governance, political economy, institutional development and economic regulation

Posts tagged ‘Ravi Shanker Prasad’

Telecom Service Providers: neutered by the “Indian” Spring

neutered

Photo credit: http://www.yourpurebredpuppy.com)

Netizens- citizens who participate in public debate from the comfort of their broadband enabled “home” space- are a furiously engaged and vocal lot in India. 800,000 of them voiced their support for the nebulous concept of “net neutrality” in an outpouring of outraged emotion, similar to the 2011 Arab Spring which put paid to Hosni Mubarak in Egypt.

Unlike the Arab Spring, this was the revolt of the empowered- internet penetration in India is barely 20%; broadband access is just 8% and 80% of net access is by the privileged owners of smart phones. This boils down to a privileged population of around 9 million netizens in a country of 1230 million citizens.

Like the Arab Spring, the netizen revolt was against a “perceived” threat to their empowerment-in this case cheap access to the internet- which ironically is also one of PM Modi’s promises to the entire Nation.

Keen to dampen the “revolt of the privileged” and anxious to avoid a political fallout, Ravi Shankar Prasad, the Union Telecommunication Minister, hastily set up an expert committee this week to advise him on the subject and publicly re-committed to make the net accessible for all. Sundry netas (politicians) waded in, some equating “net neutrality” with democracy, others raising it to the level of a fundamental right.

The occasion for this frenzy was an exceedingly well articulated and informative Consultation Paper issued by the telecom regulator-TRAI, calling for comments by April 24, on the need, or otherwise, to regulate Over The Top (OTT) services- communication service providers like Facebook or Google; content providers like You Tube or Netflicks; apps providers like Apple or Cloud computing and storage; e-commerce sites like eBay, Amazon and Flipkart. All these provide unregulated, unlicensed voice, audio or text services; they do not need to buy spectrum; they free ride over the “telecom pipes” built by telecom providers and incur none of the obligations (user data privacy, transparency in data management, data security) loaded by the government onto licensed Indian Telecomm providers- Airtel, Vodaphone, Idea, Reliance Comm, TATA Teleservices and Reliance JIO Infocomm.

The cause celebre inciting the wrath of netizens was an impending deal between Airtel-India’s largest and most internationalized mobile services provider and Flipkart-a Singapore-India e-retailer, which went from being a start-up in 2008 to a valuation of US$ 12 billion in end 2014. The deal in question was a “zero-rating” agreement under which customers would get free net access to the Flipkart services with the latter compensating Airtel directly. It is unclear if exclusive hosting of Flipkart on the Airtel network was built into the deal.

Rahul Khullar, the upright and technically brilliant, ex-babu (bureaucrat), economist who chairs the telecom regulator-TRAI, was as astonished as the Minister of telecom by the virulence and strength of the customer reaction. He went public yesterday and blamed the “noise” around the otherwise rather bland, technical concept of “net neutrality (NN)” as an outcome of “a corporate war between a media house and a service provider”.

Here are some facts drawn from the TRAI consultation paper followed by opinions.

 

Perfect NN is a mirage

First, as the TRAI paper acknowledges, perfect net neutrality (NN) is not practiced anywhere nor is it technically efficient. Management discretion with respect to the manner in which data packets move through the telecom pipes must be left with the service provider so that outcomes (service quality and cost) are optimized.

 

Inefficient “Free riders” impose costs

Second, just as democracy, by definition, requires the limitation of certain types of individual freedom in the interest of the common good, “free riding” by any player imposes an unfair cost on someone else, which has to be guarded against. This is best done by allowing the market forces of competition and prices to prevail with the government looking closely only at a single bottom line- sustainable benefit for the retail customer.

Innovation, yes but “free riding” no!

Third, for every start up like Flipkart which goes from zero to hero in five years, including by minimizing their upfront cost through “free riding”, there are Dumbo individual users like me who pay Rs 650 every month for a 3GB data plan but end up using only 1 GB. The spectrum space paid for, but left unused by us, is used by the Flipkarts to leverage their business.

Nothing wrong with that. If you are dumb you should lose out. But here is the knockout punch. Costless “free riding” works for all only till there is spare capacity in the “spectrum pipes”. The contention of the telecom providers is that the growth of data intensive OTT products (movies on YouTube and Netflicks; real time Cloud Storage; voice and audio services like Skype and Google Chat) have shrunk the space available in the “pipes” and reduced the quality of service for customers who use conventional voice and data services provided by telecom companies.  Ergo new investment is needed to upgrade the “pipes” and more spectrum is needed to accommodate the growth of data intensive 3G and 4G services.

