The war for the regulator’s heart

This piece appeared in NCAER's Margin in January/February 2006

Abstract

The telecom industry in the country continues to be one of the best performing ones in the country, far outstripping most conventional industries, let alone infrastructure ones. Naturally, then, one would assume that telecom regulation in the country is well-developed, impartial and fully functional. Nothing could be further from the truth. India’s telecom regulation continues to remain patchy and is largely held ransom to the government of the day’s whims and fancies – that remains as true today as it did when, as telecom minister in the previous government, Arun Shourie rammed home the controversial legalizing of Reliance Infocomm’s WiLL-mobile services. Indeed, over the past year or so, there has been a virtual stalemate between the Telecom Regulatory Authority of India (TRAI) and the Department of Telecommunication (DoT) which is the policy making wing of the government. So much so that Communications Minister Dayanidhi Maran recently announced a major change in policy for long distance and internet telephony without even seeking the mandatory TRAI recommendation. Major recommendations given by the TRAI over the last year have also not been acted upon.

This is not to say that there is no credible regulatory regime, but that it’s record has been quite patchy. And to the extent there have been major regulatory breakthroughs, many have been the result of long court battles initiated by the operators themselves. The ability of the regulator to rein in unfair practices by the incumbent has also been quite poor, though it must be said part of this has to do with the fact that the government itself has played no mean role in this.

Growth of industry

The telecom industry remains one of the fastest growing in the country, with around 2.5 million new customers getting added to the mobile users club every month – between 1997-98 and 2004-05, the average growth of the mobile phone business was around 80 per cent annually. Growth in the number of fixed line users, by contrast, has been much slower at around 12 per cent per annum over the same period. To put these numbers in perspective, the total number of fresh mobile phone entrants alone spend around Rs 16,200 crore in the first year of their entry, assuming a monthly telephone bill of Rs 200 and a basic handset cost of Rs 3,000 at the entry level. This sum equals around 60-70 per cent of the market for four wheelers in the country today.

As a result, most of the targets of the New Telecom Policy (NTP) 1999 have been achieved before time. The teledensity target for the year 2005 was 7, and it was 15 for 2010 – the 7 per cent target was achieved in March 2004 itself and we should be able to achieve 15 by 2006. By 2008, the TRAI’s projection is that teledensity will be 22.5, or nearly a fourth of all citizens will have a phone. Put another way, over the past few years, the annual growth in teledensity has been higher than that seen in the first 50 years of the country’s independence.

The growth, it should be mentioned, has primarily been driven by mobile phones (between 1998 and 2005, a total of over 51 million mobile subscribers were added on) while the fixed line growth has been much slower. There were 18 million fixed lines in 1998 and this grew to around 46 million by 2005. The larger share of this overall growth of fixed and mobile phone lines was achieved by private operators who put together around 45 million lines.

Growth in mobile telephony really started picking up in 2001 when the third and fourth cellular operators were allowed in, and then skyrocketed in May when Calling Party Pays was introduced and that allowed cellular phone firms to provide incoming calls free. Bringing in Reliance and Tatas at the same time also increased competition and lowered rates further.

As a result, revenues earned by the government from the sector have shot up. The exchequer earned Rs 6,236 crore from the sector in 2002-03 from license fees and service tax and this will be around Rs 17,850 crore in 2005-06 and is projected to rise to Rs 30,856 crore in 2007-08.

The rural teledensity target, however, looks like it won’t be achieved – it was 4 for the year 2010 and is currently 1.74 – since few companies are planning network expansion here and the current system of funding every line through the USO (Universal Service Obligation) fund is proving to be very slow and expensive. Till April 2005, the USO fund had collected Rs 7,254 crore but had disbursed only Rs 1,815 crore.

