Introduction
7.1 Oftel is continuing to examine the practicalities of funding universal service because, although no net costs to the universal service provider have been proven, the principle that such costs should be funded remains valid. The ideas put forward in this Chapter are therefore only provisional. Should it become necessary to institute a funding mechanism in the future, Oftel will carry out a more detailed consultation on the details of the scheme.
7.2 Some elements of universal service discussed in this document would be eligible for funding, if an undue financial burden were demonstrated:-
7.3 Other elements would not be eligible for funding in current circumstances, because they would be a responsibility faced by all operators:-
Services such as call barring and free itemised billing would not be eligible for funding in their own right, but they should be taken into account when assessing whether or not a customer were uneconomic.
7.4 In Chapter 6 Oftel concluded that there is no proven case that there is an undue financial burden on BT arising from those elements of its universal service obligations which are potentially eligible for funding. In Oftel s view, therefore, there is currently no need for funding arrangements to be set up. This position might change in the future - if an undue burden were demonstrated, it would be appropriate for funding arrangements to be put in place, so that all public operators contributed to the net costs incurred by the universal service provider. This Chapter discusses first how a universal service funding mechanism might operate, and then how such an arrangement might be used to promote competition in the delivery of universal service. Oftel does not expect that a funding mechanism will be set up before its further review of the net costs of universal service, starting in 1999. If there were a need in the future for a mechanism to be put in place, Oftel would undertake a further consultation on the full details of the funding arrangements.
FUNDING ARRANGEMENTS
Administration
7.5 Oftel has explored two possible options for the constitution and administration of funding arrangements: an Actual Fund administered by a specially created independent company or a Virtual Fund. In either case Oftel would specify the overall framework, ie the costing methodology (including the valuation of benefits), the operators that would be required to make contributions into the Fund and the basis of those contributions (see below). If there were an independent company it would collect information from operators, calculate the net costs of universal service of each operator eligible to receive payments, calculate the contribution due from each operator liable to pay into the Fund, and ensure transparency of the net cost of universal service each year and the contributions made by each operator. Contributions (payments) would be paid to (by) the independent company. Consequently, the company would have the task of managing the money balances (which would arise either from the mismatch between the timing of contributions in and payments out, or the desirability of creating a contingency fund).
7.6 The alternative to an Actual Fund would be a Virtual Fund. One approach would be for the universal service providers to produce estimates of the net cost, following the costing methodology laid down by Oftel. Information on the basis of contributions could be collected from contributing operators by Oftel, who could pass on to the universal service providers the proportion of the net cost to be paid by each operator. The universal service providers could then send out a bill to each contributing operator for the recovery of the identified net costs and organise the collection of the contribution. Operators would therefore make the payments direct to the universal service providers without the need for a central fund. An operator would be able to refer to Oftel for arbitration both the size of the net cost identified by the universal service providers and the relative size of its contribution. Details of the net costs of universal service and operators contributions would be made public each year either by the universal service providers or by Oftel. Subject to legitimate claims about the confidentiality of certain cost and revenue information, as much detail as possible about the derivation of the net costs of universal service should be available to contributing operators and other interested parties.
7.7 Oftel has concluded that its preferred option for the administration of the funding arrangements would be a Virtual Fund. In the consultative document published in December 1995 Oftel favoured the other option, a fund held by a specially created independent company. However, having explored that possibility in more depth it is concerned that such an arrangement could be unnecessarily burdensome and costly. There would be a danger of excessive administration costs, out of proportion to the size of the required payments between operators. Some operators have expressed doubts about the need for the creation of physical fund and consequent costs of managing money balances, especially if payments would be needed in advance from contributing operators (as would be the case to avoid the risk of the company being liable to make a payment to a universal service provider before it had received sufficient contributions from other operators).
Contributing operators
7.8 In Oftel s view all public network operators with an individual Telecoms Act licence could be potential contributors to the Virtual Fund. Oftel considers that, if any undue financial burden were to arise for a universal service provider, it should be funded by as wide a crosssection of operators as possible. Or, more correctly, funding should be by as wide a crosssection of telecom customers as possible, since it will be the customers of contributing operators that will ultimately pay through (slightly) higher retail prices.
Basis for contributions
7.9 The basis for the calculation of contributions depends in part upon the perspective that is adopted. It might be argued that it would be desirable to set contributions in relation to as wide a revenue base as possible: basic network revenues (retail and interconnect), including fixed and mobile voice telephony and private leased circuits (but excluding revenues from broadcast entertainment services and a few other exclusions discussed below). Another approach might be to attempt, as far as practicable, to relate the contributions to the benefits that customers receive from the provision of universal service.
7.10 If universal service were to impose a net cost on the provider, the universal service obligation would ensure that customers who might not be profitable for the universal service provider to supply would nevertheless receive a telecoms service. In such circumstances the number of customers on the network would be larger than in the absence of the universal service obligation. Other (profitable) customers benefit from the larger network by making calls to, and receiving calls from, the universal service customers and by having the ability to make calls to and receive calls from such customers.
7.11 This perspective would suggest that the revenue base should comprise a narrower set of services. For example, outgoing international calls cannot be made to universal service customers in the UK (but incoming international calls can be and so should remain in the revenue base). For a similar reason this perspective would also imply the removal of private leased circuits from the revenue base - calls to universal service customers would only be made from private leased circuits by breaking out onto the PSTN, in which case these calls would anyway be included in the revenue base. On the other hand, it can be argued that the exclusion of private circuits from the revenue base might lead them to be underpriced relative to the PSTN, which could distort competition between private circuits and the PSTN to the extent that they are substitutes. But for the majority of customers they are not currently close substitutes.
7.12 Pursuing the logic that contributions should be related to the benefits received by customers, it could be argued that contributions should be in proportion to revenues from calls specifically to universal service customers. However, this does not necessarily follow, since customers benefit not only from making calls to universal service customers, but also from the ability to make such calls. Furthermore, the approach is not currently implementable and it would generally require operators to invest in more sophisticated information systems.
7.13 Whatever set of services was considered appropriate for the revenue base, a few exclusions would be appropriate. First, whilst a contributing operator s relevant revenue includes revenues from interconnection, it should exclude interconnection payments to other contributing operators (and, if included in the revenue base, leased line payments to other contributing operators). This exclusion would ensure that the contribution to the funding of universal service made by a customer in the retail price for a call would be the same regardless of the number of network operators that were involved in conveying the call. An alternative approach to avoid doublecounting would be to exclude all interconnection sales from the revenue base. Second, the relevant revenue for calculating the contribution to be made by a universal service provider should exclude the retail revenues received from the universal service customers (and uneconomic areas and public call boxes), since these would be the customers that would give rise to the net cost being funded.
7.14 The discussion above has assumed that revenue provides the appropriate basis for calculating contributions. However, some operators have suggested that call minutes would be a preferable basis. Oftel recognises that each approach has its pros and cons and that the difference between the two could be quite significant for certain contributing operators (eg mobile operators who have much higher prices per call minute than fixed operators). Nevertheless Oftel would favour revenues as the appropriate basis. Revenue information may be easier to collect and could be more amenable to auditing. In addition, Oftel considers that an attractive feature of the use of revenues as the basis for contributions is that, generally speaking, a customer would make the same payment towards the funding of universal service for the same size of telecoms bill.
IDEAS FOR COMPETITIVE DELIVERY
7.15 If funding arrangements were to be put in place, it would be desirable to explore mechanisms for introducing competition into the delivery of universal service. This would benefit customers by ensuring that universal service responsibilities were delivered as efficiently as possible. This section discusses the practical issues that would need to be addressed for two methods of introducing competition: first, competitive tendering, under which universal service responsibilities for defined areas of the UK are subjected to an auction; and second, pay or play , under which operators may choose voluntarily to offer the defined service packages targeted at customers with affordability problems and in return receive a payment via the funding mechanism.
Competitive tendering
7.16 The idea of competitive tendering for uneconomic areas has been mentioned in previous consultative documents and has been discussed in meetings of the Competitive Delivery Working Group. Potentially, competitive tendering for areas could provide a means of testing whether or not the universal service costs of uneconomic areas and new service packages are offset by benefits. If a net universal service cost could be demonstrated in a truly competitive auction, then it would have been demonstrated that a net cost exists and setting up a universal service funding mechanism would be justified. It is, however, an important caveat that the true value would only be revealed if there were a genuinely competitive auction and no operator, especially the incumbent, held an advantage that was unrelated to greater efficiency. Such an advantage might arise, for example, from the incumbent being better informed about the costs and revenues of the area being tendered. This is an area which Oftel believes could have great potential and it is therefore discussed in some detail below.
