1. Regulatory authorities have considered a number of criteria for telecom tariffs, incorporating several objectives mentioned in Section II above. Prices are based either on costs or on demand, and could be determined at a fixed level or be allowed to be flexible (within or without a specified range). Often the cost concepts considered for cost-based pricing include a rate of return (or mark-up) together with costs, i.e., they focus on economic costs, which include a normal commercial return. This paper segregates the topics on costs and mark-up to clearly identify two different elements of the cost-based pricing, namely the cost-concept and mark-up, as well as to compare the different types of methodologies for determining a mark-up.


III.1 Prices Based on Costs


(a) Some basic concepts

2. Relevant cost-concepts include: total costs, which are equal to the sum of all fixed costs and variable costs; fixed costs are costs which do not vary with changes in output levels (or in the case of telecom, with changes in the number of calls or the volume of traffic on the network); variable costs are costs which change (or vary) with a change in the level of output.


3. These costs can be further categorized as: direct costs, i.e. costs associated specifically and only with a particular service and can be assigned directly to the production of that service; when the accounting procedures do not record all the direct cost arising due to the service under consideration, then this total has to calculated by considering all the costs which can be attributed to it directly, i.e. the directly attributable costs; indirect costs are costs whose relationship with the production of a service is indirect, e.g. maintenance cost; overhead costs are normally the fixed costs for running the operating facility; common costs are costs shared among various activities; and joint costs are costs which arise in a situation where production of more than one service takes place simultaneously and jointly.


4. Chart 5 provides a simplified picture of the different types of costs for certain telecom services, such as local calls, long-distance calls, and international calls. When the output level of a services changes, variable costs specific to that service are incurred. Similarly, direct fixed costs specific to any particular service can be identified. The other costs are either joint costs or common costs, which are costs shared between two or more services.


  1. The accounting methodology used for calculating costs can be,



  1. Some important considerations

6. As mentioned earlier, there is an increasing emphasis on basing telecom prices on costs, particularly due to a focus on enhancing efficiency. The determination of a cost-based price is, however, not a straightforward exercise. For example, there is no universal agreement on which concept of cost --- average variable cost, average total cost, long-run incremental cost, short run marginal cost, fully distributed cost, actual book cost --should be applied in specifying cost-based prices. Thus, an important question is: which concept of cost to use for setting prices?


7. Further, even if a particular cost standard is selected, other problems have to be dealt with. The choice of cost components to be considered for determining the price is another contentious issue. For example, it is not clear how the common costs should be allocated, and whether a variety of associated costs to supervise, maintain, market and account for production should be included. Furthermore, cost-based pricing methods require monitoring in order to check any tendency of the operators to report higher costs than those actually incurred. Both the estimation and monitoring of costs is a data- and skill-intensive exercise.


8. A noteworthy point is that operational costs are higher during the peak load period. Therefore, not only demand but costs also provide a basis for charging higher prices for the peak period and

lower prices for the non-peak period.


(c) Concepts of costs used for determining prices

9. Three concepts of costs that have received much attention in the context of cost-based pricing are short-run marginal costs, long-run incremental (or marginal) costs, and fully-allocated costs.


(i) Short-run marginal costs


10. Marginal costs are those costs which arise due to an increase in the output level: alternatively, these costs are the increment to total costs when an additional unit of the output is produced. These costs correspond to costs commonly referred to as "variable costs", and in Chart 5 above would include only the costs under Box A. In contrast to fixed costs, which remain unchanged after production capacity is installed, marginal costs show the resource costs required if production has to be carried out once capacity is in place.


11. A price equal to marginal costs is considered efficient because at that price the value given to the product is the same as the resource cost that would be required for increasing the output level. There are, however, certain shortcomings of a system of telecom tariffs based on short-run marginal costs. With fixed costs being a large portion of the total costs of telecom (see Annex 1), a price equal to marginal costs will not cover total costs. This is even more likely if the technology exhibits declining costs, such as for telecom. Moreover, there are measurement problems and conceptual difficulties in calculating marginal costs. A basic information needed for ascertaining marginal costs is the link between a change in the output level and in total costs.