 

The looming transmission capacity gap

Fourth, demand is exploding. Evolved services like VOIP, video services, digital services-including telemedicine, cloud computing and storage are all data intensive and requires data transmission at high speed (data streaming) with no breaks in between, unlike emails and text, which can take intermittent transmission. Text and standard content can be transmitted using a speed less than 1 Mega Bits per second (MBps). Movies in comparison require speeds of up to 10 MBps. Higher the speed more the requirement of spectrum since there are no “data packet gaps”, which can be packed in with say text, to optimize capacity usage – much like airlines pack spare seats with last minute, flexible time, low priced flyers.  TRAI reports that 36% of mobile usage in India is already for movies or video. The global average is 70%.  Growth to global levels is limited only by our quality degrading, low transmission capacity.

The government has no time bound, publicly announced plan for releasing spectrum currently hoarded by the army and by the public sector telecom companies-public gifts to them from before spectrum allocation was liberalized and its commercial value became common knowledge.

A spectrum availability gap consequently looms, as TRAI has pointed out to government. This means that TSPs have no option but to extract higher revenue efficiency from the existing spectrum for upgrading infrastructure.

 

Who should pay for OTT services?

Fifth, it is a standard rule of utility pricing that those on whose account an incremental cost is incurred must be made to pay the higher cost. Prior to the onset of internet based voice, video and instant messaging (chat) the normal share of “voice telephony” in TSP revenues was 70% and the residual 30% was data- sms, emails and other net usage.

Today, globally, the volume of SKYPE (the sms, voice and video over internet firm) traffic is 40% of conventional telephony and its nominal growth is larger than the growth of conventional telephony. Is this the end of the road then for conventional telephone companies? Not quite. In India and similar developing countries, conventional telephony has a reprieve. Until the transmission pipes can be strengthened to deliver good quality OTT services the demand for them will remain constrained. It is not uncommon, even today, for customers to resort to the fixed telephone network when mobile calls have a problem getting connected or get “dropped”.

Clearly, potential customers of OTT services, like movies, cloud computing and music should be ready to pay for the improved net services whilst those who are content with low data intensity conventional services, like text, should be insulated from the price hike. This requires progressive, price discrimination based on data use. Netizens point out that they already subscribe to TSP data plans for high speed access at prices determined by the market.

But TSPs complain that fierce competition constrains them from hiking data access charges for high speed data access. But the generous valuations put by incumbent TSPs in trying to hang onto spectrum allocations in the February 2015 auctions weakens their case of financial stress. The TRAI paper, unfortunately, does not share a marginal cost based assessment of the revenue requirement of TSPs to improve quality and enlarge access.

 

Willingness to pay

Customers should theoretically be willing to pay a higher price for accessing OTT services but India is a very price sensitive market. Over the last two decades, liberalization, fierce private sector competition and a “Nelson’s eye” regulatory approach to the quality of service, resulted in declining telecom rates and increasing usage.  This has fueled a “feel good factor”. No one wants to bust the party.

Whilst conventional TSPs have a vested interest in improving infrastructure quality their business model is under threat from technology and innovation. They have lost business in the lucrative sms market to “free” messaging services like WhatsApp (in the US it charges an annual fee of US$ 1) even though sms charges are nominal at Rs 1 per message. CARE-an Indian rating agency, estimates that the nominal revenues from sms reduced by around Rs 4000 crores (US$ 0.6 billion) in 2013.  TSP voice telephony is similarly under attack from voice OTT services. Call charges at Rs 0.5 per minute are the lowest in the world but still uncompetitive with the data cost of a VOIP call of only Rs 4 per minute. This could mean the capacity of TSPs to bear additional “pain” is limited, especially after the generous amounts bid by them in the February 2015 spectrum auctions.

That leaves only the stand-alone communication, content and app providers like Google, Facebook, Amazon – the free riders- who are unwilling to share this cost. However, their Indian clones, like Flipkart may well be willing to submit to licensing and revenue sharing just to grab market share from the entrenched global players-at least in India.

