• Home
• ICT
• Telecommunications
• Archive
• Regulation of Access
• Part II: The Options

• Telecommunications

• Archive

• TSO Deed
• Compliance
• Information Disclosure
• Ministerial Inquiry
• NZ Telecoms Updates
• Regulation of Access
• Terms of Reference

The previous section highlighted the concerns that have been raised as a consequence of the Privy Council decision and the questions that it raises for the Government's current regulatory policy towards vertically-integrated natural monopoly industries. This section evaluates the current regulatory policy in the light of these concerns.

Broad Policy Approaches For Regulation of a Vertically-Integrated Monopolist

The discussion which follows is a first-principles analysis of the broadest possible policy approaches for the regulation of vertically-integrated natural monopolies. It should be emphasised that the majority of the policies set out in this section are not under consideration by the Government. In particular, there are no plans for changing existing ownership structures (by, say, re-nationalisation). The discussion of these broad policy approaches is included here for analytical completeness.

Objectives

The Government's overall objective for the operation of the economy and the operation of markets dominated by natural monopolies, in particular, can be found in the statement of the Government's Strategic Result Areas:

"[The] establishment, implementation and monitoring of legislative frameworks for the fair and efficient conduct of business and the operation of markets, which rewards innovation, promotes efficiency and enhances investor confidence."[69]

In particular, the Government seeks to maximise the contribution of this sector to the overall growth of the economy through the promotion of economic efficiency.

Options

At the broadest level the set of policy approaches for the regulation of vertically-integrated natural monopolies may be characterised by two dimensions corresponding to:

1. controls over ownership; and
2. controls over pricing.[70]

The 'ownership controls' dimension includes, for example, state ownership of the entire integrated firm, state ownership of just the monopoly component, or enforced separation of ownership of the monopoly component. The 'price controls' dimension includes for example, controls over the prices of the essential input, controls over the price of the final product or both (or neither).

The set of combinations of these two dimensions creates a grid of possible policy categories or approaches. This is illustrated in Table 1 of Appendix C.

The Risks Of Regulation

Broadly speaking, the various policies illustrated in the grid involve either some form of state ownership, some form of breaking-up of the vertically integrated monopoly or some form of price control.

Each of these forms of intervention have significant drawbacks. State-owned firms tend to be worse at controlling costs than privately-owned monopolies because they are able to influence and be influenced by political pressures. For example, these firms may seek particular dispensations rather than take unpopular cost-cutting measures.

Breaking up a monopoly will tend to increase production costs, if there are vertical economies of scope; and transactions costs due to the introduction of arms-length dealings between the newly created firms.

Experience suggests that price control tends to distort incentives to invest, and reduce incentives to increase productivity and to innovate. It is likely to restrict the range of products offered and the range of pricing options offered by the regulated firm.[71]

The main reasons for this are as follows:

1. The lack of information available to the regulator about the cost structures of the firms they regulate[72] and about the elasticities of demand for the firms' products. There is, therefore, a risk that the regulator will set the 'wrong' access price. For example, if the access price provides insufficient revenue for the facility owner to cover costs the facility owner may have little or no incentive to maintain the facility, invest in extra capacity or invest in entirely new facilities;[73] and
2. Self-interested behaviour and lack of oversight of the body to whom regulatory authority is delegated.[74]

Inevitably, the selection of the preferred option will involve trading-off the risks of market failure against the risks of regulatory failure. In those cases where the market failure problem is not severe (such as where competition from substitutes provides sufficient competitive pressures on the monopolist) no regulation may be the preferred option.

In many industries the undesirable consequences of intervention can be kept to a minimum by permitting competition in an up- or downstream market (to the extent that it is possible) and restricting any regulatory intervention to the natural monopoly component.

There are five broad policy approaches which permit competition in the downstream market and which constrain in some way, the ability of the monopoly to exercise market power:

1. state ownership of the natural monopoly component with implicit price restraints on the essential input;
2. joint ownership of the natural monopoly component (e.g., by the users of the facility);
3. enforced separation of ownership of the natural monopoly component with price restraints on the essential input;
4. no restraints on ownership with price restraints on the essential input; and
5. no restraints on ownership with price restraints on both the essential input and the final product.

Evaluation

The various merits of these options are discussed further in Appendix C.

This document focuses particularly on the final two approaches (d) and (e). Current government policies fit with approach (d). This approach has the advantage of not requiring state ownership (with the associated inefficiencies), enforced break-up of the monopolist (with consequence losses of economies of scope), or more extensive price regulation than is absolutely necessary. It also has the advantage of preserving, as far as possible the existing 'light-handed' regulatory regime.

However, as the Government has made clear in the past, if, for whatever reason, competition fails to provide an adequate discipline on the dominant firm's use of its market power, the Government will consider the introduction of further measures such as given under approach (e), including final product price control under Part IV of the Commerce Act.

Under approach (d), the state restricts, in some way, the terms and conditions that the incumbent monopolist can legally demand for the essential input. The discussion which follows looks for solutions to the problems identified in Part I within the scope of this broad policy option. Within this approach there remain a number of more detailed alternatives to be considered such as:

1. rules for determining the access terms and conditions;
2. rules for handling social obligations; and
3. options for designing the institution that will apply the rules.

The resulting regime is unlikely to be appropriate for all access disputes in all natural monopoly industries. Therefore it is necessary, in addition, to design an institution that will act as a 'gatekeeper', deciding when the regime will apply.

The Regulatory Framework For Controlling Access Terms and Conditions

We will focus upon options that fit within approach (d) above, i.e., options for control of the terms and conditions of access to the vertically-integrated monopolist. This section presents several options for the regulatory framework that controls these access terms and conditions. These options include the current regulatory environment of reliance upon the courts and part II of the Commerce Act.

Objectives

The broad objective presented above provides little specific guidance for which to assess the various options. The following definition provides a specific set of criteria against which the regulatory framework options can be assessed.

The preferred regulatory framework is one which defines and enforces the parties' rights in interconnection negotiations in a manner that:

1. promotes economic efficiency in the relevant industry;
2. is timely; and
3. acts with a high degree of certainty and predictability.

Options

The set of policy options for the regulatory framework is characterised by two dimensions:

1. the nature of the institution for resolving disputes as they arise and, in various other ways, further defining or modifying the parties' rights as time passes; (referred to as the 'regulatory institution'); and
2. the extent of the prescriptiveness of the legislation, regulations or other principles to which the institution in (a) is required to comply, or have regard to, as appropriate (referred to as the 'principles').

The 'Regulatory Institution'

This document focuses upon four possible regulatory institutions for setting access prices and in other ways defining the parties' rights, where the parties have clearly failed to develop their own solution:

1. the courts;
2. arbitrators;
3. a permanent regulatory agency; and
4. the Government acting under delegated statutory powers.[75]

Table 2 in Appendix E compares these four regulatory institutions across a range of criteria. To summarise, statutory regulators typically have the greatest scope for imposing sophisticated regulatory solutions that require continual oversight. However, statutory regulators tend to be relatively vulnerable to influence and lobbying and may have a tendency to act in the interests of themselves and/or the regulated firms rather than the wider public interest.[76] In addition, the decisions of a regulator do not create a binding precedent.[77] On the other hand, the Courts tend to be least subject to lobbying activities and the decisions of higher courts are binding on lower courts. Arbitration lies closer to the courts to the courts in the types of remedies that can be imposed and closer to a statutory regulator in terms of vulnerability to influence and precedent value. Arbitration also, unlike the courts and a statutory regulator, does not involve a fiscal cost.

A specific mandatory arbitration option is presented in Appendix A for the purposes of focusing the comments of consultees. Two notable features of this option are that (i) the proposed regime would be timebound; and (ii) the arbitrator would have wide powers to require the disclosure of relevant information.

One of the drawbacks that has been identified with conventional arbitration is that it tends to have a 'chilling effect' on the pre-arbitration negotiations. One solution to this drawback is to constrain the arbitrator to only choose one of the final offers of the parties. This is called 'Final Offer Arbitration'. Final Offer Arbitration is discussed further in Appendix E. The comments of consultees are sought on whether or not any compulsory arbitration regime should involve Final Offer Arbitration.

The 'Principles'

At present there are no principles (in legislation or regulation, apart from the Commerce Act) that give specific guidance to the courts as to how to best define the lawful interconnection terms and conditions.

This document identifies three possibilities relating to the prescriptiveness of principles to which the regulatory institution in the previous paragraphs might be required to comply or have regard, as appropriate.

1. no principles (other than the rules in Part II of the Commerce Act);
2. broad principles;
3. detailed industry-specific principles.

As an example of broad principles, the regulatory institution might be required to have regard to the same criteria required of the /templates/MultipageDocumentPage____4596.aspxCommerce Commission in the exercise of its price control powers, namely:[78]

1. the extent to which competition is lessened or likely to be limited in the relevant market;
2. the necessity or desirability of safeguarding the interests of consumers; and
3. the promotion of efficiency in the production and supply or acquisition of the controlled service.

Regarding detailed principles, this document identifies two options, should it be desirable to set more prescriptive rules, either as an alternative or a supplement to broad policy principles:

1. Add regulation-making powers to the Commerce Act; and
2. Extend the powers of the Minister of Commerce by introducing a mechanism similar to that in s.26 of the Commerce Act. At present, s.26 requires the Commerce Commission to have regard to statements of government policy transmitted to it by the Minister of Commerce. A similar mechanism could be established by which detailed principles could be transmitted to the regulatory institution. Such a mechanism is referred to in this paper as a 's.26-type' mechanism.

The main advantage of statutory regulations is that they create greater certainty about how the detailed principles will be applied. The main advantage of a s.26-type mechanism is that it allows the Government to tailor the specific 'rules' to the particular circumstances at hand while, at the same time, permitting the regulatory institution a degree of discretion to disregard the rules should their application in a particular context lead to a perverse outcome.

If detailed access principles were to be adopted, one possible source of the statement of principles is the regulated industry itself. Each industry might be required to develop an 'Interconnection Code of Practice' which, if approved by the Government, could be issued as a s.26-type statement. The Australian gas industry has recently developed and promulgated such a code[79]. However, a similar attempt in the New Zealand telecommunications industry was unsuccessful.

Combining the 'Regulatory Institution' with the 'Principles'

The set of combinations of the 'regulatory institution' dimension with the 'principles' dimension is illustrated in Table 3 of Appendix E. The discussion below is limited to the seven combinations that officials regard as meriting further discussion. These are:

1. No principles (apart from the Commerce Act) with resolution and enforcement by the courts (i.e., the status quo);
2. No principles (apart from the Commerce Act) with resolution and enforcement by compulsory arbitration;[80]
3. Broad legislative principles with the courts
4. Broad legislative principles with compulsory arbitration;
5. Broad legislative principles with a statutory regulatory agency (such as the Commerce Commission)
6. Detailed industry-specific principles with the courts; and
7. Detailed industry-specific principles with compulsory arbitration.

Evaluation

The present regulatory regime is given by option (a). Primary reliance is placed on the parties to conclude access agreements with the courts available to deal with anticompetitive behaviour under s.36 of the Commerce Act as a backstop. S.36 only gives the courts the broadest of legislative guidance. The courts' role under the status quo is to give further guidance to negotiating parties by defining what conduct is or is not (or has or has not been) lawful.

Part I of this document raised the concern that (i) the large amount of uncertainty in the current regime, (ii) the incentive of the monopoly to exploit that uncertainty and (iii) the time taken by the courts to resolve that uncertainty, might lead to unacceptable delays before competition is able to emerge. Until that time New Zealand consumers are denied the benefits of competition.

There are two possible actions that could be taken to address these concerns. The first is to increase the speed and reduce the cost of decision-making. This might be achieved by moving decision-making to an institution (e.g., arbitration) that prescribes the terms and conditions of access, is less constrained by procedural rules and traditions and may more easily be time-bound. The second is to increase the prescriptiveness of the relevant principles, thereby reducing reliance upon the decision making body. Each of the options above takes one or both of these actions.

Appendix E contains an evaluation of the options above in terms of the objectives set out in paragraph 187. In summary:

1. The greater the prescriptiveness of the enabling legislation, the less uncertainty surrounds the precise rules of access pricing and therefore the fewer decisions that must be taken by the regulatory body. This is especially important when decision-making is slow and costly, as it is under the courts.
2. Mandatory arbitration reduces the costs and delays of dispute resolution but, when compared with an industry regulator, is constrained in the types of remedies that can be enforced. However, a regulatory agency with its own budget and wider powers of discretion, is subject to risks of pursuing its own goals at the expense of public policy goals, promoting the interests of the firms it is regulating at the expense of consumers and national welfare and adopting a more intrusive regulatory approach than may be strictly required.

Possible Interim Solutions

A couple of variants of the options mentioned above deserve further discussion, particularly as candidates for an interim solution, pending any final arrangements.

Consider first, a variant of option (e). There is at present a regulatory framework in place for the Commerce Commission (a statutory regulator) to recommend price control, subject to the broad principles contained in the Commerce Act.[81] The significant drawbacks of price control are discussed elsewhere in this document and include the potential for hindering the investment and innovation of the controlled firm. However this option has the advantage of not requiring new legislation and has the potential for having immediate effect.

Lastly, consider a variant of option (b). Recent negotiations between Telecom and Clear have seen concessions made by both sides. The remaining distance between the parties has been reduced. It might be possible to submit the outstanding positions of the parties to a form of Final Offer Arbitration to close the remaining gap. Although ad hoc, this option has the advantage of being restricted to the telecommunications industry and not requiring more extensive regulatory intervention than is absolutely necessary.

Questions for Consultees

1. Which of the options set out in paragraph 199 for defining and enforcing the regulatory environment for vertically-integrated natural monopolies would best promote economic efficiency in a manner that is timely, certain and predictable?
2. If broad legislated principles are adopted, would principles such as those set out in paragraph 195 be appropriate?
3. What are the advantages and drawbacks of communicating detailed statements of policy to the regulatory institution via Government statements (as occurs in s.26 of the Commerce Act)?
4. Should the wording guiding the 'regulatory institution' (e.g., the arbitrator) as to how much weight to put on the s.26-type statements be stronger than the 'have regard to' requirement of s.26, e.g., 'be required to comply with'?
5. What are the advantages and drawbacks of an arbitration process of the type set out in Appendix A? What are the advantages and drawbacks of Final Offer Arbitration?

Access Pricing Options

In part I, concerns were raised about the appropriateness of the Baumol-Willig rule. This section seeks to present the set of possible interconnection pricing rules and evaluate them against the Government's objectives.

Note that, in the telecommunications context, it may not be possible to select an interconnection pricing rule which achieves both the goal of economic efficiency and the goal of efficient handling of the costs of social obligations, such as the Kiwi Share. Therefore, it is appropriate to separate these two objectives. We consider here the appropriate interconnection pricing rule in the absence of any social obligations. The question of the efficient handling of the social obligation is discussed in the next section.

Objectives

In paragraph 187 we identified that the regulatory regime should (a) promote economic efficiency; (b) be timely; and (c) have a high degree of predictability. Of these, the relevant criterion for identifying which access pricing rule is most appropriate is economic efficiency (i.e., productive, allocative and dynamic efficiency).

Options

A number of possible rules have been suggested for pricing access to natural monopoly facilities including:

1. pricing at short-run or long-run marginal cost;
2. pricing at long-run average incremental cost;
3. Ramsey pricing (pricing in inverse relationship to the elasticity of demand);
4. two-part pricing (such as a high 'fixed' charge with a low 'usage' charge);
5. peak-load or other forms of capacity pricing;
6. the BW or Efficient Component Pricing Rule; (or BW less monopoly profits);
7. revenue capping rules[82];
8. (in the case of two-way networks) the rule of 'reciprocity' and related rules such as 'bill and keep'; and
9. more sophisticated pricing rules that acknowledge the information limitations of the regulator and that seek to explicitly reward the regulated firm for revealing information.[83]

Evaluation

These pricing rules are evaluated in Appendix D against the criterion given in paragraph 209. That evaluation is summarised below.

1. Although pricing at short-run marginal cost is economically efficient it usually does not provide enough revenue for the monopolist to cover total costs. In some cases the revenue-requirement of the monopolist can be met through the use of two-part pricing, charging a high fixed charge to users with a variable charge equal to marginal cost.
2. If the monopolist must recover enough revenue to cover costs, the best[84] outcome that can be achieved is to set the access price just high enough to cover the average costs of the firm. However, if the monopolist produces multiple products the average cost of the firm cannot be defined. Instead, the regulated price for each product should be somewhere above long-run average incremental costs and below stand-alone cost.[85]
3. According to Ramsey pricing, between these upper and lower bounds the price should be set higher on those products which have less elastic demands, in order to minimise the distortionary consequences of pricing above marginal cost. However, the information necessary to estimate elasticities of demand may not be readily available.
4. Where capacity constraints are likely to be a problem it may be necessary to explicitly ration capacity. One way that is suggested is through 'peak load' pricing models which suggest that the revenue necessary to cover costs should be recouped in the peak period, with price set at marginal cost in off-peak periods. This approach can be applicable in the energy sector. However, peak load markets do not exist or cannot be created in some industries.
5. In the case of two-way networks (such as interconnection of local networks in telecommunications or interconnection of collection-and-distribution networks in the NZ Postal Service) the problem is slightly different from the 'vertically-integrated natural monopoly' situation highlighted above. In these contexts, where each network is comparable in size, there is a 'bilateral monopoly'. The ability of one network to extract monopoly rents in interconnection is offset by the countervailing power of the other network. This tends to yield rules in which the required charges are 'symmetrical' or 'reciprocal'.[86]

Intuitively, although there is a vast range of more sophisticated potential pricing rules, among the set of simple 'linear'[87] interconnection pricing rules, almost all discussions focus on a particular range of prices. The top end of this range is given by the BW rule and the bottom end by long-run average incremental cost.[88]

The lower the price, i.e., the closer the interconnection price to the cost of producing the essential input[89], the greater the likelihood of entry and the greater the opportunity for competition to compete away the monopoly rents. However, given the limited information on the part of the regulator, the lower the price, the greater the risk of introducing inefficiencies.

Although, in general, such cost-based rules are more likely to achieve the objectives set out above, determining the appropriate pricing rule involves difficult and sophisticated economic analysis and will depend upon a range of factors specific to each industry and to each access pricing problem. As mentioned above, setting the 'wrong' access price may have significant adverse consequences. There simply do not exist 'bright line' rules to determine what constitutes an appropriate access price across all industries and situations. One economist[90] suggests the following list of factors that will influence the choice of access pricing rule:

1. the nature of the access facility itself, including its technology;
2. any industry-specific requirements which impinge on the ability to provide access, such as the requirements on the minimum flow through the facility;
3. the fiscal constraints on the government to fund the facility directly;
4. whether the facility is, or is likely to be, operating at capacity;
5. the likely requirements for facility expansion;
6. the ability to set access of varying qualities;
7. the predictability and stability of demand for access;
8. the nature of competition in the final goods/services market, including the presence or absence of the facility owner in that market and the regulatory constraints that exist in that market;
9. the information available about potential purchasers of access and the ability to price discriminate;
10. legal or other constraints on the use of tariffs other than marginal pricing;
11. the ease of bypass;
12. the costs of establishing a secondary market in access and the likely depth of that market;
13. the legal nature of the access pricing regime including any requirements for 'bargaining'; and
14. any differences in information available between the regulator and the access provider.

Questions for Consultees

6. Having regard to the list of factors in paragraph 214, which of the pricing rules listed below best achieves the objectives of efficiency for interconnection in order to provide (a) local telephone service; (b) long-distance telephone service; and (c) other telecommunications services, such as cellular?
   1. pricing at long-run average incremental cost;
   2. the BW or Efficient Component Pricing Rule; (or BW less monopoly profits); and
   3. (in the case of two-way networks) the rule of 'reciprocity' and related rules such as 'bill and keep'.
7. Having regard to the list of factors in paragraph 214, which of the pricing rules listed in paragraph 210 best achieves the objectives of efficiency for access to networks in other industries (such as electricity and gas)?
8. What other principles (e.g., principles relating to the technical specifications of interconnection, or unbundling of components) are necessary to achieve the objective of efficiency in the telecommunications sector? in other sectors?

Dealing With Government-Imposed Social Obligations

In part I of this paper, we identified uncertainty over the handling of Government-imposed social obligation costs (such as the Kiwi Share costs) as exacerbating the problem of uncertainty about the limits of legal behaviour more generally. This section explores options for handling these costs.

Objectives

Paragraph 187 identified that the regulatory regime should (a) promote economic efficiency; (b) be timely; and (c) have a high degree of predictability. Of these, the primary criterion for identifying which method of handling social obligation costs is preferred is economic efficiency. In relation to the social obligation costs, this can be achieved by choosing a method of handling the costs that ensures that the social obligation is met in such a way as to minimise the overall economic distortions created by the obligation. This will involve:

1. ensuring that the costs of the obligation are contributed to by all users of a natural monopoly facility in a way that does not distort competition between them; and
2. allocating the costs of the obligation on a basis which minimises the economic distortions created.

Options

As an aside, note that there are two types of social obligations: those that impose the same costs on all parties, and those that impose higher cost burdens on some specific firms than on others. The former might be called 'symmetric' cost obligations and the latter 'asymmetric' cost obligations. From the point of view of economic efficiency, only the latter are important because they may place one firm at a competitive disadvantage. This raises the risk of inefficient entry into the market and consequent losses in productive efficiency. In the presence of an asymmetric cost obligation, theory suggests that the competitor should compensate the monopolist for the (asymmetric part of the) costs incurred.

How much should the competitor compensate the firm meeting the obligation? The magnitude of the social obligation cost is typically only known to the firm meeting the obligation. That firm is likely to use this information advantage to extract as much revenue as possible from the competitor.

Steps can be taken to reduce this information advantage. For example, the firm could be required to submit information to a third party in order to obtain independent audit and verification of the costs of the social obligation.[91]

The problem of estimating the costs is complicated by the presence of joint or common costs. Any estimate of the social obligation costs will depend upon the methodology for sharing out the common costs of the firm. Furthermore, once these costs are estimated and verified, they must still be shared or 'allocated' amongst the firms in the industry.

The problems of estimating, verifying and allocating the social obligation costs can be completely avoided by the use of certain interconnection pricing rules (such as the BW rule) which permit the incumbent to preserve any existing mark-ups on its products and therefore any contributions towards the social obligations costs. The BW rule has the disadvantage however, of also permitting the incumbent to recover any monopoly profits.

In summary there are two main approaches for handling social obligation costs:

1. via interconnection pricing rules which do not require separate estimation and verification of the social obligation costs (such as the BW rule); or
2. via separate estimation and verification combined with some means of allocating the cost between competitors (whether in relation to the interconnection pricing or not).

The latter option requires the establishment of a methodology for estimation, a mechanism for verification, and a process for allocating the costs between the parties.

Regarding the methodology for estimating the social obligation costs, two possible principles that have been suggested are:

1. 'fully distributed costs (FDC)' (i.e., the common costs of the firm are allocated on some basis amongst the outputs of the firm, including those that provide the social obligation); and
2. 'avoidable incremental cost (AIC)' (i.e., the cost of the obligation is the difference between the forward-looking cost of providing the obligation in the least cost manner and the cost the firm would willingly incur in order to provide the service in the absence of the obligation[92]).

The first method has traditionally been used by regulators and is easier to implement from the company's financial accounts. Problems arise from the arbitrariness with which the common costs of the firm are allocated amongst the firm's products. As a result of this arbitrariness economists, including Professors Baumol and Kahn, reject the FDC methodology and advocate an avoidable incremental cost method. In the early 1990's, the OECD extensively reviewed the situation in telecommunications and held that the AIC method was the more appropriate method.[93]

Typically the FDC approach results in a higher cost estimate that an avoidable incremental cost approach. For example, in Australia, where one of the most comprehensive studies of a social obligation has been undertaken, Telecom Australia estimated their costs at A$850 million, while a subsequent study[94] placed the estimated annual cost at A$250 million.

Regarding the question of the basis for the allocation of the costs between the parties, economic theory suggests that these costs should be applied most heavily to those markets where demand is the most inelastic.[95] However, the measurement of elasticities of demand is difficult. It might only be possible to achieve a very approximate allocation, in line with rough estimates of the elasticities.

In the particular case of the telecommunications industry, there are various possible methods for allocating the Kiwi Share costs. For example, the Kiwi Share costs could be allocated on the basis of each competitor's pro rata share of:[96]

1. number of interconnecting minutes (i.e., number of minutes originating in the competitor's network and terminating in Telecom's network or vice versa, or both)
2. number of business lines;
3. number of business minutes;
4. total business revenue; or
5. total revenue.

Evaluation

The recovery of social obligation costs through the interconnection pricing rule (via the BW, rule say) overcomes significant difficulties involved in determining the methodology of estimating the social obligation costs, the verification of those costs and the allocation of the costs among the parties. However, as discussed above, the BW rule has certain disadvantages as an interconnection pricing rule which may outweigh the difficulties associated with separate identification of the social obligation costs.

Questions for Consultees

9. Which of the two following two options is more likely to achieve the objectives of (i) ensuring that the costs of the social obligation are contributed to by all users of a natural monopoly facility in a way that does not distort competition between them; and (ii) allocating the costs of the obligation on a basis which minimises the economic distortions created?
   1. interconnection pricing rules which do not require separate estimation and verification of the social obligation costs (such as the BW rule); or
   2. separate estimation and verification combined with some means of allocating the cost between competitors (whether in relation to the interconnection pricing or not);
10. Is there an economically efficient methodology for estimating social obligation costs? What are the advantages and drawbacks of the two methodologies ('fully distributed costs' and 'avoidable incremental costs') mentioned in the text?
11. Is there an economically efficient methodology for allocating social obligation costs among the competing networks? What methodology should be used for allocating the Kiwi Share costs among competitors?
12. How should the costs of the auditor be shared amongst the competitors?