12. Firms, however, do not observe the cost changes attributable to marginal changes in output, and thus such data is normally not available. A number of methods have been used to calculate the marginal costs of telecom operations (i.e. econometric studies, engineering planning models, optimization models, and engineering-economics models), which require considerable data but do not necessarily provide reliable estimates of short run marginal costs.


13. The task of estimating short-run marginal costs becomes even more difficult when the output comprises multiple services (or a composite basket of services). Also, the quality of information with the regulator is invariably not as good as is available to the operating entity, and thus it is difficult to ascertain the veracity of the cost data provided by telecom operators.



  1. Long-run (marginal or) incremental costs


More comprehensive information can be considered for long run marginal costs, also referred to as long run incremental costs. The long-run is defined as the time period within which additional production capacity can be installed/used.


15. Conceptually, the increase in capacity does not involve fixed costs before the capacity is installed. Thus, long run incremental (or marginal) costs cover not just the normal variable costs but also the potential fixed costs. Before the installation of the investment capacity, these fixed costs are in a sense similar to variable costs because their level is yet to be fixed and thus could be varied depending upon any change in the plans about investment. Thus, in Chart 5 above, in addition to Box A, these costs would include Box B and could even include certain parts from Boxes C and D. A number of cost categories under general overheads, R&D, general technical support, and management costs are however not added in the long run incremental costs.


16. Long run incremental costs (LRIC) are based on forward-looking costs, and would thus incorporate the effects of economies of scale and technical change. Normally, estimation of long run incremental costs relies on historical data for the past year or two and on the prospective costs arising in the future as a result of increasing the capacity of operation. Recent emphasis is to focus mainly on the forward-looking costs to determine long run incremental costs.


17. Though the inclusion of investment in the estimate of LRIC increases the coverage of total costs that is accounted for by LRIC, it raises another problem that is linked to the pattern of investment expenditures. Investment streams normally involve non-regular and large amounts of expenditure in certain years. This does not make for uniform (or similar) annual estimates of marginal or incremental expenses in the various years considered. Since the stream of annual incremental costs is uneven, using these costs as a basis for price would imply large changes in prices in different years. It has been suggested that average incremental costs be used as a proxy for the annual long run marginal (or incremental) costs in order to smoothen the fluctuations in these cost-based prices.


18. An important requirement for determining long run incremental costs is to specify the level of output for which the costs are to be assessed, the technology that would be used, as well as the time horizon for which the cost scenario has to be considered. Different versions of long run incremental costs have focused on different levels of the output to be produced and the various components of costs to be included in the cost calculations (for example, on deciding which portions of Boxes C and D should be incorporated in the cost base).


19. Different estimates of long-run incremental costs arise ranging from those calculated on the basis of a small increase in capacity to those calculated by considering a significantly large capacity enhancement. The latter estimates are higher because they cover a larger number of activities and higher output. An example of long run incremental costs which have a wide coverage is the total service long run incremental costs (TSLRIC). The TSLRIC considers an increase in the total service, and would correspond to an addition to capacity as if a completely new service facility were to be established. Alternatively, TSLRIC is the cost the firm would avoid in the long run if it stopped providing a particular service.


20. In several instances, the definition of TSLRIC is restricted to a consideration of only directly attributable costs. That definition would overlap with another concept of costs, namely, the directly embedded costs. These costs are the directly attributable costs embedded (or mentioned) in the accounts of the operator, and are thus not forward looking costs. TSLRIC, on the other hand, is based on forward looking costs.


21. To encourage efficient operation through the cost-based price, TSLRIC is estimated on the basis of the most efficient and generally available means of production. Under TSLRIC, the incremental costs include cost components that can be directly allocated to the service, as well as certain non-directly-allocable incremental costs of the shared facilities used to produce the services being considered. TSLRIC, however, does not cover all common costs and thus differs from the concept of stand-alone costs. This latter cost-concept provides an estimate of costs in a situation when the service being considered is the only service being produced (i.e., by definition, no common costs exist because there is no other service under consideration).


