FROM BUSINESS WORLD OF 5 OCTOBER 2006
Repeating
spectral follies
Telecom
Regulatory Authority of India (TRAI) has recommended a combination of reserve
pricing and royalty for the allocation of 3G spectrum. This is the beginning of
the process. The Department of Telecommunications now has to make up its mind
on what it wants to recommend. Its recommendation will go to the Wireless Preparations
Committee (WPC). The finance ministry will have its own view to put to it.
Since powerful ministries are likely to have strong views on the matter, it is
possible that it will end up with the cabinet, and that the Prime Minister’s
Office will get involved. Altogether, this is the kind of issue that our
government is rather bad at resolving and rather good at resolving badly. So
perhaps it does not matter so much that TRAI has made a bad beginning.
Still, its
proposal is not the best, and for reasons that it should know well. From the
telecommunications licence auctions in the 1990s all over the world, one lesson
emerged with glaring clarity: lump-sum charges are bad. A licence is normally
given for many years at a time. Its value depends on how the market develops
over the licence period – how fast it grows, and how intensely firms compete.
If demand booms and the suppliers are few or collude, they may rake in money,
and the value of the licence may turn out to be high. If, on the other hand,
the market bombs and competition pares down margins to the minimum, suppliers
may make little and the licence may be worthless. No one – neither the
government nor bidders – can know at the outset what will happen. Hence
bidding a fixed price for a licence is a
pure throw of dice; the chances are equal that bidders will strike gold and
that they will turn turtle. To confirm this proposition, we have the experience
of the early bidders of the 1990s in India, whom DoT’s arbitrarily high
interconnection charges made prematurely bankrupt, and whom the politicians
baled out at a price. There are even more spectacular examples the failures of
famous companies abroad.
There is another
reason why fixed sums are bad. All licences are never issued together; they are
given out over a number of years. Over the years, the market develops. Its
current attractiveness gets established, and its performance gives shape to
expectations about the future. Both influence the price bidders are prepared to
pay for the licence. So it typically happens that the government charges or
receives very different amounts for the same licence issued at different times.
And these differences in prices are inherently unjust: some bidders get
licences at bargain prices, whilst others are punished for no fault of their
own.
In better
governed countries, firms that go bankrupt just lie down and die. Their
financiers write off their losses and go on with their lives. But in our
country, those that dabble in high-value bids are big industrialists, often
financed by government banks and institutions. They do not forgive and forget.
They lobby and supplicate, bribe and hector until the government bales them
out. Hence a faulty process of licensing is bad for the quality of government.
And finally, the
government’s flip-flops establish its reputation as an untrustworthy player.
Those international investors who have the choice of governments avoid
countries with unreliable governments, or charge a premium for doing business
in their countries. In this way, the supply of foreign capital is reduced, and
its price is increased. A country like India, which would like to attract more
foreign investment, should worry about the reputation of its government – and
reputations are made of actions such as telecommunications licensing.
For these
reasons, the government should abjure reserve pricing, and go entirely by
royalty. Royalty varies with revenue, and is therefore not so bad as a fixed
price. By charging royalty, the government shares partly in the licensee’s
risk. But the risk is not eliminated; the uncertainty about the future market
remains. How is this uncertainty to be dealt with?
The government
should aim to issue licences to more bidders than are likely to survive in the
long run. They should be issued by an auction of royalty rates, but not for the
entire licence period of 30 years. Bidders should quote royalty only for the
first ten years. After that, fresh auctions should be held for five-year
periods at a time. Those licensees who do poorly will not be able to afford to
rebid and will sell out; those who do well may like to cash out at a profit and
go on to do something else. In this way, competition will be enlivened and
fresh vigour imparted to an industry that can only be oligopolistic. Its growth
will be maximized; and in the long run, growth is good, not only for the
country, but also for government revenue.