Three options, for taking us out of this conundrum, present themselves:

 

Back to the future: Minimum data access tariff

First: The nub of the problem is how to extract some of the enormous value being generated in unlicensed OTT services who free ride. TRAI declared regulatory forbearance on telecom tariffs with the onset of competition in services. The results have proved their wisdom. But TSPs complain a combination of artificially created spectrum scarcity have forced spectrum pricing up. Technological innovation has reduced margins on conventional telephony services and revenue flows instead to the OTT providers. The result is shrinking bottom lines for TSPs and cut backs on network upgrades.

One option could be for TRAI to prescribe a minimum tariff for high speed data access, based on the marginal cost of service studies, to ensure sufficient head room for network upgrade and access related investment. This minimum tariff could even be “discovered” by a reverse auction amongst TSPs-along the lines of the coal auctions for user power plants. The alternative is for TRAI do determine the minimum access tariff in the usual manner. This would force user charges for high speed data access up and better compensate TSPs.

The downside is that this option wrecks the ongoing party of the last decade with declining tariff fueling customer well-being. This may not be the best way to ring in Ache Din (PM Modi’s Good Days Are Here assurance)

Bring OTTs into the red tape tangle

Second: The government could license all OTT service providers and divert the additional funds to licensed TSPs to meet their network upgrade costs.  This would provide some financial relief to the TSPs and avoid a direct increase in customer telecom tariffs. But the License Fee would have to be significant to make a difference.

The biggest risk is that OTT providers, who by definition are “innovators”, dislike getting caught in red tape. Many of these are global players and shy of entangling with local laws. They could simply walk away. This may not necessarily be a bad thing since in China it resulted in the development of local clones. But India is presently not China. More significantly high License fees will be a bigger barrier for small local “innovators” who lack the deep pockets of foreign OTT providers.

To avoid the considerable barrier of a high upfront License Fee it could be recovered periodically linked to actual network use by an OTT provider. But this arrangement would need to be tracked by DOT or TRAI generating more red tape for OTT to tangle with.

Most important, PM Modi is in the process of defining an “International India”. Reducing transaction cost is an important part of the initiative. Putting OTTs into the muck of red tape does not align with all the assurances of enhancing the ease of “Doing Business”.

 

OTTs as bulk proxy customers of TSPs

Third, government could work on a dual strategy of massaging both the supply and the demand side of the market. On the supply side, government could announce a time bound plan to release more spectrum to ensure that the scarcity premium on spectrum becomes less burdensome for TSPs.

On the demand side, it could exercise regulatory forbearance from second-guessing the amount of “fat” there is to reallocate between customers, TSPs and OTTs. Instead it could let business arrangements prevail by encouraging mutually beneficial deals between TSPs and OTT providers, whist keeping an eye out to insulate retail customers.

In fact the Airtel- Flipkart deal was a step in this direction. Flipkart proposed to become a proxy, bulk customer for Airtel and pay it directly for the usage of all those who access its services. Small customers of Flipkart would have gained through free access.

To guard against arbitrary blocking or slowing down of access to other OTT providers, who do not choose to have such arrangements or in favour of proprietary OTT products of the TSP themselves, TRAI would need to announce a plan for exercising better oversight over data packet management to limit arbitrary and unreasonable discrimination by TSPs.

This arrangement continues the “party mood” for small consumers. It regularizes an innovative way out of the conundrum, already conceived between Flipkart and Airtel and puts pressure on global OTT players to play ball. India is the fastest growing market for the consumption of OTT products. This provides some leverage to us to lay down reasonable and light handed rules without going to the extreme step of exclusion of foreign OTTs adopted by China.  Democracy is indivisible.

 

 

 

Selling spectrum: Telecom’s Kohinoor (Reposted by the author from http://swarajyamag.com/)

spectrum

(Photo credit: http://www.intelligent-eneregy.com)

Now that the Spectrum Big Bazaar auctions are over, it’s time to sift through the smoke and dust of the financial battle and figure out who lost and who won. But before the juicy part, here are the bare facts for those who don’t spend their lives following spectrum auctions.

Limited spectrum on offer

First, the bulk of the spectrum sold in the 800, 900 and 1800 MHz frequency bands relates to licenses issued in 1995 and 1996 which are now expiring after 20 years. Very little new spectrum has been added by either the government releasing spectrum, currently reserved for the army (TRAI opines at least three times more than required) or from the cache given free to BSNL and MTNL, two government-owned telecom companies which grossly underuse their spectrum allocations because they just can’t compete with private service providers despite the public gift of free spectrum. So much for the “public Kohinoors” the government loves to hoard in public interest.