22. Considerable information is needed for calculating long run incremental costs. Nonetheless, long run incremental costs have increasingly figured as a basis for determining cost-based telecom tariffs because similar to marginal costs, these costs maintain a link with efficiency; they cover a larger portion of the costs in comparison to short-run marginal costs; and, they provide a forward looking scenario and thus incorporate dynamic elements into price formation. For India in particular, such a forward-looking perspective is likely to be highly relevant, in view of the large increase in telecom capacity and modernization that is expected in the coming years.


23. With long run incremental costs, however, the revenue earned might not cover total costs because there are a number of overheads, such as staff for billing and administration, which would normally not be included in theses costs. Prices equal to long-run incremental costs might therefore imply that the operator would not break-even. Another reason for an incremental cost-based price not breaking even is that this price is based on forward-looking costs and not on costs currently incurred. If forward-looking costs are less than current costs, then the incremental-cost-based price would not break-even. Despite this, prices based on forward-looking costs are preferred because they provide a link with the dynamic efficiencies arising in the near future.



24. The concern that a price based on marginal or long run incremental costs would not cover total costs has led to a consideration of fully-allocated costs.


(iii) Fully-allocated costs (or fully-distributed costs)


25. Fully-allocated costs methodology covers all cost components, i.e. in addition to a consideration of the cost components covered by the long-run incremental costs, this methodology requires a full allocation of common or joint costs to the individual services on the basis of some specified formula. In the various allocation methods used, the emphasis is on simplicity rather than theoretical correctness. Methods to allocate common or joint costs include, for example, allocating these costs proportional to the revenue earned by each service or to the ratio between individual output levels and the total output ("output method"); or proportional to the directly attributable costs for different services or based on the ratio between direct costs and shared costs for services ("input method"). These costs could also be allocated according to the Ramsey method, i.e. in a proportion inverse to the elasticity of demand; elasticity of demand is the proportionate change in demand for a product (or demand by a particular user group for a product) as a result of a one percent change in the price of the product (see also the discussion on Ramsey mark-up methodology in section III.2.a below).


26. The allocation methods are based on accounting principles rather than on economic ones. In addition, normally historical cost data is used to allocate costs, and such information does not bear much resemblance to the actual or prospective cost situation. Another important criticism of prices based on fully-allocated costs is that they tend to encourage inefficient operation and investment, even providing incentives for increasing costs. For example, a firm which receives a price for its product based on its total costs would tend to report to the regulator, or even actually incur, higher costs in order to be allowed higher prices. Though such (adverse) incentives are likely to be present in other cases of cost-based pricing also, they are much stronger under the fully-allocated cost methodology.


III.2 Mark-Up On Costs

27. Basically, three different criteria have been used to determine mark-up on telecom costs. One focuses on demand price and aims at minimizing the loss in efficiency that would arise if the price deviates from marginal costs. Another is to use some uniform mark-up or thumb-rule for mark-up, such as a reasonable commercial return. The third is to consider two-part pricing, which would involve either pricing initial units of use at a higher (or lower) price in comparison to subsequent units, or levying an access or rental charge while pricing the use of the service at its marginal cost.


(a) Maximizing efficiency


28. A widely acknowledged rule for determining a mark-up to minimize deviation from efficiency is the Ramsey rule. This rule states that the mark-up of price over marginal cost should be in an inverse proportion to the elasticity of demand for the product. For example, elasticity of demand of residential subscribers is likely to be higher than that of business subscribers, because the demand of the former is likely to be affected much more by a change in telecom prices. Under the Ramsey rule, this would imply that a lower mark-up should be charged to residential subscribers.


29. The Ramsey rule can be adapted to situations covering interdependent demands, consumption externalities and dynamic settings. In these adaptations, certain additional information, such as the market shares of competing firms, would be required to determine the details of the formula for an efficient mark-up on prices. Nonetheless, the mark-up would still basically be in an inverse proportion to the elasticity of demand for the service.