The Gatekeeper

The previous sections have considered options for the design of the overall regulatory environment and specific options regarding the particular access pricing rule and the appropriate handling of social obligation costs. These choices together define a particular 'access pricing regime'.

If, after careful consideration of the above options, it is determined that the net benefits of invoking an access regime in the telecommunications industry exceeds the cost, it is envisaged that such a regime would be invoked immediately for interconnection disputes arising in that sector.

However, the same access issues as have arisen in the Telecom v Clear case could arise in other vertically-integrated natural monopoly industries. If a modification to the existing regulatory framework is made to improve the regime for interconnection negotiations in telecommunications, it may be appropriate for this new regime to be available for certain access disputes arising in other sectors with the vertically-integrated natural monopoly structure.

However, the regime is unlikely to appropriate for all access disputes. Therefore, some sort of 'gatekeeper' is required. The 'gatekeeper' would decide when and to what facilities the access pricing regime would apply.

What form should the 'gatekeeper' take? This is a question of the appropriate institutional arrangements.

Objectives

The appropriate objective is the same as that identified earlier for defining the regulatory environment. The preferred option is one which determines when the access pricing regime should apply in a manner that:

1. promotes economic efficiency in the relevant industry;
2. is timely; and
3. has a high degree of certainty and predictability.

In particular, for the last criterion to be met, the preferred option should seek to invoke the regime in such a way that the regime is in force:

1. for that facility;
2. for that period;
3. for that range of disputes; and
4. for those particular parties;
5. which the net economic benefits exceed the regulatory costs.

In addition, ideally entry into the market by bona fide competitors to the incumbent would be entirely unconstrained and the administrative cost of making such decisions and the opportunities for judicial review minimised.

Options

The above analysis suggests that a facility should not be made subject to an access pricing regime unless: (i) the facility is in a dominant position in its market; (ii) access to the facility is necessary in order to compete in an up- or downstream market; (iii) the facility owner also competes in that up- or downstream market; and (iv) duplication of the facility is not economically feasible.

The question of what constitutes 'dominance' for the purposes of a declaration is a difficult question and will require careful consideration of the nature of the market. As is well recognised in other competition law contexts, mere dominance in a particular narrowly-defined market is not, in itself, sufficient.[97] Similarly, the existence of a refusal to deal is not, in itself, evidence of monopoly power. The incumbent may, for example, point to the transactions costs that would need to be incurred to protect its quality of service from the effects of third party use; to foregone economies of scope; or to capacity constraints. The legitimacy of these business justifications needs to be examined and set against any genuine exclusionary effect.[98]

As a result, substantial judgment must be involved in determining whether the facility to which access is being sought is a natural monopoly and, if so, whether an access right should be provided. It is possible to identify principles that should be considered in taking these decisions, but they are far from constituting 'bright line' rules.[99]

The question of who or what should determine which facilities qualify for an access pricing regime, is a question of regulatory institution design. As before, the set of policy options is characterised by two dimensions: the 'regulatory institution' (the courts, an arbitrator, a statutory regulatory or statutory Ministerial powers) and the legislation, regulations or 'principles' to which the regulatory institution is required to have regard. These options are summarised in Table 4 of Appendix E.

The options identified in this document for imposing the access regime in accordance with broad statutory principles are as follows:

1. the courts, subject to the Commerce Act;
2. a statutory regulatory body (such as the Commerce Commission), subject to broad principles;
3. a statutory regulatory body (such as the Commerce Commission), subject to detailed principles;
4. the Government acting under statutory powers and subject to broad principles;
5. the Government acting under statutory powers and subject to detailed principles.

