Sunday, March 4, 2012

The 2G Supreme Court Judgment - 2


Open access is the future
Flawed technical assumptions in the Supreme Court's 2G judgment, and possible ways forward for the government
Shyam Ponappa / Mar 04, 2012


The first part of this article (‘Time for a review’, BS, March 1: http://organizing-india.blogspot.in/2012/03/2g-supreme-court-judgment-1.html) dealt with erroneous assumptions, especially regarding auctions. This part covers misplaced assumptions about technology, and explores constructive alternatives going forward.

Errors in technical assumptions 
An assumption underlying the prescription of auctions is that spectrum must be assigned to operators for their exclusive use. This was how wireless evolved during the first half of the 20th century, when radio frequency interference was the predominant problem in wireless communications.



With developments in technology, some advocate open spectrum predicated on the use of “cognitive radio” or “software-defined radio”, by which user equipment avoids interference by sensing unused channels automatically. In this model, open-access spectrum is a commons.


Another approach is to use a database-driven open-access model, whereby devices register with a database, and are dynamically assigned spectrum as needed. If this were possible in 1959, when Ronald Coase first recommended auctions, it would not have been necessary to parcel out spectrum. Even in America’s developed economy, the first auction was in 1994, and it failed.1 Now, technological developments enable spectrum sharing and dynamic assignment. America’s FCC has appointed 10 database administrators for dynamic spectrum allocation, with Spectrum Bridge being the first — in operation from January 2012.
America restricts this approach to unused spectrum in the TV bands, and a portion of the 700 MHz band, called “TV white spaces” (TVWS). The UK’s Ofcom is taking similar steps, with implementation planned for 2013. While all licensed frequencies could be pooled, sharing is restricted to TVWS because of conventions and legacies, and operators’ and governments’ preference for auctions. This judgment rules out sharing, blocking other technologies if the spectrum were available.


The lure of auctions 
For markets like India, there is every reason from a technology perspective to share not only TVWS and 700 MHz, but all commercially licensed spectrum. There is a technological basis for pooled spectrum, without exclusive assignment and auctions. Yet people love auctions: liberals, because business must pay its way, and governments get revenues; conservatives, because market mechanisms substitute for government controls.2 Operators prefer exclusive assignment to the uncertainties of open access and compensation for their holdings. Governments want auction revenues. So neither governments, nor big operators, nor the uninformed public, see incentives for pursuing what is in the public interest: shared spectrum.


For Technology leaders in OECD markets, shared spectrum was not a priority, because more spectrum was available to fewer operators. For instance, in 2010, operators in many US cities had 55-90 MHz according to gigaom.com, and AT&T was using only about half its available spectrum, whereas in Delhi and Mumbai, operators had only 10 MHz.

First-come-first-served 
Can the FCFS policy be abrogated on the basis of unconstitutionality? If so, the induced turmoil and far-reaching changes in procedures required for everything from tickets for railways or airlines, state-owned assets such as land, mining concessions, even government housing (including for judges?!), and all previous licences granted by FCFS procedures, defy imagination. This urgently needs review by the Supreme Court in the public interest.

Irregularities, outcomes, contracts and cancellation 
The same 11 companies whose licences were cancelled qualified according to the FCFS principle, except that their sequence was changed, apparently through procedural irregularities. In other words, without malfeasance, the same companies would have got the licences, except for S Tel getting Delhi and someone else not. Malfeasance deserves penalisation. However, as changes resulting from irregularities are limited in the sense that the same candidates would have won, must all licences be cancelled? Is there a judicial option of annulling the award, and placing the issue before the executive for equitable resolution in the public interest? After all, it is against the public interest to induce turmoil in markets and development capabilities, which the present ruling is likely to do not only in telecom, but in other sectors like energy, mining, manufacturing and transportation. Also, if foreign companies acquired legitimate stakes in licence holders, can these contracts be nullified without proof of their malfeasance? Or could erring parties be penalised, while legitimate parties are enabled to reconstitute their position as required by law?

The way forward 
Unfortunately, it is for our discredited and dispirited government to pick itself up and dig us out of this hole. Focused, goal-oriented action on the following lines would help.

First, review petitions: A first step is structured review petitions to the Supreme Court seeking relief, without grandstanding, bluster, or abdication of responsibility.

Second, an alternative to spectrum auctions exists in open access with payment. Both public revenues as well as public usage can be well served by treating access to spectrum as an open-access right-of-way. India’s policy makers need to consider the US and the UK’s shared spectrum approach. Spectrum can be paid for as it is used, as are oil pipelines, roads, or airports and ports.3
Open access could create tremendous opportunities in India, including for other technologies, e.g., a revival of WiMAX, if Intel grasps the nettle.

Third, on the cancelled licences. This has different problem sets. 

One set comprises parties who abused the system, punishable under due process of law. 

If there are parties in a second set that did no wrong, they should suffer no penalty.

What of a subset of the first, in which a foreign partner invested legitimately and built out, provided they were within the law? If these investors acted in good faith, perhaps a legal recourse could be to place their cases before the government for resolution and rehabilitation in the public interest conforming with the laws, if need be by a dispensation from the court, or even by fresh legislation. After all, good faith investors have contractual rights. Possible solutions might be (a) to penalise the guilty partner, while absolving the innocent, or (b) cancelling the licences of the guilty, while allowing the innocent to reconstitute as required by the law.
Above all, there is need for problem-solving that is systematic, transparent and participative, with expert inputs in domains and processes, to place the sector on a firm footing.

shyam ponappa at gmail dot com


Thursday, March 1, 2012

The 2G Supreme Court Judgment - 1



Time for a review

The judgment cancelling 2G licences was based on demonstrably incorrect assumptions about auctions

Shyam Ponappa / Mar 01, 2012
This first of two articles starts out with identifying the false premises of the judgment, particularly relating to the consequences of auctions.  The next article, to be published on Sunday, March 4, 2012, addresses erroneous technological assumptions, and explores possible ways forward.