The government clearly needs to come out with a new policy initiative to jumpstart rural telephony. The TRAI has made a recommendation on how the USO fund could be used to subsidise mobile phone towers as opposed to the current policy of calling for bids for each phone line established. While this has the potential for misuse in the sense operators could take the subsidy for urban phones as well (since a large number of villages are located near urban agglomerations and so could theoretically be served by the same mobile towers), if properly implemented through inspections by third parties, it has the potential to create a lot more phone connections than the current system, and at a much lower cost. Unfortunately, the stalemate between the TRAI and the DoT has meant that this recommendation too has been put on ice.

The growth of the internet, especially broadband, has been quite poor, especially in comparison with other countries. The number of internet users per hundred people in the country rose from 0.4 in March 2003, for instance, to just 0.53 in June 2005; the same figures were just 0.02 and 0.04 respectively in the case of broadband usage.

Regulatory regime

While the spread of telephony and the dramatic fall in prices is an indirect testimony to the efficacy of regulatory policy, the biggest achievement by far stems from the principles of interconnection (the act of a user of one phone service, say Reliance, being able to talk to, or receive calls from, users of, say BSNL) first announced by Justice Sodhi, the country’s very first TRAI chief, in 1999 itself. This was then followed up by later TRAIs, and the first comprehensive Interconnection Usage Charge (IUC) order was announced on January 24, 2003 which lay down, on the basis of costs, the exact charges to be paid when a call originated from some network, the cost to be paid for carrying that call over different distances, and the charges to be paid while terminating the call on another network. These principles also stipulated that a certain maximum time frame had to be set for all interconnections to take place as, without rule and time-based interconnection, no telephony is possible.

It is this use of the cost-based principle, for instance, that saw the regulator argue that tariffs of leased lines in the country were around 70 per cent higher than they should be based on their costs as well as international benchmarks. And though it took over a year for the TRAI to finally get service providers to fall in line (the case was dismissed by the TDSAT once when VSNL argued that the TRAI had not shared its data on the matter), the sharp reduction in leased line costs that is currently taking place is owed entirely to the TRAI.

i) Incumbent’s abuse of power

While the good news on Indian telecom is that around 2.5 million customers are getting added on each year in the mobile space, the congestion on the networks has multiplied manifold. Indeed, in July 2005, the TRAI’s Quality of Service monitoring showed that there were 86 Points of Interconnection (a POI is the junction where, for instance, a Hutch phone connects to a BSNL one) where the congestion levels were as high as 10 per cent, a figure that is itself 20 times as bad as the benchmark ideal. This, however, rose to 122 POIs the next month itself.

This, in a sense, is the most evocative illustration of the problems with the country’s telecom regulatory environment – the problems are mainly with the government-owned BSNL (which controls 79 per cent of the country’s fixed lines) refusing to provide POIs and the TRAI is powerless to do anything about it. The TRAI has cited 918 cases of pending demands by private operators and of these 367 have been pending for more than a year. The TRAI had ordered that such interconnection, which is the life blood of the telecom industry, be provided within 90 days, but BSNL challenged the TRAI’s jurisdiction at the Telecom Dispute Settlement Appellate Tribunal (TDSAT) which ruled in BSNL’s favour. The TRAI has challenged this in the Supreme Court, but no date has been fixed as yet for the hearing. It appears, according to experts, that there is some ambiguity in the TRAI law, and the whole dispute can be fixed in no time if the government decides to amend the law, as they did in double quick time in the case of allowing reservations in unaided government institutions. The government, however, has not seen it fit to move on the matter and is content to let the judicial process take its time.

This, sadly, is not the only case of abuse of power by BSNL. Since there is a huge slippage in the broadband targets, the TRAI recommended that the incumbent players be forced to make available to other players their last-mile fibre going into consumers houses – this would then jumpstart the availability of broadband. While finalizing the broadband recommendations, however, the DoT simply ignored this, as a result of which, if any private player wants to lay a broadband network, he has to lay a fresh cable into people’s homes and offices. Naturally, the progress is tardy – BSNL got just 0.17 million broadband connections by September this year against the target of 1 million by December.