7.17 An extension of the idea of competitive tendering for uneconomic areas would be to put out to tender the universal service responsibility for specified geographic areas of the UK, even if the area as a whole were profitable to serve. In areas that have not been identified as uneconomic, the great majority of customer demands would be met by the market. But competitive tendering would be a way to encourage the most efficient delivery of the carrier of last resort responsibility. Furthermore, it would be one way to introduce competition into the provision of the service packages discussed in Chapter 3 (another way would be pay or play ). If there were a truly competitive auction, the extent to which the benefits of such schemes offset the financial costs would be revealed.
7.18 Public call boxes that were uneconomic for the universal service provider to provide could also be subjected to a tender process. In theory, there could be tenders for individual call boxes, but this approach is likely to be impracticable. A more practicable approach might be to hold a tender for the provision of a specified group of call boxes in a geographical area or region. Such a group could include profitable as well as unprofitable call boxes. A further approach would be to include the responsibility to provide reasonable access to call boxes as one element of the contract for the wider universal service responsibilities when tendering an area.
Potential areas to be tendered
7.19 Potential uneconomic areas that could be tendered are the remote areas where the geographical characteristics and the low population density make the area expensive to serve. The map at Figure 7.1 indicates the areas of the UK that have been found to be uneconomic (before benefits) in Oftel s latest universal service costing analysis. The map does not provide a definitive statement of uneconomic areas, because it is based on the incomplete evidence that is available. It does, however, indicate the regions of the UK where uneconomic areas might (and might not) be expected to occur. Of the uneconomic areas, the best candidates for tendering would be those where there is either competing infrastructure already in place or a good opportunity to deliver services using alternative technology at lower cost (eg fixed radio access). The concept of tendering could be extended to any distinct geographical area. Any area of the UK might include some customers that individually were unprofitable for the operator to serve. There may be a role, therefore, for a tender which would be for the residual obligations: provider of the new service packages (eg Lifeline scheme), carrier of last resort, and possibly public call box provision. Such tendering could be considered, for example, in cable franchise areas where rollout is well advanced or in areas where another operator is competing (eg Ionica, Scottish Telecom).
Tender bids
7.21 The most obvious option would be for there to be bidding for the lowest subsidy/highest payment to take on the obligation, with an explicit control on prices. If BT were to win the tender, it would anyway be subject to price control (at least until 2001) and would have to charge the same prices throughout the UK for services within the scope of the definition of universal service. If a company other than BT were to win the tender, a similar control on charges could be imposed. In addition, the winning bidder would have to offer the new service packages at the tariffs specified in guidelines agreed with Oftel.
7.22 For both uneconomic areas and other areas, companies could bid negative amounts (a bid for a subsidy) or positive amounts for the universal service obligation. The winning operator would have to compete to provide service with any other operator in the area but would be the supplier of last resort, ie would be obliged to supply basic telephony to any customer on reasonable request. Other operators providing service in the area could refuse to supply a particular customer or only do so at higher prices if they felt that customer would be costly to serve. The tender would be won by the highest bidder (if there are positive bids) or by the lowest subsidy bid, if there were only negative bids, subject to that bidder demonstrating that the requirements of universal service in the area would be met. The required subsidy would be paid to he winner of the tender using the funding arrangements that had been put in place.
Other issues
7.23 The contract to provide service would need to cover the following issues:
The detailed level of service to be supplied;
The duration of the contract. Too long a period might deter new operators from taking on the obligation, given that costs and benefits cannot be known with certainty. But too short a period could deter new investment because it reduces the potential period over which the benefits of universal service provision can be earned and could be less than the economic life of the assets invested;
The quality of service standards that would be required (eg calls lost per thousand, percentage of call boxes in good working order etc);
Contingency arrangements. If, in the extreme, the new provider s business were to fail, there would need to be contingency arrangements to ensure that customers were not left without service;
Monitoring and penalty arrangements. This would include the checks needed for Oftel to ensure that the terms of the supply contract were being met, and the action to be taken if the quality of service was not adequate.
7.24 It is quite likely that there might be relatively few bidders in any particular tender, perhaps only two or three realistic bidders, ie those operators which already have in place, or could invest in, infrastructure in the area being tendered. It would be important, therefore, for the design of the auction process to seek to minimise the chances of strategic or collusive bidding. For example, it would be desirable for a reserve price to be set, based on the estimated universal service cost of the area to the current universal service provider - if no better tender bid were received, the area would continue to be served by the current universal service provider, who would receive a payment from the funding arrangements based on the reserve price. Given the likelihood of a small number of bidders, it is possible that a single round auction of sealed bids might be preferable to a multiround auction, because of the reduced risk of strategic bidding developing. If more than one area were subject to a tender, it might be preferable to hold sequential tenders so that bidders for areas that come up for tender later on could learn from the conduct of the earlier tenders, and confidence in the process might be enhanced. On the other hand, there could be a danger that collusive bidding might be fostered by sequential tenders.