30. A mark-up which is inverse to the elasticity of demand could imply a higher mark-up for those who might greatly need the service and not have any other viable alternative. Such customers would therefore be willing to bear the cost of a higher price. If these customers are from the socially underprivileged sections of the nation, then the criteria of equity may be compromised by using a Ramsey mark-up.


31. Determination of the Ramsey mark-up requires information on costs as well as on demand for various services (or user-groups), including the responsiveness of demand to a change in its own price or to a change in the price of other services. Such information is extremely difficult to obtain or to ascertain with much accuracy.


32. Furthermore, the elasticity of telecom demand is affected by (i.e. is not independent of) the different government policies in place. Thus, a stable demand price would not be easy to calculate in a scenario with major ongoing changes in government policy, or with large changes in demand for any other reason. For this reason, rather than compute detailed demand responses to price, the Ramsey rule should be used only as a rough guide for mark-up, while keeping in mind the above-mentioned social implications of this mark-up methodology.


(b) Assuring a specific rate of return


33. Another possibility of determining a mark-up is to use a thumb-rule for the rate of return over capital or mark-up over cost, for example, a risk-adjusted "reasonable commercial return". A cost-based price which provides a specified rate of return can also be derived under a methodology termed "cost-axiomatic pricing". Under this methodology, prices are derived on the basis of certain postulates regarding the links between costs and prices, which basically aim to produce a consistent system of cost-based pricing. These postulates could include, for instance, that total revenue must include total costs plus a reasonable return; services with the same effect on costs should have the same price; a service that involves a positive cost must have a positive price; and that if a service is unbundled then the sum of the prices of the unbundled components must equal the price of the combined bundle of services.


34. The rate of return criteria for pricing has been in use for several years, most prominently in the United States. In the US, the rate of return method was based on fully-distributed costs methodology which, as mentioned above, generates perverse incentives for the telecom operator, i.e. it encourages inefficient operation and investment. Furthermore, ensuring an assured rate of return for any service increases the possibility for anti-competitive behaviour. The assured rate of return (or profit) provided to the regulated operator for any particular service could be used to cross-subsidize another product which the operator sells in a competitive market, and thus the latter product could be sold at a very low (even below-cost) price.


35. In addition, considerable amount of information is required to verify and implement the rate of return methodology, and often good quality information is not available to judge whether the regulated firm is functioning efficiently. Therefore, regulators are now increasingly relying on flexible methodologies such as price caps (see below).


(c) Two-part pricing


36. Two-part pricing involves levying an access or rental charge while pricing the use of the service at its marginal cost, or pricing different levels or units of the service at different rates (slab rates). An example of this is India’s current pattern of tariff for basic telecom, which includes a specified number of free calls and a positive tariff on all subsequent calls. The Ramsey rule for mark-up continues to be relevant for the two-part tariff. Other options for mark-up include a uniform mark-up over price, or a risk-adjusted reasonable commercial return on investment.



III.3 Subsidized prices (or a negative mark-up)

37. Subsidies in telecom are provided for various reasons. They reflect considerations such as externalities, equity, and fairness. The provision of universal service or subsidized services to specific targeted users are covered under this category. Externalities in telecom arise because this sector is an important part of the economic and social infrastructure of a nation. Also, telecom services usually involve more than one person benefiting from the use of the service, and this gives rise to externalities such as the "network externality". Network externality shows the benefit to other subscribers when a new subscriber is added to the network. This benefit arises on account of other subscribers being able to call, or being called by, the new subscriber.


38. Price subsidy in telecom could be to specific users, or to certain services (or parts of a service). Thus, rural or residential users could be subsidized, access charges or rentals for the basic telephone system might be subsidized, or local calls might be subsidized. Traditionally, specific services or users have been subsidized (or cross-subsidized) through relatively higher prices levied on other services or users. With the introduction of increasing competition in the telecom market, the higher profit segments would be under greater competitive pressure. The resultant decline in profits in these services will necessitate finding an alternative to the prevailing system of cross-subsidization. A search for such an alternative has resulted in prices increasingly becoming cost-oriented.