Another recent instance of how the government continues to play favourites is that of Virtual Private Network (VPN) services that various ISPs provide. Till recently, these were part of the normal ISP license, but when it became obvious that this was a lucrative area of business, BSNL started creating problems and refused to give leased line access to ISPs for the service despite the TRAI ruling that provision of such lines was mandatory and even laying down what the tariffs should be. Indeed, in certain cases, the incumbent operators even issued circulars instructing regional heads not to provide leased lines to customers who were using VPN services offered by rivals. The government then changed the rules overnight and declared, incorrectly, that VPN services were not part of the original ISP licenses which were free, and fixed a very high entry fee (Rs 10 crore) for such services. When the matter came to the TRAI over a year ago, it said that much lower entry fees (Rs 30 lakh) needed to be set on VPN services. The government, however, refused to move on the TRAI recommendation!

Indeed, till very recently, the TRAI was not able to get BSNL to provide details on its different services separately (account separation) which was required to ensure that BSNL was not using monopoly profits from one segment (such as fixed lines) to subsidise another area (such as mobile lines). The accounts separation was also very important in the context that the cellular industry argued (more on this later) that the Access Deficit Charge that they were contributing to BSNL’s coffers was way higher than was actually needed, and that BSNL was using this to subsidise its new cellular business. The first paper on the need to get BSNL to separate its accounts was put out in 2000.

ii) Reliance WiLL:

During the days of the WLL controversy in 2001, when fixed line license companies like Reliance began offering their customers full-blooded roaming facilities (essentially, they were offering mobile services without having paid the license fees), this was done through the use of Wireless in Local Loop (WiLL) technology. Initially, the WiLL companies said they would offer mobility only within a small area and since this was very different from the full-mobility offered by cellular companies, no license fee was charged. But since, thanks to the march of technology, it was now becoming possible for WiLL phones to offer full-blooded mobility as well, the TRAI figured out that the only way to offer restricted mobility would be through the use of a switching system called V5.2, and so it recommended that this clause be inserted in the licenses of the WiLL players.

This, however, was honoured more in the breach and Reliance Infocomm instead used a Mobile Switching Centre which is what the cellular firms used and could hence offer full mobility. The then TRAI chief MS Verma wrote to the telecom secretary Shyamal Ghosh on the matter, and was rudely snubbed for his efforts. Verma’s letter to Ghosh, in January 2001, was good 26 months before Reliance commercially launched its services, so its very obvious the government was turning a blind eye to it and was telling the TRAI to do the same.

When matters came to a boil in the appellate tribunal, there was a split verdict in August 2003. While the head of the TDSAT, the only judicial member by the way, ruled that the WiLL service should be stopped, the other two members ruled that while the services should not be stopped, a method should be found to ensure that the mobility offered was restricted – interestingly, at that point in time, the Tatas, were offering only restricted mobility. At that time, when Arun Shourie was the telecom minister, the government had to implement the order, but chose not to do so and bought time on one pretext or the other. Finally, the TRAI was petitioned on the matter and, in October, it came up with a new policy recommending Unified Access which allowed the WiLL mobile players to offer full mobility after paying a license fee equal to that paid by the fourth cellular licensee a few years prior to this. While that penalty plus license fee added up to around Rs 2,000 crore, Reliance had built up a subscriber base that was many times this by virtue of the fact that it offered (till then, illegal) mobile services while rivals like the Tatas didn’t.

iii) Calling Party Pays:

When finally introduced from May 1, 2003, this single-handedly changed the face of the mobile industry as incoming calls suddenly became free, and this caused a huge surge in sales of mobile handsets. Till then, each time a fixed line phone got called, the operator who was calling paid a certain charge for terminating the call – so, if a mobile phone user called a BSNL line, the mobile customer paid a certain charge to BSNL (calling party pays). Problem was, however, that when a fixed line user called a mobile phone, he did not make any payments to the cellular phone company. Which is why cellular phone firms charged their subscribers even for incoming calls while the new WiLL-mobile phone firms didn’t since they were seen as just an extension of the earlier fixed line phones.