7.25 If there were no other operators with competing infrastructure in the tendered area, which is likely for many uneconomic areas, then it might be difficult to find any operator to compete in a tender with BT. One possible way around this problem might be a transfer or lease of assets from BT to potential universal service providers. However, this could raise legal and practical difficulties. In Oftel s view it has no powers to enforce a transfer of assets under the Telecommunications Act 1984. Even if this were possible, unless a way could be found to prevent BT from dictating the terms on which any such transfer could take place, BT would seem to be in an advantageous position when competing against bidders reliant upon a transfer or lease of BT assets.
7.26 Potentially a large amount of disaggregated information on costs and revenues would need to be supplied to bidders. BT would be in a position of advantage visavis other bidders in formulating the size of its bid if it possessed more indepth and complete information about the area. It would be important to avoid the winner s curse , ie that, because of its information disadvantage, a bidder (other than the incumbent) would only win the tender at a subsidy that would be insufficient to cover its net costs incurred.
7.27 Oftel is enthusiastic about developing the idea of competitive tendering. It could provide benefits to consumers through encouraging the most efficient provision of universal service. It could also provide a mechanism for the existence of universal service costs, net of benefits, to be convincingly demonstrated for particular areas. This chapter has raised a number of the practical issues that would need to be addressed before competitive tendering could be implemented. These issues would need to be explored further to find the best way to tackle the potential problems identified, and to clarify the detailed nature of the arrangements that would be necessary.
Pay or play
7.28 If funding arrangements were in place, pay or play could be introduced for uneconomic customers. Operators without formal obligations to provide the service packages targeted at customers with affordability difficulties (discussed in Chapter 3) could choose voluntarily to provide such packages and thereby become eligible to receive universal service funding. To be eligible for funding, the service package introduced by the operator would need to conform to guidelines specified by Oftel. Any such funding would be netted off against the operator s contribution towards universal service net costs.
7.29 Oftel believes that pay or play has a great deal of potential to introduce competition into the provision of the targeted schemes. It would allow customers to have a choice of operator. It might also encourage innovation in the service packages offered, subject to the criteria set out in the guidelines being satisfied. Pay or play provides an alternative to tendering for introducing competition in the provision of services to universal service customers. Depending upon the circumstances, tendering might be more appropriate for one particular area, whilst pay or play might be preferable for a different area.
7.30 A large volume of cost and revenue information is required to derive robust estimates of universal service costs. It would be impracticable to carry out a fullblown costing study for each and every operator that chose to participate in pay or play. Consequently, it might be desirable to assess the payment to the pay or play operator using some of the information collected from the universal service providers. A further desirable feature of such an approach is that it might encourage participation by operators that can deliver the service packages more efficiently than the universal service providers.
7.31 The universal service cost (before benefits) of the service
package offered by the pay or play operator might be calculated
using the following formula:-
Universal service cost before benefits
=
avoidable costs of access + avoidable costs of outgoing calls + avoidable costs of incoming calls - revenues from fixed charges - outgoing call revenues - incoming call revenues
7.32 Some of the information required to implement this formula could be provided by the pay or play operator and some information could be derived from the universal service providers. For example, the pay or play operator could identify the revenues from fixed changes, outgoing calls and incoming calls that it receives (in the case of the Lifeline scheme, there would of course be zero outgoing call revenue), whilst the avoidable costs of access and calls could be estimated using information from BT. For simplicity, it might be easiest to use BT s national average avoidable cost of access. On the other hand, a more accurate estimate might be obtained by using BT s average avoidable access cost in the area served by the pay or play operator.
7.33 The appropriate payment from the Fund should attempt to reflect the net universal service cost to the pay or play operator of its participation in offering the targeted service packages. Therefore, the payment should be set after subtracting an estimate of the value of benefits from the estimated universal service cost derived from the formula above.
Questions:
10 In the event that it were appropriate to set up funding arrangements, would respondents agree with Oftel s preference for a Virtual Fund over an Actual Fund?
11 Which types of licensee would respondents consider should make contributions into a Fund, and what should be the basis of determining the size of each company s contribution?
12 What comments do respondents have on Oftel s ideas for introducing
elements of competition into the provision of universal service?
Figure 7.1 Universal service cost per exchange area
Note: Exchange areas incurring a universal service cost cover less than 0.5% of total UK lines at a total universal service cost of about £5m-10m.