39. A necessary concomitant of the move towards cost-oriented telecom tariffs would be a clearer specification of costs, i.e. a greater transparency of costs and revenues, and therefore greater transparency of the subsidies being provided. With such information, the regulator would have a better basis to consider alternative policies to fund the subsidies.


III.4 Demand-Based Prices

40. Demand-based prices provide a possibility of cross-subsidizing services with a relatively low price. While one example of demand-based price , Ramsey price, involves a consideration of both demand and costs, in general demand-based prices are determined independently of costs. A demand-based price reflects willingness of the user to pay for a service, which in turn is an indication of the advantage or value that the user derives from the product. Based on these principles, this pricing methodology focuses not on efficiency in terms of cost, but on efficiency in terms of the value given by a customer to a product (which is shown by the demand price). With such a criteria, there would be a need to impute social values on the consumption of various services by different groups of customers.


41. The problems regarding estimation of demand, mentioned above in the context of the Ramsey mark-up, continue to affect this methodology also. It is worth noting that though the price under this methodology is derived by going directly to the pattern of demand, thus bypassing any consideration of costs, it does provide an indirect basis to broadly determine the nature of the mark-up for different services.


III.5 Flexibility

42. In certain countries, regulators have started giving telecom operators greater control to fix their prices by,


    1. providing them increased flexibility for determining prices of products that are subject to regulatory price control or,
    2. not including certain product categories within the purview of price control.



43. These developments also reflect a view that the pricing system used for telephony need not be suitable for certain new value-added services in the market, for example, the internet. Flexibility of prices would help develop suitable and innovative price structures for these services.


(a) Price caps


44. Price cap is a very flexible pricing methodology, and can be adjusted to take account of a variety of objectives related to telecom tariffs. An important reason for adopting price caps has been to facilitate the process of tariff restructuring, i.e. change the prevailing pattern of cross-subsidization among services: most OECD countries use some form of price cap for regulating telecom tariffs.


45. Though the price cap is relatively simple to administer and does not require detailed knowledge of an operator's annual costs, the large number of telecom services in the market makes it difficult for regulators to intervene specifically for each individual service. Ascertaining information separately for these services is a very complex task. Under the price cap method, therefore, an upper limit on the increase in average prices (not individual prices) is specified for a selected basket of telecom services. The weights used to determine the average price for the basket are normally based on the revenue shares of different services in the basket.


46. Not all telecom services are included in the basket whose overall price is regulated through a cap. For example, no price caps are specified for services which are viewed by the regulator as being provided in a competitive market.


47. The price cap methodology allows for a consideration of additional conditions or detail. Hence, if special attention is to be focused on limiting the range of prices for certain services within the basket, the overall price cap could be supplemented by individual caps (or service band indexes) for those services. For example, in the United Kingdom, the regulated basic telephone operator must ensure that bills do not increase in real terms for each decile of residential users. Furthermore, additional conditions could be imposed along with the price cap to achieve other objectives, such as ensuring that there is no undue discrimination between different sets of customers.


48. The basic idea in the price cap methodology is to consider two elements. One is an estimate of inflation of the costs of producing a specified basket of telecom services. The other is the likely increase in the operator’s productivity for that basket of services. The former is approximated by the rate of inflation in the economy (or RPI), and the latter is characterized by a factor X. The price cap formula normally specifies that for a particular basket of telecom services, the average price cannot increase beyond a factor RPI minus X; sometimes the cap is given in terms of RPI plus a specified rate.

49. Not only does the price cap method provide considerable flexibility for pricing, but by limiting the price increase it pressurizes operators to increase efficiency of operations. It also contributes to efficiency by passing on price decreases to consumers. The caps on sub-categories serve the objectives of equity, and also limit the possibility of cross-subsidization and hence unfair competition. A point worth noting is that in general the use of price cap schemes has been associated with high or increasing profits, instead of an expected decline in profits due to a limit imposed on price increase.


50. Price caps can be adjusted periodically, and are thus expected to adapt to a changing environment. Care must be exercised in this regard because if too much adjustment is made to the price cap, or if there are too many sub-baskets, the price cap scheme becomes complicated and restrictive, and its effects become less transparent. Also, price cap schemes can cause significant distortions if applied for a long period, particularly because the relevant price and productivity changes need not be reflected in the formula used. There is thus a need to periodically evaluate the various components that underpin the price cap mechanism and to revise the actual formula from time to time. In practice, the cap is established for a limited period of time, for example, three to five years.


