The point here, is the difference that pricing policies can make in relieving traffic problems in networks. In Mumbai, you could never shake off the feeling that the system was so slow because VSNL does not discriminate between users-needs (or effective demand) for networks.

In Ireland for example, each of the main ISPs have at least two kinds of services. Take for example ESAT, one of the more popular ISPs here. ESAT runs two kinds of services. One is pitched at unsophisticated home-users who are pretty well-served by the odd couple of hours of Net-surfing in off-peak (after 6 pm) hours. The other service is aimed at small businesses and the like which need to use the Net during regular business hours.

The casual user will very soon find that if he starts using the Internet during peak hours, he ends up paying ridiculous amounts of money for doing so. Business users find, as a result, that the network is not overloaded at peak business hours.

This is a simple application of the role of consumer surplus in buying-decisions. People who value their time more, would pay more for the use of the network than others. But this is also logic that is bound to become increasingly important, if electronic commerce is to have its day in the sun.

The whole idea of electronic shopping, at least from the point of view of the consumer, is attractive because it cuts out the inconvenience of having to go out, wait in a queue, and so on. But if you find that you may have to wait for two hours to get online, you might just prefer to pop down to the corner store and get the grocery yourself. What then, would you say about the future of e-commerce? Research on e-commerce strategy has already come up with a variety of sophisticated ways to price network capacity, and reduce such network crowding. Needless to say, most of it feels that strategies such as those followed by VSNL so far, are not the way to encourage the growth of electronic commerce.

At the risk of simplifying excessively, network resource allocation strategies can be divided broadly into two categories. VSNL follows what is known as a quantity limit strategy, which basically limits the number of hours an user may have access to the network, regardless of the time the network is used. This strategy has certain advantages such as low information requirements for pricing. But these advantages are outweighed by the congestion this strategy causes. The other approach is to use differential pricing to allow users to select the quantity and quality of the service they require.

But this is not like pricing a telephone call, for example. In traditional telecom services, the carrier only needs to provide a single quality of service, to transfer calls. In electronic commerce, on the other hand, users have different needs which require different levels of service quality. For example, Telnet users might require high-speed low-volume data transfer needs, which might be quite different from the needs of online audio and video applications. If both these services were provided at the same rates, as VSNL's quantity limiting strategy might do, there would soon be an inefficient use of available network capacity.

The new pricing strategies would instead try to price services according to the consumer surplus each kind of consumer enjoyed. It might broadly try to divide the tariff for using the network into two parts. One part of the tariff would be aimed at recovering fixed connection charges, so this would really be a kind of a fixed access charge, levied on everyone who uses the network. The second part of the tariff would try to assess the externalities or the congestion cost that the user causes to the network. For example, a user who uses the network for an audio application at 11.30 in the morning, causes far more congestion than a user who uses the network to e-mail his colleague at the weekend. Accordingly the first user will be charged a much higher congestion charge than the second user.

Firms such as ESAT are not yet thinking of such advanced strategies. But they have taken the important step of realising that prices are the most efficient way we have of rationing a scarce resource. Arguably, in India, network capacity is an even more scare resource than it is in Ireland. VSNL might have to increase prices on certain kinds of users as a result. But in the long-term, this is bound to prove more economical.

For those of you interested in understanding more of the theory behind new pricing strategies, you might try reading the article by Stahl, Gupta and Whinston on the economics of network management.