But there was more to come. The next setback to the regulatory framework was a result of the way the TRAI went about its task of fixing tariffs. In its first consultation paper on Telecom Tariffs issued in November 1997, the TRAI proposed that existing rentals on basic or fixed line services should be increased by 63 to 140 per cent for different categories of subscribers, the number of free call units allowed during a bimonthly billing cycle should be reduced from 250 in rural areas and 150 in urban areas to 120 and that a uniform charge of Rs. 1,30 per extra call should be insttuted in lieu of the existing sliding scale in which call rates increase from Rs. 0.60 to Rs. 0.80, Rs. 1.00, Rs. 1.25 and Rs. 1.40 as the number of calls made increase. In addition, it proposed a reduction in long distance call charges by upto 60 per cent.
 
It should be obvious that the intent of this exercise was virtually to do away with two kinds of cross-subsidisation considered acceptable in telecom pricing strategies: first, the subsidisation of poorer, lower end users who are to be attracted into the network by lower rentals and lower call charges for less-intensive use; and, second, cross-subsidisation of local traffic with revenues from long distance traffic. While it is true that as a network matures, an effort must be made to reduce the extent of cross-subsidisation, it could not be claimed on the basis of telephone and call densities that the network in India had reached that level of maturity. Nor was the extent of reduction of cross subsidisation even in countries like France as much had been proposed by the TRAI.
 
It is not surprising, therefore, that the proposal which would have adversely affected low-end users and benefited well-to-do business subscribers was seen as a way of increasing the revenues of new private basic services operators investing in local networks through an inequitous rationalisation of the tariff structure. Despite calls for a moderation in reduction of cross-subsidisation from Parliamentarians, the TRAI chose early in 1999 to issue a note on the new tariffs and leaked to the press the fact that it had issued such a notification. This again resulted in an unnecessary brush with the Communications Minister and Parliament, which was finally resolved with a structure which reduced cross-subsidisation to a lesser degree but still provided some "relief" to new private operators and high-end users.
 
The TRAI's actions cannot even be defended on the grounds that its tariff setting principles are scientific. This comes through for example from an anlysis of its tariff setting procedures for cellular operators. The basic principle adopted by the TRAI is that rentals should cover capital costs, while call charges should cover operating costs. Our earlier discssion made clear that this does not tally with any economic reasoning on how a utility like telecom access should be priced given its characteristics
 
But what is even more disturbing is the fact that when calculating costs, there was no effort to make a normative assessment of what such costs should be. Consider the case of capital expenditures on cellular lines to be recovered through the rental. Capital expenditures consists of depreciation computed assuming a 10 year life span of equipment and weighted average interest costs placed as high as 20 per cent. Capital cost per subscriber is computed after taking utilisation of equipment into account. To make the calculation the TRAI seems to have unquestioningly relied on estimates provided to it by the cellular operators themselves. The absurdity of this comes through when we look at how estimates of capital expenditure per mobile line varied between circles and metros and even across operators in metros (Table 5).
Table 5 >>
 
To deal with variations between metros and circles the TRAI decided to base their recommendations on the metro calculations. But how is the variation within metros to be dealt with ? The appropriate procedure would have been to make a normative estimate of costs, which would have brought the estimate closer to the lowest figure in the set, since the nature of equipment used is more or less the same. The use of mormative costing procedures is routine in the case of organisations like the Bureau of Industrial Costs and Prices. Rather than opt for that procedure, what the TRAI did was to ignore the abnoramally high maximal value and then take the median value of the rest of the figures to arrive at the capital cost with which rentals are to be computed when some apropriate level of utilisation as been reached. This procedure is clearly indefensible and would have inflated capital costs as well as the derived rentals. It was on this basis that the TRAI came to the abnormally high rental figure of Rs. 400 to Rs. 484 per month, as compared to the Rs. 156 which prevailed under the original tariff structure.
 
The above instance is just quoted as an example to illustrate how there is nothing scientific about what the TRAI has done, making the debate between it, the government and users a battle between those who want "economic" costing and those who want subsidies, as it is often presented to be. The problem with the regulation and tariff setting framework which had the TRAI at the centre was that there was no internal way to monitor the "monitor". This is what explains the repeated recourse to the courts and the periodic clash between the TRAI on the one hand and the executive, the Parliament and the consumer on the other.
 
The TRAI, of course, held that its procedures were correct and that without hiking tariffs the process of introducing competition in the telecom sector would be aborted. If that is accepted, then the message which remains is that from the point of view of society and the individual consumer the price of competition is too high and therefore liberalisation is not warranted.