It was as early as 1999, that the TRAI first put out a consultation paper on CPP, but for one reason or the other, the mobile phone firms never got CPP. By 2003, when the WiLL-mobile phones began stealing a march over the cellular ones since they offered incoming calls free, the cellular firms decided to take things in their hands and cut off all points of interconnection with the WiLL phones. A national crisis followed, and when Communications Minister Pramod Mahajan read the riot act to the cellular firms, they put out full page ads in the papers exposing the policy-induced favouritism shown to WiLL-mobile firms –within a week of this, on January 24, 2003, CPP was brought in for the cellular firms as well! It’s a different matter that the new TRAI chief then delayed the launch of CPP for another few months.

iv) The ADC regime:

In May 2003, the TRAI fixed that a sum of Rs 13,000 crore was what needed to be paid to BSNL to compensate it for installing below-cost phones, and this money was sought to be recovered through a cess of sorts on all phone calls. The industry, however, pointed out that the calculations were incorrect, and within four to five months, the TRAI revised this figure down to around half. Detailed rates were then fixed of the ADC to be paid on both national and international long distance calls. With the industry continuing to protest, saying that BSNL did not deserve the subsidy, in June 2004, in another consultation paper, the TRAI came up with a range of ADC payments depending upon what rentals BSNL charged its customers – if the rental was Rs 200 per month, the ADC had to be Rs 1,402 crore and this would go up to Rs 3,436 crore if the rentals was Rs 156. In January 2005, however, the TRAI went back to the earlier figure of Rs 5,300 crore. While the TRAI now claims to have got BSNL to file separate accounts, these figures have not been made public, so there is still no way of knowing if the monies are being used to cross subsidise other BSNL businesses.

v) Reliance again

If the government and the regulator helped legalise Reliance Infocomm’s WiLL-mobile service in 2003, it did it another favour in 2004. In September or October, there was enough evidence to show that Reliance was changing the Calling Line Identification (CLI) on international calls received by its network, and replacing this with local numbers – since an ADC of Rs 4.25 per minute had to be paid on every international call either received on a network or made from it, this changing CLI had the potential of saving more than Rs 1,000 crore. When the DoT’s investigation wing first confronted Reliance with this, the company denied it; later when numbers on which this was happening were given to Reliance, the company admitted it had reserved 30,000 numbers in Mumbai, Chennai and Kolkata for the purposes of this CLI change.

Yet, the regulator refused to investigate the matter, saying it had no staff to do so – apart from the theft angle, allowing Reliance to get away with the theft meant the firm could, and was, offering cheaper long distance telephony to its subscribers and this gave it an unfair advantage over the competition and so this was clearly a matter for the regulator to take action on. While it is true, the TRAI didn’t have the requisite staff strength to do a full-fledged investigation, it never even asked Reliance how it was offering US-to-India calls for 11.9 cents (Rs 5.47 a minute) when it had to pay Rs 4.25 for the ADC, a costs of Rs 1.15 (2-3 cents) for carrying the calls to India and another 30 paise to the telephone firm in India on whose lines the calls terminated.

And while the government had jailed all small offenders who’d offered callback facilities in the past to avoid paying the high official long distance rates, it repeatedly refused to take strong action against Reliance. Even the suggestion of canceling its license was struck down, and the period of investigation limited to keep the penalties down – Reliance was not even asked to come forth with details of the accounts of its US subsidiaries/affiliates that sold the Reliance calling cards that were used for the ADC-avoidance game.

vi) Soft on cartels

While the TRAI was quick to act in the matter of leased line tariffs, it has been very slow to react to the same degree of cartelisation taking place in the long distance traffic business, leaving it to the telecom minister to publicly lash out against these companies and then finally come out with his own policy to fix things since the TRAI clearly didn’t see any problem. Right now, for instance, the tariff on a New Delhi to New York call on a mobile phone is around Rs 14 a minute whereas the costs on this are just around Rs 4-5 – Rs 2.5 for the Access Deficit Charge (ADC) and a maximum of Rs 1.5 that needs to be paid to the operator for carrying it overseas. In June, BSNL invited bids from companies to take its calls overseas and Reliance bid 74 paise for it.

