocated only in the 2.1 GHz band. Blocks of 2 ×5 MHz in this band will be auctioned. Each bidder shall be allocated only one block in each telecom service area. India must take care in allocating spectrum. Perhaps India can take its cue from Germany, which allocated spectrum in smaller blocks and subsequently, the companies can be asked to bid again, depending on consumer response. This helped the industry to roll out the operation faster as companies have to shed less amount for spectrum and it also helped the government in terms of revenue collection and safeguarding against monopoly. The lesson is that slow and meticulous planning of spectrum allocation is required, despite the pressure from lobbies like Cellular Operator Association of India (COAI) and Association of Unified Telecom Service Providers of India (AUSPI)

3 Auction Designs and Minimum Reserve Price

Spectral auction is one of the major sources of revenue available to a government. With a booming telecom sector and its net worth of approximately $ 30 billion, the contribution of communication sector to Indian GDP has increased from about 1.59 per cent in 2000 to 3.97 per cent in 2006. It becomes imperative for India to choose a proper auction design to maximise government revenue comparable to other countries like the US, UK, Germany, Switzerland, etc. At the same time, emphasis must be given to safeguard the interests of telecom operators and ensure the future growth of the telecom sector and not just keep in mind the profits from auctions.

Worldwide, the third generation licences were awarded either through a beauty contest or other auctioning methods such as sealed bidding, open bidding, simultaneous multiple round auctions and Vickrey a uction. The beauty contest is an auctioning procedure, which is not wholly dependent on the pricing. With much s imilarities to an actual “beauty contest”, here too, every contestant will have to convince the auctioneers of their ability to provide services. Based on answers to questions like – how wide the service area will be, how fast and efficiently services will be rolled out and what will be the customer usage charges – a contestant will be selected as the winner of this “beauty c ontest” and will be provided the licence.

(i) Comparative International Scenario for 3G Auction: Some of the methods followed by various countries and their profitability rating is shown in Table 1 (p 19).

The advantage of a beauty contest, when compared to open auctions, is that the operators will be able to work on 3G sideby-side, even as they keep on focusing on the provision of mobile services to those who are still unconnected and remain a priority. Also, financial problems due to direct auctions (where bidders quote extremely high prices) and major delays in rollout after 3G auctions, a scenario that occurred in Europe after 3G auctions, can be

INDIAN INSTITUTE OF MANAGEMENT AHMEDABAD (Coordinator) REQUIRES THE FOLLOWING PERSONS FOR NAIP PROJECT (Assessment of impact of climate change on water-energy nexus in agriculture under canal irrigation system)

Research Associates and Senior Research Fellows:

Qualification: Master/PhD (or equivalent) in Hydrology, Economics, Engineering, Science or equivalent. Relevant experience preferred.

Tenure: Minimum one year, extendable based on annual performance review.

Age Limit: Senior Research Fellow: 35 years for men, 40 years for women Research Associate: 40 years for men, 45 years for women (Relaxation of 05 years for SC/ST and 03 years for OBC candidates)

Salary: Linked with the positions as laid down in the NAIP rules/ guidelines.

Apply to: NAIP Project, Wing-12A, Indian Institute of Management, Vastrapur, Ahmedabad - 380 015, latest by 31st January 2009.

Office Assistant: Qualification: B.Com, other conditions as above.

Please see for details.

january 17, 2009


Table 1: Spectral Auctioning Design (Various Countries)

Country Method of Licence Distribution Key Aspects Year S&P* Profitability Rating

Finland Beauty contest First country to award 3G licences. It was not price based. Four operators 1999 2000 Average Excellent
Japan France UK Beauty contest Beauty contest + auction Auction Three operators $ 4.5 billion per licence for the first two licences and later auctioning for remaining two licences Five licences – four for established 2001 Very good
operators and only one for a new operator 2000 Very good

c ollusion problems. The organisers of most of the auctions after the UK’s auctions failed to give enough attention to attracting entry and magnified their problems by permitting joint-bidding agreements prior to the auctions. The German and Austrian auctions demonstrated the vulnerability of ascending auctions to collusive behaviour

US Simultaneous multiple round auction Reserve price set to $ 10 billion 2004 Good

* Standard and Poor’s (S&P) is a division of McGraw-Hill that publishes financial research and analysis.

deterred. To develop a spectrum auctioning plan for 3G we should refer to the countries who have already auctioned 3G spectrum. For example, the UK ran the world’s first 3G auction in March 2000 for four licences. The problem faced was that there were also exactly four incumbent 2G mobile phone operators who had the advantages over any other bidders of existing 2G brand names and customer bases to exploit and lower costs of building 3G networks as they already had a 2G infrastructure, which they could back on. Therefore, the auctioneers were concerned that an ascending auction might deter other firms from entering and bidding strongly. So the government planned to run a hybrid of the ascending (English) and sealed-bid (Dutch) auctions and called it an “Anglo-Dutch” auction. The intention behind this was that the sealedbid stage would induce some uncertainty about which four of the five finalists would win and entrants would be attracted by the knowledge that they had a chance to make it to the final stage. So the sealed-bid stage would attract entry and so also raise revenue, while the ascending stage would mean less loss of efficiency than that might result from a pure sealed-bid auction.