51. The application of price caps has not been a static tool, and the content of the formula and the coverage of the basket subject to price caps have changed, as required. In practice, the assessment of likely productivity gains is difficult, and the factor X in the price cap formula has been the result of close negotiations. Thus, informed pro-active process of regulation has been necessary to make the cap work effectively for increasing the productivity level.


(b) Coverage by (or exclusion from) price regulation

52. Current trends in some countries show that higher the level of competition in a market, greater the likelihood of the regulator relying on the market for fixing prices, subject to safeguards regarding anti-competitive behaviour. While it is difficult to determine the extent of competition required for excluding a service from price regulation, adequate competition would normally require more than two similarly placed operators in the market. It would be useful to develop specific criteria for establishing thresholds that might trigger the withdrawal of a service from a price cap basket. Furthermore, the competition policy framework might need to be strengthened to ensure that collusion or predation in the market can be satisfactorily addressed.


(c) Price floor and ceiling to check abuse of market power

53. Unfair competition is a major concern in the telecom sector, which normally has a dominant incumbent operator in the market. To address unfair competition, it is necessary to devise some criteria for a price floor. One suggestion in this regard has been to use long run incremental costs as a floor price.


54. It is sometimes argued that low introductory prices need to be offered for a new service. In such a situation, special care would need to be given to competition policy-related issues.


55. Another concern regarding abuse of market power relates to a high price being charged to customers. Price ceilings based on fully-allocated costs or stand-alone costs have been considered for addressing this concern.


(d) Various options of price combinations

56. Another example of price flexibility is the option of different price-service combinations offered to customers. Of the different options provided, the customers choose those best suited to their own needs. Examples of these options include different combinations of access and usage price, or different combinations of telecom tariffs for various services. The provision of discounts for high volume of usage could also be considered under this category of pricing.


III.6 Conclusions

57. The discussion above shows that the methodologies used by regulators emphasize efficiency, flexibility, fair competition and social objectives. This raises several questions and involves making choices.


(a) Need to choose which services should be subject to price regulation


58. To begin with, a regulator needs to determine which services should be subject to price control and which should be left outside the purview of such control. Furthermore, the nature of the regulation or control might differ among different services which are subject to price regulation. For instance, certain services (such as essential services) could be subject to closer price scrutiny and control, including a specification of the price level. Others might be subject to price control, but only in terms of price floors and ceilings.


59. Thus, important questions in this regard include,


(i) What should be the basis for categorizing services into those subject to price regulation, and others? Should it be, for example, basic/non-basic telecom services, or voice/non-voice telecom services? If basic/non-basic services, then what should be the coverage of basic services for this purpose?


(ii) Should the services not subject to price regulation be those services which are considered not essential? Or, should non-essential services be categorized into those subject to weaker price regulation and others not subject to such regulation? If yes for the latter query, then on what criteria should such a categorization take place? What should be the definition of essential services in this context?


(iii) Should a specific price level be determined for certain services. If yes, which services should be subject to this type of price regulation?


(iv) Which services should be subject to floor and ceiling prices?


(v) Should the type of price regulation depend on the extent of competition in the market? If yes, then what is the link between the type of price regulation and the extent of competition?


(vi) What is adequate competition? Should one use a thumb-rule that three or more operators in the market result in adequate competition for reconsidering the pricing methodology to be applied to the service provided by these operators?



(b) Methodology for fixing a specific level of the telecom tariff

60. A specific price level can be fixed on the basis of costs or demand. Alternatively, a cost-based price could be determined together with a mark-up which might be based on demand or on some other criteria mentioned earlier. As brought out above, among the methodologies that can be used for fixing a specific price level, long-run incremental costs methodology is most suitable for achieving the objective of operational efficiency. In this context, the concept of total service long run incremental costs (TSLRIC) was highlighted, because these costs cover a large portion of total costs (TSLRIC is the cost the firm would avoid in the long run if it stopped providing a particular service).