On a Delhi-Mumbai call, the costs are around 1.3 a minute (30 paise ADC, 70 paise for the carriage cost, and another 30 paise for termination fees that have to be paid to the operator on whose network the call is made to) but the tariff is around double this.

Indeed, when the regulator was formulating the rules for a new Universal Service License (USL) which would allow users to do any kind of service with just one license, it’s curious that it invoked a principle it had never invoked before – that the existing long distance players had invested good money and so needed to be given more time before new players came in! In the event, the TRAI decided that a new USL license would cost Rs 107 crore in year one (a national long distance one today costs Rs 100 crore while and international one costs Rs 25 crore) and going down progressively to Rs 102 crore in year 2, Rs 92 crore in year 3, Rs 72 crore in year 4, Rs 32 core in year 5 and Rs 30 lakh after that – so, it wouldn’t be before a few years that there would have been real competition in the sector according to the original TRAI plan.

vii) Spectrum screw up

The two most important things in the telephony business, anyone will tell you, are the interconnection regime and availability of spectrum, or the frequency on which the phone companies transmit their signals (the latter applies not just to mobile telephony, but also to fixed line telephony where the last mile connectivity is being done spectrally). In the case of the first, as we’ve seen, the TRAI’s ability to rein in the incumbent is poor thanks to the government continuing to protect BSNL. On spectrum issues, the TRAI’s track record is one of having worsened matters.

While the TRAI first said there was enough spectrum to go around, and used this to bring in new players when WiLL-mobiles were being regularized and allocated them fresh spectrum (India is perhaps the only country in the world that has 6-7 players in each telecom circle), the TRAI’s current view is that there is a huge spectrum shortage. The regulator rebuts this charge of having created the spectrum shortage on two counts. First, it says that when the new players were brought in, in 2003, it didn’t look as if the growth of mobile phones, and hence the need for spectrum, would be as high as it is today. Second, the spectrum that the new players (like Reliance) used could not have been used by the existing players who used GSM cellular technology – so whatever else, the TRAI’s actions didn’t reduce the spectrum available for the existing operators.

The TRAI recommendations on Unified Licensing in October 2003, which is what brought Reliance and the Tatas in, however, say something quite different. For one, it says mobile phone growth ‘has accelerated from around 3 lakhs (sic) subscribers per month in May 2002 to almost 2.26 million subscribers per month in May 2003 … In March 2003, the wireless subscriber base was 13 million, which has almost doubled in last seven months … the expected wireless subscriber base by December, 2005 will be 100 million.’ As a matter of fact, the mobile subscriber base will be around 65 million by the end of the year, but surely this makes a mockery of the current claim that, two years ago when it brought in more players, the TRAI didn’t expect this kind of explosive growth – indeed, the TRAI’s projections two years ago showed the spectrum shortage in 2005 was going to be pretty serious and yet it got in more players. The claim that the WiLL-mobile firms use a completely different spectrum from the GSM lot is also incorrect as WiLL-mobile companies were allotted additional spectrum in the 880 MHz to 889 MHz frequency in which GSM equipment also works.

Curiously, the same paper says ‘additional spectrum is now being made available by Ministry of Defence and the existing contractual commitments to existing cellular and WLL players can easily be met, leaving out a balance for more players.’ Today, the TRAI’s stance is that there isn’t anymore spectrum to give since the Ministry of Defence is not vacating the spectrum!

