3 Interconnection(1)

��

��

��

��

��

��

��

��!��

��"�� #!��$��

�� !��%��

�� &��

'��(�� %�� )��

�� %��!��

��!��"#�$%��&� "��'()(*�*�'*'�

��

��

"��+��%��,��!!��%��,��-��.�/��

�� !��"��# ��$ %�� $��

��

& '��"�� &��( �)��*��!�� (��! +��# !��, �� ,��

��

��

�-� ��"�� !�� "

�-� �� .�� #� �$�� !�� %�� !�� &�� %�� '�� (�� #)�!� �� *�+��,��

�-� /�� !�� (�� !�� -�� .� ��$��%� .��/� ��

�-� ��)��"��!�� 0�� $*��$�� !�� 1�$��,��,�%� .��/� �� '�� 2��2��1�$��,�� ,� �� '��' #3��!�� -��- 4�� *��0�� 5�� " 4�� *�� ,��0��

��

(�0��6�)��7�8 ��9�*�� 6�)��7�8 �� :�5�� 6�)��7�8 ��*��:�,��*�!�� ,�� -6�)��7�8 ��#�� -6�)��7�8 #)��*�;��$�� '6�)��7�8 ��*�%� �� 19 �-6�)��7�8 ��$��1�$��,��,�%� .��/� �� ,��0�� "

/��;��7�8;��1�$��,��<�2��2�� ;��7�8;��1�$��,��<��.��2��2�� ;��7�8��,�1�� 2��1�� ;��7�8��0��=��7��,�6� �� !��

��1��$��7�8 �� *�� !�� "��$��7�8 !�� 0�� (�� $��7�8 ��!�� $��7�8 !,0�� ,�(��,0�� 1�$��,�� $��7�8 !�� ;��,�!�� 2��2��1�$��,�� $��7�8 � �� ,� �� "��$��7�8 ��9�*�� 4�� *��0��

��

��

��

��!��"��

#��$��$��%��#�0��$�,��

��,/��1��!#��%2

��-$$/��-��3�12��/��1114$--��24-�

��/��,/��2��-��1��/!!��-��!#��!�!��,��%�

%2��&��-$$/��-��5&��6

��$�/��,��%��#��%��-��-��2��1114��,��#4�7!8�-��7��/��%�

3 - 1

MODULE 3

INTERCONNECTION

3.1 Interconnection Principles

3.1.1 The Importance of Interconnection

Interconnection of telecommunications networks hasbeen important for a century, but never more so thantoday. Originally, operators, such as PTTs and theNorth American Bell companies, interconnected withneighbouring operators. However, these operatorsretained monopolies over all networks andequipment in their geographic serving areas. Fordecades, few other types of interconnectionoccurred.

Beginning in the 1970s, customers began tointerconnect a growing range of terminal equipmentand private network facilities to the incumbentoperator’s facilities. With the liberalization oftelecommunications markets over the last fewdecades, effective interconnection arrangementshave become key to the operations of anincreasingly wide range of services. These servicesinclude local, long distance and international fixed,mobile and satellite services, providing everythingfrom basic voice telephony to high speed Internetconnectivity to Internet multimedia services.

Competition is the key to the growth and innovationof today’s telecommunications markets. Intercon-nection is a critical factor for the viability of

competition. For most of the history oftelecommunications, operators and governmentadministrations negotiated with each other to set theterms of interconnection without regulatory interven-tion. The emergence of competition has changedthis. Incumbent operators have little incentive tomake things easy for their new competitors, andmost of the bargaining power in negotiations lieswith the incumbents.

Strategic anti-competitive behaviour on interconnec-tion matters by incumbents has retarded orprevented competition in many telecommunicationsmarkets around the world. Incumbents can engagein a wide range of behaviour to frustrate effectivecompetition. For example, they can charge exces-sive rates for interconnection, refuse to build ormake available adequate interconnection capacity,and refuse to unbundle network elements orservices necessary for efficient interconnection. Newentrants in telecommunications markets have little tooffer in negotiations to remove these barriers tocompetition. Today, there is a consensus amongtelecommunications experts and policy makers thatdecisive and informed guidance by regulators is re-quired to pave the way for effective interconnectionarrangements.

Interconnection is an important consumer issue.Telecommunications users cannot communicate

Telecommunications Regulation Handbook

3 - 2

with each other or connect with services theydemand unless necessary interconnection arrange-ments are in place. Interconnection of a multitude ofdifferent types of networks has brought tremendousbenefits to consumers and businesses around theworld in the last decade. Without efficient intercon-nection arrangements, services such as directinternational dialing, all Internet-delivered services,automated teller machines and e-commerce wouldnot be possible.

Increasing network interconnection will continue toimprove the convenience and utility of telecommuni-cations service for users around the world in thenext decade. Inadequate interconnection arrange-ments not only impose unnecessary costs andtechnical problems on operators - they also result indelays, inconvenience and additional costs forbusinesses, consumers and, ultimately, for nationaleconomies.

According to ITU’ surveys, Interconnection-relatedissues are ranked by many countries as the singlemost important problem in the development of acompetitive marketplace for telecommunicationsservices, interconnection has been a highlycontentious issue in Europe. Almost half of allcountries in the Asia-Pacific region indicated thatinterconnection issues were a top regulatory priority.While fewer countries in the Arab states (20%) andthe Americas (30%) pointed to interconnection as aregulatory priority, the general level of networkcompetition was still low in those regions. That ischanging. The importance of interconnection issueswill increase in all regions as network competitiondevelops.

This Module examines the arrangements that mustbe put in place between operators, and the stepsthat can be taken by regulators, to facilitate effectiveinterconnection.

3.1.2 Scope of Interconnection Issues

Interconnection is defined in different ways in thedifferent regulatory and policy regimes that deal withit. A good recent definition is included in the 12 July2000 proposed European Commission Directive onaccess and interconnection:

“interconnection” means the physical and logicallinking of public electronic communicationsnetworks used by the same or a differentundertaking in order to allow the users of oneundertaking to communicate with the users of thesame or another undertaking, or to accessservices provided by another undertaking.Services may be provided by the parties involvedor other parties who have access to the network.(Article 2 – CEC(2000d))

This definition differs from others in that it includesinterconnection of networks used by the sameundertaking and not just networks of differentoperators. The proposed Directive also differs fromsome other regulatory interconnection regimes inthat it includes a separate concept of “access”,defined differently from interconnection:

“access” means the making available of facilitiesand/or services, to another undertaking, underdefined conditions, on either an exclusive or non-exclusive basis, for the purpose of providingelectronic communications services. It coversinter alia:

➢ access to network elements and associatedfacilities and services, which may involve theconnection of equipment by wire or wirelessmeans;

➢ access to physical infrastructure includingbuildings, ducts and masts;

➢ access to software systems, including opera-tional support systems;

➢ access to number translation or systemsoffering equivalent functionality;

➢ access to mobile networks, in particular forroaming; and

➢ access to conditional access systems fordigital television services.

Interconnection is a specific type of accessimplemented between public network operators.Access in this Directive does not refer to accessby end-users.

Module 3 - Interconnection

3 - 3

Interconnection

The last sentence of the definition is important. Itdistinguishes the Commission’s use of the term“access” from its normal meaning, which relates toend-user access, for example in the terms “accesslines” or “network access service”. Despite this po-tential confusion, the types of inter-operator “access”listed in the Commission’s definition are very impor-tant in the context of interconnection.

The types of “inter-operator access” listed in theCommission’s definition are treated as an integralpart of “full” or “efficient” interconnection in otherjurisdictions. They may also be considered as“supplemental” or “ancillary” forms of interconnec-tion. These types of access arrangements aretypically addressed in interconnection agreementsentered into between experienced operators.

Whatever the regional or local definition of intercon-nection, the matters included in the Commission’sproposed definition of “access” must be dealt with aspart of a comprehensive approach to interconnec-tion. In this Handbook, therefore, we will deal withthis type of “inter-operator access” in detail, as anintegral part of full interconnection.

3.1.3 Interconnection Issues

Commercial, technical and operational arrange-ments must be made to facilitate interconnectionbetween network operators. A number of issuesmust be agreed upon by the operators, or deter-mined by the regulator, in order to finalize thesearrangements.

The major commercial issues of concern to newentrants are generally related to the cost of intercon-nection. In North America and Europe, for example,up to 50% or more of the total costs of some long-distance operators have been paid out in intercon-nection charges to local operators. Suchinterconnection charges are particularly significantfor operators that rely heavily on resale or that mustpay a subsidy or contribution component as part ofinterconnection charges. The practice of combiningsubsidies and cost-based charges is widelydiscouraged, for the reasons set out in Section.3.3.5.4. Even without a subsidy component, the levelof interconnection charges is often an importantfactor in determining the financial viability of a newtelecommunications service provider.

Interconnection costs are certainly not the onlymajor issue. Various technical and operationalissues are also critical to both incumbent and newoperators. Box 3-1 lists some of the most importantinterconnection issues encountered in manycountries.

3.1.4 Regional Interconnection Rules

In recent years, the development of regional tradingareas and the implementation of multilateral tradeagreements has accelerated the liberalization ofinterconnection policies.

A leading example is the 1997 European Intercon-nection Directive (97/33/EC). It contains rulesspecifically aimed at liberalizing nationalinterconnection regimes. The Directive requiresinterconnection arrangements to be public and non-discriminatory. It also requires interconnectioncharges to be cost-based. Related EU Directivessupplement and amend the European interconnec-tion regulatory framework. These Directives includeobligations on special access (98/10/EC) andprovision of leased transmission capacity(92/44/EC).

The provisions of the European Directives related tointerconnection are fairly general in nature. Thisapproach permits adaptation to the EU’s differentnational legal regimes and regulatory frameworks.The European Commission has taken additionalsteps, beyond the Directives, to improve intercon-nection arrangements. One such step is thepublication of “best current practice” interconnectionrates. These interconnection rates are significantlylower than those of some member countries, sug-gesting that these countries should take action tomeet international cost benchmarks. Another majorstep was the recent adoption of rules and aproposed regulation to require unbundling of thelocal loop. These rules are discussed later in thisModule.

The European Commission has also reviewed itsinterconnection-related Directives. As previouslyindicated, on 12 July 2000, the Commissionpublished a proposed new Directive on access to,and interconnection of, electronic communicationsnetworks and associated facilities (COM(2000) 384).The proposed new Directive seeks to respond to the


3 - 4

convergence phenomenon by covering a broaderrange of electronic communications networks andservices. It also contains some new and differentprinciples. However, under the proposed newDirective, the key provisions of the three previous(above-noted) Directives will continue to be legallybinding on European Union Member States,pending further reviews.

Other multilateral organizations have also developedinterconnection guidelines. For example, the Asia-Pacific Economic Co-ordination (APEC) Telecom-munications Working Group has developed aFramework for Interconnection. Unlike the EU

Box 3-1: Some Key Interconnection Issues

Framework and Procedural Issues

➢ Adequacy of regulatory guidance for interconnection negotiations

➢ Availability of interconnection with incumbent operators for various types of services

➢ Access to standard interconnection terms with incumbent operator

➢ Independent and timely dispute resolution

➢ Non-discriminatory access to interconnection facilities and services

➢ Access to PSTN network specifications (including planned network changes)

➢ Treatment of Universal Service, Universal Access or Access Deficit Charges

Commercial Issues

➢ Level and structure of interconnection charges; basis for calculation (i.e. type of costs used tocalculate charges, revenue sharing, bill and keep, etc.)

➢ Unbundling of interconnection charges for different network components and related services

➢ Resale of network facilities and services

➢ Payment for network modifications to facilitate interconnection

➢ Confidential treatment of competitive and customer information

Technical and Operational Issues

➢ Open network standards and technical compatibility

➢ Location of Points of Interconnection (POI)

➢ Access to signaling systems, advanced digital features, billing system, operations support systems(OSS), call-related databases and other software to provide advanced services

➢ Access to unbundled network components, including local loops

➢ Equal ease of customer access to competitive networks (e.g. customer dialing parity)

➢ Access to numbers and implementation of number portability

➢ Collocation and sharing of infrastructure (e.g. buildings, poles, conduits, ducts, towers)

➢ Quality of interconnection, including availability of sufficient interconnection capacity to avoidcongestion, and to ensure the timely provisioning of interconnection services and facilities


3 - 5

Interconnection

approach, this framework is not binding on APECmembers. The APEC framework is intended toprovide principles, examples of interconnectionapproaches in APEC economies, and other usefulinformation to assist in the development of nationalinterconnection policies. Similarly non-bindingapproaches have been taken in interconnectionprinciples published by other regional organizations,such as CITEL in Latin America.

3.1.5 Multilateral Interconnection Rules

The 1997 WTO Agreement on BasicTelecommunications (formally known as the FourthProtocol of the General Agreement on Trade inServices or GATS) was the first widely acceptedmultilateral trade agreement to include bindinginterconnection rules. These rules were included inthe so-called Reference Paper, an informal textcontaining regulatory principles negotiated amongWTO Members. The Reference Paper becamelegally binding on WTO Members that attached it aspart of their “additional commitments" in their GATSSchedule of Commitments on telecommunicationsmarket access. The Reference Paper was attachedin whole or with minor modifications by 57 of the 69signatories to the Fourth Protocol. Six additionalsignatories elected to list some of the principles intheir Schedules, but not the entire document.

All WTO Members have the option of undertakingthe obligations of the Reference Paper in theirGATS Schedules on interconnection or othermatters, whether or not they participated in theFourth Protocol. As of late 1999, a total of 64 WTOMember governments had committed to theinterconnection obligations of the Reference Paper.This increase from 57 was due to the submission ofcommitments by seven more countries since theFourth Protocol. Of these, four WTO Membersattached the Reference Paper to telecommunica-tions commitments they made after the Protocolnegotiations ended and three countries attached it tothe GATS Schedules they filed upon accession tothe WTO. Most of the nearly 30 additional countriesseeking accession to WTO are expected to alsocommit to the Reference Paper and its interconnec-tion obligations.

The most important interconnection-related rules setout in the WTO Regulation Reference Paper are

summarized in Box 3-2. The full text of theReference Paper provides more detail than the box.

The paper’s central principles are non-discrimina-tion, transparency, and the availability of reasonableinterconnection terms, including cost-oriented ratesand unbundled access, from "major suppliers". Theconcept of "major suppliers" in the Reference Papercan generally be assumed to refer to operators witha dominant position vis-à-vis essential infrastructureor market share. Thus, at present, the Paper’sinterconnection disciplines would most commonlyapply to monopoly or former monopoly fixed-lineoperators.

The Reference Paper was designed as a set ofgeneral rules or principles to be observed, ratherthan as detailed prescriptive guidelines on how theprinciples are to be implemented. This approachmakes the paper adaptable as telecommunicationsmarkets evolve, and provides flexibility for applica-tion to different legal systems and regulatory inter-connection frameworks.

Box 3-2: Interconnection Rules of WTORegulation Reference Paper

Interconnection With “Major Suppliers” must beassured:

➢ At any technically feasible point in thenetworks

➢ In a timely fashion

➢ On non-discriminatory and transparentterms (including quality and rates)

➢ Sufficiently unbundled to avoid charges forunnecessary components

➢ At non-traditional interconnection points ifrequestor pays charges

Procedures

➢ Procedures for interconnection to majorsuppliers must be made public

Transparency

➢ Agreements or model interconnection offerof major supplier must be made public


3 - 6

As a practical matter, therefore, more detailedguidance is essential to turn the general ReferencePaper principles into workable interconnectionarrangements, agreements, national regulations orregulatory directives. The experience of other coun-tries can provide valuable precedents in this regard.

When the GATS Agreement on BasicTelecommunications came into effect on 15February 1998, many signatory countries did not yethave detailed interconnection rules in place. Somestill do not. Given the general nature of theReference Paper principles, it will be a challenge formany countries to develop sufficiently detailedinterconnection regimes to put “flesh on the bones”of their GATS obligations.

Before examining the details of interconnectionarrangements, the following sections of this Modulewill review the basic principles underlying most inter-connection rules.

3.1.6 Interconnection Principles

3.1.6.1 Providing Advance RegulatoryGuidelines

There continues to be a regulatory debate about therelative advantages of providing ex ante or advanceinterconnection guidelines versus ex post regulation.Proponents of the ex post approach generally favournegotiation of interconnection agreements betweenoperators, with recourse to regulatory dispute reso-lution or competition law remedies, if negotiationsfail.

Several years ago, there were more advocates ofthe ex post approach, particularly outside of NorthAmerica, than there are today. This approach wasbased on the belief that regulation should be mini-mized in competitive markets. Many regulatorsrecognized that the financial, technical andoperational details of interconnection arrangementscould be complex. They considered that incumbentoperators and new entrants would generally have amuch better understanding of these arrangementsthan regulators. They were also concerned thatinappropriate regulatory intervention in interconnec-tion matters could impose high costs on the sector.

For these reasons, a large number of regulators andtelecommunications experts promoted industrynegotiation as the main approach for developinginterconnection arrangements. Ex ante regulatoryintervention was discouraged. The focus of regula-tory attention was on dispute resolution, in the eventindustry negotiations broke down.

In recent years, there have been increasing doubtsabout the effectiveness of the ex post approach.There appears to be a growing consensus thatadvance regulatory guidelines, or even specific inter-connection rules, are necessary to facilitatesuccessful negotiations. This view has beenexpressed recently by the European Commission, inits 12 July 2000 proposed Directive on access andinterconnection. The Commission stated:

“…there is a consensus that ex-ante sectorspecific rules will continue to be neededalongside competition rules to regulate accessand interconnection, until such time as there isfull and effective competition in all segments ofthe market.” (CEC (2000c))

This view has long been held by regulators andpolicy-makers on the other side of the Atlantic.During the 1980s and 1990s, US and Canadianregulators issued a series of detailed guidelines anddecisions on most aspects of interconnection withdominant operators, including interconnection ratesand technical terms and conditions. The more inter-ventionist approach of the North Americanregulators appears to have led to more unbundlingof network services, more competition, and arguablymore service innovation and growth.

The issues of negotiating interconnection arrange-ments and approaches to regulatory intervention arediscussed in detail in Section 3.2.2 of this Module.

3.1.6.2 Focus Interconnection Obligationson the Incumbent Operator

One generally accepted means of minimizing regu-latory intervention is to limit imposition ofinterconnection obligations to dominant incumbents.In practice, this is the most effective and efficientmeans of utilizing limited regulatory resources.


3 - 7

Interconnection

This approach is sometimes subject to criticism byincumbent operators. They argue that this approachamounts to regulatory “handicapping” and construc-tion of “non-level playing fields”. Others suggest thatuniversal imposition of interconnection obligationswould provide more interconnection opportunities forall operators.

However, this is a minority view. The consensusview is that universal imposition of interconnectionobligations on all operators, large and small,generally amounts to over-regulation. In principle,only firms with a dominant market position have theability to establish interconnection termsindependently of competition. Non-dominant com-petitors would find it difficult to independentlymaintain excessive interconnection rates, ordiscriminatory conditions. Other service providerswishing to interconnect could avoid such unfavour-able interconnection arrangements by interconnect-ing with a competitor, including the dominantsupplier. Over time, as markets become increasinglycompetitive, it may be possible to deregulate moreinterconnection arrangements, including those ofonce-dominant operators. However, in the transitionperiod to full competition, a degree of asymmetricregulation is required in order to level a playing fieldthat is tilted in favour of incumbents.

For these reasons, the regulatory approach to inter-connection in this Module focuses on interconnec-tion arrangements with dominant incumbentoperators.

This approach is consistent with the ReferencePaper of the WTO Agreement on BasicTelecommunications, which only imposesinterconnection obligations on dominant operators(i.e. “major suppliers”). It is also consistent with theEuropean Commission’s 12 July 2000 proposedDirective on access and interconnection. Theproposed Directive aims to expand the scope of itsinterconnection framework to a wider range ofelectronic communications networks. However, onlydominant operators will be subject to the ex anteregulatory obligations proposed by the Commission,such as mandatory interconnection, resale,collocation, etc.

3.1.6.3 Transparency

Transparency is a major policy objective of multilat-eral trade agreements as well as the nationaltelecommunications policies of many countries.While there is a lot to be said for protecting theconfidentiality of business agreements in a competi-tive marketplace, interconnection with dominantincumbents is generally considered an exception.

Confidential treatment of interconnection arrange-ments would provide incumbents with an opportunityto act strategically to thwart competitors. Forexample, such operators could enter into confiden-tial interconnection agreements that provideunfavourable interconnection arrangements withcompetitors, and more favourable ones withaffiliates. Dominant operators could also limit thefunctionality of the types of interconnection offered,levy excessively high charges, and otherwise actstrategically to limit competition.

Transparency of interconnection arrangements is aneffective means of discouraging anti-competitivestrategic behaviour by dominant operators. It iseasier for regulators to detect and remedy suchbehaviour if interconnection arrangements are madepublic. Publication of agreements also makes iteasier for regulators and all industry participants tocompare interconnection rates, terms and condi-tions. Transparency also assists in developingindustry standards and benchmarks, as well as bestpractices on operational and administrative issues.

Many countries require publication of referenceinterconnection offers or model interconnectionagreements. To further promote transparency, someregulators maintain public registries of interconnec-tion agreements, or require publication of agree-ments by operators. In some cases, interconnectionagreements are available over the Internet.

Where interconnection agreements are made public,various mechanisms can be used to protectconfidential commercial information. For example,Indian legislation requires the regulator to maintain aregistry of interconnection agreements. However, atthe request of parties, the regulator may direct thatparts of an agreement be placed in a confidentialportion of the registry. In such cases, a summary of


3 - 8

the confidential parts must be made publicly avail-able.

3.1.6.4 Non-Discrimination

Avoidance of discrimination is a central objective ofmost interconnection policies. Discrimination ininterconnection arrangements can take severalforms. One form involves discrimination by a domi-nant operator in interconnection arrangementsentered into with several different new competitors.For example, new entrant B may obtain betterarrangements than new entrant C. Such discrimina-tion is relatively easy to detect if interconnectionagreements are public.

It should be noted that interconnection arrange-ments may vary from one competitor to anotherwithout being “unduly” or “unjustly” discriminatory.The two competitors may have voluntarily agreed todifferent arrangements, for example, to suit theirdifferent operating conditions. The real test, there-fore, should not be “discrimination” in the sense of“differences” in interconnection arrangements. Thetest should be “unjust”, “undue” or “unfair” discrimi-nation, in the sense that an interconnectingcompetitor is placed at a significant disadvantage asa result of less favourable interconnection arrange-ments.

The other major form of discrimination is oftenharder to identify. It involves the provision of morefavourable interconnection arrangements by adominant firm to its own operations or its affiliatesthan to competitors. Disputes or complaints aboutthis form of discrimination are often difficult forregulators to resolve. For example, it is sometimesimpossible to grant a competitor exactly the sametype of interconnection arrangements as it is possi-ble to provide to an internal operation.

Various approaches have been developed to identifyand resolve cases of discrimination of the secondtype. Since interconnection arrangements need notbe identical, the objective of preventing undue dis-crimination has been described as one ofdeveloping “comparably efficient” interconnectionarrangements.

Some incumbents discriminate against competitorsby treating them as “customers” rather than “peers”

or “co-carriers”. This approach often leads to higherprices and inferior interconnection arrangements.Regulators should generally insist that intercon-necting carriers should be treated on an equal andreciprocal basis, as peers and not customers.

One type of discrimination can be fatal to theprospects of competition. It involves providinginsufficient network capacity to interconnectingoperators, as compared to an incumbent’s ownservices. Network congestion can be a deadly anti-competitive barrier. Regulators must sometimesintervene to ensure non-discriminatory rationing ofnetwork access and transport facilities. They mustoften also ensure that established PSTN operatorsconstruct sufficient capacity to handle growingdemand that can be expected in a competitive tele-communications market.

One regulatory approach to reduce, or at least assistin the identification of, discrimination between adominant firm and its competitors involves the es-tablishment of structural or accounting separationsor divestiture. Under structural separationapproaches, a dominant firm is required to move itscompetitive operations into a separate affiliatedcompany, with separate management, accountingrecords, etc. Divestiture involves selling all or part ofthe separate affiliate to other persons. Accountingseparations involve setting up separate accountingrecords only, and not actually requiring the estab-lishment of a separate legal entity for the competitivebusiness. These approaches are discussed inSection 5.3.3 of Module 5 – Competition Policy.

Another less interventionist approach that iscommonly used by regulators and competitionauthorities to prevent undue price discrimination bya dominant firm is an “imputation approach”. Suchan approach is applied to vertically integratedsuppliers. Such suppliers include operators that pro-vide a retail service, like local telephone accessservice, on a competitive basis, and also provide awholesale service, like international telephoneservice, on a monopoly basis to itself and othercompetitors.

Under an imputation test, a vertically integratedsupplier would be required to include the sameamount it charges to its competitors for internationalservice in its own retail rates, and to add an amount


3 - 9

Interconnection

sufficient to cover its additional costs of providinglocal services. Imputation tests are discussed underthe heading Vertical Price Squeezing in Section5.3.4. of Module 5.

3.1.6.5 Cost Orientation

Interconnection principles, such as those set out inthe Reference Paper for the WTO’s Agreement onBasic Telecommunications and the EuropeanUnion’s Interconnection Directive, require intercon-nection charges to be “cost-oriented”.

There are various reasons for specifying that inter-connection charges should approximate costs.Without a cost-based standard for setting intercon-nection charges, an established monopolist ordominant operator would have an incentive todemand a high price for terminating calls thatoriginate on a new competitor’s network. Similarly, adominant operator would have an incentive to paylittle or nothing to the competitor to terminate calls

originating on the dominant operator’s network. Inthe absence of regulatory intervention, some newcompetitors might have little choice but to acceptsuch a deal or remain unable to interconnect.

Serious problems can result from a dominant firmcharging competitors interconnection prices that aresignificantly above cost. First, it deters market entryand the development of competition. Second,customers of the competitors will ultimately have topay for these excessive charges. Third, the exces-sive prices can provide a pool of revenues that thedominant firm can use to subsidize losses, forexample losses incurred as a result of predatorypricing action taken by the dominant firm to drivecompetitors out of a market.

The approaches used by telecommunicationseconomists and regulators to calculate interconnec-tion costs, and telecommunications costs generally,are discussed in Section 3.3 of this Module, inModule 4 and in Appendix B of the Handbook.

Box 3-3: Summary of Widely Accepted Interconnection Principles

➢ Terms of interconnection should not discriminate unduly between operators or between a dominantfirm’s own operations and those of interconnecting competitors

➢ Interconnection should be permitted at any technically feasible point, but the requesting operatorshould pay any additional costs of non-standard interconnection

➢ Interconnection charges should generally be cost-based (i.e. the evolving best practice specifiesthat the cost standard should be forward-looking long-run incremental costs; there is normally amark-up to cover forward-looking joint and common costs)

➢ Cost inefficiencies of incumbent operators should not be passed on through charges tointerconnecting operators

➢ Where reciprocal interconnection and costs can be expected to be reasonably balanced, bill andkeep arrangements are an efficient alternative to cost-based interconnection

➢ Regulatory guidelines and procedures should be prescribed in advance, to facilitate interconnectionnegotiations between operators

➢ Standard terms and procedures should be published for interconnection to dominant operators➢ Interconnection procedures and arrangements should be transparent➢ Interconnection arrangements should encourage efficient and sustainable competition➢ Network elements should be unbundled, and charged separately➢ Charges related to universal service obligations should be identified separately, and not bundled

with interconnection charges➢ An independent regulator (or other third party) should resolve interconnection disputes quickly and

fairly


3 - 10

3.1.6.6 Other Interconnection Principles

A number of other interconnection principles havebeen proposed and adopted by regulators, policymakers and trade organizations. In many cases,these are variations on the same themes. Box 3-3summarizes widely accepted interconnection princi-ples.

3.1.7 Contents of InterconnectionAgreements

The contents of interconnection agreements varyconsiderably. Much depends on the regulatory

framework. If the existing regulatory frameworkprovides sufficient detail on the terms and conditionsof interconnection, then interconnection agreementscan be shorter. The same is true if an incumbentoperator, or an industry group, has publisheddetailed interconnection tariffs, technical standards,procedures, etc. which can be incorporated into anagreement. In other cases, interconnection agree-ments must be more comprehensive.

Bearing these variations in mind, Table 3-1 providesa list of the possible contents of a “typical” intercon-nection agreement.

Table 3-1: Contents of a Typical Interconnection Agreement

Contents Detail and Comments

Interpretation

Recitals ➢ “Whereas” clauses add historical and legal context to assistunderstanding by future readers of agreements

Definition of Key Terms ➢ Terminology varies significantly among different countries andoperators

➢ It is important to ensure compatibility of terminology to the localenvironment when adapting interconnection agreements from othercountries

➢ Definitions in other documents may be referenced, e.g. definitions inlaws or regulations, regulatory guidelines, ITU definitions

Scope of Interconnection

Description of Scope andPurpose of Interconnection

➢ Different types of interconnection agreements have differentpurposes (e.g. two local networks, local to longdistance/international, fixed-to-mobile, mobile-to-mobile, local ISP toISP backbone)

➢ The purpose of some interconnection agreements is to providetermination services or transit services; others involve provision ofunbundled facilities, etc.

➢ Interconnection architecture (annotated diagrams)

Points of Interconnection and Interconnection Facilities

Points of Interconnection(POI) and Related FacilitySpecifications

➢ POI locations (e.g. exchanges, meet points) usually listed in anappendix; may be modified from time to time. Typically includesexchange types and street addresses

➢ Specific POI facility locations (e.g. digital distribution frame; manholesplice box)


3 - 11

Interconnection

Table 3-1: Contents of a Typical Interconnection Agreement (cont’d)

➢ Description of network facilities to be interconnected (e.g. OC-3 fibreoptic terminals with interconnecting single-mode optical fibres)

➢ Specify capacity and/or traffic volume requirements

➢ Indicate which party is to provide which facilities (include diagram ofPOIs and interconnected facilities)

➢ Technical specifications, for example:

➢ Calling Line Identification (CLI) specs

➢ Other advanced digital feature specs, e.g. call forwarding, callername ID, etc.

➢ Basic and ISDN call control interface specs

➢ Local Number Portability (LNP) query-response network specs

Signaling Interconnection ➢ Specify type of signaling networks/standards (e.g. CCS7)

➢ Signaling POIs locations to be specified (i.e. Signal Transfer Pointsor STPs)

➢ Point Codes to be specified

➢ Technical interface specifications (e.g. signaling links to be dedicatedE-1 or DS-1 transmission facilities; operating at 56 kbps)

➢ Diagram of signaling interconnection architecture

Network and Facility Changes

Planning and Forecasts ➢ Requirement for mutual notification of network changes and capacityforecasts, for example:

➢ traffic forecasts for each POI

➢ local number and portability requirements

➢ area code saturation and changes to increased digit phonenumbers

➢ default and redundant routing arrangements

➢ Periodic network planning reports may be specified

Facility Ordering Procedures ➢ Specify rights and obligations of each party with respect to orderingand provisioning of interconnection facilities (including unbundlednetwork elements – see below).

➢ Confidentiality requirements and procedures to ensure same

➢ Ensure no anti-competitive use of order information (e.g. no contactswith end users; competitive service divisions of operator receivingorders)

➢ Specify points of contact (e.g. Interconnection Service Groups; E-mail addresses, etc.)

➢ Specify order format and procedures (e.g. standard order forms maybe utilized in paper or electronic (EDI) format)


3 - 12


➢ Procedures to expedite specific orders

➢ Co-ordination process for migration of customers between operators(e.g. coordination of cut-overs to prevent or minimize serviceinterruptions to end users)

➢ Procedures for ordering operator to arrange for all equipmentinstallations and changes at end-user premises

➢ Order confirmation and order rejection procedures, timelynotification, notification of additional charges, etc.

➢ Order completion notification and reporting requirements

Traffic Measurement and Routing

Traffic MeasurementResponsibilities andProcedures

➢ Describe party responsible; measurement and reporting procedures(see billing procedures below):

➢ Rules for routing of different types of traffic, if any (e.g. Bill and Keeplocal traffic that is to be terminated reciprocally without charge maybe carried on “Bill and Keep” trunks; traffic to which terminationcharges apply may be carried on other trunks, e.g. transit trunks,national traffic trunks, etc.)

Infrastructure Sharing and Collocation

Sharing of Infrastructure,Procedures and Costs

➢ Availability of poles, conduits, towers, rights of way, etc.

➢ Procedures, if any, for determining available capacity; procedures forallocating capacity among requesting operators (e.g. first come/firstserved)

➢ Prices and/or costing method

➢ Provision and pricing of supplementary services (electrical power,security systems, maintenance and repairs, etc.)

➢ Sub-licences on property of third parties (e.g. right of way owners,municipal and other public and private property owners, whereinfrastructure is located), insurance and indemnification for damages

Collocation ➢ Availability of actual or virtual collocation (e.g. for transmissionfacilities on exchange premises); list of addresses where collocationis available; procedures for determining available space; reservationof expansion space

➢ Prices and/or costing method for collocated space

➢ Provision and pricing of supplementary services (e.g. electricalpower and emergency backup power, lighting, heating and airconditioning, security and alarm systems, maintenance and janitorialservices, etc.)

➢ Procedures for ensuring access to and security of collocatedfacilities (notification; supervised repair and provisioning work and/orseparated premises, etc.)


3 - 13

Interconnection


➢ Negotiation of other lease and/or licence arrangements, includingissues of sub-licences on property of third parties (e.g. buildingowners, right of way owners, municipal and other public propertyowners), insurance and indemnification for damages

Billing

Scope of BillingArrangements andResponsibilities

➢ May include different arrangements, for example:

➢ Operators billing each other for interconnection services (e.g.termination) and facilities (e.g. unbundled loops and other networkelements)

➢ Performance of billing functions by some operators for others (e.g.local operators billing end-users for long distance or internationaloperators, ISPs, etc.)

Billing Procedures ➢ Interconnection billing media – discs, tapes, paper and/or electronic(EDI) transfers; format and software specifications

➢ Guidelines for production of interconnection billing outputs, including:

➢ Applicable industry standards (e.g. CABS, BOS, SECABS, usedwith or without modifications)

➢ Billing data format and data elements

➢ Standardized codes and phrases

➢ Billing schedule

➢ Customer Service Record (CSR) provision, including:

➢ details to be supplied by provisioning local operator (e.g. record ofinterconnection elements used, including circuit and other (e.g.DSLAM) equipment identification numbers)

➢ media (e.g. tape, paper, etc.) and schedule for delivery

➢ other requirements to facilitate efficient verification and billing ofend-user by non-provisioning operator

➢ Retention periods for billing data

Payment Terms andConditions

➢ Billing fees and related charges.

➢ Payment terms and conditions, including late payment penalties;service disruption credits, etc.

Billing Disputes andReconciliation Procedures

➢ Contact details for reconciliation and billing queries

➢ Responsibilities to provide back-up records

➢ Notification of billing disputes

➢ Initial resolution procedures (e.g. escalation to more seniormanagement)

➢ Final resolution (referral to arbitration, regulator or courts)


3 - 14


Quality of Service/Performance and Trouble Reports

Quality of Service ➢ Service performance standards may be specified in appendix, forexample:

➢ Average time for provisioning interconnection circuits

➢ Percentage of interconnection cut-overs made on scheduleddates

➢ Comparative provisioning performance for competitors and self(or affiliates)

➢ Switching and transmission quality measures on interconnectedcircuits (e.g. probability of blockage at peak hours, transmissiondelay and loss – consider referencing ITU-T recommendations

Testing and Maintenance ➢ Right to make reasonable tests, and to schedule serviceinterruptions; procedures to minimize disruption

Trouble Reports ➢ Procedure for trouble reports; notice periods; response timestandards

➢ Duty to investigate own network before reporting faults tointerconnecting operator

➢ Responsibility for costs incurred to second operator in investigatingfaults subsequently found to exist in first operator’s network.Calculation of charges (labour, etc.) for investigating trouble reports

System Protection andSafety Measures

➢ Responsibilities of parties to take necessary precautions to preventinterference with, or interruptions of, other parties’ networks orcustomers

Interchange and Treatment Information

Data Interchange Format ➢ Method and format of data interchange between carriers, includingdata interfaces, software, forms, etc.

Data to be Exchanged ➢ Specify all data types and systems for which data is to beinterchanged, for example:

➢ New facilities and service orders, network changes and forecasts,billing, etc. (see above)

➢ Number allocations and other data required for call routing andlocal number portability (where applicable, e.g. where LNP systemis operated by incumbent operator rather than an independentparty)

➢ Customer listings in directories and databases

➢ Access to network databases, for provision of advanced services


3 - 15

Interconnection


Access to and use ofCustomer Information

➢ Confidentiality procedures for customer information, including:

➢ Establishment of separate interconnection services group withsecure data (password protection for electronic files; locks fordata rooms and filing cabinets, etc.)

➢ Confidentiality forms to be completed by all relevant employees(penalties and bonding optional)

➢ Procedures to ensure protection of customer privacy

Access to and use ofOperator Information

➢ Confidentiality procedures (see customer information procedures –above)

➢ Intellectual property rights

Equal Access and Customer Transfer

Equal Access Procedures ➢ Procedures depend on equal access approach, e.g. carrier pre-selection; casual selection. Detailed procedures normally incumbentfor carrier pre-selection, including:

➢ Customer authorization requirements (signature on prescribedform, clear choice requirements)

➢ Authentication and measures to prevent unauthorized customertransfers (slamming)

➢ Penalties for unauthorized customer transfers

➢ Methods of reporting customer transfers (contact points and data tobe provided)

➢ Order confirmation procedure (format, medium, etc.)

➢ Schedule to implement transfers

➢ Procedures to implement transfers

➢ Dispute resolution process (e.g. escalation through seniormanagement, arbitrator and regulator); information to be provided indispute resolution process

➢ Procedures for dealing with disputed customers (which operator maycontact customer, information to be provided to and/or obtained fromdisputed customers)

Ancillary Services

Operator Assistance ➢ Types of operator assistance services to be provided, includingdirectory assistance, translation services, fault report routing, etc.

➢ Call handling and operations procedures

➢ Fees and billing procedures


3 - 16


Other Ancillary Services ➢ Subscriber listings in telephone directories

➢ Information and billing inserts

➢ Repair and maintenance services

➢ Other services provided by one or other operators to increasemutual operating efficiencies

Termination

Grounds for Termination andRestrictions

➢ Termination may only be permitted subject to certain restrictions (e.g.regulatory approval for termination of interconnection by incumbentoperator)

➢ Grounds for termination by incumbent may include:

➢ Regulatory or court orders

➢ Bankruptcy, insolvency, receivership, etc.

➢ Cessation of business

➢ Fewer, if any, termination restrictions in competitive markets, and bynon-dominant operators

Termination Procedures ➢ Advanced notice requirements

➢ Payment of non-recoverable interconnection costs incurred bydisconnected operator

➢ Computation and payment schedule for disconnection costs

➢ Dealings with end-users, communications restrictions, etc.

➢ Disconnection cutover procedures.

Other Provisions

Force Majeure ➢ List of conditions for which non-performance of interconnectionagreement obligations will be excused

Assignment ➢ Rights of assignment and restrictions on same (e.g. consent orregulatory approval requirements)

Applicable Laws ➢ Agreement to be governed by, and interpreted in accordance with,the laws of relevant jurisdiction

Regulatory Approvals ➢ Specify regulatory approvals required for effectiveness and/orrenewal, amendment, termination, etc. of agreement

Breach of Agreement ➢ Remedies and penalties

➢ Liabilities, indemnification and limitation of liabilities

Legal Interpretation ➢ Standard provisions for legal interpretation and enforcement ofagreement (e.g. entire agreement clause, effect of unenforceableterms, cumulative rights and remedies, etc.)


3 - 17

Interconnection


Dispute Resolution ➢ Procedures for resolution of disputes under agreement that are notspecifically dealt with elsewhere. For example:

➢ Good faith negotiations, time schedule for same, escalationthrough management levels

➢ Referral to regulator, arbitrator or court (e.g. of different types ofissues)

➢ Selection of, and procedures for, arbitration

Term ➢ Duration of term

➢ Renewal rights and procedures

Amendment ➢ Review and re-negotiation procedures

➢ Impact of regulatory changes

3.2 Interconnection Procedures

3.2.1 Establishing InterconnectionArrangements

A variety of different approaches have been used toestablish interconnection arrangements. The mainapproaches are listed below. Combinations of theseapproaches have been used in different countries atdifferent times.

➢ Regulatory prescription (ex ante) of intercon-nection arrangements.

➢ Negotiation between operators.

➢ Establishment of general regulatory guidelinesfor operators to negotiate.

➢ Regulatory mediation to facilitate operator-negotiated agreements.

➢ Regulatory prescription (ex ante) of defaultinterconnection arrangements, for example,based on other jurisdictions, that will apply ifnegotiations fail.

➢ Regulatory decisions to resolve interconnectiondisputes.

➢ Independent arbitration or mediation of inter-connection disputes.

➢ Regulatory review, variation and approval ofnegotiated arrangements.

Active industry participation is necessary to developpractical interconnection arrangements However,there has also been a growing consensus that it isnecessary to have regulatory involvement to provideadvance guidelines for operator negotiations and toresolve disputes. Different approaches to balancingindustry participation and regulatory intervention arediscussed in the following sections.

3.2.2 Negotiation of InterconnectionArrangements

In many countries, industry negotiation has been themain approach to establishing interconnectionarrangements. As previously discussed, there aregood reasons for this. Operators understand theirnetworks and operational requirements better thanregulators, and they have the technical informationrequired to implement effective interconnectionarrangements.

However, without regulatory intervention and direc-tion, interconnection negotiations do not usuallyproceed successfully. Incumbent operators are


3 - 18

generally suspicious that interconnecting operatorswill seek subsidized access to their extensiveexisting networks. Indeed, interconnection at almostany price is less expensive for a new entrant thanduplicating major parts of the PSTN. However, thepurposes of interconnection include minimization oftotal network costs, and speedy introduction of com-petition and rollout of new services, such asbroadband access services. Interconnectionobligations must often be imposed on incumbents,whether or not they agree with them, in order topromote sector development.

Some incumbents may also act strategically duringthe course of negotiations to implement arrange-ments that can effectively prevent or hindercompetitive entry. Consequently, regulators mustfind ways to overcome incumbents’ reluctance tointerconnect their network to new competitors’networks on efficient, cost-based terms andconditions.

Despite encouragement from governments andregulators, the reality is that dominant incumbentshave little incentive to enter into agreements thatexpedite competitive entry by interconnectingoperators. Incumbent operators hold all thebargaining power in negotiations. New entrants havelittle to offer in exchange for favourable interconnec-tion terms. They can promise market expansion,which should benefit all operators. However, mostincumbents see this benefit as being outweighed bythe loss of existing markets to new entrants.

Delays and failure have characterized many inter-connection negotiations. In some of these situations,regulators subsequently realized that delays anddisputes could have been resolved by appropriateregulatory intervention. For example, regulatorscould have applied benchmarks or best practicesfrom other countries. In other cases, while negotia-tions did produce interconnection agreements, thesewere sometimes one-sided, costly and inefficient.Sometimes, new entrants accepted one-sidedagreements as the only means available to start upbusiness and avoid bankruptcy.

As a result of this experience, many regulators andinterconnection experts have concluded that it isgenerally impractical to direct dominant incumbentsto negotiate interconnection agreements with new

entrants, without adequate regulatory guidance. Exante regulatory direction and ongoing supervision ormediation are generally required for operators tonegotiate reasonable interconnection agreementson a timely basis.

3.2.3 The Regulator’s Role inInterconnection Negotiations

Once it is decided that regulators should play a rolein promoting the successful conclusion of intercon-nection negotiations, the next question is: how canthe regulator intervene most effectively? Regulatorshave a variety of tools available to expedite negotia-tions and to assist in the successful completion ofinterconnection agreements. Some provenregulatory approaches are described below.Variations and combinations of these approachescan be used in some cases:

➢ Establishing guidelines in advance ofnegotiations – As indicated in Section3.1.6.1, there is a consensus that ex anteinterconnection guidelines are a necessaryand effective means to promoting good in-terconnection agreements. The task ofdeveloping such guidelines has been madeeasier for newer regulators due to thegrowing number of published interconnectionprinciples and guidelines established byother regulators. The increasing availabilityof precedent interconnection agreementsand the development of “best practices” andbenchmark interconnection charges in othercountries also make it easier for regulators toestablish such guidelines. The remainingsections of this Handbook also discussapproaches that can be used in establishingex ante guidelines.

➢ Setting default interconnection arrange-ments in advance of negotiations –Regulatory interconnection guidelines areusually fairly general. As a result, there areoften disputes among operators about howbest to apply guidelines. This can causedelays and impasses, and the need forfurther regulatory intervention. One approachto deal with this issue, is for the regulator topublish default interconnection arrangementstogether with guidelines. If the negotiations


3 - 19

Interconnection

fail, the default arrangements will apply.Such an approach was adopted for someinterconnection issues by the US regulator inits landmark 1996 interconnection order.

In the case of a first interconnection agree-ment with an incumbent, it may be difficult fora regulator to establish appropriate defaultarrangements. The regulator may need toreview the issues in depth, obtain informa-tion and submissions from the operators, etc.before it is in a position to establish defaultarrangements. However, defaultarrangements will usually be easier toestablish for subsequent agreements.

As with guidelines, published interconnectionagreements and the development of “bestpractices” and “benchmark” interconnectioncharges in other countries is making it easierfor regulators to establish defaultarrangements. Benchmarking has beenused extensively by the EuropeanCommission, and at the international level,such as in the US-Japan bilateral telecom-munications negotiations.

Finally, if there is a concern about the appro-priateness of the default arrangements, theregulator can provide a “sunset” clause fortheir applicability. In other words, theregulator can indicate that the defaultarrangements will cease to have effect after,for example, one year. That will provide timefor a more detailed review between the timenegotiations fail and the sunset of the defaultarrangements.

➢ Establish deadlines for various stages ofthe negotiations – Deadlines should be setat the outset of negotiations for completion ofvarious steps or deliverables. For example,the incumbent might be asked to produce aproposed interconnection agreement in 30days. Alternatively, deadlines can beproposed as soon as it appears delays willoccur. Consequences of the failure to meetthe deadlines can include regulatory inter-vention to impose an agreement andindependent mediation or arbitration.

Another option that is sometimes proposedis final offer arbitration. In final offer arbitra-tion, an independent arbitrator must selectone of the final offers put forward by twodisputing parties. In theory, this provides anincentive for the parties to make reasonableoffers. In practice, this approach is generallyinappropriate for interconnection negotia-tions, due to the number of issues involved,their complexity, and to the regulatory goal ofdeveloping efficient and non-discriminatoryarrangements. The regulatory goal is notsimply to establish an interconnectionarrangement, but to establish a good one.

➢ Establish Industry Technical Committees– Bilateral or multilateral industry committeesare often the best forum for establishing thedetails of interconnection arrangements. Ifnegotiations are proceeding smoothly,incumbents and new entrants may take theinitiative to delegate the details of technicalinterconnection arrangements to workinggroups or committees. However, in somecases, it may be necessary for the regulatorto take the initiative to ensure appropriatetechnical committees are established. Ineither case, it is usually good practice to setdeadlines for reports by such committees.

Depending on the degree of co-operationbetween operators, representatives of theregulator may also be able to play a usefulrole on the committees. They can often fa-cilitate agreement on interconnectionarrangements, suggest alternativeapproaches when there is an impasse, andotherwise mediate the discussions. In somecases, it will be necessary or useful for theregulator to retain expert consultants toassist in this role, and particularly inassessing the merits of conflicting positionsof operators.

Sometimes industry technical committeework can drag on for months or years. Insuch cases, the committees actually slowdown the process of reaching interconnec-tion agreements. Delays can result from theestablishment of committees with rigid workschedules, lack of familiarity with intercon-


3 - 20

nection technologies on the part of theregulatory participants, unnecessary processconcerns, and other factors. The regulatorshould be flexible and willing to adoptalternative approaches to ensure that theindustry technical committee processproduces results on a timely basis.Alternatively, in some cases, the processshould be abandoned, and otherapproaches adopted.

The industry technical committees estab-lished under regulatory supervision inCanada have generally been consideredvery successful. The CanadianInterconnection Steering Committee (CISC)and its sub-committees included participationfrom interested industry firms, as well as rep-resentatives of the regulator. CISC wasestablished after a regulatory decision thatprovided ex ante guidance on the terms andconditions of interconnection. However muchdetail remained to be determined by CISC. Ittook about 2 years to reach agreement onmajor issues, and regulatory interventionwas required from time to time. However,CISC managed to achieve consensus onmany important interconnection issues. TheCISC committees continue to deal withongoing issues that arise, for example, inconnection with new types ofinterconnection.

➢ Incentives to complete interconnectionarrangements – A carrot can be moreeffective than a stick. Various incentives canoften be provided to conclude interconnec-tion agreements. Incumbents depend onregulators for approvals or actions that cansometimes be linked to the successful con-clusion of interconnection arrangements.

An example of this approach can be found inCanada. In 1984, the incumbent operators(the “wireline operators”) were licensed toprovide new cellular telephone services. Atthe same time, licences were issued to anew entrant cellular operator. As an incen-tive, the incumbents were prohibited fromstarting up their cellular services until theyhad completed interconnection agreements

with the new entrant. The arrangements thatapplied to the new entrant would also applyto the incumbents’ own cellular operations.This “no head start” rule proved to beeffective. Mutually acceptable agreementswere quickly concluded. The incumbent op-erators did not want to delay the introductionof their own cellular services.

In developing positive incentives for incum-bents to complete interconnectionagreements, regulators must take care toensure that they do not create incentives fornew entrants to stall or frustrate the negotia-tions. In the Canadian example discussedabove, for instance, if the new entrants hadnot been ready to start up service, theymight have delayed start up by theincumbents by stalling completion ofagreements. Regulators must provideincentives for both sides to completenegotiations.

Finally, the prospect of receiving compen-satory interconnection charges can providean incentive for incumbents to concludeinterconnection agreements. Mostincumbents focus on short-term loss ofmarket share to competitors. However, thosethat take the longer view, and build appropri-ate network facilities, can earn significantinterconnection revenues as a result of thenew traffic stimulated by their competitors.

➢ Appoint mediators or arbitrators – Wherenegotiations fail, or where they are likely tofail, success can often be achieved byappointment of a mediator or arbitrator. Thetwo are different in that arbitrators areempowered to make binding decisionswhere an agreement cannot be reached.Mediators can provide additional information,develop compromises, propose alternatives,and persuade. However, they cannot imposetheir own decision on the negotiations.

It is possible for regulators or regulatory staffto act as mediators and arbitrators. However,this in not always the best approach,particularly in the case of inexperiencedregulators and staff. Interconnection is a


3 - 21

Interconnection

complex area, and the costs of delays andimproper regulatory intervention can be high.There is a growing body of internationalinterconnection “know-how”. Experiencedindependent interconnection experts canoften add valuable experience. They canrecognize issues from other countries,suggest options for unresolved issues, andotherwise save time. In addition, the use ofoutside experts maintains the independenceand credibility of the regulators. The regula-tors can act as a final decision-maker in theevent the mediation process fails. They canalso review the final decision of an arbitrator,if necessary.

One or more of the foregoing regulatory approachesis usually required to promote the successful con-clusion of interconnection negotiations. Whateverthe approach, it is important for regulators to beproactive in establishing interconnection proceduresand guidelines that will promote the negotiation ofeffective interconnection agreements. Further,where negotiations fail, regulators must be preparedto take steps to bring them to a successful conclu-sion.

3.2.4 Dispute Resolution

In most countries, it is the regulator’s role to resolveinterconnection disputes. The WTO RegulationReference Paper requires signatories to theAgreement on Basic Telecommunications toestablish an independent dispute resolutionmechanism. The Paper requires recourse to anindependent domestic body to resolveinterconnection disputes within a reasonable time.This may be the regulator or another independentbody.

In practice, regulatory dispute resolution can be adifficult task. Most regulators will normally be lessinformed than the operators on the details of inter-connection. The risk of making an unsatisfactorydecision deters many regulators from wading intointerconnection disputes.

However, regulators must resolve disputes in adecisive and timely manner, or competition andsector development will be retarded. If informationon local costs is insufficient, international bench-

marks can be applied. Other practices applied inforeign jurisdictions can provide useful precedents.Discussions with other regulators and assistancefrom expert advisors can facilitate the regulators’task.

If interconnection negotiations fail, an operator,usually the new entrant, may apply to the regulatorto resolve the interconnection dispute. There is nosingle best approach to resolving a complex inter-connection dispute, but some approaches are betterthan others. Table 3-2 suggests some approachesregulators may use in resolving interconnectiondisputes.

The WTO Regulation Reference Paper defines anindependent regulator as follows:

“Independent Regulator” - The regulatory body isseparate from, and not accountable to, anysupplier of basic telecommunications services.The decisions of and the procedures used byregulators shall be impartial with respect to allmarket participants.

As discussed in Module 1, the degree of independ-ence of regulators varies in different countries. Insome countries, the regulator is a governmentministry, or a government agency that also hasresponsibility for the operations of a state-ownedincumbent. Many observers would not consider sucha regulator independent for the purpose of resolvinginterconnection disputes. While such a regulatormay technically be in a separate organization fromthe incumbent, it has similar interests. Both are partof the government telecommunications bureaucracy.Both may consider the financial and operatinginterests of the incumbent as their prime concern.

In such cases, other independent dispute resolutionbodies should be considered, possibly using someof the approaches set out in Table 3-3. These mightinclude an independent arbitrator or mediatoracceptable to both parties. One option is to have anindependent dispute resolution body established bya senior branch of government (the executive orlegislature). This body need not be set up as acostly, permanent bureaucracy. It can be staffed ona temporary basis with independent domestic andinternational telecommunications experts. Anotheroption is to request an international agency with re-


3 - 22

sponsibility in the telecommunications sector, suchas the ITU or The World Bank, to appoint orrecommend an independent dispute resolutionexpert or panel to assist in the domestic disputeresolution process.

3.2.5 Ex Ante Regulatory Guidance

In some countries, regulators have prescribeddetailed interconnection conditions before intercon-nection arrangements are made. Examples are the1996 US and the 1997 Canadian interconnectionorders for competitive local operators. In these

countries, lengthy regulatory interconnectionproceedings were held before the rulings weremade. Input was obtained from incumbents, newentrants and other interested members of the public.In the end, detailed decisions were issued,specifying many of the approaches and specificrates, terms and conditions on which interconnectionshould occur.

This experience produced a wealth of information,analyses and insights into interconnection issues.However, the work effort required to produce a de-tailed set of interconnection rules should not be

Table 3-2: Approaches to Resolving Interconnection Disputes

Improving the informationbase for decision-making

➢ Require parties to clearly define areas of agreement and dispute

➢ Send written information requests to operators to clarify disputed issuesand provide information for interconnection decisions

➢ Require written argument (with supporting facts and research, ifnecessary) to assist in clarifying the issues in dispute

➢ To increase transparency, consider making the arguments (but notconfidential business data) available for comment by other interestedparties and the public

➢ Consider inviting other interested parties (e.g. other interconnectingoperators, service providers, or user groups) to comment on the issues

Obtaining expert assistance ➢ Hire an experienced interconnection expert to assist in clarifying theissues, formulating information requests, and providing general adviceto the decision-makers

➢ Consider appointing a mediator (or, if the parties agree, an arbitrator)

➢ Use outside parties for informal mediation, arbitration, informationgathering or other participation in the negotiations. This approach isparticularly useful in countries where direct regulatory involvementwould “taint” the legality or politically prevent it from making an unbiasedfinal decision.

Improving accuracy andcredibility

➢ Consult with other regulators on their experience in similar cases

➢ Review decisions and interconnection agreements approved by otherregulators

➢ Consider circulating a draft of the decision to resolve the dispute to thedisputing operators and other interested parties. Their comments shouldbe made public. Comments and corrections can improve the accuracyof the final decision.


3 - 23

Interconnection

underestimated. Moreover, these lengthyinterconnection proceedings did not produce the“final word” on interconnection arrangements. Inboth Canada and the US, there have been lengthyfollow-up proceedings before the regulators and inthe courts. In Canada, much of the detail ofinterconnection arrangements was left to a numberof industry technical committees led by regulatorystaff. This CISC process (which is referred to above)produced very useful results, but it took about 2years to resolve most of the issues.

It should be recognized that interconnection is adynamic issue. The types of telecommunicationsinfrastructure and services are constantly changing.As a result, interconnection requirements continue tochange as well. Where regulators prescribe inter-connection arrangements, they should be viewed asflexible rules that should evolve with telecommuni-cations networks and markets.

3.3 Financial Terms ofInterconnection

3.3.1 Interconnection Charges

Interconnection charges often account for a verysignificant part of the costs of new telecommuni-cations operators. This is particularly the case withnew entrants that do not own end-to-end networks.The level and structure of interconnection charges

are, therefore, major determinants of the viability ofoperators in a competitive telecommunicationsmarket.

Over the years, a variety of approaches have beenused to calculate interconnection charges andgenerally to determine the financial terms of inter-connection. In this Section, we first consider thegeneral approaches that have been used todetermine interconnection charges. Later in theSection, we review specific types of interconnection-related costs that are often treated in specific ways.Examples are start-up costs, costs of interconnec-tion links and collocation and infrastructure sharingcosts.

3.3.2 Approaches to Setting InterconnectionCharges

This Section reviews the general approaches thathave been used to determine interconnectioncharges. While there is no single correct approach,there is a consensus among telecommunicationsand trade experts that the best approaches are cost-based. However, other approaches have their meritsin some circumstances. Table 3-3 provides anoverview of the main approaches used to determineinterconnection charges. Readers interested in moredetail on the costing concepts and economictheories underlying them should refer to Appendix Bof the Handbook.

Table 3-3: Main Approaches to Interconnection Charges

Approach Description and Examples Comments

ForwardLookingIncrementalCosts

➢ Charges based on forward-lookingcosts of facilities and services providedto interconnecting operator (usuallyestimated over the long run, i.e. LongRun Incremental Costs or “LRIC”)

➢ Examples: Australia, Canada, theHong Kong SAR of China, Chile, andUS local operators

➢ Variations of LRIC include LRAIC,TSLRIC and TELRIC. Theseapproaches include different elementsof fixed and common costs (e.g.

➢ Generally accepted as best practice

➢ Approach sends most efficient pricesignals; based on current technologyrather than existing book assets

➢ Closest approximation of costs in afully competitive market

➢ Requires study and some cost anddemand estimates.

➢ Usually leads to lower interconnectionrates; this stimulates competition butprovides lower revenues to incumbent


3 - 24

Table 3-3: Main Approaches to Interconnection Charges (cont’d)

overheads, and fixed-servicecosts).that are excluded fromtraditional LRIC analyses. Thesevariations are growing in acceptanceas “best practices”. They are describedin Appendix B of the Handbook

operator

➢ May be substantially out of line withactual book costs of inefficientincumbents

Can be inappropriate if end-user prices areseriously unbalanced (e.g. set well belowcosts and below interconnection charges

HistoricalAccountingCosts

➢ Charges based on the accountingrecords of the operator supplying theinterconnection facilities or services

➢ Generally includes an assignment ofdirect costs and an allocation ofcommon costs booked in theaccounting records

➢ Examples: UK, 1995 Japanesesystem, and Sweden

➢ Common practice; less favoured byregulators and experts today

➢ Less efficient since historical costswere often incurred less efficiently thanthose based on current technology andoperational circumstances (e.g.privatization)

➢ Accounting records often misstate realvalue of assets: based on subjectiveaccounting policies and politicaldecisions regarding investments

➢ Usually requires study to as-sign/allocate booked cost tointerconnection facilities and services

SenderKeep All(SKA)

(Bill andKeep)

➢ No charges payable betweeninterconnecting operators fortermination of each other’s traffic

➢ Typically, each operator pays for itsown facilities up to the point ofinterconnection, plus charges for anyunusual costs incurred by the otheroperators to accommodate its traffic

➢ Examples: Indian, US and Canadianlocal operators, and Indonesianregional operators

➢ Works best where the two operatorsare similarly situated and exchangeapproximately the same amount oftraffic (e.g. for interconnecting localoperators)

➢ Charges can apply to compensate fortraffic imbalances

➢ Without such charges, SKA can retardfinancing and development of rural orother services, where there is animbalance of traffic (i.e. moreincoming)

➢ Was the main model for interconnec-tion of ISPs in many markets.However, this is changing as largerISPs, with substantial backbonefacilities and reach, increasingly treatsmaller ISPs as customers rather thanpeers

RevenueSharing

➢ Typically, new entrants pay theincumbent operator a share of theirrevenues from interconnected services(or all services)

➢ In some revenue-sharing arrange-ments, no additional charges are

➢ This approach is simple – no need forcost studies to determineinterconnection charges

➢ Generally considered non-transparent

➢ Potentially inefficient and anti-


3 - 25

Interconnection

Table 3-3: Main Approaches to Interconnection Charges (cont’d)

payable between interconnectingoperators for termination of eachother’s traffic; in others, additionalcharges do apply for direct intercon-nection costs (e.g. transmission links,interconnection interfaces)

➢ Examples: Thailand, Indonesia, andChina

competitive (i.e. when excessiverevenue shares are paid)

➢ Sometimes prescribed by governmentsor PTTs as the only basis on whichinterconnection will be permitted in anotherwise closed market; sometimestreated as a “tax” for doing business ina country. May be a transitional step toa more efficient approach

InterconnectChargesbased onRetail Prices

➢ Interconnection charges based onprices to end users

➢ A discount is sometimes applied forinter-operator charges. This can beestimated based on the avoided costsof the supplying operator (e.g. retailbilling and marketing costs).

➢ Examples: US local resale prices, pre-1995 Japanese approach

➢ Difficult to estimate appropriatediscount – may lead to inefficiency (i.e.high discount discourages constructionof competitive facilities; low discountundermines financial viability ofcompetition)

➢ Specifically rejected in some jurisdic-tions (e.g. Hong Kong, China whichdifferentiate “carrier-to-carrier” chargesfrom retail rates)

OtherNegotiatedInterconnectCharges

➢ Interconnection charges have beennegotiated between operators basedon a wide range of other approaches;some principled, many arbitrary

➢ Example: International accountingrates, and some reseller agreements

➢ Efficiency of charges depends on howclosely they approximate efficientcosts; many negotiated chargesinclude implicit subsidies betweenoperators and customers

➢ Level of negotiated charges oftendepends on the bargaining power ofthe operators

3.3.3 Comments on Different Approaches

Internationally accepted interconnection principlesgenerally require interconnection charges to be cost-based or “cost-oriented”. This is the case with theinterconnection principles of the WTO’s Agreementon Basic Telecommunications and the EuropeanUnion’s Interconnection Directive. Cost-basedpricing of interconnection services is consistent withbest practices adopted by regulators in mostcountries. This issue is discussed further in Section3.1.6.5.

Forward-Looking Costing Approaches

There remains a fair amount of debate in regulatorycircles about the best approaches to use to calculate

interconnection costs in different circumstances.However, today most regulators and expertsgenerally agree that the ideal approach for calculat-ing the level of interconnection charges would beone based on forward-looking costs of supplying therelevant facilities and services. This ideal is usuallyimplemented by means of some variant on the long-run incremental cost (LRIC) approach. Thisapproach has been entrenched in the regulations ofsome countries (e.g. India) and the laws of others(e.g. the US).

The major variations of the LRIC approach that havebeen most widely accepted by regulators andexperts are:


3 - 26

Long Run Average Incremental Costs (LRAIC) -A long-run costing approach that defines theincrement as the total service. It differs fromtraditional marginal and incremental costmeasures by including allowance for the fixedcosts specific to the service concerned: “service-specific fixed costs”. The European Commissionhas adopted this approach.

Total Service Long Run Incremental Costs(TSLRIC) - This approach, developed by theFederal Communications Commission (FCC) inthe USA, measures the difference in costbetween producing a service and not producingit. TSLRIC is LRIC in which the increment is thetotal service.

Total Element Long Run Incremental Costs(TELRIC) -This approach, also developed by theFCC, includes the incremental cost resulting fromadding or subtracting a specific network elementin the long run, plus an allocated portion of jointand common costs.

Other variations - There are other variations onthe LRIC approach. In Canada, for example, theregulator uses an incremental cost approach(Phase II Costing) and adds a mark-up toapproximate forward-looking fixed and commoncosts. Other regulators have developed differentapproaches.

A well-designed LRIC-type approach provides anestimate of the costs of an operator to provide inter-connection in a fully competitive market. An LRIC-type calculation generally starts by estimating thedirect costs incurred by an operator in providing theinterconnection services in question. These costsare calculated over the “long run”, usually at leastten years, in order to average out the inherently“lumpy” nature of the investment costs of intercon-nection facilities in the year they are introduced.

In addition to the directly attributable costs, LRIC-type calculations generally include a capital costcomponent. This component is intended toreimburse the operators for the costs of financingthe interconnection facilities, since these costs arenecessarily incurred by the operator providing thefacilities.

As can be seen from the preceding descriptions, themost widely accepted LRIC-type approachesgenerally include a reasonable allocation of joint andcommon costs. Such costs can also be calculatedon a forward-looking basis, to approximate the costsof an efficient operator. Joint and common costs are,by definition, not directly caused by the interconnec-tion services, but are nevertheless incurred by anoperator in connection with its interconnectionfacilities and services. Common examples of suchcosts are the salaries of the president, managingdirector or legal counsel of the operator. By includingcapital, joint and common costs, a LRIC approachcan approximate costs in a competitive market,while providing reasonably full compensation to theoperator supplying the interconnection – assuming itoperates efficiently.

Further descriptions of the methods used tocalculate long-run incremental costs, includingLRAIC, TELRIC and TSLRIC are included inAppendix B and in Module 4.

While variations on the LRIC approach areconsidered the best practices by most experts, thereare practical limitations on their applicability. Someof these are listed in Table 3-3. Some of these limi-tations are particularly significant in countries withless developed telecommunications sectors. Forexample, if local retail telecommunications rates areset well below costs, setting interconnection pricesat LRIC may not permit a new, local services entrantto run a viable business. The new entrant’s inter-connection costs may exceed its retail prices. Whilerate rebalancing is the long-term solution to thisproblem, in the short term interconnection rates mayneed to be discounted in order to permit competitionto emerge. There are other practical problems withthe application of LRIC-type approaches in someenvironments.

Other Approaches

The applicability of the non-LRIC-type approacheslisted in Table 3-3 depends on the circumstances ofdifferent countries. The comments in the Tabledescribe strengths, weaknesses and other consid-erations. Several other comments follow.

Modifications are often made to the variousapproaches to attempt to compensate each operator


3 - 27

Interconnection

more closely for costs resulting from its interconnec-tion. An example is the Sender Keep All (Bill andKeep) approach. As indicated in Table 3-3, thisapproach is appropriate where the two operators aresimilarly situated and exchange approximately thesame amount of traffic. Thus, it is often used for in-terconnection of local operators in the same city orneighbouring regional operators.

The Sender Keep All approach may be modified toadd charges to compensate for traffic imbalances.For example, operator no. 1 may receive and termi-nate more traffic from operator no. 2 than it sends tothat operator. Operator no. 1 will then usually incurhigher costs as a result of the interconnection thanoperator no. 2. To compensate for this imbalance,operator no. 2 may pay a cost-based interconnec-tion charge to operator no. 1 for every minute oftraffic it sends that exceeds the traffic it receives.

A word or two about revenue sharing approaches.An element of revenue sharing may be appropriatein some cases to distribute surplus revenues afterpayment of cost-based interconnection charges.However, in some cases, revenue shares paid toincumbents have included a wide range of compo-nents, ranging from interconnection costs to a“licence fee” for operating in a jurisdiction or“compensation” to an incumbent for loss of businessto new entrants, or fulfilment of universal serviceobligations.

The latter three components are typically not costbased. They are usually not transparent and are notrecommended in any jurisdiction where the regulatorwishes to improve efficiency in the telecommunica-tions sector. These approaches can be subject toabuse. For example, excessively high revenue-sharing arrangements have been imposed in somejurisdictions in a short-sighted attempt to earn addi-tional operator or government revenues. The effectis to prevent efficient competition.

If revenue-sharing schemes must be used, thenregulators should consider identifying each compo-nent of the revenue share separately. This includes,for example, shares to pay for cost-based intercon-nection charges, for concession or licence fees, etc.This approach adds transparency and allows for thegradual elimination of revenue-sharing componentsthat are not cost-based. Universal service charges

should be dealt with by means of a separate charge,not a revenue-sharing formula. Issues related touniversal service and universal access charges arediscussed in detail in Module 6.

Table 3-3 does not provide an exhaustive list of theapproaches to calculating interconnection charges.Other approaches exist. One example is the EfficientComponent Pricing Rule (ECPR), which basesinterconnection charges on the net incrementalcosts of interconnection, plus the “opportunity costs”or margin lost by the incumbent as a result of traffic“taken” by the new entrant. This approach has beendiscussed among academics and consultants, buthas generally not been accepted by regulators as areasonable option.

Finally, interconnection charges are sometimesindexed or “price capped” to determine futureincreases (e.g. for a five or ten year period). Suchapproaches provide certainty to interconnectingparties regarding their level of future costs orrevenues.

3.3.4 Specific Interconnection Costs

3.3.4.1 Start-up Costs

The network infrastructure of most incumbentoperators was designed to function on a monopolybasis. In the transition to a competitive telecommu-nications market, some modifications are usuallyrequired to the operator’s switching and transmis-sion facilities and related software to permit efficientinterconnection among multiple operators. Forexample, switches must be programmed torecognize and route traffic to telephone numbers onthe network of interconnection operators. Additionalnumbers must often be allocated and equipmentmodified to deal with them. These modifications areoften referred to as “start-up costs”, since they arerequired at the outset to permit interconnection.

Regulators in different countries have treated start-up costs in different ways. Some take the view thatnew operators are the beneficiaries of interconnec-tion, so they should pay all start-up costs. In theextreme, this approach is applied not only tointerconnecting transmission circuits, but to allmodifications and upgrades to an incumbent’snetwork required to facilitate interconnection. Some


3 - 28

new operators accept this approach as the only onethat will provide them with interconnection,particularly in countries with state-owned PTTs.However, this approach has disadvantages. It canimpose a heavy financial burden on a new entrant,shift costs of network upgrades from incumbents tocompetitors, and ultimately lessen the chances ofviable competitive entry.

A different approach that is more pro-competitive innature has been adopted by a number of countriessuch as Canada. The approach is based on theassumption that competition is introduced to benefitall telecommunications users and the economy ingeneral. Interconnection start-up costs are seen as adirect result of the policy decision to open a marketto competition. It is also recognized that the costsincurred by all operators will, market conditionspermitting, generally be borne by telecommunica-tions users.

Therefore, some basis is developed to apportioncosts among established and new operators on theassumption that they will generally pass these onthrough user rates. A specific surcharge may beconsidered, but may not be adopted for politicalreasons. One method of apportioning costs is on thebasis of the projected use of telecommunicationsservices (including interconnected services) in thefuture. A formula can be established to adjustcompensation between operators in case actual usediffers from projected use of telecommunications orinterconnected services.

Under this approach, the incumbent will generallybear a large share of start-up costs. Someregulators regard this approach as necessary orappropriate to facilitate competition. Understandably,this approach is generally opposed by incumbents.

3.3.4.2 Interconnection Links

Different approaches have been adopted toapportion the costs of the physical links betweeninterconnecting operators. Such links includetransmission lines or radio links that carry the inter-connecting circuits. They also include the ducts,towers, manholes and other support infrastructure,as well as the modifications that are required to thetransmission-related facilities (e.g. cross-connects

and distribution frames) in order to accommodatethe interconnected circuits.

One approach is to require the new operator to paythe entire cost of the transmission links and relatedfacilities. This approach is based on the theory thattransmission facilities are being added and themodifications made solely for the benefit of the newoperator and its customers. If this approach isadopted, incumbents should not be able to recoverany more than the actual costs of the transmissionlinks and related facilities. Sophisticated costingapproaches are not required. Normally, these costsare easily tracked through expense invoices, relatedlabour costs and overhead. As a general principle,the costs should not exceed fair market costs forinstalling the links. Incumbents may have anincentive to inflate charges for such links, and regu-latory oversight may be required to ensure chargesare based on market costs.

One method of ensuring charges for interconnectionlinks are not inflated is to give the new operator theoption of installing the links itself, including work onthe premises of the incumbent. Specifications forsuch work can be subject to discussion at a jointtechnical committee with a dispute resolutionmechanism. Work on its premises can be monitoredby the incumbent to avoid arguments aboutimproper work or sabotage.

As with start-up costs (see discussion in previousSection), interconnection links are a necessaryprerequisite for the development of a competitivemarket. Taking this view, regulators may consider itappropriate to apportion the costs of such linksbetween incumbents and new entrants, based onthe assumption that end users of all operators willultimately benefit.

The simplest, and probably most common methodof apportioning costs of interconnection links is tohave each operator pay the costs of its interconnec-tion links up to the Point of Interconnection (POI).Since POIs are often located in or near theexchange of the incumbent, this method can imposesignificant costs on a new operator. However, underthis approach, the new operator can decide how toconfigure its network to limit its costs.


3 - 29

Interconnection

3.3.5 Structure of Interconnection Charges

The structure of charges for interconnection oftenvaries from country to country. These variationsreflect a number of factors, including differences inthe telecommunications infrastructure, policydifferences and varying levels of effort on developingcost and price structures. Price structures need notbe complex to be efficient and fair. In many cases,simplicity is best. However, with some effort, a pricestructure can be developed that levels the playingfield for all operators and facilitates more efficientinterconnection.

Box 3-4 sets out some basic principles for an effi-cient interconnection price structure.

Operators, regulators and telecommunicationsexperts have long discussed how best to refine tele-communications pricing structures to improveefficiency. Many of the principles applicable to othertelecommunications prices also apply to the struc-

ture of interconnection charges. Several examplesare given below.

3.3.5.1 Fixed and Variable Charges

As a general principle, interconnection chargesshould reflect the difference between fixed andvariable costs of interconnection. For example, thefixed costs of providing a dedicated network accessline (loop) are best recovered through a fixedcharge. On the other hand, where the costs ofnetwork components, such as telecommunicationsswitches, are traffic sensitive, they are bestrecovered through usage charges. Usage chargesare usually based on time (minutes). In the case ofinterconnection of Internet backbone operators andInternet Service Providers, charges are often basedon capacity (bits of traffic).

While it is not always practical to implement thisprinciple, doing so is consistent with efficient pricingtheory. Distinguishing between fixed and variablecosts in the charges for interconnection components

Box 3-4: Principles for Efficient Interconnection Price Structures

➢ Interconnection charges should be cost-based (ideally based on long-run average incrementalcosts, including cost of capital, plus a reasonable markup to cover forward-looking joint andcommon costs)

➢ Where information is available, costs should be based on the current replacement costs of assets(discounted to their remaining service life); in the absence of such costs, depreciated book value ofassets is sometimes used

➢ Interconnection charges should be sufficiently unbundled so that an operator seekinginterconnection need only pay for the components or services it actually requests

➢ Where the costs of a particular component vary significantly in different locations, theinterconnection charges should be disaggregated (e.g. costs of access lines may be higher in ruralareas (where they are typically longer) than in cities)

➢ Charges should not include hidden cross-subsidies, particularly of an anti-competitive nature (e.g.charges for monopoly-supplied network components should not be inflated to a level well abovecosts in order to fund below-cost provision of competitive components). This principle is adopted inthe WTO Regulation Reference Paper.

➢ The structure of interconnection charges should reflect underlying costs. Thus, fixed costs shouldbe covered by fixed charges, variable costs by variable charges. Peak and off-peak charges shouldbe set where there is a significant difference in costs.


3 - 30

will send the right price signals. For example, therewill be less incentive to overuse usage-sensitivenetwork components if they are priced based onusage, rather than on a flat monthly charge.Establishing a price structure that reflects underlyingfixed and variable costs should lead to a moreefficient use of those components.

3.3.5.2 Peak and Off-Peak Charges

Peak and off-peak pricing differentials have beenused for retail pricing of telecommunicationsservices for many decades. Charging higher ratesfor usage in peak hours provides users with anincentive to call in off-peak hours. Advantages of apeak/off peak pricing structure include:

➢ reduced peak-hour congestion;

➢ reduced demand to build new infrastructure tomeet peak traffic loads;

➢ increased overall network utilization; and

➢ improved quality of service.

The same principles of peak and off-peak pricing areoften incorporated in interconnection charges. If theyare not, then interconnecting operators will have noincentive to charge higher rates to their end-usersduring peak hours. The result can be a migration ofpeak-hour traffic to new entrants, who will thenimpose higher costs on incumbent’s that must buildthe infrastructure to support the higher peak-hourloads.

Good regulatory policies, such as those adopted inHong Kong, China specifically provide that thestructure of interconnection charges must reflect thebehaviour of underlying costs. Thus, the “carrier-to-carrier” charging principles in Hong Kong encourageinterconnection charges to reflect both fixed/variableand peak/off peak cost differences.

3.3.5.3 Unbundled Charges

In an increasing number of countries, telecommuni-cations policies require incumbent operators toprovide competitors with access to unbundlednetwork components. This approach is supported bythe WTO Regulation Reference Paper, which states

that major suppliers must provide interconnection ona basis that is sufficiently unbundled so that asupplier need not pay for network components orfacilities that it does not require for the service to beprovided.

In keeping with their WTO commitments, or gener-ally because it is good policy, many regulators haveissued directives requiring unbundled charges. Forexample, in India, a regulation was issued in 1999by the Telecommunications Regulatory Authority ofIndia (TRAI) that states that “No service providershall be charged for any interconnection facility itdoes not seek or require”. (TRAI (1998a))

3.3.5.4 Universal Service and ADC Charges

In many countries, incumbent operators incurdeficits in carrying out uneconomic universal serviceobligations (USO) or universal access obligations.Beneficiaries of these social obligations generallyinclude high-cost service areas, such as remotevillages or low-income customers. In somecountries, however, deficits are not incurred by theincumbents to perform specific universality. Rather,the deficits are incurred as part of a policy ofmaintaining low access charges for all customers.These are usually referred to as Access DeficitContributions (ADCs) to distinguish them fromUniversal Services Obligation (USO) payments thatgenerate revenues for more targeted socialpurposes.

In a monopoly environment, ADCs are often paidfrom services priced above costs (e.g. internationalrates or business services) to access costs that arepriced below cost. In the case of the incumbent,ADCs may be explicit, or implicit in unbalancedrates. Traditional telecommunications policies oftenprevent “rebalancing” of the prices to more closelyreflect their costs. New interconnecting operatorsoften do not have similar universal service obliga-tions or access deficits. Accordingly, they are oftenasked to contribute to USO payments or ADCs ofthe incumbent.

There are a number of ways of dealing with thisissue. These are discussed in detail in Module 6. Asindicated in that Module, the best practice for regu-lators is to levy any USO or ADC charges separatelyfrom interconnection charges. As demonstrated in


3 - 31

Interconnection

this Module, the underlying concepts and calcula-tions for interconnection charges are very differentfrom those underlying USO and ADC charges.

If USO or ADC charges are established, it is clearlya good practice to identify them as separate frominterconnection charges. Blending the two chargesremoves transparency from the interconnectionprocess. Separate charges permit regulators tocomply with the requirement of the WTO RegulationReference Paper that USO charges beadministered in a transparent, non-discriminatoryand competitively neutral manner. Please seeModule 6 for a more comprehensive discussion ofUSO and ADC issues.

3.3.6 Internet Interconnection Charges

Over the past decade, the Internet has changedfrom a co-operative to a commercial communica-tions medium. It has also changed from a relativelysmall education and research-based data network toa network that accounts for more traffic than voicetelephony in several countries today. Thistransformation of the Internet has changed the basisfor interconnection charges among ISPs and be-tween ISPs and the operators of the large capacitybackbone telecommunications networks that carryInternet traffic.

Originally, many ISPs regarded themselves asequals or “peers”. They generally entered into Billand Keep interconnection arrangements. Underthese ‘peering’ arrangements Internet networksexchanged traffic without levying charges or payingfees to each other. The underlying premise forpeering arrangements was that Internet networks ofsubstantially similar size and traffic volumesbenefited more-or-less equally from interconnection,and incurred generally similar costs.

Over time, some Internet Protocol (IP) networksexpanded their coverage to national and global lev-els. Some network operators developed intospecialized IP backbone operators, carrying largevolumes of Internet traffic for long distancesbetween ISPs and Internet hosting services. Thesebackbone network operators generally provide ‘tran-sit’ services. Transit services involve thetransmission of Internet traffic between two or moreISPs and Internet hosts. Providers of Internet transit

services may or may not provide any Internetcontent or access services themselves. Some ISPswith larger networks also provide transit services, inaddition to standard Internet interconnectionarrangements.

ISPs generally interconnect with each other and withInternet backbone providers at Internet ExchangePoints (IXPs). These are sometimes referred to asNetwork Access Points (NAPs), although that termis becoming less common. IXPs have switchingequipment and routers that permit interconnection ofthe various Internet networks using the IXP. As withthe Internet generally, IXPs are evolving intoincreasingly multifunctional, and commercial opera-tions, that charge fees for an increasingly widerange of services, rather than just facilitating ‘free’interconnection of ISPs. Many IXPs now providecollocation services, providing space as well asequipment for Internet routing, transmission, web-hosting and other services. Separate, market-basedcharges are usually levied for such services. As withmost Internet-related services, these charges aregenerally unregulated, except where they areprovided by a dominant incumbent operator.

The transition of the Internet to a more commercialmedium, with large disparities between the sizesand functions of Internet networks, has changed thestructure of Internet interconnection charges. Insome cases, interconnecting ISPs still exchangetraffic with each other as ‘peers’ on a Bill and Keepbasis. Under this arrangement, each ISP typicallypays its own costs of transmission, routing and otherequipment, or shares the costs on a negotiatedbasis.

However, such peering arrangements are becomingless common, particularly where different types orsizes of Internet operators interconnect. There,asymmetrical charges have become the norm. Thebackbone network operator, or the larger ISP,usually charges the smaller ISP or local accessprovider for interconnection and transit services. Thebasis for such interconnection charges is oftensimilar to those found in other parts of the telecom-munications industry. Charges are typically basedon one or more of the following variables:


3 - 32

➢ traffic flow or usage, based on the increasingcapacity of Internet routers and other equipmentto measure traffic;

➢ imbalance of traffic flows between ISPs;

➢ distance or geographical coverage;

➢ number of points of interconnection; and

➢ other cost-based interconnection charges.

All of these charging variables are related to costsincurred by the ISP providing the service, or at leastproxies for such costs. This trend toward cost-basedinterconnection charges is consistent with develop-ments in other telecommunications services.

One anomaly in the trend toward cost-based Inter-net charges has been related to the traditionallyheavy reliance on US-based ISPs and Internetbackbone providers by ISPs in other countries. Dueto the early lead of the US-based Internet industry,and the heavy concentration of attractive Internetweb sites in the US, many ISPs in other countrieshave paid US ISPs for transportation to and from theUS to their home country. There have often been noreciprocal charges paid by US ISPs for traffic to theinterconnecting ISPs in other countries. Thisimbalance has become a hot policy issue within theITU and other international organizations. WithinAPEC, for example, Australia and various Asiancountries have complained that current costs ofinterconnecting with North America are too high andthat it is inequitable that Asian networks are notcompensated for their costs in carrying trafficgenerated by North Americans.

In April 2000, ITU Study Group 3 adopted Recom-mendation D.iii on International InternetInterconnections:

“Noting the rapid growth of Internet and Internetprotocol-based international services: It is rec-ommended that administrations involved in theprovision of international Internet connectionnegotiate and agree bilateral commercial ar-rangements applying to direct internationalInternet connections where each administrationwill be compensated for the cost that it incurs in

carrying traffic that is generated by the otheradministration.”

The US and Canada have opposed this recommen-dation. They argue that the North American bias ofInternet routing will decrease over time, ascompetition and market developments reduce costsand increase Internet facilities in other regions. TheUS, in particular, has long argued that the Internetshould remain unregulated in most respects. Theproposed resolution was considered at the ITU’sWorld Telecommunications StandardizationAssembly in Montreal in October 2000. After muchdiscussion, the Assembly adopted arecommendation that calls for arrangements to benegotiated and agreed upon on a commercial basiswhen direct Internet links are establishedinternationally. The new recommendation does notprescribe any particular costing approach; thusoperators are free to determine the approach to beused in implementing it. This recommendation hasbeen referred to as a framework for futurediscussions. The US and Greece stated that theywould not apply this recommendation in theirinternational charging arrangements.

Local interconnection charges are also important tothe viability of ISPs. Local Internet access providerswill be principal beneficiaries of the move to unbun-dling of local loops, which is discussed in Section3.4.6 of this Module. Unbundled local loops can beused by ISPs to provide DSL-based high speedInternet services on more favourable terms thanthose currently available in most markets.

In a number of countries, cable television networksprovide an efficient and highly successful form ofhigh-speed local Internet access. These ‘cablemodem’ services have generally been provided onlyby the serving cable TV operator. This has given thecable operator a strong position in ISP marketscompared to other ISPs without high-speed capabili-ties. Several countries have considered whether torequire cable operators to interconnect with otherISPs to provide them access to high-speed cablenetworks.

In Canada, the CRTC has ordered major cableoperators to grant other ISPs access to their highspeed networks at a discount from retail ISP rates.In the US, the FCC has not, to date, taken similar


3 - 33

Interconnection

action. Some US cable operators have entered intoagreements with ISPs to access their high-speednetworks on an exclusive basis, thus making accessunavailable to competitors. This appropriateness ofsuch exclusive arrangements is under considerationby the FCC.

3.3.7 Interconnection with Mobile Networks

As indicated in various places in this Module, mobileoperators must obtain interconnection with incum-bent operators of the PSTN in order to ensure theviability of their services. In general the interconnec-tion principles and practices described in thisModule apply to interconnection by mobile operatorsto the PSTN. However, certain differences apply tointerconnection with mobile operators.

Historically, regulators devoted much less attentionto mobile services than fixed services. Mobileservice was priced at a substantial premium to wire-line service. As a result, mobile service was viewedas a discretionary or even a luxury service whereconsumers did not need much in the way of regula-tory protection. As well, mobile service was offeredcompetitively in many countries, with the expectationthat market forces rather than regulators would bethe prime force in setting prices. Mobile operatorswere not perceived as possessing market power inthe same way as fixed operators.

However, the role of mobile services has changed inrecent years, leading to increased regulatory interestand attention:

➢ The consumer rates for mobile service havedeclined in both developed and developingcountries. The combination of rate decreases,the fact that consumers like the flexibility of mo-bile service, and improvements in mobiletechnology (such as longer battery life) havecontributed to an enormous increase in thenumber of mobile users. Indeed, in some coun-tries, the number of mobile users now exceedsthe number of fixed users. Thus, for many,mobile service is no longer a luxury – it is theprime way in which they access the PSTN.

➢ Some less developed countries have begun todevote much more attention to fostering thegrowth of mobile service, as they realize that

implementing mobile infrastructure can bequicker and less capital intensive than buildingthe type of ubiquitous wireline networks that arefound in most developed countries.

➢ All countries have come to appreciate therevenues that can be realized by auctioningmobile wireless spectrum. Bidders will take thedesign of the regulatory environment intoaccount as they assess how much to bid.

When mobile service was first introduced, mostcountries adopted Calling Party Pays (CPP)arrangements. Under CPP, the person that origi-nates a call is the one that pays for it, whether itoriginates on a mobile or fixed-line telephone. Aperson who makes a mobile-to-fixed call pays themobile operator at the retail rate. The mobileoperator, in turn, pays the fixed operator aninterconnection charge that is relatively small whencompared to the retail rate. Usually, theinterconnection charge is invisible to the mobilecaller. However, the situation is quite different for afixed-to-mobile call. Because the interconnectioncharge paid by the fixed operator to the mobileoperator is relatively large, the fixed operator willwant to recover it from the caller who makes the call.Accordingly, the fixed operator will charge asubstantial surcharge for fixed-to-mobile calls, withthe surcharge (less an administrative charge) beingpassed on to the mobile operator. The mobileoperator does not charge its customers for callsreceived from the PSTN.

CPP has not been adopted in countries such as theUS and Canada, where most local calls on thePSTN are not metered, but charged at a flat monthlyrate. These are referred to as Receiving Party Pays(RPP) or Mobile Party Pays (MPP) environments. Ina RPP country, the mobile customer pays both formobile-to-fixed calls and for fixed-to-mobile calls.However, the customer on the fixed network paysthe same amount to call someone whether on thefixed network or on a mobile network. Interconnec-tion between the fixed and mobile operators isgenerally on a reciprocal basis, either bill-and-keep(also referred to as sender-keep-all) or mutual com-pensation at the same interconnection rates that arefound in fixed-fixed interconnection arrangements.


3 - 34

A number of countries that do not have CPP areconsidering a switch to it, or have done so. Forexample, Mexico introduced CPP in April 1999. Thismove is partly motivated by evidence of highermobile subscriber growth rates in CPP countries. SriLanka has announced its intention to change toCPP. The transition to CPP affects subscribers of allnetworks in a market, including PSTN subscribers.Their bills will be increased since they will becharged for calls to mobile subscribers. Accordingly,the transition normally involves regulatory supervi-sion to ensure, among other things, that PSTNsubscribers are adequately notified of increasedcharges that will appear on their bills.

Because fixed-to-mobile calls are so much moreexpensive than fixed-to-fixed calls in a CPP country,many countries have distinct dialling prefixes forfixed-to-mobile calls. In that way, consumers under-stand that they will be charged a premium for fixed-to-mobile calls, and it is obvious when such chargingtakes place.

In recent years, some observers have expressedconcern about the level of CPP charges for fixed-to-mobile calls. The ITU’s Trends 2000 Report, whichfocuses on interconnection, points out that inEurope, where CPP arrangements prevail, the aver-age fixed-to-mobile interconnection rate was USD0.21 per minute for a three minute call. Thiscontrasts with mobile-to-fixed interconnection ratesof USD 0.01 per minute for local interconnection,0.014 for single transit interconnection and 0.02 fordouble transit interconnection. The ratios of fixed-to-mobile and local mobile-to-fixed rates range from alow of 8.7 in Norway to a high of 34 in France. Thereport suggests that asymmetrical regulation offixed-line and mobile operators may have resulted ininflated mobile termination charges under CPP.

Some observers believe that the high level of CPPcharges for fixed-to-mobile calls is due to a combi-nation of two factors, market failure and regulatoryinattention:

➢ The market failure arises because there is littlecompetition in fixed-to-mobile rates. Mobileoperators often compete vigourously onsubscription and mobile-to-fixed rates, servicelevels and coverage, but they rarely compete onfixed-to-mobile rates. Such competition does

sometimes arise, for example in Finland, wheremobile operators have reduced fixed-to-mobilerates in line with mobile-to-fixed rates. Incountries where there is a monopoly fixedoperator, the fixed operator has little incentive toreduce fixed-to-mobile rates. Even in countrieswith competing fixed operators, there seems tobe little evidence of competition to reduce fixed-to-mobile rates.

➢ The regulatory inattention arises because, asexplained earlier, mobile service was historicallyviewed as a discretionary or even a luxuryservice that appealed to a narrow segment ofusers. In many countries, mobile service wasoffered competitively, and rates were set bymarket forces. Unlike the fixed networks, regu-lators did not have good cost data for mobilenetworks. Without cost data, the regulators werenot in a position to determine if fixed-to-mobilerates might be higher than necessary.

The result of these two factors is that fixed-to-mobilerates in some countries have remained at high levelseven as mobile-to-fixed rates have declined sub-stantially due to reduced costs and vigourouscompetition.

An examination of the fixed-to-mobile rates that arecharged to the customers of a fixed operator leadsto an examination of the interconnection chargeslevied by the mobile operator to the fixed operatorfor the termination of a call on the mobile network.Few countries have examined the costs of mobiletermination and applied these costs in setting inter-connection charges. One country that has recentlymade such an attempt is the United Kingdom. In a1998 report, the Competition Commissiondetermined that fixed-to-mobile termination rateswere substantially above cost. In 1999, OFTELordered that rates be substantially reduced to aceiling of 11.7 pence per minute, and that the ceilingbe further reduced by 9% per year (after inflation) fortwo years thereafter. OFTEL will be considering iffurther pricing action is needed following this period.

High mobile interconnection rates may be reducedby competition over time. However, as mobile serv-ices catch up with and overtake fixed networks,there is likely to be more regulatory scrutiny of highmobile termination rates, particularly where they are


3 - 35

Interconnection

thought to be set at levels that are significantlyabove cost.

3.4 Technical and OperationalConditions

While financial arrangements are important to thedevelopment of interconnection arrangements, thetechnical and operational conditions determine howefficient and “seamless” interconnection is from theusers’ perspective. These conditions can alsodetermine whether competition in a particular marketwill succeed or fail.

The most important technical and operationalconditions are neither complex nor difficult to under-stand. At a minimum, regulators should develop anoverview of the key technical and operationalconditions in order to resolve disputes that may arisein interconnection negotiations.

3.4.1 Provision of Information byIncumbents

3.4.1.1 Availability of Agreements or Offers

The advantages of transparent interconnectionarrangements are discussed in Section 3.1.5.4. Thesimplest way to encourage transparency is torequire publication of interconnection agreements oroffers of incumbents. In this regard, the WTORegulation Reference Paper requires signatories toensure that a major supplier will make publiclyavailable either its interconnection agreements or areference interconnection offer.

The advantages of publication of interconnectionagreements or standard offers include:

➢ Publication facilitates interconnection by existingand potential new entrants. It allows them toobtain basic interconnection terms andconditions without lengthy negotiations orregulatory orders;

➢ It discourages undue discrimination by adominant operator (or by both parties to anagreement) that may not be readily detectableby regulators if filed in confidence;

➢ It facilitates comparisons of interconnectionrates, terms and conditions among major op-erators; and

➢ It assists in developing industry standards,benchmarks and best practices.

The disadvantage of mandatory publication of inter-connection agreements is that it breaches thenormal confidentiality of commercial agreements.However, this disadvantage can be mitigated inseveral ways. One is to permit deletion ofcommercially sensitive information from filed agree-ments. This can include proprietary network orservice information and related costs. In such cases,a confidential filing with the regulator is normallyrequired. Another approach is to require only thefiling of standard agreements or offers (“referenceoffers”), rather than all executed agreements.

For the reasons discussed in Section 3.1.5.2, thefiling of interconnection agreements between non-dominant operators is not generally required. TheWTO Regulation Reference Paper requirespublication of agreements with major suppliers, or areference interconnection offer with them. A numberof countries with well-developed regulatory regimes,for example Denmark and the UK, only require thepublication of interconnection agreements ofincumbents.

There is often no telecommunications regulatoryrequirement for publication of interconnectionagreements between smaller operators. However,these are increasingly being made public to complywith the securities laws of some countries. In thesecountries, securities regulators require companiesthat issue shares to the public to disclose their mate-rial contracts. Examples of such agreements can befound on the EDGAR Web Site in the US.Agreements between new entrants can provideinsight into interconnection arrangements in lessregulated markets.

3.4.1.2 Network Specifications

Interconnected networks must be technicallycompatible. A new entrant must, therefore, haveaccess to technical specifications of the network ofthe incumbent with which it will interconnect. Simi-larly, the incumbent requires information on the


3 - 36

technical characteristics of an interconnectingoperator’s network. For example, it will be importantfor both operators to know the types of switching,routing and transmission equipment used by theother, signalling protocols, number of circuits and theprojected volume of traffic to be exchanged.

Sufficient information is required to permit the inter-connecting operators to design their own networksto provide efficient connectivity between eachother’s customers. Regulators should ensure thatincumbents and new entrants do not withhold infor-mation necessary to ensure efficient interconnectionarrangements for both sides.

Operators should not be permitted, for example, towithhold necessary information on the grounds thattheir standards and specifications are proprietary. Ifnecessary, some technical information could beexchanged under non-disclosure agreements. Inpractice, however, this is impractical and canfrustrate interconnection of future networks. Thetelecommunications sector is evolving towards moreopen standards, and this is a trend that regulatorsshould encourage. Open standards are oftendeveloped through industry committees with regu-latory observers or mediators. In keeping with thispractice, regulators should encourage interconnec-tion operators to establish technical committees todevelop specifications, protocols, and procedures forthe interconnection of their networks.

In many cases, incumbent operator networks havenot been designed to anticipate interconnection withother operators. Accordingly, some network modifi-cations are often required to permit interconnection.Treatment of such network modifications or “start-upcosts” is discussed in Section 3.3.4.1.

3.4.1.3 Network Changes

Telecommunications networks are dynamic. In mostcountries, networks are constantly changing as newswitching and transmission facilities are added, newsoftware and features are installed, and newprotocols adopted. The most obvious example is thecurrent transition from circuit-switched to packet-switched networks, such as Internet Protocolnetworks, to carry both data and voice traffic.However, the network plans of operators change

regularly in response to technological development,market and budget considerations.

Over time, as the networks are modified, it is goodpractice for regulators to require that networks ofdominant incumbents evolve into more opennetworks.

3.4.2 Treatment of Competitor Information

Monopoly or dominant providers of local telephoneservices, and certain other monopoly services, are ina position to collect competitively valuable informa-tion on their interconnecting competitors. A typicalsituation might involve a local monopoly operatorthat receives orders from a long distance competitorto install leased local lines to interconnect with thecompetitor’s POP. The monopolist would know thatthe competitor had located a relatively heavy longdistance user (probably a business or governmentuser) that had sufficient traffic to require a leasedlocal line. In the absence of competitive restrictions,the monopoly could send a salesperson from its ownlong distance division to offer a discount or otherincentive to the customer to persuade it not to use itscompetitor’s services.

Abuse of such competitive information is subject toregulatory restrictions in many countries. The Refer-ence Paper on Regulation that forms part of theWTO’s Agreement on Basic Telecommunicationsattempts to prohibit such activities. The ReferencePaper requires signatories to maintain “appropriatemeasures” for the purpose of preventing major sup-pliers from engaging in anti-competitive practices.One of the practices identified is using informationobtained from competitors with anti-competitiveresults.

A national example of a prohibition against competi-tive misuse of information can be found in theGeneral Licence issued by the Irish regulator.Condition 20 of that licence deals with misuse ofdata in the following terms:

“The Licensee shall not make use of network ortraffic data, traffic profiles or any other data ofany nature, and which are not otherwise publiclyavailable and which become available to theLicensee directly or indirectly either as a result ofentering into interconnection arrangements or


3 - 37

Interconnection

otherwise as a result of carrying telecommunica-tions messages, in such a way which, in thereasonable opinion of the Director, would undulyprefer the interests of any business carried on bythe Licensee or an Affiliate or place personscompeting with that business at an unfairdisadvantage.” (OTDR (1998))

A good approach to preventing abuse of competitiveinformation is the establishment of an Interconnec-tion Services Group (ISG). This is sometimes calleda Carrier Services Group. The idea is to establish aseparate organization within the incumbent operator,whose role it is to handle interconnection-relateddealings between that operator and interconnectingoperators. For example, all orders by interconnect-ing carriers for interconnection links, additionalcapacity and customer access lines would besubmitted to the ISG. The ISG will process theorders.

Safeguards will be put in place to ensure that infor-mation obtained by the ISG is not used for improperpurposes. For example, where a new entrant ordersan access line from the incumbent operator to servea new customer, the ISG should not pass thatinformation on to the marketing department of theoperator to try to “snare” or “win-back” the customerbefore the access line is installed. Confidentialitysafeguards should include codes of conduct withmandatory suspension or termination of employeeswho “leak information”. Separate office space,locked filing cabinets, audits and other measurescan help ensure confidentiality of ISG information.

3.4.3 Treatment of Customer Information

Monopoly providers of local telephone services arein a position to collect information on their custom-ers. Such information may include names,addresses and telephone numbers, as well asinformation on monthly billing levels, calling patterns,percentage of calls unanswered, etc. Customerinformation of this type can be very valuable in mar-keting new services. For example, customers withvery long calls may be heavy Internet users to whomInternet services can be successfully marketed.Users with many missed calls make good customersfor voice-messaging services. Customers with highinternational calling would be good targets to tie up

in long-term contracts if a competitive internationalservice operator is about to be licensed.

In some countries, including the US and Canada,regulatory restrictions are imposed on the use ofcustomer information. Some of these rules areaimed at protecting the privacy of customers. Forexample, customers typically do not want the worldto know what phone numbers they call.

Another example of a regulatory restriction is foundin the European Union data protection directives andin related laws of EU Member States. These lawsimpose specific obligations on telecommunicationsservice providers regarding the use that can bemade of billing and other customer data, including aprohibition against using such information to markettelecommunications services to customers unlessthe customer has consented to that use of its data.Other countries have implemented, or are consider-ing similar consumer protection rules.

Other restrictions are aimed at preventing anti-competitive use of customer information gathered bymonopoly operators that have competitive opera-tions or affiliates. Such rules may require amonopoly local operator, for example, to share anycustomer information that it provides to itscompetitive operations or affiliates with intercon-necting operators or other direct competitors in thesame business line. For example, if a local monop-oly operator’s long distance services division collectsinformation to identify heavy Internet users to help itsInternet division sell services, it would be required toprovide the same information to competitive InternetService Providers.

These restrictions are based on the assumption thatthe local monopoly service provider is in a position tocollect the information solely due to its monopolyposition. Distribution of this type of information canbe handled through an Interconnection ServiceGroup (see Section 3.4.2).

3.4.4 Points of Interconnection

The interconnection policies of many countriesrequire incumbent operators to permit interconnec-tion with their networks at any technically feasiblepoint. This policy is reinforced by the WTORegulation Reference Paper, which requires


3 - 38

signatory countries to ensure interconnection at anytechnically feasible point with their major suppliers.

Interconnection agreements and regulatory ordershave established different interconnection points indifferent countries. Box 3-5 provides examples oftechnically feasible interconnection points that havebeen prescribed by regulators or established in in-terconnection agreements.

The definition of technically feasible interconnectionpoints is not static. Telecommunications networkscontinue to evolve. As new technologies, such asthose based on the Internet Protocol and digitalsubscriber loops, are rolled out, it is becoming tech-nically feasible to interconnect networks at differentpoints. Therefore, interconnection agreements andregulatory directives should not prescribe limitationson the points of interconnection that will be permit-ted. It should be open to interconnecting operatorsto propose interconnection at different points asnetworks evolve.

The costs of interconnection incurred by bothoperators will vary depending on the points of inter-connection. Incumbents will sometimes proposestandard points of interconnection of their networkswith other operators. These standard points of inter-connection may be set out in the “reference

interconnection offers” major suppliers are requiredto make available pursuant to the WTO RegulationReference Paper.

In some cases, new entrants may wish to intercon-nect at points other than the standard points. In suchcases, the Reference Paper provides that suchinterconnection should be made available uponrequest. However, the requesting party may berequired to pay charges that reflect the cost of con-struction of necessary additional facilities.

A variation on the theme of interconnection at non-standard points can be found in a recent regulatorydecision in the United Kingdom on Third Generationcellular services. The UK regulators have recentlyruled that new Third Generation cellular networksshould have access to earlier generation cellularnetworks at points around the country, by means ofa compulsory roaming arrangement. This example isset out at Box 3-6.

3.4.5 Access to Unbundled NetworkComponents

In an increasing number of countries, telecommu-nications policies require incumbent operators toprovide competitors with access to unbundled

Box 3-5: Examples of Technically Feasible Interconnection Points

➢ The trunk interconnection points of local and national tandem exchanges (most common point ofinterconnection or POI)

➢ The national or international circuit interconnection points of international gateway exchanges➢ The trunk side of local exchanges➢ The line side of local exchanges (e.g. at the main distribution frame (MDF) or Digital Distribution Frame

(DDF)➢ Cross-connect points of any exchange➢ “Meet points” at which operators agree to interconnect➢ Signaling transfer points (STF) and other points outside of the communications channel or band, where

interconnection is required for CCS7 or other signaling to exchange traffic efficiently and to access call-related databases (e.g. a Local Number Portability (LNP) database).

➢ Access points for unbundled network components➢ Cable landing stations


3 - 39

Interconnection

network components. Unbundling generally refers tothe provision of network components on a stand-alone basis. Unbundling permits interconnectingoperators to access a single unbundled componentwithout an obligation to buy other components aspart of an “interconnection service”.

There are many possible types of unbundlednetwork components. The policies of some countriesrequire provision of certain features, functions andservices on an unbundled basis – as well as certainphysical facilities. These features, functions andservices may be associated with transmission orswitching facilities. They may also be associatedwith software facilities, such as databases that

support the efficient provision of telecommunicationsservices. Examples include access to directoryinformation databases, operator services andsubscriber listings in telephone directories.

In this Module, we will use the term “networkcomponents” to refer to both physical networkfacilities and these “non-physical” features, functionsand services. Box 3-7 lists examples of unbundlednetwork components.

Unbundling of the local loop is a special case ofunbundling that is currently being addressed byregulators in many countries. It is dealt with in moredetail in the next Section.

Box 3-6: Compulsory National Roaming in the UK

Background:

As part of the process leading to the licensing of “Third Generation” cellular wireless networks in the UK,Oftel and the Department of Trade and Industry ("DTI") dealt with the issue of compulsory roaming. Theregulators determined that any existing wireless network operator which participated in the auction toobtain spectrum for Third Generation network services would be required to accept a licence modificationobligating the operator to negotiate an interconnection agreement to provide national roaming access tonew entrants. The aim was to prevent incumbent operators from using their existing wireless networks toan unfair competitive advantage while new entrants built up their networks and territorial coverage. Ineffect, DTI and Oftel determined that access to earlier generation networks was an essential facility to bemade available to new entrant competitors. (The concept of essential facilities is discussed in the nextSection.)

The Nature of Roaming:

Roaming is typically an arrangement between wireless network operators or services providers to allowaccess by one service provider's customers to the network or services of another service provider locatedoutside the service area of the first service provider. Roaming arrangements require the implementation ofsubscriber authorization and billing systems. They also require appropriate technical and spectrumcapacity arrangements to be at all points of access by customers of roaming operators.

The Requirements of National Roaming:

DTI and Oftel intend to make what was previously a system of negotiated interconnection among non-competing wireless operators a compulsory arrangement between incumbents and a new entrant. Nationalroaming is to be made available on a non-discriminatory basis. Oftel will deem the incumbent to have costsof roaming services equal to the rates for roaming services charged to competitors. Oftel will then includesuch deemed costs in determining whether the service charges of incumbents are sufficient to cover costsand make an adequate return. National roaming services will not be available to a competitor before thecompetitor has achieved network roll-out covering at least 20% of the UK population, and may expire anytime after 31 December 2009. Roaming charges are to be determined on a “retail minus” rather than “costplus” basis (meaning that roaming charges will be derived from end user charges, less a discountreflecting elements of cost not incurred in providing the roaming service rather than an end user service).


3 - 40

Decisions as to what components to unbundle andhow to unbundle them are sometimes left to nego-tiations between operators. According to theJapanese interconnection policy, for example,unbundling should be promoted as much aspossible through a process which takes intoconsideration the opinions of carriers other than theincumbent. However, the Japanese policy alsoindicates that the regulator should be involved ifnegotiations fail. In practice, for the reasonsdiscussed below, negotiated unbundling arrange-ments are generally unsatisfactory in the long run.The incumbent has little incentive to unbundle itsnetwork sufficiently to permit competitors to operatevery effectively.

Rationale for Unbundling

The purpose of unbundling policies is to lowereconomic and technical barriers to competitive entry.The large capital costs of building duplicatenetworks raise a significant barrier to entry.Competitors may not be willing or able to finance theconstruction of complete networks. However, theymay be willing to build parts of such networks. Forexample, they may build certain switches, inter-exchange transmission facilities, and access lines ina limited number of locations. If the regulatory

framework permits, competitors can then obtainother network components, such as switchingcapability and access lines in other locations, fromthe incumbent. This permits new entrants to mixtheir self-built network components with those of theincumbent in an efficient manner.

The ability to mix self-built network components andthose of the incumbent will increase the viability ofthe business case for competitive entry in manycountries. Thus, competition will emerge where itotherwise would not. The use of the incumbentoperator’s network components by competitors willoften be transitional. Over time, the competitor willbuild more of its own facilities and become a full-fledged facilities-based operator.

Many incumbents are unwilling to provide competi-tors with access to unbundled network componentsunless they are required to do so by regulation.While the issue is still controversial in somecountries, and among some experts, mandatorynetwork unbundling is becoming more common.

Unbundling Policies

The trend to unbundling was given a strong impetusin the WTO Regulation Reference Paper. The

Box 3-7: Some Possible Unbundled Network Components and Services

➢ Network access lines (local loops and related functions)

➢ Local switching functions

➢ Tandem switching functions

➢ Inter-exchange transmission (e.g. between local and tandem switches)

➢ Access to signaling links and signal transfer points (STPs)

➢ Access to call-related databases (e.g. line information, toll-free calling and number portability databases)

➢ Central office codes (NNXs)

➢ Subscriber listings (in telephone directories and directory databases)

➢ Operator services

➢ Directory assistance functions

➢ Operations support systems (OSS) functions


3 - 41

Interconnection

Reference Paper states that major suppliers mustprovide interconnection on a basis that is sufficientlyunbundled so that a supplier need not pay fornetwork components or facilities that it does notrequire for the service to be provided. While thisstatement is supportive of unbundling policies, it isquite general. It provides little guidance for thedevelopment of national unbundling policies.Unbundling policies are still in the early stages ofdevelopment in many countries.

Unbundling policies have developed in the US,Canada, Australia, Singapore, Hong Kong and othercountries, including, more recently, the EU. The newregulatory framework for electronic communicationsservices proposed by the European Commission on12 July 2000 provides a strong new impetus forimplementation of national unbundling policies.Particularly significant in this regard, is the EU’s newregulation on local loop unbundling, which will comeinto force on 31 December 2000.

Unbundling has also been required in other EUregulatory documents. Article 7(4) of the EUInterconnection Directive provides that interconnec-tion charges must be sufficiently unbundled so thatan applicant for interconnection is not required to

pay for anything that is not strictly related to theservice requested. Similarly, Article 7(4) of theRevised Voice Telephony Directive (Directive98/10/EC) states that:

“Tariffs for facilities additional to the provision ofconnection to the fixed public telephone networkand fixed public telephone services shall, inaccordance with Community law, be sufficientlyunbundled so that the user is not required to payfor facilities which are not necessary for theservice requested."

Advantages and Disadvantages of Unbundling

There are some disadvantages to a full-scale man-datory unbundling policy. In particular, it can act as adisincentive to the construction of competitivenetwork components, and the development of truefacilities-based competition. However, thedisadvantages appear to be outweighed by theadvantages. Moreover, the potential disadvantagescan generally be avoided if the pricing and otherterms of the unbundling guidelines are properly set.The main advantages and disadvantages of amandatory unbundling policy are summarized inTable 3-4.

Table 3-4: Advantages and Disadvantages of Unbundling

Advantages➢ Reduces economic barriers to entry, by

allowing new entrants to construct somecomponents of their networks and obtain othercomponents from the incumbent operator

➢ Encourages innovation, since new entrants cancombine new technologies (e.g. ADSL and IPdata/voice switches) with components ofexisting networks (e.g. access lines)

➢ Avoids unnecessary duplication of components(e.g. access lines in remote areas,transmission tower space)

➢ Facilitates access to rights of way, towers, etc.by new entrants (in many countries it can bevery time consuming and expensive to obtainsuch rights)

Disadvantages➢ Reduces incentive for construction of

competitive network facilities (depending onthe availability and price of unbundledcomponents)

➢ Can enrich the new entrant at the expense ofthe incumbent operator (if unbundledcomponent prices are set below costs)

➢ Requires detailed regulatory intervention andtechnical co-ordination


3 - 42

Regulatory Approaches to Unbundling

Given the potential disadvantages of a mandatoryunbundling policy, some regulators have adoptedmodified approaches to such a policy. Theseapproaches are intended to achieve someadvantages and avoid some disadvantages of poli-cies that require unbundling of all networkcomponents. Some of these approaches may besummarized as follows.

➢ Transitional Unbundling Requirements -Access to certain types of unbundledcomponents may be required for a limited periodof time. This approach can apply, for example,to access lines (loops) in urban areas.Unbundling of access lines might be required forthe first five years after a market opening. Thus,competitors can use the incumbent’s accesslines to “jumpstart” competition. However, theywill have to construct their own access lines byyear five, in order to maintain networkconnections with their customers. In theory, thisapproach will encourage the development ofcompetition in the short term. At the same time itshould promote development of completefacilities-based competition over the mid tolonger term. Local loop unbundling is describedfurther in the following Section of this Module.

➢ Selective Unbundling Requirements - Someunbundling policies distinguish between networkcomponents. They require unbundling of someand not others. Unbundled access may berequired only for certain types of components.For example, unbundled access may berequired for network components in caseswhere construction of duplicate componentswould cause environmental damage or publicinconvenience. Thus, incumbents might berequired to provide access to towers, poles,conduits, ducts, aerial access lines and insidewiring, where a proliferation of such facilitieswould degrade the environment, disrupt publicroads, and/or otherwise inconvenience thepublic. The same may be true of access lines orswitching facilities in architecturally or culturallyimportant areas. Such access might be requiredover the long term as well as the short term.

Many countries are still developing policies onnetwork unbundling. Unbundling policies vary fromcountry to country, depending on the conditions oflocal telecommunications markets. It is arguable thatmandatory unbundling is less desirable in countrieswith very limited telecommunications networkinfrastructure and large pent-up demand. In suchless developed countries, mandatory unbundlingmay reduce the incentive to build much-needed newinfrastructure. On the other hand, in some lessdeveloped countries, the business case for newentry may not be viable without mandatoryunbundling. Each telecommunications marketshould be carefully assessed to determine the roleunbundling policies should play in sector develop-ment.

3.4.6 Local Loop Unbundling

Mandatory unbundling of local loops is increasinglybeing used as a regulatory tool to accelerate com-petition in local access markets. Around the world,telecommunications network competition hasdeveloped most rapidly in the long-distance andinternational markets. Local access markets aregenerally less competitive. Wireless services cur-rently provide an alternative means of localnarrowband access in many markets, andbroadband competition is starting. However, wirelineservices still provide the main means of local accessaround the world. There, high entry costs and lowmargins have discouraged competition.

Competition in local access is increasingly seen asan important policy objective. One reason is theperceived need to provide more competition in high-speed access markets in order to accelerate the rollout of Internet, e-commerce and video services.Many regulators and policy makers see such com-petition as necessary to maintain or increase thecompetitiveness of their national economies.

Regulators have now mandated unbundled accessto local loops in a range of different economies. Atone end of the income spectrum, these countriesinclude the US, Australia, Canada, Singapore andthe EU members. Unbundled loop access has alsobeen mandated in a number of middle incomecountries, such as Mexico and the Slovak Republic,as well as in lower income countries, such asAlbania, Guatemala, Kyrgistan and Pakistan.


3 - 43

Interconnection

Types of Local Loop Unbundling

Local loop unbundling regimes typically requireincumbent operators to provide access to their localloops to competitors. Other third parties, such ascustomers, may sometimes also obtain unbundledaccess. Access to local loops is provided at a pointof interconnection somewhere between the networktermination point on the customer premises and theline-side of the access network operator's localswitch. From this point of interconnection, the com-petitor will obtain dedicated or shared access to thelocal loop. The competitor will thus be able to usethe loop as a direct transmission medium betweenits network and the customer's premises.

Various technical options are available for local loopunbundling. In its proceedings on unbundled accessto the local loop in early 2000, the EuropeanCommission’s DGIS focussed on three main optionsfor access to local loops:

➢ Full unbundling of the local loop (unbundledaccess to the copper pair for competitiveprovision of advanced services by third parties);

➢ Shared use of the copper line (unbundledaccess to the high frequency spectrum of the

local loop for the competitive provision of DigitalSubscriber Loop (DSL) systems and services bythird parties); and

➢ High speed bit stream access (provision ofxDSL services by the incumbent).

Although different approaches are possible, thesethree are the main ones in use today. Each of themis described in greater detail below.

Full Unbundling (Copper Loop Rental)

Full unbundling can provide new entrants withaccess to raw copper local loops (copper terminatingat the local switch) and sub-loops (copper terminat-ing at the remote concentrator or equivalent facility).In the case of unbundling at the local switch, the linkbetween the main distribution frame (MDF) and thelocal switching equipment on the incumbent’spremises is re-routed and connected to the newentrant’s switch. The new entrant takes over theoperation of the local loop.

Figure 3-1 illustrates this type of full unbundling of alocal loop. The illustrated case assumes that the

Figure 3-1: Full Unbundling – Local Loop

L o c a l lo o p

L in k re - ro u te d fro m in c u m b e n t’ss w itc h to n e w e n tra n t ’s

MDF

In c u m b e n t’slo c a l s w itc h

T o n e we n tra n t ’s

s w itc h

P S T N

P S T N

Source: Adapted from CEC (2000b)


3 - 44

customer has decided to change telecommunica-tions service suppliers. The local loop that previouslyconnected the customer to the incumbent’s switchhas been re-routed to connect it to the new entrant’sswitch. The new entrant will then use the unbundledlocal loop to provide an alternative local accessservice to that previously provided by the incumbent.

Figure 3-2 illustrates full unbundling in a case wherethere are two local loops to a customer’s premises.One loop is unbundled by the incumbent and re-configured to connect the customer to the newentrant’s network. The other loop continues toconnect the customer to the incumbent’s network. Asimilar approach would apply where there are threeor more loops to a customer’s premises. In eachcase, the customer could decide how many loops itwanted connected to different operators. Theapproach illustrated in Figure 3-2 would be usedwhere a customer wants to retain its basic telephoneservice with the incumbent. It can do so and, forexample, at the same time have a dedicated con-nection to a new entrant’s xDSL services to accesshigh-speed data services (e.g. Internet or videoservices).

Full unbundling of the type illustrated in Figure 3-1and Figure 3-2 essentially involves rental of adedicated copper loop by the incumbent to a newentrant. Such copper loop rental provides the newentrant with direct access to and use of the copperloop. This allows new entrants to operate their ownend-to-end transmission systems. Such operationalcontrol can be important to ensure the integrity andquality of high-speed services.

Although Figure 3-1 and Figure 3-2 indicate that thepoint of interconnection is at the distribution framewhere the copper loop terminates, it is also possibleto locate the point of interconnection at a remoteconcentrator unit (remote line unit).

Shared Use of the Copper Loop

An alternative means of providing access to the localloop involves shared access rather than exclusiveaccess by a new entrant. In this form of unbundling,the incumbent and the new entrant provide servicesover the same loop.

Figure 3-3 illustrates one form of sharing the localloop. In this case, the customer will continue to

Figure 3-2: Full Unbundling – Two Local Loops

S eco ndL oca l loo p

M

D

F

In cum bent’slo ca l sw itch

P S T N

X D S Lm od em

N e w en tran t’sD S L accessm ultip lexe r

(D S L A M )

N ew en tran tP C or o ther

cu s tom er eq u ip m e nt

F irst loca l loo p



3 - 45

Interconnection

Figure 3-3: Shared Use of Copper Loop Using Splitter

sp litte r

te lLoca l loop

A D S Lm odem

N ewe ntran t

Incum be nt’slo cal sw itch P S TN

M

D

Fdata

P C or other custom erequ ipm ent

D S L accessm ultip lexer(D S LAM )


receive basic PSTN services from the incumbent,and at the same time, receive DSL access servicesfrom a new entrant. As illustrated, a splitter is locatedbetween the MDF and the incumbent’s local switch.The splitter is connected to both the incumbent’sswitch and to a DSL access multiplexer (DSLAM)connected to the new entrant’s high-speed network.

As indicated, the splitter separates telephone anddata traffic. Thus, the voice frequencies of the loopcontinue to be used by the incumbent. The non-voice frequencies are made available to the newentrant to provide high-speed services. In effect, thisarrangement provides unbundled access to the highfrequency spectrum of the local loop for the com-petitive provision of Digital Subscriber Loop (DSL)services by new entrants.

Shared use of copper line can provide a cost-effective solution for some customers. For example,it permits a customer to retain the incumbent as itstelephone service provider, and at the same time,select a new entrant to provide high-speed Internetservice over the same loop.

High-speed Bit Stream Access

A third approach to providing access to the localloop involves provision by an incumbent of a high-

speed bit stream to new entrants. To do this, theincumbent would install a high-speed access link tothe customers’ premises and then make it availableto other operators to enable them to provide high-speed services. Provision of bit stream accessservices requires provision of both the transmissionmedium (e.g. copper cables, coaxial cables andoptical fibre cables) and the transmission system(e.g. synchronous digital hierarchy transmission onoptical fibres and xDSL transmission on coppercables).

In the case of high-speed bit stream access, thepoint of interconnection will usually be at the incum-bent’s local switch, but circuits could be back-hauledto points of interconnection further up the switchinghierarchy. Technically, bit stream access can beprovided to any transmission system, since it onlyrequires reservation of a specified bandwidth, ratherthan dedicated use of a physical loop. This accessarrangement does not entail any unbundling of acopper pair. Rather it uses the higher frequencies ofthe copper local loop, as in the case of shared useof the copper line.

Providing high-speed bit stream service can beattractive for incumbent operators as it does notinvolve physical access to copper pairs. As a result,for example, it would not hinder the progressive


3 - 46

modernization of the local access network byreplacing copper with fibre.

Figure 3-4 illustrates the provision of high-speed bitstream access by an incumbent. In this example,two customers obtain high-speed data services fromtwo different service providers, the incumbent and anew entrant. At the same time, the incumbentcontinues to provide basic PSTN services to bothcustomers.

The three means of access to the local loop referredto above are not necessarily mutually exclusive.Where regulators mandate local loop access, theymay require or permit incumbent operators toprovide one or more alternative forms of access.

Advantages and Disadvantages of Unbundlingthe Local Loop

The main reason regulators have required incum-bents to unbundle their local loops is to promotecompetition and innovation in access and advancedhigh-speed services. However, there continues to bean active debate on the merits of mandatory loopunbundling. There remain arguments against it, aswell as for it. Table 3-5 summarizes the pros andcons of mandatory loop unbundling.

Implementation of Local Loop Unbundling

Different approaches may be used in mandating andregulating local loop unbundling. The appropriateapproach will often depend on the state of competi-tion in the relevant market for local access. Possibleapproaches include:

➢ Mandatory loop access without specification ofthe type of access arrangement. In this case, itis likely many incumbents will choose to offer bitstream access, which enables them to retaingreater management control and possibly obtainhigher access charges from competitors. Thedisadvantage of this approach is thatcompetition may be delayed. Incumbentoperators will have little incentive to accelerateimplementation of bit stream accessarrangements, at least until they are positionedto provide competitive services.

➢ Requiring bit stream access only (see previouspoint – same considerations apply).

➢ Requiring all three forms of access describedabove, except where the incumbent can

Figure 3-4: Provision of High-Speed Bit Stream Access

Custom er ofincum bent’s

dataservices

Loca l loop

Local loop

P STN

H igh speed bits tream serv iceprovided to oneor m ore newentran ts

splittersMDF

DS LAMopera ted

byincum bent

C ustom er ofnew entrant’sdata serv ices



3 - 47

Interconnection

demonstrate significant problems with dedicatedloop rentals.

➢ Requiring all three forms of access in some orall national markets.

Various other regulatory approaches to unbundlingmay be developed.

Local loop unbundling may be a transitionalphenomenon in some areas. Unbundling of loopsmay be required, for example, to facilitate competi-tion in the short term. This will enable new entrantsto roll out service rapidly, while they are constructingalternative access networks in the areas where thereis sufficient demand.

Implementation of local loop unbundling continues tobe a novel issue for regulators in many countries. Amajor source of experience to date is the United

States. In the US, the 1996 TelecommunicationsAct requires incumbents to offer access to unbun-dled network elements and to making retail servicesavailable at wholesale prices. The US regulator hasstated that “[p]reventing access to unbundled localloops would either discourage a potential competitorfrom entering the market in that area, therebydenying those consumers the benefits of competi-tion, or cause the competitor to constructunnecessarily duplicative facilities, therebymisallocating societal resources” (FCC, First Reportand Order in the Matter of the Implementation ofthe Local Competition Provisions in theTelecommunications Act of 1996). The FCC andUS state regulators have subsequently taken furthersteps to facilitate loop unbundling.

As of June 1999, approximately 685,000 loops hadbeen provided to competitors in the US as unbun-dled network elements. This represented an

Table 3-5: Arguments For and Against Local Loop Unbundling

Pros Cons

➢ Accelerates introduction of local accesscompetition, including xDSL access

➢ Reduces incentive to build alternative accessnetworks and more sustainable facilities-basedcompetition

➢ Accelerates competition, service innovationand roll out for high speed services, including:

➢ Internet services

➢ Video services (including interactive ones)

➢ E-commerce

➢ other data services

➢ May undermine investment in alternativeaccess networks (wireline and wireless)

➢ May complicate modernization of incumbentoperators’ networks (e.g. if some access loopsare dedicated to competitors use)

➢ Requires prolonged and detailed regulatoryintervention compared to facilities-based accesscompetition

➢ Avoids duplication of access networks, andincreases network operating efficiencies

➢ Requires more technical co-ordination betweenoperators compared to facilities-based accesscompetition

➢ Provides new revenue streams to incumbent(which may or may not exceed existingrevenues from loops, depending on tariffs)

➢ Reduces disruption of streets andenvironment due to construction of newaccess networks


3 - 48

increase of 180 percent over the previous year. Inaddition, competitors had collocation arrangementsin exchanges covering 60 percent of all lines in theUS (compared with 32 percent of all lines theprevious year). By the end of 1999, competitors hadprovided 117,000 xDSL lines, up from 1,500 lines in1997, while incumbents provided 386,000 DSL lines,up from 32,000 at the end of 1998. Competitors hadinstalled over 1,400 data switches, a fivefoldincrease over 1997. Recent estimates suggest thatabout 60% of the US population had access to DSLat the beginning of 2000, with 25% located in citieswith four or more DSL providers.

In July 2000, the European Union adopted aRegulation on Unbundled Access to the LocalLoop. The regulation will be binding on dominantoperators in EU Member States, as of 31 December2000. Issuance of the regulation is based on theassumption that providing access to the local loop toall new entrants will increase the level of competitionand technological innovation in the local accessnetwork, and in turn stimulate the competitive provi-sion of a full range of telecommunications servicesfrom simple voice telephony to broadband services.The regulation is aimed, in part, at ensuring that theEU does not fall further behind the US in thedeployment of high speed access and the advancedservices it enables.

The European regulation requires dominantoperators to provide physical access to third partiesat any technically feasible point of the copper localloop or sub-loop. The third party can locate andconnect its own network equipment and facilities atsuch points (i.e. at the local switch, concentrator orequivalent facility) in order to deliver services to itscustomers. Dominant operators are required tomake unbundled loop access available to thirdparties under transparent, fair and non-discriminatory conditions. In addition, the regulationprovides that the dominant operators must providecompetitors with the same facilities as they provideto themselves or their associated companies, andwith the same conditions and times. Regulators aregiven authority to intervene in pricing issues andresolve disputes in connection with the regulation.

Experience in other jurisdictions suggests thatregulatory guidance is required in determining thepricing (and costing) of unbundled local loops.

Operator negotiations, or unilateral price setting byincumbents can result in anti-competitive pricing.Where advance regulatory guidelines are notestablished, ex post regulatory intervention will oftenbe required. A recent Australian case illustrates thepoint. In early August 2000, the Australian regulator,the ACCC, found that prices imposed by thedominant operator (Telstra) on competitors for localloop access were too high.

3.4.7 Sharing of Infrastructure andCollocation

Extensive infrastructure is required to build tele-communications networks. Key supportinginfrastructure includes poles, ducts, conduits,trenches, manholes, street pedestals, and towers.Sharing of such infrastructure can significantlyincrease the efficiency of telecommunications supplyin an economy. The same is true in the case ofsharing building space in exchanges to permit two ormore operators to “co-locate” their cable and radiotransmission facilities and related equipment.Collocation permits direct (or near-direct) access toexchange switches and local access lines.

Availability of infrastructure sharing and collocationcan significantly decrease barriers to competitiveentry. The acquisition of rights of way and otherpermits required to build pole lines or towers, digtrenches or install ducts and conduits can be verytime consuming and expensive. In some countries,only government entities, such as the incumbentoperator, have clear legal authority to obtain rights ofway, occupy public property or expropriate privateproperty. Sharing of infrastructure and collocationcan reduce costs for the new entrant, and at thesame, time provide additional revenues to incum-bents.

An added benefit is reduced environmental impactand public inconvenience. Competitive entry intotelecommunications markets has led to a prolifera-tion of cellular and microwave towers, aerial polelines and road trenches in many countries. Thisresult has become an increasing concern for manymunicipalities and other local administrations.

Some regulators require incumbents to permit infra-structure sharing and collocation of a new operator’stransmission facilities in their exchanges. Other


3 - 49

Interconnection

operators, including new entrants, are frequentlyrequired to cooperate as well, at least in the sharingof infrastructure that is seen to be environmentallydegrading, such as towers. In some countries, thirdparties that own support infrastructure, such aselectrical power utilities, are also encouraged toparticipate in sharing arrangements.

In some jurisdictions, sharing of infrastructure occurswithout regulatory intervention. Both sharing partiescan benefit from the arrangements. In these jurisdic-tions, sharing of infrastructure is often seen as amatter to be freely negotiated between operators.However, as with other interconnection issues, thereis often an asymmetrical market situation. In somecases, incumbents resist sharing their infrastructure.In these markets, regulatory intervention will berequired to implement efficient sharing and colloca-tion arrangements.

Table 3-6 lists steps regulators can take to promotesharing of infrastructure and collocation.

Once there is clear regulatory direction that infra-structure sharing and collocation must be permitted,operators are sometimes able to negotiate mutuallyacceptable sharing arrangements. In many othercases, however, regulatory direction or disputeresolution has been required to finalize sharingarrangements. Regulators seeking to expeditesharing arrangements may want to provide advanceguidelines on such arrangements, after taking intoaccount the views of incumbents and new entrants.

Some of the main issues that have arisen in relationto infrastructure sharing and collocation are:

➢ Rationing of space between incumbents’ futurerequirements and current and futurerequirements of various new entrants;reservation of future expansion space for eachoperator.

➢ Pricing of facilities, and costing basis for thesame.

➢ Access and security arrangements for variousoperators’ equipment. Collocation premises ofdifferent operators are usually separatedphysically (e.g. by wire mesh) and locked.

➢ Appointment and supervision process for mutualcut-overs and work affecting more than oneoperator’s facilities. Payment and rates for thesame.

➢ Provision and pricing of ancillary services suchas electrical power and back-up power, lighting,heating and air conditioning, security and alarmsystems, maintenance and janitorial services,etc.

➢ Negotiation of other lease and/or licencearrangements, including issues of sub-licenceson property of third parties (e.g. building owners,right of way owners, municipal and other publicproperty owners), insurance and indemnificationfor damages.

3.4.8 Equal Access

On a level competitive playing field, telecommunica-tions users should be able to access the services ofnew entrants as easily as those of incumbentoperators. Without equal ease of access, newentrants will find it difficult to attract customers. Whileaccess need not be exactly equal, accessing acompetitor should not be significantly more difficult.

In the early days of long-distance competition inCanada and the US, for example, customers wereoften required to dial up to 20 or more extra digits toroute calls to new entrants’ networks. This significantdifference in access was due to the historical designof the PSTN. The operators’ switches had beenprogrammed for a monopoly environment. The addi-tional digits were required to permit the operators’switching software to identify the new entrant towhich the call should be routed as well as to providebilling details for the customer. It is not surprisingthat the new entrants initially found it difficult to en-courage customers to switch services from theincumbents.

Over time, many incumbents and telecommunica-tions equipment manufacturers redesigned theirswitches and related software. These facilities arenow far more adaptable to the requirements of amulti-operator environment. Dialling parity is easy toachieve with the right software package. This hasmade it much easier to implement equal access.


3 - 50

However, changes in incumbent procedures and theregulatory environment are also required to facilitate

equal access in a previously monopoly environment.

Table 3-6: Steps to Promote Infrastructure Sharing and Collocation

DevelopRegulatory Policy

➢ Publish a regulatory policy encouraging infrastructure sharing and collocation

➢ Encourage local authorities, such as municipal governments to support andfacilitate infrastructure sharing

➢ Encourage reciprocity of infrastructure sharing (i.e. new entrants should berequired to size and build their facilities to permit sharing with incumbents andother operators)

➢ Require incumbent operator to publish a standard offer and price list for accessto key infrastructure components: poles, ducts, conduits, tower space, etc.

➢ Incumbents should be required to provide information on the location ofinfrastructure, and capacity available for sharing (e.g. excess capacity in ducts,towers, etc.)

➢ A joint committee of operators should be established to plan infrastructurecapacity, co-ordinate permits from local authorities and improve the mutualefficiency of the infrastructure provisioning process

➢ Operators should be able to reserve capacity in advance on reasonable terms

Price of Sharedand InfrastructureCollocation

Regulators should encourage development of clear pricing guidelines (the followingguidelines are illustrative only)

➢ Normally, incumbents and other operators should be able to recover at leasttheir direct incremental costs of sharing, plus reasonable overheads

➢ Additional price components may be subject to negotiation and regulatorydispute resolution

➢ Prices for collocation and infrastructure sharing should generally be unbundledso that the operator requesting access is only required to pay for the services ituses

➢ Cost of new infrastructure should be shared among 2 or more operators inproportion to their use of the infrastructure (e.g. number of antennae locatedon a microwave tower)

➢ Costs of increased capacity and re-location of infrastructure should be sharedamong those that benefit from such works. Where an incumbent operatorreceives no benefit from works required to accommodate a new entrant, itshould normally not pay, unless and until it benefits from such works. Analternative approach is to allocate the costs among sharing operators based onuse, with a surcharge for the operator that requests the work.

➢ Future sharers of infrastructure should reimburse early entrants forexpenditures that benefit them

RegulatorySafeguards

➢ Shared infrastructures should be made available to all operators on a non-discriminatory basis. This includes the owner of the infrastructure. Capacityshould normally be provided on a first come, first served basis. The regulatorshould approve rationing schemes for scarce capacity.


3 - 51

Interconnection

Table 3-6: Steps to Promote Infrastructure Sharing and Collocation (cont’d)

➢ New entrants (or other operators) that do not use ordered infrastructurecapacity within a set time period should be required to return it. A penalty forexcessive orders may also be appropriate

➢ Operators that provide shared infrastructure should record and have availablefor regulatory review: provisioning times for their own operations andcompetitors

➢ Physical separation of infrastructure (e.g. by walls or fences) may bewarranted where necessary to prevent sabotage, but operators should beencouraged to share in the most efficient manner

There are basically two approaches to providingequal access:

➢ Call-by-call customer selection – Customersselect the operator of their choice for each call.They usually do this by dialing a short code orprefix for their selected operator. For example, inColombia, customers dial 09 to route nationalcalls through TELCOM’s network, 05 to routethem through Orbitel’s network, and 07 forETB’s network. The main requirements toprovide this type of equal access on an efficientbasis are:

➢ Trunk-side interconnection by new entrantsto incumbent switches.

➢ A numbering plan that allocates equivalentnumbers to the incumbent operators andnew entrants (For example similar accesscodes for long distance and internationalcompetitors; and equivalent blocks of accessnumbers for local and mobile operators).

➢ Provision of basic signalling services by in-cumbents to new entrants including CallingLine Identification (CLI); answer anddisconnect supervision.

➢ Appropriate billing and audit arrangements topermit direct billing by each operator or bill-ing by one and remission to the others. Forexample, the local operator might do allbilling and remit long distance charges to theother operators.

➢ Operator pre-selection – Under this approach,customers select a operator for some or all oftheir calling. For example, an operator otherthan the incumbent might be selected for alllong distance and international calling. After theselection is made, all calls from these customerswill be routed to the operator of choice until theirselection is changed. The main requirements forthis type of equal access are:

➢ Trunk-side interconnection by new entrantsto incumbent switches.

➢ Switch software features to identify customerselections and to route and bill callsappropriately to the selected operator.

➢ Appropriate billing and audit arrangements topermit direct billing by each operator or bill-ing by one and remission to the others. Aswith the call-by-call approach, the localoperator might do all billing and remit longdistance charges to the other operators.

The implementation of equal access has beenuneven around the world to date. It is available, forexample in Argentina, Australia, Canada, Chile,Hong Kong, and the US, but unavailable to date inmany other countries. Equal access is morecommon for international and local services but lessso for long distance services. In some countries,equal access is unavailable due to limitations ininstalled switching and software facilities. In others, itis due to delays in implementing a numbering planthat allocates equivalent numbers to competitors. In


3 - 52

some, regulators have simply not seen equal accessas a priority.

Market experience in more open markets has dem-onstrated that there is considerable inertia amongtelecommunications customers. Regulators thatwish to expedite the development of fully competitivemarkets will, therefore, want to consider equalaccess as a useful approach.

3.4.9 Quality of Service to InterconnectingOperators

It is good regulatory policy to require incumbentoperators to provide a reasonable quality of inter-connection services and facilities. Without such apolicy, it would be possible for an incumbent tofrustrate a competitor’s ability to provide competi-tively attractive services. For example, if anincumbent connected its own new customers’circuits within days, but delayed connection of acompetitor’s customers’ circuits for months,customers in a hurry would likely choose theincumbent’s services.

The WTO Regulation Reference Paper deals withquality of interconnection with major suppliers insignatory countries. It requires interconnection to beensured under terms and conditions that are no lessfavourable than those provided for their own similarservices. Interconnection must also be no lessfavourable than that provided to a major supplier’ssubsidiaries, its other affiliates or to non-affiliatedservice suppliers.

Similar types of policies in many countries require“non-discriminatory” interconnection by anincumbent. In practice, it is very difficult to ensurethe implementation of such policies. Many intercon-nection complaints of new entrants deal withunequal quality of interconnection as between theincumbent’s services and their own.

The practical tools available to a regulator topromote high quality interconnection are:

➢ Establishing interconnection quality of servicemonitoring requirements;

➢ Monitoring complaints seriously, and estab-lishing significant penalties for clearly unequalservice quality; and

➢ Establishing an independent InterconnectionServices Group within the incumbent’sorganization.

Quality of interconnection services can be monitoredby an Interconnection Services Group (ISG) (seeSection 3.4.2). The ISG should measure quality ofservice to interconnecting operators, and compare itto the incumbent’s self-provisioning. For example, itshould ensure that new circuits ordered byinterconnecting operators are provisioned, onaverage, within the same number of days as internalorders.

Table 3-7 provides examples of interconnectionquality of service measures. Where interconnectionservice problems are serious enough to warrantregulatory supervision, regulators can monitor thesemeasures. Regulators may also establish amonitoring regime in advance, to prevent problems.A monitoring regime may require reports fromincumbents on two types of quality of serviceperformance:

1. Absolute performance based on establishedstandards or international benchmarks, and

2. Relative performance by the incumbent in pro-viding interconnection facilities to itself and tointerconnecting operators.

Interconnection policy in some countries may requirean incumbent to provide superior interconnectionservices to interconnecting operators under somecircumstances. For example, it may be useful torequire an incumbent to provide interconnectingoperators with higher quality service than it normallyprovides for its own services – if the interconnectingoperator is willing to pay for the difference. Such anapproach has applications in industrialized countriesseeking to promote the provision of advancedtelecommunications services.


3 - 53

Interconnection

Table 3-7: Some Key Interconnection Quality of Service Measures

Provisioning Measures ➢ Average time for provisioning interconnection circuits and otherinterconnection facilities and services (including unbundledcomponents)

➢ Percentage of installation appointments met for competitors’ serviceinstallations

➢ Average time for processing changes in customers from incumbentoperator to competitor (in an equal access regime)

➢ Percentage of repair appointments met for competitors

➢ Comparative provisioning performance for (1) competitors, (2)affiliates, and (3) self-provisioning (including measures such as thoseset out in the previous points)

Switching and TransmissionQuality Measures

➢ Probability of blockage in peak hour on interconnecting circuits

➢ Transmission delay (ref: ITU-T recommendation G114)

➢ Transmission loss (loudness – ref: ITU-T recommendation P76)

➢ Noise and distortion (ref: ITU-T recommendations, including Q551-554, G123, G232, G712, P11)

➢ Other transmission quality standards (e.g. for digital services ref: ITU-T recommendations G821 re: bit errors and timing, and G113 re voicecoding problems, and for both analogue and digital services ref: ITU-T recommendations G122 re: echo and loss of stability; and P16 et.al re crosstalk).

This type of policy can also be useful in lessdeveloped countries. In many less developedcountries, the quality of service provided by anincumbent is below international standards. This lowquality of service is often due to financial constraintson the incumbent. In such cases, regulators shouldbe willing to promote improvement of the quality ofservice provided to a new entrant, provided the newentrants pays for it. For example, a new entrant maybe willing to pay for new trunk circuits between thepoint of interconnection at a congested customerservice exchange and a tandem exchange.

Such payments can be a win-win situation for theincumbent and new entrants. Arrangements of thistype are best negotiated between incumbents andinterconnecting operators. However, someregulatory supervision may be required to ensurenew entrants do not have to pay excessive charges.Similarly, the regulator may need to ensure that the

incumbent does not require payments from newentrants to construct facilities to improve theincumbent’s competitive advantage, as a conditionof providing an adequate quality of service.

3.4.10 Quality of Interconnected Services

The previous Section discussed the provision ofservices by incumbents to interconnecting operators.Regulators in most countries are also concernedwith the broader issue of the quality of service to thepublic. Many regulators established quality ofservice reporting systems during the time serviceswere provided in their countries on a monopolybasis.

To deal with the emergence of competition, somecountries have apportioned responsibility forproviding a prescribed quality of service amonginterconnecting operators. For example, in the UK,


3 - 54

the regulator prescribed maximum delays forinterconnecting operators. The purpose of thesemaximum delay standards was to ensure callsbetween operators met national transmission speedstandards. Customer PBX equipment at each end ofa call was allocated 5 milliseconds (ms); originatingand terminating local network operators 3 ms each;and the long distance network operator 7 ms, for atotal maximum delay of 23 ms.

Other countries have taken a more deregulatoryapproach. They have not imposed quality of servicereporting requirements on new entrants. This

approach is based on the assumption that newentrants will not be able to attract and retaincustomers if their quality of service does not matchor exceed that of the incumbent operator. Based onthe same approach, it should be possible to removeregulatory quality of service requirements fromincumbents once competition is well established andthey lose their market power.

As competition develops, it should be possible formore and more regulators to take the latterapproach. Regulation of service quality can then beleft to the market, rather than to regulators.

Documents

3 Interconnection(1)