61. Further, even for those services which might not be subject to any price regulation, it would be necessary to provide for a mechanism to deal with situations of unfair competition. This may require fixing certain price level as a floor.


  1. Thus the questions in this context include,


  1. Is long run incremental cost an appropriate concept to use for determining cost-based prices? If yes, would it be better to focus on a wider coverage of long run incremental costs, such as TSLRIC?
  3. Would demand-based pricing be a relevant basis to use instead? If yes, would this be so for all situations/services or only for some of them? What should be the criteria for making this decision?
  5. Does the fact that new entrants would be coming into the market and that there will be a substantial increase in telecom capacity in particular imply that TSLRIC is an appropriate concept for cost-based pricing?
  7. Should the floor and ceiling prices be linked to certain concepts of costs? If yes, which concepts (e.g. certain version of long-run incremental costs, stand-alone costs). If not, what should be the basis for choosing floor and ceiling prices?
  9. Should a price floor be defined for all services in order to address the issues of unfair competition? If not, for which types of services should such a price floor be provided? If such a price floor were to be specified, should a rebuttable presumption be that prices lower than the floor result in unfair competition.


(c) Mark-up On Tariff

63. The next step is to consider what mark-up should be used to make the telecom activity attractive for investors. Of the different types of mark-up considered above it appears that due to data problems, the Ramsey mark-up can at best provide a rough guide. A mark-up which assures a risk-adjusted reasonable return for the operator is easier to implement. It should also be noted that while the Ramsey rule links the mark-up with the value given to a product by the consumers, the uniform mark-up rule continues to link it to costs. With a uniform mark-up, the ratio of different prices continue to exhibit the resource cost (or opportunity cost) ratio.


64. With regard to costs, it would be worth considering the consistency conditions that were mentioned earlier in the context of cost-axiomatic pricing. These include, for example, that services with the same effect on costs should have the same price; if a service can be unbundled then the sum of the prices of the unbundled components must equal the price of the combined bundle of services; the price of an unbundled service should not be more than the price of a bundled service containing the unbundled service; and that any change in a cost-based price must be on the basis of recognizable changes in costs.


65. Another issue arising in the context of mark-up would be whether the mark-up should be based on the value (or use) of a service for the subscriber. This would imply using a mark-up derived on the basis of a demand-based price.


(d) Price caps

66. The above-mentioned pricing methodology for determining a specific price level would in effect be similar to a price based on fully allocated costs. Hence, it would be necessary to guard against the dis-incentives generated by such a system, namely the tendency of the operator to become inefficient.


67. Furthermore, since an increase in the local call rate could be a sensitive issue, another important consideration could be whether or not a sub-basket for price caps should be considered for local calls, and whether a stricter limit on price increase should be applied to that sub-basket


68. Therefore, it would be important to consider


  1. Is there a need to supplement the above-mentioned price methodologies with a price cap mechanism such as RPI minus X mentioned above?
  3. If a price cap mechanism were to be used, then which services (methodologies) should it apply to. For example, should it apply to both tariffs specified in terms of a specific level as well as the tariff specified in terms of floor and ceiling?
  5. Which telecom services should be subject to sub-price caps, and what should be emphasized in these sub-categories when the price cap is being decided for them?


(e) Deficit might remain even after applying a mark-up

69. Even after a mark-up, a deficit of cost over revenue might still remain. Such a deficit could arise for two reasons. One, due to the costs incurred for implementing social objectives (such as the universal service obligation, preferential access or call charges for specific user groups, or a specified number of free calls for each billing period). Another reason for the deficit could be that the concept of cost that is used for determining the telecom tariff would not cover total (or fully-allocated) costs, even if there were no costs incurred to meet the social objectives. As discussed in Annex 3, the second type of deficit could be covered by a higher mark-up on tariff or an increase in rental. To the extent that this cannot be done, the deficits arising due to both these reasons could be considered as deficit incurred on account of social objectives.


70. An important point to consider in the context of subsidies to telecom is that if teledensity is very low (such as in India), there is likely to be a considerable positive network-externality when additional lines are installed. This would suggest that a lower price should be charged compared to a situation where such externalities do not exist, e.g. a situation where universal coverage has been achieved. Such externalities provide one of the reasons for focusing on universal services provision.


71. Some important consideration in this regard are,


(i) Should actual costs or forward-looking costs be used to calculate the fully-allocated costs considered for deriving the second type of deficit mentioned above? The use of forward-looking costs is likely to maintain a link with efficiency and future developments. Is this an objective important enough to over-ride other considerations?


(ii) Should the mark-up be applied before the calculation of the deficit, or after determining the extent to which the second type of deficit mentioned above can be covered through a rise in price or rental?


(iii) Regarding the issue of lower telecom tariffs for certain users, there is a need to consider to whom subsidies should be provided, and how much. For example, should there be an upper limit on the increase in the local call rates?


(f) Methods to fund the deficit arising for social reasons

72. It is important to clearly specify the extent of the subsidies to be provided for social reasons, in order to provide a better basis for policy-formulation in this regard. Thus, there is a need to enhance the transparency of subsidies. Experience in some countries (such as Australia and Chile) suggests that the extent of these subsides is likely to be less than initially anticipated, even if different methods are used to calculate the subsidy levels. In Australia, a study was conducted to calculate the amount of such subsidies required. In Chile, the authorities asked private companies to submit tenders showing their assessment of how much subsidy they needed in order to operate in areas otherwise considered unprofitable.


73. There are a number of methods to fund the deficit. One is to create a fund through a levy or tax revenue, or to use the license fee for financing this fund. If a levy were imposed for financing this fund, then the price of certain telecom services might need to be increased to accommodate this "additional cost".


74. Another possibility of covering the deficit is through an increase in telecom tariff. This price increase could either be a flat rate or be an increase which incorporates a discount on the volume of calls.


75. Yet another possibility is to raise the monthly or bi-monthly rental for access to the network. An increase in rental to cover such a subsidy, however, could adversely affect the demand for telecom (particularly from relatively poor subscribers), and might be against the objective of rapidly increasing teledensity. Furthermore, a rental subsidy might be justified in view of the externalities of network expansion.


76. On the other hand, the Ramsey rule suggests that rental could be increased without much welfare loss by focusing on users whose demand elasticity is low with respect to the amount of rental charged. In that context, it is often mentioned that business users could be charged more than the residential users, because the elasticity of demand of the former for telecom is less than the latter’s elasticity of demand. Thus, the relevant questions are whether rental should be increased, and if the answer is positive, then by how much and for which user groups.


77. It might be difficult to obtain the (demand-based) information required to properly consider the question of whether or not rentals should increase. In particular, it might be difficult to adequately identify and separate business and residential subscribers. Another option in such a situation could be to provide flexibility of choice through price options offering different combinations of rental and usage price. For instance, a high rental combined with lower usage fee could be one price option, and a lower rental combined with higher usage fee could be another option. It would be necessary to consider whether such a flexible option should be preferred to other schemes involving an increase in rentals.


78. The importance of the interconnection price should also be borne in mind in this context. Both the interconnection price and telecom tariff need to be considered together for assessing the financial situation, and hence the subsidy required, for the telecom operators.


(g) Need for detailed account-keeping and unbundling of services

79. More detailed account-keeping would be required if it is decided that cost-based tariffs should be used. Even otherwise, detailed costs-accounting would be required for enhancing the transparency of subsidization. In addition, there would be a need to "unbundle" the provision (or pricing) of various telecom services so that the costs and revenues of different services could be considered separately. Detailed and unbundled information on costs is extremely important also for properly considering cases of unfair competition. Unbundling limits the possibility of unfair competition by preventing hidden cross-subsidization of a service which is "bundled" with another service.


80. In this regard, it is important to consider to what extent, and which, services should be unbundled, and whether consistency of pricing should be maintained between a service when it is sold in an unbundled form or together with another service.

back to objectives


return to india_gii's home page