Indeed, having brought in the WiLL-mobile players on the basis of their so-called spectral efficiency (the claim was that they used the scarce spectrum resources five times as efficiently as the cellular players), the TRAI has now done another volte face and says equal spectrum should be granted to both technologies. The TRAI also wants this spectrum to be allocated free. This has two problems. First, if WiLL is genuinely more spectrally efficient, given the WiLL firms more spectrum will make them better off vis a vis the cellular firms since the expansion of mobile networks is getting held up due to a serious spectrum shortage – since the debate about which technology uses spectrum more efficiently will never get resolved satisfactorily, perhaps a better solution would be to auction the spectrum. Two, while more spectrum can be allocated to the WiLL firms immediately, this will allow them to begin offering the next generation of mobile services, called 3G, which essentially allow much higher data transfer speeds (it’s like the difference between a dial up and broadband internet speeds), the spectrum needed to allow the cellular firms to offer 3G services is still not available as the defence services are yet to vacate it.

The road ahead

Apart from the issue of ensuring discipline in the interconnection regime and getting BSNL to fall in line, to fix the ADC mess which unfairly subsidises the public sector BSNL, and coming up with a solution to get more phones in rural India, another critical competition issue that the TRAI needs to move on is that of the carrier access code – again, this is an area where the successive governments have played foul and have not allowed the TRAI to instill discipline in the market. In markets like the US, for instance, subscribers are free to choose which long distance carriers they want their calls to be carried on through the use of a Carrier Access Code (CAC). So, for instance, you could have an Airtel phone, but you find that Reliance Infocomm is offering a lower-cost long distance call to Mumbai from Delhi. So, if you had CAC, and Reliance’s CAC was 9876 say, all you would need to do would be to dial 9876 from your Airtel phone, get on to the Reliance network, and then dial the Mumbai STD code and phone number. That’s how competition unfolds in the sector – right now, you just have to accept the rates offered by your access provider, in this case Airtel, and that’s why long distance tariffs are so high in the country.

On July 24, 2002, it appears, the TRAI had first issued instructions that CAC be implemented, but BSNL said it was not ready at that time, and it was given an extension by the government. The TRAI recently set a new deadline for the end of November, but that too has passed without CAC getting implemented. CAC is critical, for while the new long distance policy announced by Minister Maran will dramatically lower entry fees in the long distance market (an international and domestic long distance license will cost just Rs 2.5 crore apiece and various rollout obligations for both have also been done away with), no newcomer can offer such services unless he has access to the ultimate consumer, and unless consumers can access the new service provider through CAC, their existing service providers will never let them do this.

The other big change that is required to be legislated is in connection with what is known as Unified Licensing that would allow anyone to provide any service. A cable TV provider, for instance, could provide fixed line telephone services. This is what the TRAI in fact recommended in its USL paper, but the government has still to act upon it. The problem here, however, is that the TRAI kept the cost of a USL very high, at around Rs 1,600 crore for an all-India license, so even if the government accepts the principle, the cost will keep most people out.

Along with the long distance policy, Minister Maran, for instance, also announced that genuine internet telephony would be allowed in the country as opposed to the restricted PC-to-PC telephony that is currently permitted – so, you can call up your cousin in Mumbai on his phone using internet telephony using services such as those offered by companies such as Vonage and AT&T in the US – it is because internet telephony is becoming so hot that auction site eBay paid $2.6 bn for Skype and promised another $1.5 bn if it grew to certain levels by 2009. The problem, however, is that internet telephony has been allowed only for access providers – if a new company, say Sify, wants to provide such telephony, it will first have to pay Rs 1,600 crore to get an access license! Clearly the government needs to fix this. Since the cost of an access license has really been determined by the cost of a mobile phone license, what the government needs to do is to offer the license for free, but auction the spectrum. So, if someone wants to offer fixed lines through laying of cable through the ground, he can get the license for free. If, however, he wants spectrum for mobile telephony, he will just have to bid for it. It’s a simple solution, but no one in either the government or the TRAI appears to be doing much about it.

ends