The sealed-bid stage would also make collusion harder. The UK auction team made a deliberate strategy of being the first to auction and the fact that planning continued for three years before the actual year of auctioning, served as add-on time for a sustained marketing campaign to attract entrants. The version of an ascending auction that was used was widely judged a success; nine new entrants bid strongly against the incumbents, creating intense competition and record-breaking revenues of 39 billion euros.

Whereas Germany auctioned 12 blocks of spectrum from which bidders could create licences of either two or three blocks,

Economic & Political Weekly

january 17, 2009

for example, four firms could win large three-block licences or six firms could win smaller two-block licences. This contrasted with the previously discussed auctions in which all the licences were of a predetermined (though not always identical) size. As always, firms could win at most one licence each. The 12 blocks were sold by a simultaneous ascending auction, much like the previously discussed auctions. The point of the design was to let the number of winners be determined by the bidders who might have information unavailable to the government about, for example, the engineering advantages of large versus small licences. But such an auction’s outcome is driven by bidder’s profits, not by consumers or social welfare. Since the bidding in the British auction had already revealed a lot about bidder’s relative valuations of different licences, it would have been wiser to fix the number of licences in advance. The auction also proved vulnerable to collusion and entry problems: only seven bidders participated. The entry of weaker bidders was perhaps discouraged by the ascending design, as in other auctions after the UK’s.

Discussion on spectrum auctions can never be complete without discussing the Nigerian model (though it is 2G auctioning). The reason is that the Nigerian GSM auction, by virtually all reports, was considered to be a success. The final licence price of $ 285 million per licence was substantially more than the government had expected to raise. The Nigerian GSM auction surpassed virtually everyone’s revenue expectations and raised more revenue per capita than the 3G auctions in the Netherlands and Italy.

A key determinant of the success of the European telecom auctions was how well their designs attracted entry and discouraged collusion. The sequencing of the a uctions exacerbated the entry and during the auctions and there were also rumours of collusion in the ascending auctions in Italy, the Netherlands and Switzerland. All these problems were aggravated by the auctioneer’s failure to use the information from the UK auction to set sensible reserve prices. The auction design is not something that is ready-made for all.

The ascending design that worked very well for the UK worked very badly in the Netherlands, Italy and Switzerland. The experience of the various countries in auction provides valuable insight into the challenges faced by regulatory agencies as auctioneers. Moreover, the auction serves as an excellent example of why an allocation mechanism such as an auction needs to be evaluated by more than a single outcome if it is to be used for subsequent resource allocations in the future. Thus, the experiences of these countries should be studied with utmost interest by government and regulatory agencies in India.

(ii) 3G Spectrum Auctioning Design for India: The auctioning of 3G spectrum as proposed by the DoT recently, called “controlled ascending e-auctioning”, which emulates the 3G auctioning followed by FCC in the US. In this method of auctioning, the base price is fixed prior to the auction, depending on the metro and the circle. Based on the experience of successful 3G auctions in certain countries, the DoT has proposed that the reserve price for a block of 2 × 5 MHz in the 2.1 GHz band should be 0.5 per cent of GDP, which in the case of India, would come to $ 0.5 billion or about Rs 2,100 crore, which is twice that recommended by TRAI. Thus, the reserve price will be as seen in Table 2 (p 20).

The DoT has proposed that when the number of bidders left is equal to the number of blocks of spectrum being auctioned, in any service area, the auction will end. All the bidders will have to match the bid of the highest bidder. In case they do not match, then that block would be


Table 2: Proposed Reserve Price for 3G Auctions in India

Circle Reserve Price (Rs Crore)

Mumbai, Delhi and Category ‘A’ 160

Chennai, Kolkata and Category ‘B’ 80

Category ‘C’ 30

offered to the next highest bidder at the highest bid price. If any block is left vacant, then the block would be reauctioned. This would ensure that all successful bidders pay the same amount for the same 3G spectrum blocks.

A comparison of the proposed Indian 3G auction design with that of other countries is essential to help us frame better designs for auction. The proposed framework for 3G auctions in India projects a minimum

0.5 as the ratio of revenue from spectrum allocation to GDP. Also, with 260 million mobile phones and a mobile teledensity of over 80 per cent, if US could generate $ 10 billion, India, with over 230 million mobile phones and mobile teledensity of 22 per cent and a scope of exponential growth for a decade, clearly argues a higher spectrum pricing. The Indian ministry of finance, which is expected to take a share from the revenue earned through 3G auctions, hence, wants higher fees for spectrum. Thus, it is highly vital that the auction design, for 3G services in India, must be reviewed with at most care.

