The supposed tension between market pricing of spectrum and the need to keep the price of services low is emerging as a pivotal point in the formulation of the National Telecom Policy (NTP) 2011. One view is that auctioning spectrum will increase the cost of operators, thereby the price paid by consumers and the pace of diffusion. There is, however, a different position, too, on the relation between spectrum costs, market price, and diffusion in the Indian context.
In the case of mobile services, the fixed cost is incurred in stages: the upfront cost of the minimum necessary block of start-up spectrum comes first, followed by the cost of physical infrastructure. It is, therefore, possible that if the cost of spectrum is higher than expected, operators may reduce the spending on physical infrastructure, especially for the less viable portions of the circle. If the market is growing and highly competitive, the spending on physical infrastructure in the viable portions is likely to remain the same.
There will not be much difference in prices in the two scenarios. This is because after the network is rolled out, the spectrum cost and the cost of network become sunk costs and will not affect the price which will be determined solely by the variable costs and revenues which, in turn, depend on the operating costs, degree of competition and demand in the market. Since the roll-out in the viable portion of the market is unlikely to be affected by a higher spectrum cost, the price in that market will not change. The price in the unviable portion of the market will mimic the price in the viable portion. The low impact of higher spectrum cost will be even more clearly observed if additional spectrum charges are imposed with retrospective effect, as in that case most of the fixed investments would have already been made.
If the upfront cost increases to the point that fewer operators decide to enter, it is possible that prices could rise on account of the higher market power of those who do enter. However, this possibility of higher prices is based on the higher market shares making a material difference in the level of competition. The economist William Baumol in 1982 introduced the notion of “contestability” as a more general conception of market competitiveness than a market share- based measure. Introducing a robust Mobile Virtual Network Operator policy, promoting tower and spectrum sharing are some steps that would make the telecom market more contestable. In 2008 the entry of the CDMA (code division multiple access) operators in the GSM (global system for mobile communications) space would probably have been sufficient to cause the steep fall in prices that was observed with the other entrants not causing much material difference in the degree of competition.
Distributing the available spectrum to a fewer number of operators would have the effect of increasing the average holding of spectrum per operator and reducing the requirement of access infrastructure, especially in “capacity-constrained” urban areas. Thus a suitable chosen higher cost of spectrum along with a larger allocation of spectrum to a fewer number of operators could leave total fixed costs unaffected, making the situation identical to the one with a lower spectrum cost and a higher number of operators, provided that the reduction in the number of operators does not affect the competitiveness of the industry, as it would not be given that we currently have 14-15 licensed operators per circle.
In today’s scenario, charging market prices and allowing trading will allow the efficient operators to emerge with the higher spectrum holding dampening the expansionary effect on fixed costs. Market prices need to be charged not merely for the allocation of contracted spectrum, but also for the incremental spectrum. Currently, the sole means of charging for incremental blocks are spectrum usage charges.
Spectrum usage charges are not a sunk cost, but a variable cost like a sales tax. Like all such taxes, they have the effect of increasing prices. Further, the spectrum usage charges in India increase at higher holdings of spectrum. This makes them more distortionary than the ordinary sales tax, even one which increases with increasing revenue, because they penalize increases in revenue that are accompanied by an increase in the usage of only one of the inputs, i.e., spectrum. Hence, the production decision is biased in the direction of increased use of physical infrastructure.
If spectrum is being charged for at market prices and the level of competition is high, spectral efficiency will automatically emerge from market forces. Hence, it is best if the current spectrum usage charge is eliminated or assessed as a low, constant percent of revenue. Similarly, the licence fee should also be converted into a constant, low levy indexed to administrative costs. Note that the new regime will not reduce the government’s revenue, but change the form of the revenue from levies on turnover to one-time charges.
Finally, let us recall that promoting universal access to telecom services in rural areas is not hampered by spectrum availability, but by physical infrastructure. The USO (universal service obligation) fund is supposed to address the viability gap in this regard. Thinking about spectrum is hard enough. We could do without extraneous objectives.
Rohit Prasad is on the faculty of MDI Gurgaon.
Comments are welcome at firstname.lastname@example.org