It’s implausible that business is not shaping the politics of 3G spectrum auction. After all, it’s in the interest of some operators for 3G to happen now and in the interest of others for it to stay in limbo for some time.

GSM operators
Bharti, Vodafone, Idea

Auction please—now. They are facing a 2G spectrum crunch, which is choking their voice business. With the government short on 2G spectrum, they are banking on 3G to ease the pressure.

Dual-technology operators
Reliance, Tata

They have sufficient 2G spectrum to drive their voice business on GSM and get data users on CDMA. They might still take part to stock up on this finite resource, but every delay works in their favour.

Government operators
BSNL, MTNL

They got 3G spectrum in 2009, with the price to be determined in the public auction. With their profits in free fall, a delay suits them. Plus, it extends their first-mover advantage, not that they have done much with it.

New operators
Uninor, Aircel, MTS, Loop and others

They have adequate 2G spectrum. Since they are currently investing to set up 2G voice networks, they will be hesitant to commit more to 3G. A delay gives them time to catch up with older players.

***

The 3G Ministerial Face-off

The government has made a meal of the 3G spectrum auction. Even as the ministers indulge in blame games and turf wars, the Prime Minister’s Office (PMO) maintains a studied silence.

Pranab Mukherjee
Finance Minister

Was pushing for an auction in 2009-10, as it meant Rs 30,000 crore to mend government finances. Squabbling ministers have all but killed that plan.

A Raja
Telecom Minister

After the 2G licence farce, Raja was under pressure to get 3G right. He blamed the delay on the Defence Ministry: the armed forces weren’t vacating 3G spectrum.

AK Antony
Defence Minister

Was all set to vacate spectrum. But offended by Raja’s broadside, Antony became obstinate: defence will vacate only after it gets an alternate network, currently being built by BSNL.

Veerappa Moily
Law Minister

Dragged into the buck-passing when Raja sought his ministry’s opinion on holding the auction without spectrum in hand. His verdict: it may end up in the courts.

Manmohan Singh
Prime Minister

Brought in Sam Pitroda to resolve the telecom-defence impasse. Pitroda’s road map would have led to auctions this fiscal and defence vacating spectrum by August. But no one’s listening.

***

Just who is directing traffic at the auction for 3G spectrum? Telecom Minister A Raja should have been the man. But the man who sank to new levels of infamy in 2008 by hastily handing out licences at a tenth of their market value seems to have lost his zeal. He pleads helplessness and passes the blame on to the Defence Ministry, which has to vacate 3G spectrum for the auction to happen. The Defence Ministry says it’s a question of national security and that unless the armed forces get an alternate network, it won’t release the spectrum. Easily done. It’s an inter-ministerial conflict that could have been resolved by the Prime Minister, but hasn’t. Not yet, at least.

In the ministerial merry-go-round, operators wait for spectrum, consumers wait for new technologies and services, and the Finance Minister waits for money to balance his books. “It is not just about telecom,” says Hemant Joshi, telecom leader for Deloitte in India. “There are lots of inter-linked opportunities in sectors like education and healthcare going abegging.” If the country is waiting and everyone is losing, who is benefiting? A top official of a leading GSM (a telecom technology) operator says the strings are being pulled and confusion is being orchestrated by a leading CDMA (a competing technology) operator.

“Personality clash cannot be the only reason. There are behind-the-scenes actors advising them (the ministers) not to change their stance,” he says. Retorts his counterpart from the top CDMA operator: “It is more of an operational issue within the government.”

It’s classic GSM versus CDMA: two polarised voices, speaking on condition of anonymity, with an eye on a larger prize. This fight is not about 3G services—data and multimedia—as it doesn’t have a strong business case yet. This fight is about control of the sector. Specifically, it’s about control of spectrum—the finite resource of radio waves that enable calls and data transfer—and how best to align it to business interests.

What iron ore and coal mines are to a steel maker, spectrum is to a telecom company. The amount of spectrum determines how many subscribers an operator can service, the voice quality of its network and the valuations it commands. As these five policy flashpoints dating back to the beginning of the GSM-CDMA rivalry show, the fight is as much about how much spectrum an operator can get as it is about how it can stop others from getting it. And, at the moment, the sector as a whole is losing out.