4 Conclusions

From the regulatory and policy perspective, spectrum auctions ensure efficient usage by allocating it to those entities that value it the most, while also generating revenues for governments. However, auctions may lead to unexpected outcomes as, for example, when regulatory agencies have inadequate market information, there may be a mismatch between expected and actual bidder behaviour or auctions may be poorly designed. The key challenge before regulatory agencies is to design auctions in such a way as to meet the objective of fostering competition while, at the same time, ensuring that bidders can effectively use the spectrum for their business. The pressure lobbies like COAI and AUSPI will try to pressure the DoT and TRAI on spectrum allocation as well as reserve price but the DoT must do extensive analysis before actually detailing key aspects of auction design, it is also vital to realise that the telecom industry can be a growth driver for the country, provided its growth is monitored and r egulated in a systemically designed regulatory environment. Auctioning of spectrum is one element of the telecom ecosystem; the actual aspects will be the deve lopment of the infrastructure and support system for the roll-out of the 3G services.

The 3G spectrum auctions to be held in India will have a great impact in the country’s booming telecom industry. Though a best auction design would lead to maximise the revenue earned by the government, it is important that the auction design safeguards interests of all the stakeholders. The main factors of concern are the telecom industry, its subscriber growth, average r evenue per user, tele-density and the customer. On the issue of new and foreign operators, entry is an apprehension that operators, especially those who missed out on the latest round of 2G licences, will drive up the bidding price in their eagerness to get into the market. This will result in

  • (a) high tariffs for the 3G services, and
  • (b) deprive the existing operators from providing the 3G services. Thus, while designing the 3G auctions for India, the DoT and TRAI must keep in mind the interest and the future of telecom industry and hence must sketch an optimum framework for auction design which need not be the best.
  • Visiting Scholars Program 2009

    Overview: The Visiting Scholars Programme (VSP) has been instituted in IGIDR as part of its outreach activities with the objective of supporting high quality research being conducted by Ph.D scholars and Junior Faculty Members of colleges and universities in India and abroad. The scholarship will enable up to five selected scholars to spend three months at a stretch at IGIDR anytime during April 1, 2009 - March 31, 2010, to work on a concrete research proposal that can be satisfactorily completed during the three-month period.

    Eligibility: At the time of application, applicants should possess a Masters degree in any field and be registered for Ph.D. on a topic broadly involving any area of economics including energy and environmental policy. Applicants are required to submit a curriculum vitae (CV), a research proposal and two letters of recommendation. One of the letters of recommendation should be from the concerned Ph.D supervisor in the case of a Ph.D student, and the principal of the College/University in the case of a Junior faculty member. Recommendation letters should clearly give the contact details of the recommenders. Candidates should clearly mention in the application the preferred three-month period of stay at IGIDR.

    Research Proposal: The proposal should involve economic issues including those related to energy and environment. It should be strictly within 1,500 words, and its scope should be such that the project can be completed in three months. Proposals in both theoretical and empirical issues will be considered. The research proposal should be accompanied by a signed certification by the applicant as well as by the concerned thesis supervisor/principal, stating that the research proposal is original work of the applicant.

    Selection will be based on a joint evaluation of the CV, letters of recommendation and the quality of the research proposal submitted at the time of application. Proposals will be evaluated on the basis of a double-blind refereeing process.

    Emoluments: Selected scholars will be paid Rs. 9000/- a month towards scholarship, reimbursed to and fro three-tier AC train fare, and provided free accommodation on campus for a stretch of three months. Foreign scholars residing outside India, however, will have to bear the transportation expenses on their own.

    Requirements: Selected scholars will be required to submit a progress report of the research project at the end of six weeks and a complete research report along with a seminar presentation of the research findings at the end of the three-month stay. Failure to submit a progress report on time may lead to the termination of the scholarship programme. Once the research project is completed and published, the scholar should acknowledge IGIDR in an appropriate way.

    Application Deadline & Notifications: Applications can be sent electronically or by post to reach IGIDR by January 31, 2009. Letters of recommendation should be sent by the concerned persons directly to IGIDR via email/post. The subject head for email and top of the envelope for post should indicate APPLICATION FOR VISITING SCHOLARS PROGRAM 2009. Selected applicants will be notified via email/post by February 15, 2009. Those selected will have to accept the offer latest

    by February 28, 2009.
    Address for Communication: Visiting Scholars Program 2009,
    Indira Gandhi Institute of Development Research (IGIDR)
    General A. K. Vaidya Marg, Goregaon (E), MUMBAI 400 065, INDIA
    Email: ; Telephone: +91-22-28416200

    january 17, 2009

    Dear Reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here

    Back to Top