Fiscal success

Second, the government secured firm bids worth Rs 110,000 crore (US$ 17.8 billion) as revenue from auction—25 or 33% paid upfront and the rest to be paid in 10 annual instalments after a moratorium of three years.

Spectrum in the 900 band garnered an incredible 120% more than the reserved price. TRAI had predicted and service providers moan, that they have paid through their nose just to hang onto spectrum they had till now.

The 1800 band secured an auction value only 14% above the reserved price. TRAI had predicted that the small amount and non-contiguous spectrum on offer would negatively impact valuation despite the 1800 band being a fallback option for supporting service in the preferred and scarce 900 band. The 800 band secured a respectable 64% above the reserve price value.

The big fight for 900 band spectrum

Third, the 900 band was the star of the auction. Despite comprising only 46% of the spectrum auctioned, it secured the lion’s share of revenue—71% of the total auction value. The 1800 band, comprising 27% of the spectrum auctioned, did the worst with a revenue share of only 9% of the auction value. The 800 band, used mainly for data transfer, balanced its share in spectrum auctioned of 27% with a share of 21% of revenue in auction value.

The institutional short circuit

Fourth, let’s look at the institutional arrangements within which the auctions happened. Telecom Regulatory Authority of India (TRAI) has an overwhelmingly “advisory” role. The real decision making power on spectrum pricing; quantum of spectrum to be sold and the constraints to be imposed on the buyers of spectrum—right to swap, share or leverage spectrum financially, all vest with the Minister in the Department of Telecommunications (DOT).

In this sense, TRAI is very similar to the low profile Central Electricity Authority which is an adjunct of the Ministry of Power and advises it on techno-economic matters. Note that the similarity extends even to the nature of the entity. Both are “authorities” not “commissions” like latter-day utility regulators, including the electricity regulators, are titled.

This institutional arrangement has advantages. It reduces the risk for TRAI in offering innovative advice. After all, should the CAG or the CBI come calling with their inevitable investigations, usually once the government has changed, TRAI can turn around and coolly assert they decided nothing, being mere advisors to DOT. Also, this arrangement liberates TRAI from considerations of political economy- all of which weigh-in for government.

The arrangement also encourages TRAI to be transparent and participatory in its functioning which is a boon for ”Telecom watchers” who are not industry “insiders”.

TRAI’s consultation papers are drafted painstakingly. They provide a wealth of information and outline options for decision making. Most importantly, TRAI encourages open participation. None of this happens in the DOT.

Once an issue reaches the “in box” in DOT, the iron curtain crashes down; information put out is terse; formatted to confuse the “outsider” and as minimalist as Swedish furniture. The cumbersome “Right To Information” route becomes the only way to get access. But even then, the rationale for the many commercial decisions the DOT takes, is never disclosed. TRAI itself has, courageously, voiced its frustration on this “zipped lips” policy of DOT.

Undeniably, DOT could be much more.

transparent and “un-babu like”. Equally, TRAI has to learn to play ball. Adopting a strictly “technically correct” approach is not what an “advisor” is paid to do. In fact, by “islanding” itself, TRAI has cut its significant technical expertise off, from the deliberations in government around the “second best” option, which is the kernel of public decision making.  The correspondence between TRAI and the DOT around its recommendations for the 2015-16 spectrum auctions bears out this institutional short circuit.

The bottom line is that, whilst the existing institutional arrangement could work better, what exists is far better than deciding everything in the cozy confines of the DOT. TRAI Chairs and its members have been outstandingly progressive in safeguarding public interest.

Fiscal success but a tunnel vision?

Fifth, is the question how successful has this particular auction been? Undeniably, from the purely fiscal point of view, no one—hopefully, least of all the CAG—can quibble. But the issues TRAI raised, and which remain unanswered, are pertinent.

Should “public silver” be sold only for short-term gains or should there be a wider strategic objective? Telecom adds 3% to the GDP, so a 30% growth in telecom adds a significant 1% point to national growth and could meet one-eighth of the target of 8% increase in GDP. The bulk of the incremental, “good-jobs” are in the information technology space and depend on robust growth. Identifying needs and monitoring the delivery of social services are strongly dependent on a pan-India telecom network being available. Consider also that in a vast, poor country like India, where 25% of the citizens are domestic migrants who remain culturally, strongly linked to their home village or town and it becomes easy to view telecom access as a necessity for improving the quality of life.