As an example of broad principles, it could be required that in consideration of whether or not to invoke an access pricing regime for a particular facility, the regulatory institution may have regard to whether or not:

1. the facility owner has a dominant position in a market;
2. the facility owner also competes (or may potentially compete) in the up- or downstream market which the firm seeking access wishes to enter (or continue to participate in);
3. it is economically infeasible for the competitor to provide the facility itself now or in the foreseeable future;
4. the facility has sufficient capacity to meet the demands of the competitor;
5. the parties have made proper efforts to secure a private agreement, but without success;
6. the likely economic benefits resulting from regulation of the access terms and conditions are likely to substantially outweigh the economic costs.[100]

   Factors to take into account when considering item (f) include:

   1. the magnitude of any monopoly profits being earned in the up- or downstream market by the facility owner and the likely development of substitutes;
   2. the availability of an access pricing rule that is:
      1. efficient; and
      2. easily implemented in the access pricing regulatory framework (e.g., can be applied by an arbitrator if necessary);
   3. the ability of the access pricing regulatory framework to efficiently and easily control other terms and conditions of access;
   4. the quality of the information needed to implement the pricing rule and to regulate the other terms and conditions of access;
   5. the potential for any undesirable side-effects or consequences that have arisen as a result of an access pricing regime in this or any other sector.[101]

In order to meet the objectives defined above, it is necessary that the regulatory regime also specify the scope of the disputes for which the access pricing regime will apply and possibly also the specific competitors for whom the access pricing regime will apply.[102]

In addition, in order to avoid the risk that the regime was invoked simply to reduce end-user charges, an invocation of the access regime would have to distinguish between competitors and customers of the monopolist. In the case of telecommunications, for example, the access regime might be restricted to those firms for whom a substantial portion of their business is the switching of telecommunications networks and for whom 95% of the telecommunications traffic is carried on behalf of third parties.

Evaluation

These options above are evaluated in Appendix D in accordance with the objectives listed in paragraphs 235-237.

Given the fact that invoking the access pricing regime may have sizeable distributional consequences, the regulatory institution will inevitably be subject to intensive lobbying and influence activities. Therefore it is particular desirable that the regulatory institution remain objective and not be unduly influenced by such lobbying.

The main conclusions are as follows:

1. Although the courts are relatively immune to influence activities, they may not have the expertise to consistently make accurate, predictable, efficient declarations.
2. Although governments have access to substantial amounts of expertise, they are more subject to influence activities. To a certain extent, this can be constrained by placing limiting principles in the enabling legislation and by widening the number of authorisations which must be obtained before a request for an access pricing regime is granted.
3. Statutory regulators lie between these extremes.

Whichever of the options above are chosen, it may not be possible to define the principles or the 'threshold' in such a way that the regime is invoked only for genuine access problems.

If the threshold that must be met before the access regulation regime is invoked is too high, consumers may be denied the benefits of competition. On the other hand, given the magnitude of the distributional issues involved, if the threshold is set too low, downstream competitors to the monopolist are likely to engage in prolonged and intensive lobbying and influence activities in an attempt to invoke the regime. Even putting aside the inefficiencies that are likely to arise from regulation itself, given the similarities in cost structures and grades of service that are likely to arise when common facilities are used, consumers could be deprived of real diversity in the market place.[103]

Even if it were possible to set the threshold accurately, it may not be possible to limit the exercise of discretion by the regulatory institution, which may not always be exercised in the greater public interest. Experience in other countries suggests that even where an access regime was established with the best of intentions, it is not always possible to limit its abuse.

Given the risks of delegating the authority to invoke the regime, it may be preferable not to grant that authority at all. In particular, it may be preferable to, for example, establish such a regime for the telecommunications industry only or, alternatively, not establish such a regime at all.

Questions for Consultees

13. Is it possible to delegate from the Government the authority to invoke an access pricing regime? Do the risks outweigh the benefits?
14. Which of the options set out below best meets the objective of promoting economic efficiency subject to timeliness, certainty and predictability, taking into account any possible regulatory costs? In particular, is the judgment about when to invoke an access pricing regime best made by the Crown?
    1. the courts subject to the Commerce Act;
    2. a statutory regulatory body subject to broad legislative principles;
    3. a statutory regulatory body subject to detailed legislative principles;
    4. the Government acting under statutory powers and subject to broad legislative principles;
    5. the Government acting under statutory powers and subject to detailed legislative principles.
15. Is it possible to define a threshold, for determining which disputes should have access to a new access regulation regime, that meets the objectives set out in paragraphs 235-237? Do the principles set out in paragraphs 243-244 meet these objectives? If not, what principles might define such a threshold?
16. Is it necessary to distinguish formally between bona fide downstream competitors and other end-users or customers in the telecommunications industry for the purposes of determining access to a new access regulation regime? Does the suggestion in paragraph 246 satisfactorily make this distinction?