The Supreme Court judgment of February 2, 2012, cancelling 122 2G licences needs a detailed review. This is because it is based on faulty premises relating to economics, finance and technology. If the Supreme Court entertains review petitions on this judgment, it is imperative that the judges be aware of these false premises, and that they be correctly informed regarding these issues. This article gives a few instances of such errors and explores the logic of auctions.

First, as an example of an error, the judgment states, “Spectrum has been internationally accepted as a … renewable natural resource which is susceptible to degradation in case of inefficient utilisation.”

Time for a review
The fact is that spectrum is not renewable, nor is it degraded. Spectrum is completely unaffected by use, unlike the degradation of land or water through use. However, use of a particular range of frequencies in a given space and time can block another user’s effective access to the same spectrum in that space and time — hence the need for considering efficient societal use.

Second, the judgment states that “the Government of India has already taken a decision to ... allot the same [spectrum] by auction”, quoting Telecom Minister Kapil Sibal. The fact is that the government had not announced such a policy decision before the judgment.

Third, the judgment prescribes auctions as being in the public interest. Are they?

The assumption that auctions are in the public interest warrants a detailed review. Amidst a cacophony of confused opinion based on little knowledge and less understanding, here is the evidence:

a) Maximum public revenues: auctions or revenue share?
Revenue from TelecomAssume for a moment that public revenues are indeed the appropriate measure in the public interest. What does the evidence show? An estimate from the Telecom Regulatory Authority of India (TRAI) in 2005, of auction fees foregone after the transition to revenue-sharing, was Rs 19,314 crore from March 1999 to March 2007. In fact, actual revenue-share collections by March 2007 amounted to double that number, or Rs 40,000 crore. Further, the amount collected by March 2010 was Rs 80,000 crore.













Sources: Auctions - TRAI, 2005:

Revenue Share: CAG, 2010: 
http://www.performance.gov.in/sites/default/files/departments/telecom/CAG Report 2009-10.pdf

These data demonstrate that over seven and 10 years, revenue-share collections far exceeded auction fees foregone. Over the entire life-cycle (20 years or more with extensions?), the revenue-share collections will overwhelm even the Comptroller and Auditor General’s (CAG’s) imaginary lost revenues.

b) Public interest: revenues, or access and usage?

What is really in the public interest — revenue collections or the benefits of usage? The CAG report and the clamour for auctions assume that revenue collections reflect the public interest. However, the draft National Telecom Policy 2011 (NTP-2011) states as its first objective: “Provide high quality, affordable and secure telecommunication services to all citizens.” It states that revenue generation will be secondary.

In other words, the policy objective is to provide the benefits of accessible, affordable services to users, not to maximise revenues collected. This was the first time the government unequivocally stated an objective that appeared emphatically in the public interest. The Supreme Court has thus far seen it differently, although this has nothing to do with upholding the law.

The confusion is made worse because the preponderance of literature is by “auction experts” focusing on high fees — and not at all on the services that should have followed but didn’t, because the capital went into the auctions instead of building service capability. A notable exception is a more balanced study of spectrum auctions worldwide that considers social gains as well as fees — which estimates social gains at an overwhelming 240:1 (“What really matters in spectrum allocation design”, Thomas W Hazlett and Roberto E Munoz, April 9, 2010: http://ideas.repec.org/p/reg/wpaper/372.html).

c) Are auctions in the public interest?

There was one successful auction in India in 2001 – because the market was dead – for a fourth mobile operator per circle. Other auctions in India and abroad resulted in the failure of network rollout and services, but were hailed as successes because of high auction fees. For cases of “operation successful, but patient dead”, read on.

Auction failures
  • US, 1994: The first US auction netted huge bids. Soon after, a number of “successful” bidders declared bankruptcy. This was repeated in the 1995-1996 “C”-Block auctions.
     
  • India, 1994: This auction in 1994 was followed by chaos from overbidding and default. The sector recovered only after many years, when the bids were set aside in favour of revenue-sharing with NTP-99. It took almost a decade before a reduction in revenue share (lower fees) and tariffs (calling party pays) led to explosive growth in mobile telephony from mid-2003.
     
  • UK, 2000/European Union, 2001 (3G): Considered a spectacular success, netting about $35 billion in the UK, followed by high bids in Austria, Germany and Italy that netted over $100 billion, these auctions raised about ten times the amount expected. The markets collapsed thereafter, and the bidders couldn’t service the debts incurred. Companies have taken a decade to recover, moving cautiously even now on 4G.
     
  • India, 2010 (3G and broadband wireless access): Hailed as a success, with over Rs 1,00,000 crore bid, lacklustre performance has followed, as companies struggle with the “winner’s curse” of paying too much to corner spectrum.
Auction experts have written disparagingly of “failures” (low fees) in countries like the Netherlands, Switzerland, Sweden, and non-auction countries like South Korea, Japan and Finland (until 2009). However, these disparaged countries have the best broadband services, according to a 2010 study by Saïd Business School at Oxford (http://www.sbs.ox.ac.uk/newsandevents/releases/PublishingImages/3 - Broadband quality ranking - by economic development.jpg). That is not surprising, considering that the capital was invested in service delivery, instead of in vying for spectrum.


The next article, which will appear on Sunday, March 4, 2012, will cover erroneous technological assumptions, together with policy issues and possible approaches to a constructive resolution of the gutted prospects of our telecommunications and broadband services sector.


                                                              shyamponappa at gmail dot com