Flashpoint 1: CDMA’s Backdoor Entry

The GSM-CDMA conflict began in 2002, when Reliance decided to make a charge in telecom, and bought a fixed-service licence for Rs 410 crore. However, instead of the conventional fixed phones, it launched a service that fell between a fixed service and a mobile service. It was called wireless in local loop (WLL) and was on the CDMA platform. It was like a cordless phone, except its signal range wa­sn’t a radius of 20-50 metres, but a few kilometres. Think Connaught Place (CP) or Nariman Point.

Call Drop

The price wars are hurting the sector. Growth will come from data and value-added services. For GSM operators, 3G is a must.

Further, Reliance kept increasing its WLL signal range. So, CP became Delhi, which became the National Capital Region, which became India as the company stitched up its roaming links. Tata did exactly the same thing, with a WLL service called Walky.

The GSM operators complained that Reliance and Tata were offering a mobile service in the garb of a fixed service—while paying a fraction of what they had paid in licence fees. Reliance and Tata, they charged, had made a backdoor entry into the mobile business. They took the two companies to court, in what was to be the first of many skirmishes between GSM and CDMA.

Flashpoint 2: CDMA Goes Frontdoor

Even as the court case was going on, in 2003, DoT offered a single licence for all services (fixed, mobile, Internet, STD and ISD). The price of the all-India, all-service licence was Rs 1,651 crore—what the fourth mobile operator paid in 2001. In spite of the business maturing in two years, the government didn’t increase the licence fee.

Reliance migrated to an all-service licence, paying Rs 1,096 crore as entry fee. It also paid a penalty of Rs 526 crore—an admission of its WLL misrepresentation. The government said: let’s start over, everyone pays Rs 1,651 crore. If an operator had paid more over the years (as Bharti and Vodafone had), the excess would be adjusted against revenue-sharing dues in coming years. And if an operator had paid less (Reliance and Tata), it would have to make good the shortfall. “Policies in the sector are made on the fly,” says Mahesh Uppal, Director, Com First (India), a telecom consultancy. “When problems arise, a patch-up is done.”

All the operators settled their accounts, but the GSM-CDMA fault line had deepened. While GSM was the leader, CDMA was the challenger. While GSM was marketed as a service for the classes, CDMA was pitched as a service for the masses. While GSM was suited for voice, CDMA’s strength was data (and it could also do voice).

Flashpoint 3: Race For Spectrum

These contrasts shaped how each operator ran its business and what policy stance it took. As mobile telephony finally took off, operators realised they would need more spectrum to accommodate their rising subscriber numbers. And since spectrum was finite, they figured, too much of it could never be a bad thing. More so, since it was free. The race for spectrum was on.

In 2001, the government had decided that a GSM operator would get 4.4 MHz on award of a licence plus 1.8 MHz after it crossed a certain subscriber threshold. For CDMA, which was a more efficient technology, the entry-level spectrum was fixed at 2.5 MHz (with 2.5 MHz to follow).

Policies in the telecom sector are made on the fly. And when problems arise, a patch-up is done.Mahesh Uppal, Director, Com First (India)

With the subscriber-linked policy, the number of subscribers became the key metric for an operator. New subscribers at any cost became the new business plan. Initially, it suited the GSM players because they had a head-start over the CDMA players. But CDMA responded forcefully, with schemes that played the price card aggressively. Reliance shook the market with its Monsoon Hungama scheme. GSM operators followed with ‘lifetime validity’ schemes.

In 2006, a fake and inactive subscriber scam surfaced in Punjab, establishing that operators were inflating numbers to bag spectrum. A subsequent inquiry conducted by the Department of Telecommunications (DoT), based on April 2009 numbers, showed that against the 380 million subscribers reported by operators, the actual subscriber base was not more than 300 million—about 25% less. The rest were multiple, inactive and fake connections.

Flashpoint 4: New Licences

The race for spectrum took a new turn in August 2007, when Trai ruled there was no need to cap the number of operators. And, more importantly, it assured spectrum to the new licencees on a first-come, first-served basis. Within a month, DoT was flooded with 575 applications. (One circle equals one application; so, if Unitech applied for an all-India licence, it would have put in 22 applications.)