Clearly, no one wants a situation where funds from spectrum sale in the hands of the government, come at a huge cost hurting the growth of private telecom services or adversely affecting their affordability.

The problem is that both government and TRAI are prone to fall back on the “cost-of-service”—heavy handed model of regulation, which entails examining the pockets of the private service providers, either to ferret out if they are making too much profit (DOT) or too little profit (TRAI) and somehow adjust the profit by either squeezing service providers (DOT) or conversely compensating (TRAI) through spectrum pricing.

Undealt with legacy issues compromise the future

The Spectrum Usage Charge (SUC) is one such additional levy, over and above the spectrum cost. It is levied as a proportion of the Adjusted Gross Revenue of a service provider. Actually, there should be no usage charge at all, since spectrum is now auctioned and bidders reflect a part of their expected surplus in the bid. The user charge is a revenue generating instrument, left over from the days when spectrum was “allocated” administratively. TRAI has repeatedly suggested that it be brought down transparently from 5% to 3% before an auction, to induce bidders to reflect the reduced cost into a higher bid. But DOT has resisted this move without ascribing a reason for its obduracy.

More importantly, now that the auction is over, reducing the user charge retrospectively is not an option. Any reduction in license cost will now go directly into the pockets of the service providers. Worse, those bidders who may have “inside” knowledge of a probable post-auction reduction, would have been enabled to bid lower on the basis of this “privileged information”. A classic case attracting the allegation of potential “crony capitalism”.

The current Minister of Telecom is too savvy a person to fall into this trap. But political office is transient with too little institutional memory. In comparison, service providers bid their time better than a leopard and the institutional memory of an elephant. It is an unequal match which can only go against the public interest.

Why go to Nagpur via Kanpur?

But even on the issue of technical efficiency to achieve a narrow fiscal objective, TRAI had raised “red flags” which merit more serious consideration. TRAI had warned of a price war if additional spectrum was not made available for new comers in 900 or 1800 bands. If newcomers have to fight it out with incumbents, clearly, bidders will scrape their revenue models to the bone. This is what has happened.

In the 900 MHz band, the bids are more than 2.2 times the reserve price. Was there really that much undiscovered fat in the revenue models of the service providers? Clearly, if there were such huge margins then the auction has achieved what markets are supposed to do– cut the producer surplus to optimum levels and pass the benefits to customers.

But there could also be a fine print to this bidding madness. DOT has still to take decisions on the property rights of the new spectrum owners. We know that they can share spectrum in the same band (900 and 1800 MHz being considered as one band) with other service providers, but on what terms? When and how will they be able to trade it with spectrum in other bands within and outside their License Areas? When does government intend to implement it’s “in-principle approval” to create a post-auction market for spectrum trading? If not, why not?
All these are “untapped value pools” for enhancing producer surplus post the auctions. Are we sure that this discretion will be used well?

More importantly, recent history in natural resource management in India shows that the use of administrative discretion immediately raises “red flags” and can invite an adverse comment from the CAG or CBI investigations on the charge of “crony capitalism”.

The spectrum auction is for 20 long years. Does this mean that if the Minister or his officers err on the side of complete probity and safety, it effectively seals our economic future at a sub-optimal equilibrium level till the next auction two decades hence? And only because they failed to take these crucial decisions upfront prior to the February auction?

Of course, the Indian bureaucracy has a delightful habit of springing welcome surprises. It is entirely likely that they will find a way out of this conundrum. But TRAI’s observation remains relevant—why go to Nagpur via Kanpur?

Who won and who lost

And finally the juicy part—who won? The Finance Minister won big time. He got around Rs 9,000 crore in FY 2015 to plug part of our fiscal deficit, courtesy the generosity of the winning bidders, who paid up part of the upfront payment before time to meet the March 31st  fiscal deadline.

The DOT lost the opportunity to burnish their image and came out instead looking like bumbling, inefficient, indecisive babus.

TRAI covered itself in short-term glory by being outspoken—almost pugilistic and technically sound.

The handful of six private service providers, being canny business people, frankly don’t care how the government wants to play the game, so long as they win out in the end—which some of them are likely to.

Customers are on a roll with the fizz of tailored talk and data plans and an ascending escalator of bundled services to care too much one way or the other.

And the nation—well, any outcome which helps reach the tight fiscal deficit target helps India get good risk ratings; cheaper credit and most importantly international recognition that PM Modi means business. All’s well that ends well.

Tag Cloud

%d bloggers like this: