|
|
|
|
 |
|
|
|
|
|
|
|
 |
 |
 |
|
||||
| Policy review of two-part charging - 1 September 2003 | |
||||||
 |
 |
 |
|
||||
 |
|
|
|
 |
|
||
|
|
|
|||||
|
A
statement issued by the Director General of Telecommunications Chapter 1 – The consultation Chapter 2 – Review of responses to the consultation document Chapter 3 – Conclusions Annex A – List of responding parties S.1 In January 2003, the Director General of Telecommunications (the 'Director') issued a consultation document, Policy review of two-part charging. This document presented the initial findings of Oftel’s policy review on whether a system of two-part charging ('2PC') should be introduced into the wholesale interconnection charging regime. S.2 The purpose of undertaking this policy review was to assess the likely benefits and costs of introducing a system of 2PC and to consider its potential impact on industry and the interconnection regime. S.3 There are two key elements that make up the interconnection cost of a call. These are the cost of setting up a call (call set-up) and the cost of maintaining the connection for the duration of the call (call duration). Under 2PC a separate charge would be made for each of these elements as compared to the current system where only a single charge is made. S.4 Under the current arrangements, BT’s interconnection charges are made on a pence per minute (ppm) basis. The costs of call set-up and call duration are averaged to create a single ppm charge. The effect of this averaging is that the charge for shorter than average calls is lower than it would be under a system of 2PC and the charge for longer than average calls is higher than under a system of 2PC. This effect evens out over all calls, implying that BT recovers its costs appropriately. However, where operators purchase interconnection for calls that are predominately longer or shorter than the average duration, BT’s current charges for interconnection could be either higher or lower than if charged for under a system of 2PC. A move to 2PC would introduce a separate charge for the set up of a call plus a ppm charge for call duration and could result in wholesale charges for each individual call that more efficiently align cost recovery to its cause. S.5 The Director took the position in the consultation document that 2PC should only be introduced if the benefits of introducing it significantly exceeded the direct costs of implementation, given the presence of further intangible and non-quantified costs. The factors that could result in such further costs are discussed in Chapter 3 of the consultation document and Chapter 3 of this statement under the heading Policy Issues. They include uncertainties in determining a two-part charge (ie the split between call set-up and call duration), the relationship between unmetered and metered Internet access, whether the benefits of implementing 2PC at the wholesale level would be passed on to consumers, and the regulatory resources in Oftel/Ofcom and among operators that would be required to implement 2PC at a time of already considerable regulatory change. S.6 Oftel’s analysis in the consultation document showed that, over a period of five years, there could be a welfare benefit in excess of £5 million from 2PC. The size of the benefit would depend on the assumptions in the analysis, in particular the ratio of set-up costs to duration costs. As a result, Oftel undertook a sensitivity analysis on this and other assumptions in order to gauge their effect on the quantification of the potential welfare benefits of introducing a system of 2PC. S.7 Based on the evidence available to it at the time, Oftel estimated that the cost to BT and other operators of implementing 2PC could be in excess of £12 million over a period of five years. Oftel also noted that the individual costs faced by operators would vary substantially depending, for example, on the current state of each operator’s billing system, its scale and complexity. S.8 Therefore, in the consultation document, the Director concluded that, on the evidence available to him, the benefits of introducing 2PC did not outweigh the costs. Indeed, according to Oftel’s sensitivity analysis, it appeared that the costs were likely to exceed the benefits even on an 'optimistic' calculation of the likely welfare benefits of introducing 2PC. S.9 In addition to asking for general comments, Oftel posed several specific questions in the consultation document to which responses were invited. These questions included asking in detail about operators’ likely costs of implementing 2PC and questions regarding whether the welfare benefits of 2PC were likely to be passed on to consumers at the retail level. S.10 The Director received a total of 13 responses to the consultation document. Of those to express a definitive view (all but two responses), only three respondees (NIACT (if the cost to industry was not prohibitive), Energis and Easynet) favoured the introduction of 2PC. Energis and Easynet called for the introduction of 2PC for Number Translation Service ('NTS') calls only. S.11 In its response to the consultation, BT estimated its potential costs of implementing 2PC at £5 million over a period of five years, compared to an estimate of £4.5 million used in the consultation document. Some other operators attempted to quantify their potential costs of implementing 2PC. These estimates varied considerably but, taken as a whole (including consideration of BT’s estimate of its costs) together with other comments regarding implementation costs, it is the Director’s conclusion that the figure used in the consultation document as an estimate of the cost to industry of implementing 2PC was a reasonable one in the context of the cost-benefit analysis. S.12 The Director has concluded that there is unlikely to be a significant net benefit associated with the implementation of 2PC. Indeed, the evidence available to him suggests that the implementation of 2PC would be more likely to result in a net cost. Therefore, it is the Director’s view that wholesale changes to the current interconnection charging arrangements to implement a system of 2PC are not appropriate. S.13 This statement discusses the responses to the consultation and sets out the Director’s conclusions. Chapter 1 briefly reviews the consultation document and the Director’s views at that time, Chapter 2 reviews the responses to the consultation and Chapter 3 sets out the Director’s conclusions following the completion of the consultation.
Chapter 1 – The consultation1.1 Oftel has discussed the issue of 2PC, at a general level, on a number of occasions:
1.2 However, each time the subject has been raised, very few operators have shown interest in furthering the discussions. As a result, BT has rejected requests for 2PC from individual operators on the grounds that a broad industry consensus needs to be gained before changing existing charging practice. 1.3 In January 2002, Energis referred BT’s refusal to apply 2PC to Number Translation Service ('NTS') calls to Oftel. Oftel’s consideration of the dispute led to the conclusion that NTS traffic could not be treated in isolation from other traffic types and that Oftel could not realistically make a direction to resolve the dispute in isolation. The policy review that followed resulted in the consultation document of 27 January 2003. This document assessed the likely costs and benefits of implementing 2PC. The consultation document also investigated the practicalities of implementation, including some of the less tangible costs and likely knock-on effects to other parts of the regulatory framework. Successful calls1.4 In the consultation document, Oftel’s position was that it did not favour adopting a system that would encourage charging retail customers for unsuccessful call attempts (even though they impose a cost on the network provider) because callers are not in a position to regulate their unsuccessful call attempts, even with full information on cost available to them (see note below). Therefore, it is unclear that charging for unsuccessful calls would promote economic efficiency and result in consumer benefit. On this basis, the consultation document only considered introducing 2PC for successful calls, with no separate charge for unsuccessful calls. It was proposed that costs incurred in unsuccessful calls should be recovered from the generality of successful calls. All call types1.5 Under the current single part charging system, the costs of call set-up and call duration are averaged to create a single ppm charge. This average charge leads to appropriate cost recovery at an aggregate level by BT. If BT is to continue to recover its overall costs, then any reduction in the charge for long duration calls resulting from the introduction of 2PC, must be accompanied by an increase in the charge for short-duration calls. As such, there is no objective reason why changes to the interconnection charging mechanism should be restricted to a subset of short-duration calls (eg higher charges on NTS voice calls which would offset lower charges on long duration NTS calls to the Internet), or to call origination, rather than to all calls including call termination. Benefits and costs1.7 The consultation document acknowledged that a move to 2PC could result in wholesale charges for each individual call that more efficiently align cost recovery to its cause, which would better reflect the economic principle of cost causation. 1.8 The consultation document used six principles of cost recovery suggested by Oftel and adopted by the Monopolies and Mergers Commission in its report on telephone number portability of November 1995 (see note below). These principles are (1) cost causation; (2) distribution of benefits; (3) effective competition; (4) cost minimisation; (5) practicality; and (6) reciprocity. 1.9 The consultation document considered that the key issues in respect of whether or not to introduce 2PC are associated with cost causation, practicality and the costs of implementation. It therefore went on to compare the likely costs and welfare benefits of 2PC and the practicalities of implementation. Calculating the welfare benefits of 2PC 1.10 In order to assess the possible welfare benefits of 2PC, Oftel analysed data on average call duration together with information on the processing costs of call set-up and call duration provided by BT. Oftel used this information to estimate the cost savings for different types of call:
1.11 This estimation was done using a systematic calculation based on the following five steps (see Chapter 2 of the consultation document for a full explanation):
1.12 These calculations showed that, over a period of five years, the implementation of 2PC could result in a cumulative welfare benefit in excess of £5 million when all call types (excluding FRIACO) are aggregated. Calculating the costs of 2PC 1.13 The consultation document noted that BT had commented that its implementation costs for 2PC are likely to be significant. It said that important changes to its wholesale billing systems would be needed as it would be required to collect and rate volume data on call set-up separately from call duration and ensure that such data was of billable quality. 1.14 The 'ballpark' estimate of its costs of implementation that BT provided to Oftel for the purposes of the consultation document was that these costs could be as high as £2.5 million for the first year and £500,000 per year thereafter, giving an undiscounted total of £4.5 million over five years. 1.15 In respect of the cost to other operators of implementing 2PC, the consultation document took the view that other operators would need to modify their billing systems in order to collect and rate data on successful call attempts in order to be able to validate BT’s bills. The costs for each operator could vary substantially depending on the current status of their billing systems and decisions such as whether to introduce 2PC at the retail level. 1.16 At the time of preparing the consultation document, Oftel had not attempted to collect information from operators as to their likely costs of implementation, but estimated the total cost to all other operators (apart from BT) as in excess of £8 million over a period of five years. Oftel asked operators to provide estimates of their costs (over a period of five years) of implementing 2PC as part of their responses to the consultation document. Conclusions on benefits and costs 1.17 The Director’s conclusion in the consultation document was that the implementation of 2PC could result in a net cost to industry of £7 million over a period of five years. 1.18 Oftel also performed sensitivity analysis on its calculations of the potential welfare benefits of 2PC on the following three parameters:
1.19 According to this sensitivity analysis, the largest increase in the figure for the welfare benefits of implementing 2PC came from increasing all set-up costs by 50 per cent (with a corresponding reduction in duration costs to maintain overall cost recovery). This led to an estimate for the welfare benefits over five years of implementing 2PC of approximately £8.3 million. The consultation document noted that, even on this most 'optimistic' calculation of the welfare benefits, the implementation of 2PC could still result in a net cost to industry of almost £4 million over five years. Policy issues1.20 The consultation document noted that the following policy issues required consideration:
The process of determining a two-part charge 1.21 The consultation document examined the division of costs between call set-up and call duration in other countries that have already adopted 2PC. Very wide variations in the ratio of the call set-up charge to the duration charge for single tandem call origination were found to exist. Oftel’s opinion was that such variation was unlikely to reflect real differences in network design between the different countries. Uncertainty surrounding the most appropriate way to split costs between call set-up and call duration leads to uncertainty in respect of the magnitude of benefits that would be obtained by implementing 2PC. The relationship between metered and unmetered Internet access 1.22 Oftel considered that FRIACO should not be included in any introduction of 2PC (see paragraph 1.6). It was important for Oftel to consider future trends in metered and unmetered (ie FRIACO) Internet access when assessing the possible welfare benefits of introducing 2PC. Oftel stated that it did not expect metered access to increase significantly in the future and Oftel’s calculation of the welfare benefits of introducing 2PC was based on current volumes for metered Internet access. If metered Internet access volumes were to decline, the future benefits of introducing 2PC would be less than estimated in the consultation document and vice versa. Therefore, the consultation document asked stakeholders to give their views on future trends for both metered and unmetered Internet access. Passing the benefits to consumers 1.23 Oftel noted that the economic benefits of 2PC depend on consumers facing prices that better reflect costs. Oftel considered that competitive pressures may drive cost savings on long duration calls through to the retail level and that the higher costs of short duration calls in a 2PC regime may also be reflected at the retail level. Oftel also noted that lower income (below £17,500 per annum) households were less likely than higher income (over £30,000 per annum) households to access the Internet at home. Therefore, it was likely that lower income households make a higher proportion of shorter than average duration calls and so the introduction of 2PC could introduce a mechanism for transferring welfare benefits from consumers with lower income levels to the comparatively better off. Stakeholders were asked to comment on whether they would introduce 2PC at the retail level if it were introduced at the wholesale level. Regulatory, commercial and operational resources 1.24 The consultation document noted that the implementation of 2PC would require significant regulatory resource to adjust charges and charge controls across most of the interconnection regime. Given the scale of changes introduced by the new EU Directives and the Communications Bill (as it then was), Oftel felt that weight must be given to the costs of the regulatory effort (both in Oftel/Ofcom and industry) needed to introduce 2PC. Stakeholders were asked to comment on their regulatory resources that would be required for the introduction of 2PC. Notes: The Monopolies and Mergers Commission used this argument in their report Cellnet and Vodafone: Reports on references under section 13 of the Telecommunications Act 1984 on the charges made by Cellnet and Vodafone for terminating calls from fixed-line networks dated 21 January 1999 (see paragraphs 2.458 to 2.472). The full text of this report can be viewed at www.competition-commission.org.uk/rep_pub/reports/1999/421cellnet.htm#full. MMC Report entitled Telephone number portability: A report on a reference under section 13 of the Telecommunications Act 1984 dated 14 December 1995. The full text of this report can be viewed at www.competition-commission.org.uk/rep_pub/reports/1995/374telephone.htm#full.
Chapter 2 – Review of responses to the consultation document2.1 The Director received a total of 13 responses to the consultation document. Annex A lists the names of all parties who have responded. Non-confidential versions of the responses can be viewed at www.oftel.gov.uk/publications/responses/2003/twopc0103/index.htm. 2.2 Of those to express a definitive view (all but two responses), only three respondees (NIACT (if the cost to industry was not prohibitive), Energis and Easynet) favoured the introduction of 2PC. Energis and Easynet called for the introduction of 2PC for Number Translation Service ('NTS') calls only. For the purposes of this statement, NTS is taken to refer to the regime behind calls to numbers identified in the National Telephone Numbering Plan as a Special Service number or a Premium Rate Service number as well as calls to 0500 Freephone numbers and excluding calls to 0844 04 numbers for Surftime Internet access services and calls to 0808 99 numbers for FRIACO. Successful calls2.3 All but one of the responses agreed with Oftel that charging retail customers for unsuccessful call attempts is undesirable. The reasons for Oftel's view are explained briefly in paragraph 1.4 above and also in Chapter 1 of the consultation document. One respondee, Mr. P.R. Welham, suggested that a charge for unsuccessful calls in certain contexts (eg telemarketing or TV initiated dial-in polling) should be made at a wholesale level and could, depending on the commercial considerations of the relevant operators, pass through to the retail level, for example to larger commercial users. His reasoning is that unsuccessful calls in these contexts are "customer-controllable". 2.4 Oftel notes that the respondee is not proposing that the consumers actually making the calls are charged for unsuccessful call attempts. Oftel does not favour charging end-users in the circumstances described above for the same reasons that it does not favour charging for unsuccessful calls in general: even if consumers know that a charge will be made for unsuccessful calls, it is difficult for them to modify their behaviour in the light of this price signal as they cannot control or know for certain whether the call they are about to make will be successful or not. It is therefore unclear whether charging for unsuccessful calls will promote economic efficiency and result in consumer benefit. The respondee is arguing that there is an exception to this general premise as commercial customers such as those organising television dial-in activities are in a position to gauge the proportion of calls that will be unsuccessful, and should therefore be liable for these costs. Oftel notes this point but considers it to be a separate issue from the general one of charging for unsuccessful calls at the wholesale level. All call types2.5 Easynet and Energis disagree with Oftel’s position in the consultation document that, if the case for the implementation of 2PC is made out, then it should be introduced at the wholesale level for all call types excluding FRIACO (see paragraphs 1.5 to 1.6 above and also Chapter 1 of the consultation document for a full explanation of Oftel’s view). 2.6 Both Energis and Easynet advocated the introduction of 2PC for NTS calls only. They argued that NTS calls have different characteristics to other call types which justifies applying a separate charging mechanism and that, in fact, NTS calls have historically been treated differently for interconnection charging purposes. 2.7 Easynet described how NTS calls were already treated differently from other call types through the use of the Network Charge Differential ("NCD") methodology introduced by Oftel in November 1999 (a sliding scale giving a percentage uplift above single tandem charging according to the number of points of connection an operator has with BT). Easynet compared the proposal to replace the NCD method with automatic, per call, billing based on BT’s Inter-Network Call Accounting ('INCA') system with the introduction of 2PC and estimated that the costs to both BT and operators of introducing the INCA-based charging system would be greater than for 2PC and that the work necessary for both could be completed in the same timeframe (INCA is the name of BT’s wholesale billing system for calls. It measures the amount of BT’s network used in conveying calls to another operator’s network and applies the appropriate wholesale charges for each conveyance element. INCA has not been used for wholesale NTS charging primarily because of its inability to recognise the difference between calls which originate on BT’s or other originating operators’ networks. This is being addressed and it is likely INCA billing will replace NCD in the future). 2.8 Easynet also state that it is reasonable to make the assumption that it is possible to consider each ring fenced call type (eg NTS calls) in isolation from an accounting point of view. Easynet believe this would have no cost effect on other call types ie the effect within the NTS community would be neutral in respect of introducing 2PC for NTS interconnect calls only. 2.9 Oftel disagrees with this analysis. The Network Charge Controls are designed to achieve an appropriate cost recovery at the aggregate level by BT. Applying 2PC to NTS calls only would require an adjustment to the remaining one part charge for all other calls in order to ensure that BT continued to achieve an appropriate cost recovery. This is because the current ppm charges are calculated using an average call duration that is based on all calls. Removing NTS calls from the calculation would significantly change the average call duration of all other calls due to the disproportionately long nature of NTS calls (due to Internet calls). Therefore, the ppm charge applied to all calls other than NTS would have to be adjusted to reflect the change in average call duration. 2.10 Oftel considers that the result of the cost-benefit analysis applies to Easynet’s proposal. Applying 2PC to NTS calls only would not result in a cost-benefit analysis in favour of introducing 2PC. Oftel’s calculation of the benefits of introducing 2PC was based on implementing 2PC for all calls. If 2PC were introduced for NTS calls only, the potential welfare benefit would be less than that estimated by Oftel in the consultation document because 2PC would be applied to a smaller number of calls. Furthermore, it is not obvious to Oftel that introducing 2PC for NTS calls only would result in significant savings as regards implementation costs. For example, in its response, BT considered that introducing 2PC for certain call types only might increase its costs because of the need to maintain two different charging regimes. Benefits and costsSix principles of cost causation 2.11 Some respondees made comments on the application of six principles of cost recovery suggested by Oftel as a framework for discussion of the relevant issues. (1) Cost Causation 2.12 Global Crossing considered that the proposition that retail customers should not be charged for unsuccessful calls meant that the 2PC model considered by Oftel does not fully reflect the principle of cost causation. 2.13 Although Oftel accepts this assertion, it considers that the reasons for not charging for unsuccessful calls outweigh the alternative. Indeed, as the consultation document noted (at paragraph 1.4), the economic benefits achieved when price reflects cost derive from the consumer facing an appropriate price signal to modify their behaviour. Charging for unsuccessful calls would constitute such a price signal but consumers are not in a position to modify their behaviour as generally they will not know, when initiating a call, whether it will be successful or not. It is unclear, therefore, that charging for unsuccessful calls would promote economic efficiency and result in consumer benefit. Responses to the consultation indicated general agreement on this point and it is therefore the Director’s position in this statement that, were 2PC to be introduced, there should only be a charge for successful calls and the costs incurred in unsuccessful calls should be recovered from the generality of successful calls. (2) Distribution of benefits 2.14 Oftel considered in the consultation document that the introduction of 2PC would benefit consumers of longer than average calls (assuming that lower costs were passed on to consumers). Oftel also stated in the consultation document that consumers in lower income groups were less likely to access the Internet at home than higher income consumers. Global Crossing asserted that the introduction of 2PC would therefore not benefit lower income households as a high proportion of their calls are shorter than average length voice calls. Energis took a different view of the same issue, stating that the introduction of 2PC would stimulate more Internet access among lower income groups due to lower costs. 2.15 On a related point, some operators felt that certain consumers would be particularly dis-benefited by the introduction of 2PC. Credit card validation, which involves a high volume of very short duration calls, was cited as an example where costs could increase very substantially, with such increases ultimately being passed on to consumers. 2.16 Oftel considers that the key issue is whether consumers as a whole would benefit from the introduction of 2PC. However, Oftel notes that the distribution of benefits arising from the introduction of 2PC would probably not be equal across all consumers and that, on the basis of current calling patterns, consumers in lower income groups would probably receive the smallest share of the benefits. (3) Effective competition 2.17 Responses generally agreed with Oftel’s position in the consultation document regarding the application of this principle, that competition is unlikely to be materially affected by a move to 2PC. Global Crossing was of the view that the introduction of 2PC might have an adverse effect on competition because operators whose businesses are principally targeted at voice traffic (eg operators offering business to business managed network services) could face substantially increased costs. It was argued that these costs could not be passed on to business customers due to the fixed price and relatively long duration nature of contracts between operators and business customers. The result could be the exit of operators from relevant markets and a corresponding decrease in competition. 2.18 While Oftel notes such arguments, it has not seen substantive evidence that would lead it to believe that 2PC would materially effect competition given that competitors in the same markets would be likely to face the same changes in their costs as a result of the introduction of 2PC. (4) Cost minimisation 2.19 Energis argued that 2PC may encourage cost minimisation in certain markets. Energis proposed the hypothesis of more efficient management of modem banks by ISPs to avoid unsuccessful login attempts once a call has been connected. Energis also cited potential benefits to call centre operators from answering calls rather than prompting callers to try again later. 2.20 Oftel notes that Energis has not attempted to provide any quantification of these kinds of potential benefits. Oftel does not currently believe that these types of efficiencies would significantly impact on its cost benefit analysis. (5) and (6) Practicality and reciprocity 2.21 BT provided some detailed comments on this issue. It considered that any lack of universal take-up of 2PC by all interconnecting operators would result in practical difficulties and increased complexity in the interconnection charging regime, for example in respect of transit charging if a terminating operator and originating operator had not both adopted 2PC. Another problem could be that non-reciprocity might create incentives for artificial inflation of traffic. 2.22 Similarly, BT identified a number of practical difficulties if 2PC were not introduced for all call types. Oftel notes the existence of such potential problems and maintains its view that if 2PC were to be introduced, it should be for all call types excluding FRIACO. 2.23 BT also stated that set-up costs vary (both absolutely and relative to duration costs) according to the number of exchanges a call goes through. Therefore, a single set-up charge for all calls would not accurately reflect the principle of cost causation. A system which measures set-up costs on a call by call basis would require an increased quantity of pricing reference data, with an associated need to modify the algorithms used in relevant billing and financial modelling systems. Oftel understands that the relationship between the set-up cost and the total cost will vary depending on the proportion of total costs associated with switching rather than transmission. 2.24 BT also identified other practical issues that would be associated with the introduction of 2PC, including how to treat internationally originated and operator-assisted traffic. 2.25 Oftel agrees that the introduction of 2PC would give rise to a number of practical issues that would have to be addressed. It may also be the case that any non-reciprocity between operators as regards 2PC could increase operators’ operational costs. Calculating the welfare benefits of 2PC 2.26 Responses did not fundamentally challenge Oftel’s methodology or calculations in respect of the potential welfare benefits of introducing 2PC. 2.27 BT did provide comment on Oftel’s calculations. BT argued that there were a number of implicit assumptions in the welfare analysis that could lead to the overall benefits of 2PC being over-stated. 2.28 Specifically, BT stated that there were two central issues not fully taken into account in Oftel’s welfare analysis. The net effect of both may offset each other to a degree and they are unlikely to demonstrate that the benefits of 2PC outweigh the costs. 2.29 Firstly, BT believed that an exponential demand function (see note below) has the following drawbacks:
2.30 Secondly, BT argued that it is unlikely that much of the impact of 2PC at the wholesale level would be carried down to the retail level. Therefore, Oftel’s estimates of the welfare benefits for consumers (ignoring the issue of exclusion) may be overstated. 2.31 Regarding the first point, Oftel is aware of the drawbacks pointed out by BT. However, Oftel chose the exponential demand function as it was the optimal one given the objective of measuring consumer surplus. While extensions to the model may change the magnitude of the welfare gains, Oftel agrees with BT that such changes would be unlikely to have a significant impact on the overall cost-benefit equation. 2.32 BT’s second point is addressed in detail at paragraphs 2.57 to 2.64 of this statement. 2.33 Energis argued that the introduction of 2PC for all calls could result in users of voice telephony altering their usage according to new price signals with the potential for certain types of call, for example "social, national" to increase in duration as the costs of continuing a call is reduced. 2.34 Demand effects manifest themselves in terms of changes in call duration and/or call volumes. The model used in the consultation document to calculate the possible welfare benefits of 2PC models demand effects by changes in call volumes (ie total number of minutes) rather than call duration. Modelling demand effects using call duration would be an extension to the welfare model through further iterations in price and quantity changes. Although this may increase the estimate of welfare gains, Oftel believes that this change in welfare is unlikely to be significant enough to change the overall cost-benefit equation. 2.35 Energis also argued that costs to providers of solely voice traffic may not be increased to any great extent if 2PC were introduced, given that Energis believed providers are over-paying for conveyance at present. 2.36 As noted above (see paragraph 2.9) Oftel considers that the Network Charge Controls are designed to ensure that there is an appropriate recovery of costs by BT when calls are considered in the round. It therefore remains the case that the cost benefit analysis demonstrates that the costs of implementation outweigh the benefits. 2.37 In conclusion, Oftel considers that none of the responses provided substantial argument or evidence that would persuade Oftel to alter materially the methodology or calculation of the potential welfare benefits of introducing 2PC. Calculating the costs of implementing 2PC Direct costs incurred by BT 2.38 In its response, BT provisionally estimated its system and operational costs for implementing 2PC to be of the order of £5 million over the first five years (Oftel notes that this is £500,000 more than the estimate provided by BT for the consultation document). BT stated that these costs have been estimated assuming that 2PC applies to all calls on a reciprocal basis and is introduced with immediate effect ie avoiding a phased roll-out or partial implementation. 2.39 BT considered that its costs are likely to be higher if these assumptions are changed. For example, if 2PC were to be introduced on a non-reciprocal basis, BT would expect to incur additional costs associated with managing different price lists and dealing with operator disputes. 2.40 BT stated that it would also expect further costs in addition to those outlined above, for example costs incurred in revising network baskets and product definitions. 2.41 In their responses, both Energis and Easynet questioned the validity of the costs estimates provided by BT for the consultation document. Both companies thought that BT’s costs to implement 2PC would be considerably less than BT’s estimates. Indeed, Energis stated that the cost to BT would be relatively minor. 2.42 Easynet argued that BT’s current billing system could cope with 2PC. Easynet said that BT’s billing system is capable of measuring both total minutes and the number of calls accurately for billing purposes due to the existence of fixed fee services. 2.43 Easynet said it was surprised by BT’s estimate (for the consultation document) of £2.5 million costs in the first year for implementing 2PC. Easynet pointed out that the NTS Focus Group has been informed by BT that upgrades to the BT billing engine to support INCA features would cost in the region of £1 million. In Easynet’s view, the implementation of INCA billing as proposed is more complicated than the implementation of 2PC. Easynet and Energis expected ongoing costs for BT to be negligible rather than the £500,000 per year estimate provided by BT in the consultation document. 2.44 Oftel noted in the consultation document (at paragraphs 2.34 to 2.39) that BT’s current billing system does provide a record of the number of calls as well as the number of call minutes. This means that BT has a basis from which to start in respect of modifications to the system that would be required in the event of the implementation of 2PC. Oftel does not necessarily accept the premise that the implementation of INCA billing is more complicated that the implementation of 2PC. Implementing 2PC would require a number of changes to BT’s billing systems that are different to those necessary for INCA. For example, the introduction of 2PC would require different modifications to the algorithms used in BT’s billing systems to calculate charges than those needed for the new INCA regime. Although Oftel is not able to reach a firm conclusion on this point without further detailed technical analysis of the changes to BT’s billing system needed to implement 2PC, the key point is that even if BT’s costs of implementation are lower than it currently suggests, there is a considerable margin between the costs and the benefits. Given that the cost-benefit analysis does not take into account various intangible costs, it does not appear feasible to Oftel that errors in BT’s estimations would be of sufficient magnitude to alter the analysis such that the benefits of 2PC would outweigh the costs. Direct costs incurred by other operators 2.45 Despite specific questions in the consultation document, a significant number of responses did not attempt to quantify the costs that might be incurred in implementing 2PC. 2.46 Where operators did attempt to quantify their costs, the estimates varied. Orange stated that it is likely that its costs could well exceed £1 million. At the other end of the scale, another respondee considered that its costs would be minimal as it has already made changes to its billing, pricing and accounting systems due to the implementation of 2PC in another country. Energis argued that many operators would be able to modify their billing machinery to support 2PC without significant additional costs. 2.47 Easynet said that changing its billing systems to accommodate 2PC would require little resource. Easynet also made the point that it would not expect to have to contribute to BT’s costs of modifying its billing system to implement 2PC. 2.48 Although, Oftel has not received many estimates of the costs of implementing 2PC from operators, it has not seen any evidence to suggest that the likely costs of introducing 2PC would be significantly different from those estimated in the consultation document. Any difference is unlikely to be of a magnitude sufficient to alter the cost benefit analysis such that the benefits of 2PC would significantly outweigh the costs of implementing it. In this regard, Oftel notes that the total direct costs estimate in the consultation document (approximately £12 million) would have to be reduced by nearly 60 per cent for it to be less than Oftel’s central estimate for the welfare benefits of introducing 2PC (approximately £5 million). Policy issuesProcess of determining a two-part charge 2.49 The consultation document described wide variations in the ratio of the call set-up charge to the duration charge for single tandem call origination in other countries that have implemented 2PC (see paragraph 3.2 of the consultation document). 2.50 Energis commented that the process of dividing or attributing costs between call set-up and call duration is possible to achieve. Energis said that the Monopolies and Mergers Commission has previously taken information from BT on the elements of network usage and relevant costs during set-up and ongoing connection (see note below). Energis also argued that comparisons with other European countries are of limited benefit as none of these countries have enjoyed a competitive telecommunications market for as long as the UK. 2.51 Crucially, Oftel has undertaken sensitivity analysis on this issue – see Table B1 of Annex B to the consultation document. By significantly varying the ratio of set up costs to duration costs, the overall welfare benefit does not dramatically change. Hence, Oftel concludes that differences in the attribution of costs between duration and set up are unlikely to significantly alter the overall outcome of the cost benefit analysis although it remains the case that doubt over the exact split in costs between call set-up and call duration introduces uncertainty into the magnitude of the benefits that might be realised from the introduction of 2PC. Relationship between metered and unmetered Internet access 2.52 The majority of operators who commented on this issue expected demand for metered Internet access to decline over time, mainly as a result of increased growth in broadband and FRIACO services. However, operators expected this to be a gradual process over a number of years. 2.53 Energis commented that there will continue to be a market for long duration metered Internet access calls and that this market may be encouraged to evolve with the introduction of 2PC. Both BT and Energis commented that they would not expect there to be any material impact on FRIACO of reduced charges for long-duration NTS calls. 2.54 Oftel’s position is that it has incorporated the effect of price changes on demand in the welfare analysis in the consultation document. Additionally, Oftel undertook sensitivity analysis on the relevant demand elasticity value (Table B2 of Annex B to the consultation document), both by flexing it upwards (where demand will increase by a relatively greater amount in response to a given change in price) and flexing it downwards (where demand will increase by a relatively smaller amount in response to a given change in price). It is clear from this analysis that different demand scenarios do not significantly impact on the cost benefit outcome. Furthermore, the fact that most operators expect demand for metered Internet access to fall over the coming years lends weight to the position that the welfare benefits from the introduction of 2PC were not underestimated in the consultation document. 2.55 However, Oftel notes that its own consumer research does not entirely support the operators’ views that demand for metered Internet access is falling. Oftel’s latest quarterly residential consumer research shows that the use of narrowband metered packages in UK homes with Internet access had remained relatively stable over the previous year, whilst the use of narrowband unmetered packages had fallen. 2.56 Oftel’s latest quarterly SME research shows that the use of narrowband metered packages amongst UK SMEs had actually increased slightly during the previous year, whilst use of narrowband unmetered packages had fallen. However, some caution is required as before May 2003, figures were based on any package used, whereas the survey in May 2003 was based on the main connection type used. In any event, the evidence available does not suggest that demand for metered Internet access is set to rise significantly in the future, such that the potential welfare benefits from the introduction of 2PC have been significantly underestimated. Passing the benefits to consumers 2.57 There were mixed views on whether operators would introduce 2PC at the retail level if it were introduced at the wholesale level. BT’s response made reference to qualitative research of residential consumers via focus groups that indicated strong resistance among consumers to the idea of a set-up charge, with consumers preferring a minimum call charge. Energis suggested that operators may well introduce 2PC at the retail level if it were introduced at the wholesale level. 2.58 Oftel notes that respondees generally agreed with Oftel’s conclusions in the consultation document (at paragraph 3.7) that the change in cost structure imposed by the introduction of 2PC at the wholesale level would be reflected in adjusted charges at the retail level. 2.59 Easynet argued that it has faced considerable difficulties in moving between price points within the NTS number ranges. One of the reasons Easynet gave is the loss of customers when, for example, a dial-up Internet access number is changed. Easynet suggested the introduction of 2PC as a remedy to this issue because long duration calls (such as those for metered Internet access) would become cheaper on average and there would not be a need for operators to move between price points. 2.60 In response, Oftel considers that operators attempt to move between price points in response to competitive pressure. Assuming the benefits of 2PC are passed on to consumers, a move to 2PC would impact on all operators terminating long duration NTS calls. Therefore, the incentives and disincentives for all operators to shift between price points to gain competitive advantage will remain the same regardless of the introduction of 2PC. 2.61 Easynet and Energis also provided several further arguments regarding the passing of the welfare benefits of 2PC through to consumers. Easynet said that additional revenues resulting from the introduction of 2PC will be passed on to consumers in the form of enhanced features or increased provision of value added content. Easynet also made the point that any additional revenue would be revenue that would have been received had prices been reflective of cost and that depleted revenue streams have limited the features that have been economically viable to provide. Easynet also said that where services are provided behind 0845 numbers, the benefits of 2PC would be passed to customers in the form of enhanced services. Easynet considered that operators of longer duration calls are dis-benefited by the current regime and this should be rectified by 2PC. In the same vein, Energis argued that the implementation of 2PC would introduce the possibility of more innovative tariff packages for consumers. 2.62 Oftel considers that, depending on the extent of competitive pressures in downstream markets, it is possible that additional revenues could be passed on to consumers. This is one of the central assumptions of the welfare model. However, there is no guarantee that this will be the case. The key point is that, as indicated by the welfare model, even if revenues are passed downstream (the best case scenario), the benefits are still unlikely to significantly outweigh the costs of implementation. Whilst it may be true that operators of longer than average duration services may have additional revenues under a system of 2PC, it can equally be argued that, under 2PC, operators with shorter than average duration calls will find their revenues to be lower. 2.63 Regarding the argument that lower revenues limit features, Oftel considers that this is not necessarily the case as operators have had the option of positioning their calls on more expensive retail price points to allow for the greater costs of providing additional features. The introduction of 2PC should not be seen as a way of correcting perceived inadequacies in the current NTS regime and the inability of terminating NTS operators to control their revenues. Oftel will soon publish a consultation document on options for NTS terminating operators to achieve revenue certainty and this consultation will examine several ways (in addition to migration) to resolve problems created by the current mechanism for retail pricing of 0845 and 0870 calls. The introduction of 2PC would not be an appropriate solution to the problems of revenue uncertainty for terminating NTS operators and Oftel’s assessment of whether or not to implement 2PC is correctly based on a consideration of the potential net benefit to all consumers. 2.64 In respect of the introduction of more innovative tariff packages for consumers, Oftel also notes this argument but considers that there is no guarantee that this would be the case. Oftel also notes, however, that if 2PC were introduced, consumers would still ultimately bear the costs of implementation. Most importantly, as noted above, even if revenues are passed downstream, the cost benefit analysis still arrives at the conclusion that the costs of implementation will outweigh the potential welfare benefits. Regulatory, commercial and operational resources 2.65 Most operators considered that implementing 2PC would consume significant regulatory and operational resources. 2.66 BT noted that, inter alia, regulatory and operational resources would need to be allocated to overhauling existing reciprocity agreements and replacing all current rates. BT estimated the impact on its resources on the operational side could be in the region of £400,000 initially, with an additional £650,000 to £700,000 per annum on an ongoing basis. 2.67 Energis argued that the cost of regulatory and industry resource to implement 2PC will be less than for other regulatory changes such as local loop unbundling, CPS or wholesale line rental. 2.68 Oftel has primarily assessed the case for the introduction of 2PC on its merits but it must also consider such a change in the context of the other major regulatory changes that are currently ongoing. Whilst it is possible that the cost of implementing 2PC may be less than for other regulatory changes, this, of itself does not impact upon the results of the cost benefit analysis for 2PC. Also, in two of the cases cited by Energis (CPS and local loop unbundling), Oftel had no discretion as to whether or not regulatory resource should be expended on these projects, as both CPS and local loop unbundling were required to be implemented by EU Directives. Other policy issues raised 2.69 Easynet made the point that if it is decided that minimum call charges on short duration calls should be passed on at the wholesale level to prevent over-recovery by BT (an issue currently under consideration by the NTS Focus Group), then having 2PC at the wholesale level would make it less attractive for "unscrupulous operators" to set up deliberately short duration calls behind low priced NTS number ranges in order to receive minimum call charge fees at the wholesale level. Furthermore, Easynet add that any decision to pass on such charges would de facto be a type of 2PC functionality. 2.70 In relation to the activities of "unscrupulous operators", Oftel would expect other operators to maintain best working procedures to minimise the occurrence and impact of such behaviour. However, in the absence of evidence of a significant problem directly attributable to the lack of a 2PC charging system, Oftel does not consider that this issue impacts upon its cost benefit analysis carried out in the consultation document. Furthermore, Oftel does not consider that the passing on of minimum call charges for short duration calls at the wholesale level would necessarily either be or require 2PC functionality, as such a charge would not necessarily be based on a cost directly attributable to call set-up only. Oftel considers the introduction of such a charge to be a separate issue (with a specific context) from that of the general introduction of 2PC. 2.71 Both Energis and Easynet have argued that current conveyance charges are based on an out of date average call duration of approximately 2.4 minutes rather than 3.7 minutes. As a result, they argue that BT is significantly over-recovering its call set-up costs. Both have put this forward as an argument for the introduction of 2PC and also urged Oftel to investigate this as a distinct issue. 2.73 Oftel made it clear in the consultation document that it is aware that the current system means that longer than average duration calls are effectively relatively overpriced compared to shorter than average duration calls. However, Oftel considers the key point to be that the current Network Charge Controls are designed to allow BT to make an appropriate cost recovery when all calls are considered together. The introduction of 2PC would not change this methodology. Therefore, this argument does not alter the cost benefit analysis for 2PC put forward in the consultation document. 2.74 Furthermore, when forecasting future values of variables relevant for setting the Network Charge Controls, the Director took account of many factors including the impact of Internet traffic (see note below). Easynet and Energis are arguing that call duration assumptions used in setting the Network Charge Controls should reviewed in isolation from other variables in respect of which differences between outturn and forecast values may offset any impact that call durations being longer than (implicitly) forecast has on overall cost-recovery by BT. On this basis, Oftel does not consider the Easynet and Energis argument to be justified. Effectively, Easynet and Energis would require the Network Charge Controls to be completely re-opened. The Director would only consider re-opening the Network Charge Controls in exceptional circumstances and would require compelling evidence that this action was necessary. Notes: In this type of function demand increases with price as a direct inverse function of the price level itself. Increasing elasticity with price is generally intuitively reasonable and perhaps better than assuming a constant demand elasticity, although this particular function implies a relatively strong change in elasticity by price itself. However, it provides a simple formula for deriving consumer surplus. See the MMC report British Telecommunications plc: A report on a reference under section 13 of the Telecommunications Act 1984 on the charges made by British Telecommunications plc for calls from its subscribers to phones connected to networks of Cellnet and Vodafone dated 21 January 1999. The full text of this report can be viewed at www.competitioncommission.org.uk/rep_pub/reports/1999/422bt.htm#full. See www.oftel.gov.uk/publications/pricing/pcr0101.htm#Chapter%204:%20Financial%20modelling%20-%20main%20assumptions for an explanation regarding values for variables relevant to setting the Network Charge Controls.
Chapter 3 – Conclusions3.1 The Director has fully considered the representations received in response to the consultation, and assessed the views and evidence available to him. On this basis, he concludes that a system of 2PC should not be introduced at the wholesale level for interconnection charges between operators. 3.2 The Director maintains his view that 2PC should only be implemented if the benefits of it significantly outweigh the costs, given the presence of further intangible and non-quantified costs. Furthermore, he considers that he is supported in this view by the overwhelming majority of responses received to the consultation. 3.3 As concluded in the consultation document, Oftel does consider that there are clear welfare benefits associated with a system of 2PC. However, on the evidence available to it, Oftel considers that implementing a system of 2PC would require industry to incur significant direct costs. As a result, it is the Director’s conclusion that there is no significant net economic benefit associated with the implementation of 2PC. Indeed, based on the evidence available to it, Oftel concludes that the implementation of 2PC would be more likely to result in a net cost over a period of five years. Oftel’s sensitivity analysis shows that even on a very optimistic calculation of the benefits of 2PC (obtained by increasing all set-up costs by 50 per cent with a corresponding decrease in duration costs to maintain overall cost recovery), the estimate of direct costs incurred by industry in the consultation document would have to decrease by around 25 per cent (from approximately £12 million to £8 million) for there to be a possibility of a net benefit over five years from implementing 2PC. Thus, the Director is able to conclude, on the evidence available to him, that it is unlikely that the benefits of 2PC significantly outweigh the costs. 3.4 The Director also notes that, at a time of considerable regulatory change, the implementation of 2PC would undoubtedly consume scarce regulatory resources in Oftel/Ofcom and across industry. In the absence of significant net benefit, the Director would re-iterate his view that wholesale changes to the current interconnection charging arrangements to implement a system of 2PC are not considered to be appropriate.
Annex A – List of responding parties
|
|
||||||
|
|
||||||
 |
 |
|
|||||
 |
|
||||||
