Thursday, October 11, 2012

The Supreme Court Delivers

Now, the spectrum and licence issues need resolution

On September 27, 2012, the Supreme Court of India delivered the opinion of a bench comprising five Judges on the Presidential Reference regarding the auction of 2G spectrum.

Shyam Ponappa / Oct 04, 2012

The Supreme Court’s opinion on the Presidential reference* dismissed two preposterous claims. One is that it is beyond the ambit of Parliament and the government to formulate economic policies. The second is that the government must allocate resources only through auctions. It’s like the end of a self-destructive nightmare. True, our heedless kleptocracy as a society of rogue politicians, bureaucrats, defence personnel, and complicit citizens, led to this pass. Even so, the anarchic “destructionism” of these claims is as reprehensible as the kleptocracy they seek to tear down. Fortunately, the Supreme Court opinion rose above the populist clamour.

There’s still a mess to clear. The big picture is that the Supreme Court left its decision on spectrum auctions unaddressed. In matters of detail, some points need resolution based on facts. These are discussed below to dispel prevalent myths.

Myth 1: Auctions maximise govt revenues

“Auctions may be the best way of maximising revenue…”: paragraph 116 of the opinion. This contravenes the evidence after the National Telecom Policy -99 (NTP-99), that revenue-sharing maximises government revenues as well as public benefits. It also ignores the many auction failures.

Consider the evidence: auction revenues foregone were estimated at under Rs 20,000 crore for 1999-2007, because the sector was mired in losses and was unable to provide services effectively or pay those dues. By comparison, actual collections from revenue-sharing by March 2007 were more than double, at Rs 40,000 crore. Collections by March 2010 were Rs 80,000 crore. Current annual contributions to government revenues may be about Rs 18,000 crore on Adjusted Gross Revenues estimated at Rs 1,40,000 crore, plus taxes, amounting to perhaps Rs 36,000 crore.

Re public benefits, access to telephony grew from a few million users in 1999 to about 700 million today (excluding around 250 million shadow subscriptions).

An ameliorating caveat in paragraph 12 states: “…if the State arrives at the conclusion … that maximum revenue would be earned by auction of the natural resource in question, then that alone would be the process”, and this is expanded in paragraph 119:

“Where revenue maximisation is not the object of a policy of distribution, the question of auction would not arise. Revenue considerations may assume secondary consideration to developmental considerations.”

This has not prevented erroneous conclusions in the press that auctions are the only valid process, notwithstanding that the conditions stipulated in the order, eg, that government’s actions be “fair, reasonable, non-discriminatory”, were always operative, if not adhered to in instances of abuse, as in the 2G scam.

Myth 2: Maximum govt collections are in the public interest

Government collections as the public interest criterion may work for colonial powers extorting revenues from subject states, or possibly for utopias whose political economy is so balanced that such cross-subsidisation works. Developing economies like India presumably can and should seek the welfare of their people. The same populists crusading for maximum government collections accuse governments of corruption and waste. This doesn’t provide a coherent approach to infrastructure, where each capital-intensive sector is configured to deliver a specific service. For instance, the energy sector has to deliver power, while telecommunications must deliver communications services. Neither can be expected to deliver toilets or water. Yet, many well-intentioned people seem to nurture such irrational expectations.

The spectrum and broadband link

The first prerequisite for broadband is high-speed connectivity. The second is reasonably priced services. Our objectives are, therefore: (a) a broadband network, (b) available anywhere (c) at reasonable prices. Our networks are deficient, however, particularly in rural and semi-urban areas. A host of factors are responsible, ranging from limited public sector network rollout, combined with a private sector focus on the most lucrative urban centres, with incentives skewed to voice telephony. Applications need connectivity based on networks that require spectrum.

Problems and solutions

Consider an application like distance education. The need is for networks and services of high quality (followed by the additional requirement of content). What is apparent is that such applications cannot be effective without the connectivity. So we’re back to the need for networks, of fibre where feasible, and wireless elsewhere. This brings us back to the need for spectrum.

Reviewing facts

As regards the facts relating to the 2G judgment deserving review:
The solution the Supreme Court has not considered is that operators need only to use spectrum, for which they can be charged a fee. The evidence of widely available Wi-fi shows that innovation and usage thrive if spectrum is available. The Supreme Court, the government, and the public need to recognise that allocating spectrum to operators is only one way to use spectrum.

There need be no alienation of spectrum at all, if policies allow open access and charge fees. Then, spectrum could be used like any infrastructure network, eg, airports, highways, or rail, on payment of usage charges. The sharing could be in at least two ways. Operators could pool spectrum for collective use. For this, (i) regulations must allow pooling/active facilities sharing, and (ii) operators must agree on terms and procedures. Another way is for mandatory spectrum sharing using the database-driven systems being implemented in the US by Spectrum Bridge and Telcordia. Similar deployments are planned in the UK, the European Union, and in Singapore. The TV white space is shared because this range is available for sharing, and not because other bands cannot be shared.

There are immense societal costs of duplication in capital investments in multiple networks, including the last-mile spectrum access, of operators using dedicated networks with limited passive facilities sharing (such as towers), compared with the benefits of open-access to common networks, if policies changed. These would employ active facilities sharing (equipment, and not just construction) to reduce capital equipment, construction costs, energy for towers, carbon emissions from a more limited physical network, possibly reduced radiation from a rationalised network with small cells with lower-powered equipment, and the multiplier effect on the finite available spectrum.

Enormous productivity benefits could accrue through ICT applications in infrastructure such as smart grids for energy, transportation, education, healthcare, and government services, as well as many commercial applications.

The Supreme Court could also uphold contractual obligations, by discriminating against actual transgressors in the 2G spectrum allocation, while rehabilitating those who operated within the law.

shyam[no space]ponappa at gmail dot com