The top 9% of a firm’s sub- scribers account for 29% of its revenues and 45% of margins. That’s why RCom and Tata are eyeing post-paid customers.Kunal Bajaj, Managing Director, BDA Connect India

In January 2008, Minister Raja awarded 121 licences. There were five new companies among the allottees. He also awarded GSM licences to Reliance and Tata. The price for an all-India, all-service licence was still Rs 1,651 crore—what it was in 2001. The number of subscribers had increased from 1.1 million to 210 million in the intervening period, the cost of setting up a telecom network had halved, but still the government didn’t see a case to increase the cost of a licence.

The older GSM operators lost out in the realignment. “The government had said it would give spectrum first to existing players who had applied, then to players who had given the money and got the licence, and then to cross technology,” says a top official at an old GSM operator. “But it happened in reverse order.”

Even as the older GSM operators queued up, Reliance and Tata leapfrogged them, and walked away with 4.4 MHz of spectrum. As did the five new players. To make matters worse for GSM operators, Trai recommended trebling subscriber numbers for incumbents to get more spectrum. “This will push my eligibility back by several years,” says the GSM official. The exponential increase proposed either means spectrum was being given out liberally previously or the norms have been tightened unduly to squeeze the existing players.

Flashpoint 5: The 3G Impasse

Today, the older GSM players need spectrum—fast. With 2G spectrum drying up, and a pile of players waiting, they can’t bank on it. So, they are keen on the 3G auction. Once again, the CDMA players, who have now became dual-technology players, have a position that is diametrically opposite to the older GSM camp.

Both Reliance and Tata already have sufficient CDMA spectrum. Armed with new GSM spectrum, they are increasing their presence in the GSM space—in the past year, they have matched Bharti and Vodafone in subscriber additions. The more 3G is delayed, the more it works in their favour, on both the voice and data businesses.

If the spectrum crunch persists for the older GSM players, the voice quality and call drops on their networks could get worse—and compel subscribers to shift. Meanwhile, Reliance has opened a new front against them. It is now saying that under the licence norms, only 6.2 MHz of spectrum is free. Above that, operators should pay. It’s a position that suits Reliance because its spectrum usage is below 6.2 MHz. But it could weigh heavily on Bharti and Vodafone, which have 10 MHz in some circles. If they have to pay for this additional spectrum, it will compromise on their ability to pay in the 3G auction, if and when it does happen.

“It will happen,” says an official at a leading telecom-technology company. “The process has gone too far ahead to go back.” If so, the question is how much of an inroad Reliance and Tata can make in the interim. Both companies are looking to increase their share of post-paid subscribers, which is currently just 1%. For Bharti, the figure is 4.7%. “The top 9% of subscribers account for 29% of revenues and 45% of margins,” says Kunal Bajaj, Managing Director, BDA Connect India, a telecom consultancy. “On the GSM side, the Tatas’ execution has been brilliant, but Reliance seems to have fallen midway,” says Bajaj.

By focusing on subscriber additions on their new GSM network, both Reliance and Tata are clearing space in their CDMA networks for data services. Both companies have launched high-speed broadband (up to 3 Mbps) on USB dongles, on their CDMA networks. Here, their average revenues per user (ARPUs) are Rs 1,000-1,500 per month—7-10 times their voice ARPUs. The GSM players can’t match that because GSM doesn’t support similar data speeds on 2G networks. Another reason for 3G.

Once 3G happens, older GSM players will redistribute their spectrum usage. They will woo high-value, post-paid subscribers with 3G services. So, the longer the wait, the greater the threat for them. The whispers are that dual-technology operators are two months away from offering the same high-speed data service on a handset, which could attract post-paid subscribers.

The 3G auctions will change a lot, much of it to the consternation of the dual technology and newer GSM players. “It will set a price for the spectrum,” says Uppal. “No one will dare give it for free.” The only players who will be comfortable paying that kind of money will be the older GSM players, perhaps the two dual-technology players.

Every delay in the 3G auction postpones that process of discovery. It also causes direct damage to the industry. As it is, the industry’s revenues are in decline because of the pricing war. In spite of a 36% increase in subscribers between December 2008 and September 2009, revenues are stagnant. A similar trend is seen in the government’s tax and revenue-sharing collections from the sector (see graphic: Call Drop). Customers are being denied access to the latest technologies. The GSM-CDMA war on these lines could prove to be costly for everyone.