Thursday, February 3, 2011

Spectrum Auctions: 'Jhatka' or 'Halal'?

The choice is between sudden death and a slow one

Shyam Ponappa / February 3, 2011

Why do people advocate spectrum and licence auctions? Is it because they think auctions work? Is it the appeal of an ideology, like capitalism or socialism? Or is it because governments often collect large sums, and auctions seem fair (in a market-driven sense) and transparent? Theorists apparently cannot find better ways to allocate spectrum or licences, despite the alternative of technical and financial short-listing followed by a lottery. Yet, while desiring high government collections, people really want reasonably-priced good infrastructure, and continue to rail against government waste. Let’s review some so-called “successful” auctions and what followed.

1994: The US spectrum auction

Prior to 1994, the US used to allocate spectrum on demonstrated capacity and merit (“beauty contests”). The spectrum auction in 1994 netted record bids. The Federal Communications Commission chairman reportedly said: “Auctions have proven once again to be a success not only by awarding licences to those that value them most, but also by decreasing the national debt.” Then disaster struck, with a number of “successful” bidders declaring bankruptcy. As BusinessWeek put it in 2010 with the benefit of hindsight, “... over time, beauty contests have delivered fewer problems and higher value to society than have airwave auctions.”1

1994: India telecom licences

In 1994, India auctioned telecom licences. Chaos followed owing to overbidding and default. Thereafter, the sector struggled from one contention to the next, with the government and operators deadlocked by 1998. The New Telecom Policy of 1999 provided a breakthrough, tossing aside the auction bids in favour of shared revenues. After the percentage share was reduced to reasonable levels, and “Calling Party Pays” halved tariffs in 2003, mobile services grew exponentially to over 725 million subscribers by 2010. Interestingly, the Telecom Regulatory Authority of India estimated that auction fee foregone till March 2007 was over Rs 19,000 crore, whereas actual revenue collections were double, at Rs 40,000 crore; by March 2010, the collections were 80,000 crore.

2000: The UK 3G auctions

The 3G auction in the UK was hailed as a spectacular success, reaping bids of about $35 billion.

2000: The France and Germany 3G auctions

Germany followed, netting $67 billion, and the finance minister quipped that the auction was for unexpected revenue to pay the national debt. France demanded a flat fee of $4.5 billion per licence.

The dotcom bubble burst in March 2000, followed by communications and technology companies a year later, and the bidders went into a tailspin. The collapse nearly bankrupted not only British Telecom owing to the enormous debt it incurred for the bids, but the entire industry worldwide. The economic slump that followed made it impossible for firms to pay off high debts, as their interest payments increased while their ratings fell.

A contrarian move in France is noteworthy for its prescience and insight. CEO Martin Bouygues (pronounced “Bweeg”) of the third mobile operator, Bouygues Telecom, refused the government’s demand of $4.5 billion as the fee for a 3G licence, making it the only mobile communications company in Europe with no investment in 3G. Mr Bouygues’ letter in May 2000 appeared on the front page of Le Monde, asking: “What should I tell my employees? … That we have a choice between a sudden death and a slow one?” While his opposition was ignored, by 2002, the French government dropped its asking price by more than 85 per cent to induce Bouygues to accept a 3G licence.

In terms of results, the auction “failures” – the Netherlands, Switzerland, Sweden, and “non-auction” countries like South Korea, Japan and Finland (until 2009) – have the best broadband services.2

Kapil Sibal’s appointment as India’s telecom minister has brought hope, with prospects of radical improvements in infrastructure, especially broadband, with a clean hand. Mr Sibal’s recent pronouncements on a new telecom policy, however, raise the spectre of another deadlock. Here are two examples: (a) “Adequate spectrum will be provided to all service providers.”

This is feasible not through slivers of spectrum for many operators, but only if there is a common carrier access, that is, all operators can access spectrum for a reasonable fee. There is no indication of what “adequate” means, nor of pooling or sharing spectrum.

Let’s hope the domain experts have been heard and not shouted down on “adequacy”. For instance, the Telecom Equipment Manufacturers’ Association had recommended that two blocks of 50 MHz each in the 698-806 MHz band be allocated to facilitate the development of wireless equipment and services. Large blocks of contiguous spectrum offer far more efficient capacity than many narrow bands. For local innovation, to get low costs, we have to think of adequacy in these terms, and not slivers of 4.4 MHz or 6.2 MHz.

(b) “Spectrum henceforth will be awarded only on a market-based mechanism.”

If the criterion for success is high bids and not delivered services, in effect, this means auctions, and the result is likely to be dismal. Those enamoured with auctions focus on the success of bids, ignoring the purpose of spectrum/licence allocation, which is service delivery resulting in consumer surplus (societal benefits).

If the operators choose to roll over and accept authoritarian decrees, a deadlock may develop again as it did in 1998 between the government and the public interest.

The government’s choices include:

  • a genuine effort at developing comprehensive and integrated policies for reasonably priced services, while carrying along stakeholders;
  • a cosmetic effort, letting stakeholders vent, and then issuing arbitrary decrees that leave a mess. For example, too many operators with fragmented spectrum; or
  • attempting a political or populist fix, seeking to make the United Progressive Alliance look good, the Opposition look bad, bleeding all operators to avoid accusations of a sell-out, and still leave a mess.

The first alternative is in the public interest; the second and third are not. The issues that need comprehensive transformation are spectrum and network sharing for service delivery at least cost. The government and Mr Sibal have the opportunity to choose an approach resulting in excellent delivery including broadband at reasonable prices.

1 'Spectrum Auctions Hurt Mobile Consumers', Bengt Nordstrom, Businessweek, August 17, 2010:

2 'Broadband quality ranking - by economic development', Pierre Verdi, October 27, 2010:

No comments: