· cost pricing of international services discourages ... to make trade-offs between these objectives in the ... that prices are more closely aligned with costs

��

��

��

��

��

��

��

��!��

��"�� #!��$��

�� !��%��

�� &��

'��(�� %�� )��

�� %��!��

��!��"#�$%��&� "��'()(*�*�'*'�

��

��

"��+��%��,��!!��%��,��-��.�/��

�� !��"��# ��$ %�� $��

��

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

��

��

�-� ��

�-� &""��)�� !�� "�� #�$�� "

�-� ��!�"�� .�� % "�� #&� '�� (�$ )��*�� +�� (�� (�$ ,��- .��/�� 0&�� '��/ �� #&� �� )�� #&� �� )�� 1��! (�$$�� 2�� (�$$�� 1 �� '��/�� 3 �� !��" �� % �� (�$ �� 1

�-� ��!�"�/�� .�� "�� #&� 45�� )�� "�� 6�� +�� 2�% ��

�-�- ��0��!�"� ��

�-�-$ ��&��

&""��1��2��*!,��0��

&""��1��2��'��(��0�� 3

��

��(�1��'�5 � �- 7��/�� 3'�5 � �- �*$�� '�� (�$ )��*�� +'�5 � �- 8�� *$�� (�� 0�� '�5 � �- '�� (�$ )��*�� 8�� '�5 � �- �� (�� )�� '�5 � !- 0�� )�� 1'�5 � 1- �� )�� )�� 45�&�� (&�� 3'�5 � 3- #�� )�� "'�5 � "- 9�% �� '��/�� (�� )��5�� : 45�*$�� 1'�5 � �+- 45�*$�� 8��5$�� $� �� (&�� (��

�� $�� "'�5 � ��- 7&� �� 45�� (�� (&��; �+'�5 � ��- �4(� #�� (�*$�� <�&��

��4��)�� - ��5 �� 4(� #�� =

�� .�� (�� )�� - ��5 �� 4(� '�� (&�� #�� )�� - �� (�$ �� #��* 0�� * �"3" � �""� �3)�� - �� (�$ �� #��* 0�� * �""� � �""� �3)�� - �4(� '�� #�� '��/� �!)�� !- �4(� �� #�� '��/� �!)�� 1- ��5 �� 4(� #�� = �� :

#&� >��& �� ? �1)�� 3- ��5 �� 4(� �� (&�� #�� 1)�� "- �� '�� "

��0��#�� - #�$�� 1#�� - �� 4�*�� #)� �� &� 8�� #�� - 0 ��**�� '��& #��*@ � �� (�$ �� #�� - A )�� 2�� (�$ �� #�� - A )�� (�$ �� 8� ��#�� !- ��**�� '��&*��/ 4�*�� A )�� BCD ��#�� 1- �� (�� 2�� (�$ �� !#�� 3- �� (�� (�$ �� &�*�� 8� �!#�� "- 6 )�� 45�*$�� E�&�� &�*� ��#�� +-4�*�� 8�� &� #�� <�� :

'�� 3#�� -�� +

��

��

��

��!��"��

#��$��$��%��#�0��$�,��

��,/��1��!#��%2

��-$$/��-��3�12��/��1114$--��24-�

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

%2��&��-$$/��-��5&��6

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

4 - 1

MODULE 4

PRICE REGULATION

4.1 Introduction

This Module discusses price regulation in the tele-communications sector. Before reading the Module,readers may want to review the section on theeconomic rationale for price regulation in the tele-communications sector that is found in Appendix Bof the Handbook. As indicated in Section 1.1 ofAppendix B, price regulation is normally justifiedwhen telecommunications markets fail to producecompetitive prices.

In this Module, we look more closely at the specificobjectives of price regulation and at the regulatoryapproaches used to achieve those objectives. Thebasic approaches to price regulation have evolvedwith the transformation of the telecommunicationssector from monopoly to competition. As regulatorshave increasingly recognized the benefits of compe-tition, they have adapted price regulation to takeadvantage of those benefits.

Today, price cap regulation is the most widelyaccepted form of price regulation in the sector.Because of its pre-eminence, a substantial part ofthis Module is devoted to price cap regulation.Before dealing with it, however, we discuss the ob-jectives of price regulation and review otherapproaches to price regulation, particularly Rate ofReturn (ROR) regulation and its variations.

4.1.1 Objectives of Price Regulation

Good price regulation mimics the results of efficientcompetition. However, price regulation may haveadditional objectives. The objectives of price regula-tion may be grouped into three broad categories:

➢ Financing objectives;

➢ Efficiency objectives; and

➢ Equity objectives.

Financing Objectives

An important objective of price regulation is toensure that regulated operators are permitted toearn sufficient revenue to finance on-goingoperations and future investments. The minimumamount of revenue associated with the financialobjective is often referred to as the operator’s“revenue requirement”. To mimic the effect of acompetitive market, the revenue requirement shouldideally match the amount required by an efficientoperator to finance its operations and investments.This aspect of the financial objective may beconsidered as setting a revenue “floor” for efficientoperators.

Telecommunications Regulation Handbook

4 - 2

Some traditional forms of price regulation, includingRate of Return regulation, do not allow operators toearn revenues in excess of their revenue require-ments. This aspect of the financial objective isassociated with preventing excessive revenuesassociated with monopoly or dominant marketpositions. It is discussed in greater detail in Sections1.1 and 1.2 of Appendix B of the Handbook. Thisaspect of the financing objective, which may be con-sidered a revenue “ceiling”, has been relaxed undersome specific conditions in other forms of priceregulation, particularly price cap regulation.

Efficiency Objectives

It is generally accepted that price regulation shouldpromote efficiency in the supply of telecommunica-tions services. However, efficiency can be measuredin different ways. Three main aspects of efficiencyare discussed below.

Allocative efficiency is achieved when theprices of services reflect their relative scarcity. Inan efficient market, prices will equal the marginalcost of producing each service. In the telecom-munications sector, prices of international andlong-distance services have traditionally been setsignificantly above their costs while local calls arepriced below theirs. This is viewed as anexample of allocative inefficiency. The above-cost pricing of international services discouragesconsumption of such services. On the otherhand, pricing local calls below cost encouragesconsumption beyond the level at which local callscan be economically provided. A more detaileddiscussion of allocative efficiency is presented inSection 1.2 of Appendix B of the Handbook.

Productive efficiency has two related aspects.One aspect relates to the most efficient mix ofinputs (capital, labour, etc.) for a given level ofoutput. Some forms of price regulation canreduce productive efficiency. Rate of Return(ROR) regulation, for example, is generallyviewed as encouraging operators to use an inef-ficiently high level of capital for its level of output.A second aspect of productive efficiency requiresthat the services be produced as efficiently aspossible, that is by minimizing all inputs. Therelated concept of x-efficiency describes a situa-tion in which an operator’s costs are not

minimized because the actual output from thegiven inputs is less than what could be achieved.

Dynamic efficiency is achieved when resourcesmove over time to their highest value uses. Suchuses include efficient investment, improvedproductivity, research and development, and thediffusion of new ideas and technologies.Dynamic efficiency involves the movement fromone type of efficient use of resources to anothertype of efficient use of resources.

Equity Objectives

Equity objectives motivate many regulatory deci-sions on telecommunications prices. Equityobjectives generally relate to the fair distribution ofwelfare benefits among members of society. Tele-communications regulators are primarily concernedwith two different aspects of equity in the regulationof prices:

Operator-consumer equity relates to thedistribution of benefits between consumers andthe regulated operator. For instance, manypeople would not consider it equitable thatmonopoly operators be allowed to earn highprofits for an extended period of time without im-proving or extending service. In this regard, theaim of many regulators is to ensure that thesavings that result from improved technologicalinnovations are shared equitably between theoperator and consumers. Price cap regulationincludes a mechanism for consumers to share inthese productivity gains.

Consumer-consumer equity relates to thedistribution of benefits between different classesof telecommunications consumers. For example,in Colombia, consumers in lower socio-economicbrackets pay less for the same local telephonesubscription services than consumers in higherbrackets. This approach implements a govern-ment policy aimed at improving consumer-consumer equity.

Balancing the Objectives of Price Regulation

The main challenges of price regulation involve thedesign and implementation of low-cost and effectiveregulatory approaches that induce the regulated

Module 4 – Price Regulation

4 - 3

Price Regulation

operator to achieve the socially desirable objectivesdiscussed above. Regulation imposes a burden onthe economy in the form of direct costs totelecommunications operators for enforcement andcompliance. It may also place indirect burdens onconsumers in the form of loss of choice of operatorsand/or services. A practical objective in the design ofprice regulation approaches should be to impose theleast burden necessary to achieve their purposes. Ata minimum, benefits of price regulation should justifyits costs.

In practice, there is often disagreement over tele-communications price regulation because the threebroad regulatory objectives, financial, efficiency andequity, can conflict with one another. Some peoplewill place more importance on one objective thanothers. This means that the regulator will often haveto make trade-offs between these objectives in thecourse of implementing price regulation.

4.1.2 Rate Rebalancing

This Section contains a brief discussion of pricerebalancing, or rate rebalancing, as it is morefrequently called. This important topic is dealt with ingreater detail in Appendix 4-1 of this Module.

The term “rebalancing” refers to moving the pricesfor different telecommunications services moreclosely in line with the costs of providing eachservice. Currently, telecommunications price struc-tures in many countries are highly unbalanced, withsome services priced well above costs and othersbelow costs. Telecommunications costing isdiscussed in detail in Section 1.4 of Appendix B ofthe Handbook

Prices of telephone connections, monthly subscrip-tions, and local calls have traditionally been setbelow costs in many countries. Resulting deficitshave been subsidized by higher-than-cost longdistance and international calling prices. Some ofthe historical reasons for these traditional pricingstructures are discussed in Section 4.2.2.

Unbalanced price structures are not sustainable in acompetitive environment. New competitors willgenerally enter those market segments where profitmargins are highest, such as long distance and in-ternational calling. Incumbent operators will

therefore be under pressure to reduce subsidies orrisk losing customers in the more profitable marketsegments. Traditional unbalanced price structuresare also inefficient in that higher-than-cost pricesencourage uneconomic entry by high-cost opera-tors. Lower-than-cost prices discourage economicentry, even by low-cost operators.

Costs of different telecommunications services havebeen decreasing at different rates as a result oftechnological developments. This has furtherunbalanced telecommunications prices. Where tele-communication markets are open to competition,prices of different services will tend to move towardstheir costs. However, in monopoly or non-competitive environments they may not, and theregulator may be required to take steps to ensurethat prices are more closely aligned with costs.Efficient monopoly pricing, and related matters, suchas Ramsey Pricing, are discussed in Sections 1.1and 1.2 of Appendix B of the Handbook.

A significant amount of rate rebalancing hasoccurred in many industrialized countries in recentyears. Comprehensive price comparisons havebeen conducted by the OECD for its 29 membercountries since 1990. The effects of rebalancingcalls in member countries is presented in Figure 4-1.As this figure illustrates, since 1990, the averageprice of local calls in OECD countries has risen bymore than 30%. In contrast, the average price oflong distance calls (110 km and 490 km calls) hasdecreased by about 30% over the same period.

Figure 4-2 shows the effect of rebalancing on pricesfor business services. Over the 1990-1998 period,fixed charges (connection and subscription)increased by over 20% and usage chargesdecreased by over 20%, for an overall weightedreduction of about 12%. Note that overall teledensityin the OECD countries has increased steadily,despite rebalancing. The relationship betweenrebalancing and consumer welfare is discussedfurther in Module 6.

These two figures and those contained in Appendix4-1 indicate that rate rebalancing has producedlower overall prices for most consumers in a majorityof the countries surveyed. However, this is not theonly benefit of rebalancing. Rate rebalancing willalso increase social welfare by moving prices closer


4 - 4

Figure 4-1: Index of OECD Tariff Rebalancing by Distance, including Local Calling

50

60

70

80

90

100

110

120

130

140

150

1990 1991 1992 1993 1994 1995 1996 1997 199850

60

70

80

90

100

110

120

130

140

150

27 km 110 km 490 km Local Calling

Notes: All indices set to 100 in 1990 Average weighted by number of access lines Calculation based on PPP’s expressed in USD

Source OECD (1999)

Figure 4-2: Index of OECD Business Charges and Teledensity

6 0

7 0

8 0

9 0

1 0 0

1 1 0

1 2 0

1 3 0

1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 9 9 7 1 9 9 8

T o ta l C h a r g e s U s a g e C h a rg e s F ixe d C h a rg e s T e le d e n s ity

Notes: All indices set to 100 in 1990 Average weighted by number of access lines. Calculation based on PPPs expressed in USDSource: OECD (1999)


4 - 5

Price Regulation

to costs. This is illustrated in more detail in Appendix4-1, and in other studies that have examined rebal-ancing in different countries. Rate rebalancing willprovide benefits to the economy in addition toproducing lower overall prices. Therefore, there is astrong case to be made for rate rebalancing, withour without the introduction of competition.

4.2 Approaches to Price Regulation

4.2.1 Introduction

Different approaches have been developed over theyears to regulate telecommunications prices. Some,involving rules-based approaches, are designed toprovide stability and certainty, as well as achievingregulatory objectives. Others have been more adhoc and discretionary.

This Section begins with a discussion of twocommon pricing approaches: traditional discretion-ary price setting and Rate of Return regulation. ThisSection is followed by a discussion on incentiveregulation. In our analysis we consider how well thethree approaches achieve the broad objectives ofprice regulation: namely the financing, efficiency,and equity objectives.

4.2.2 Discretionary Price Setting

Traditionally, in many countries, price regulation wasfocussed heavily on social objectives as well asfinancial or economic ones. This was particularlytrue where the government operated the telecom-munications network. Under such circumstances,prices were usually set to promote consumer-to-consumer equity objectives. In many countries, therewas little or no analysis of the economic impacts ofsuch policies.

Where discretionary price regulation existed, orcontinues to exist, it is usually characterized bybelow-cost prices for connection, subscription andlocal calls. The shortfall is made up by higher-than-cost international call prices, and sometimes alsohigh long-distance prices.

The frequently-stated objective of this type of pricingis to promote affordability of basic telephone serv-ices. This type of pricing may also incorporate thevalue of service principle. Simply stated, this

principle assumes that a prospective buyer will pay aprice that is related to the value derived from theservice and that telephone services are morevaluable to some classes of customers than toothers. Accordingly, businesses are often chargedmore than residential customers for the sameconnection and subscription services. It is assumedthat businesses are major users of international andlong-distance services, and that they value suchservices highly. Accordingly, higher rates arecharged for such services.

Discretionary price regulation approaches in manycountries were interventionist. Often the governmentor the Minister in charge would micro-manage thePTT’s pricing structure, severely reducing its abilityto function as a normal business enterprise. In somecases, telephone prices were increased to make upgovernment budget deficits, without extensive con-sideration of the economic or social impacts of suchincreases.

In some countries, traditional discretionary priceregulation failed to generate enough revenue to paythe operating costs of the incumbent operator or tosupport network upgrades and expansion. As aresult, the operator’s revenue requirement and thefinancial objective of regulation were sometimes notmet.

In some jurisdictions, telephone revenues of state-owned operators were treated as part of generalgovernment revenues. Expenditures of the state-owned operator, including those for investments, areincluded in the general government budget. Poorgovernment fiscal management made it impossibleto meet a PTT’s revenue requirement. Such anarrangement deprives the operator of the capitalrequired to upgrade its network. It can also reducethe incentive for the operator to innovate and reducecosts, which hurts the dynamic efficiency objective.In practice, such operators often have poor perform-ance and over-staffing, which means that theproductive efficiency objective is not met either.

Long-term capital investments should make up alarge part of the costs of a telecommunicationsoperator. However, cash-strapped governmentssometimes extract cash from state-owned operatorsto finance other government priorities. This has beenmore common where there was no explicit rules-


4 - 6

based regulatory regime that requires prices to beset to meet a revenue requirement calculated to in-clude long-term capital investments. Enough cashmay be left for the operator to meet its day-to-dayoperating requirements, but not enough to upgradeor expand the network.

Where this has happened, the result has been anundersupply of telecommunications services andwaiting lists for service. In some countries, telecom-munications prices have been increased solely tomeet general government revenue requirements,without regard to the specific revenue requirement ofthe telecommunications operator. Instead ofimproving telecommunications service, the proceedsof telephone rate increases have sometimes beenused to meet a wide range of other governmentpriorities, from subsidizing postal services to payingthe armed forces.

In some cases, it is said that local telephone ratesare kept at low levels to maintain affordability ofservices for low-income subscribers (i.e. to meetconsumer-consumer equity objectives). In reality,however, the initial telephone users in most emerg-ing economies are not the poor. With low prices, therelatively privileged group of telephone users end uppaying much less than it can afford. At the sametime, the operator cannot expand the network toprovide service to other users. This undermines theoperator-consumer equity and consumer-consumerequity objectives. As a result, most of the poorhouseholds, especially in rural areas, receive nosubsidy at all because they have no access. Insummary, experience has shown that discretionaryprice setting approaches have seldom achievedtheir social or economic goals, at least on a long-term basis.

Traditional discretionary price setting approacheshave usually resulted in inefficient price structures.Table 4-1 summarizes the main differences betweenprices that typically result from discretionary pricesetting and the types of cost-oriented prices thatwould result from competition.

A detailed discussion of telecommunications costs isprovided in Section 1.4 of Appendix B.

4.2.3 Rate-of-Return Regulation

Rate of Return (ROR) regulation is a rules-basedform of price regulation. Unlike discretionary pricesetting, ROR regulation provides an operator withrelative certainty that it can meet its revenuerequirement on an ongoing basis. The essence ofROR regulation is simple. First, the regulatedoperator’s revenue requirement is calculated. Thenthe operator’s individual service prices are adjustedso that its aggregate service revenues cover itsrevenue requirement.

In calculating the revenue requirement, the regulatorfirst reviews the operating costs and financing (e.g.debt service) costs. Typically there is some regula-tory scrutiny to ensure that the costs werenecessarily and prudently incurred in order toprovide the regulated services. If not, they may bedisallowed from the “rate base”. The operator will notbe entitled to increase its prices or rates to recoversuch disallowed costs.

The next step in calculating an operator’s revenuerequirement is to determine its rate of return. Inorder to allow the operator to remain financiallyviable, and to attract new capital for its operations,ROR regulation permits the operator to recover notonly its direct operation and financing costs, but alsoa fair return on its rate base. The regulator deter-mines an appropriate rate of return on capital for agiven time period (typically one to three years). Thisreturn is generally based on a review of financialmarket conditions, plus any additional operator orindustry-specific issues (industry or operator risk,operator specific taxation issues, etc.).

Based on the approved rate of return, a revenuerequirement is calculated (i.e. total revenues thatmay be generated in a given period). The revenuerequirement is to be recovered from the sum of allservices provided. If an operator earns more than itsallowable rate of return, the regulator will requireprice reductions to bring the operator’s rate of returndown to the allowable level. Conversely, if theoperator does not meet its allowable rate of return, itwill request price increases to raise its revenues.


4 - 7

Price Regulation

Table 4-1: Typical Result of Discretionary Price Setting

Service Discretionary Price Setting Efficient Cost-oriented Pricing

Connection Very low price: typically below $50.Waiting list used to ration demand.

Related to the incremental costs of providing the line.

Subscription Relatively low price: typically below$3/month. Network congestionused to ration demand.

Related to the incremental costs of local service,including the local exchange switch and the "localloop" portion of the network. Local service costs varysignificantly across different service areas, based ondensity and other factors. Higher charges levied onbusinesses due to their higher demands for main-tenance and service quality.

Local Calling Very low, unmetered or non-existent local call charges.

Calls charged per minute and in some cases withadditional call set-up surcharge. Discounts for off-peak calling and special promotions.

DomesticLong-distanceCalling

High charges with multiple callzones. Longest distance typicallycharged at a multiple of 20 or moretimes local call rate.

Calls charged per minute with possible reductions forduration of call. Discounting during off-peak periods.Ratio between longest-distance call and local call inrange of five to one or less. Tendency to distance-insensitive or “postalized” prices.

InternationalCalling

Generally very high, especially todistant countries. Accounting rateskept high and number of outgoingcircuits kept low to generate netsettlement payments.

Calls charged per minute with possible reductions forduration of call. Discounting during off-peak periods.Ratio between international and national callstypically in excess of 3 to 1, but coming down due toaccounting rate reform.

Source: Adapted from ITU (1998a)

ROR regulation is designed to equate an operator’stotal revenues with its total costs. It is generally notdesigned to equate revenue for any particularservice to the cost of that service. As a result, it doesnot specifically address the structure of prices. Inpractice, where ROR regulation is applied, thestructure of prices generally tends to fall somewherebetween cost-oriented prices and the prices thatresult from discretionary price setting.

Weaknesses of ROR Regulation

The weaknesses of ROR regulation are summarizedin Box 4-1. The main weakness is that it does notprovide operators with a strong incentive to operateefficiently by reducing their operating costs. Theycan usually recover most if not all of their coststhrough rate increases, and they are not permitted toretain additional profits earned by reducing theircosts. As a result, ROR regulation does not promote

the efficiency objectives of price regulation as wellas other forms of regulation.

The perceived inefficiencies of ROR regulation mustbe put into perspective. The reality is that operatorsin some industrialized countries performed relativelywell under ROR regulation for nearly a century, tak-ing advantage of gains in technology and sharingthe benefits with their customers in the form of lowerprices. Nevertheless, because of the identifiedweaknesses, many regulators in industrializedcountries have been introducing forms of incentiveregulation instead of ROR regulation.

Concerns about the inefficiencies of ROR regulationarose in industrialized countries after extensivenetworks had been constructed. The most importantobjective in many developing countries is to buildnetwork infrastructure to meet unsatisfied demand.


4 - 8

Box 4-1: Weaknesses of Rate of Return Regulation

Lack of Incentive to Minimize Costs

➢ In ROR regulation, the operator’s prices are set at a level sufficient to cover its costs. This is whyROR regulation is often referred to as “cost plus regulation”. From a dynamic perspective,therefore, the operator has little incentive to reduce its rate base or its operating costs. Incompetitive markets, where the market determines price levels, an increase in costs will reduceprofits. Therefore cost containment is a major objective of operators in a competitive market.

Lack of Innovation/Productivity Improvement

➢ Over time, ROR regulation of a monopoly operator will lead to a lower rate of productivity improve-ment than would occur under effective competition. ROR regulation does not provide the operatorwith a strong incentive to increase its productivity.

Capital Bias – The Aversch-Johnson Effect

➢ ROR regulation provides incentives to increase the amount of capital that the operator invests. Thehigher the capital expenditure, the higher the rate base, and the greater the total return theoperator can earn. It therefore encourages the operator to use an inefficient input mix. Theoperator will have an incentive to use an inefficiently high capital/labour ratio for its level of output.This result is often referred to as the Aversch-Johnson effect, named after two economists whodescribed it. The effect is an indication that productive efficiency is not being maximized.

Cost of Regulation

➢ ROR regulation requires the operator and the regulator to spend significant amounts of time andmoney. The rate base must be repeatedly calculated by the operator and reviewed by theregulator, the cost of capital must be recalculated, and so on. Rate reviews or hearings must beheld on a regular basis, incurring costs to the regulator, the operator, and other participants in theprocess.

Interventionist Nature of ROR Regulation

➢ The regulator is required to review many aspects of the operation and management of the firm in adetailed manner. This includes scrutiny to prevent rate base “padding”. Over time, this type ofdetailed regulation may place a regulatory burden on the firm that impedes its ability to function asa normal business enterprise.

Inadequacy for Transition to Competition

➢ ROR regulation operates relatively slowly, and generally does not allow operators the pricingflexibility they need to respond to competitors’ actions.

➢ The introduction of competition in some parts of the telecommunications sector, combined withcontinuing ROR regulation in monopoly segments, means that vertically-integrated operators havean incentive to engage in anti-competitive practices (e.g. anti-competitive cross-subsidization).

This will typically require a very large capital invest-ment. As a result, the concern about ROR regulationemphasizing capital investment is not as significanta concern in developing countries. The political andeconomic environment in many developing countries

minimizes the differences between ROR and incen-tive regulation. In fact, any economically sustainableform of rules-based price regulation would bepreferable to the ad hoc forms of discretionary price


4 - 9

Price Regulation

setting currently practised in some developingcountries.

4.2.4 ROR-Incentive Regulation

The term ROR-incentive regulation is generally usedto describe variations on ROR regulation that weredeveloped in different US states to respond toperceived weaknesses in traditional ROR regulation.ROR-incentive regulation has enjoyed limitedpopularity in other parts of the world.

Incentive regulation provides inducements and pen-alties that encourage an operator to meet regulatorygoals.

The different types of incentive regulation generallyshare the following elements:

➢ The operator often participates in setting goalsor performance targets.

➢ The operator is given more flexibility than undertraditional ROR regulation. The regulatortypically does not prescribe specific manage-ment actions. For example, the operator may berewarded for reducing its operating costs but nottold exactly how to reduce these costs.

➢ The regulator restricts some activities of theoperator.

➢ Rewards and penalties established by theregulator motivate the operator to performefficiently.

4.2.5 Types of ROR-Incentive Regulation

In this Section, we summarize some of the incentive-based regulatory schemes that have beenimplemented in the US telecommunications industry.These forms of regulation typically replace traditionalROR regulation.

Banded Rate-of-Return

Under this form of incentive regulation, regulatorsestablish a range (or band) of authorized earnings.Prices are set to generate earnings that fall withinthe authorized range. When only a narrow band ofearnings is permitted, the operator’s incentives are

similar to those created by traditional ROR regula-tion. A broad band of earnings can create strongerincentives for the operator to reduce operating costsand improve operations. For instance, rather thanset the rate of return at 12%, the operator might beallowed a return of between 10% and 14%.

Rate Case Moratoria

Rate case moratoria can be implemented byagreements between a regulator and an operator tosuspend regulatory scrutiny of the operator’searnings for a fixed period. This form of incentiveregulation is often used at the beginning of a transi-tion to price cap regulation. It gives the regulatedoperator an incentive to lower operating costs, sinceit may retain higher earnings during the transitionperiod.

Earnings-Sharing

Under an earnings-sharing plan, the operator mayretain higher earnings. However, earnings in aspecified range are shared with consumers.Typically, these plans are set up with differentsharing ranges based on a prescribed ROR. Thesesharing ranges can differ substantially from plan toplan. In one example of this type of plan, theregulated operator keeps 100% of the earnings upto 10%, the operator and consumers split earningsbetween 10% and 14%. The operator’s earnings arecapped at 14%.

4.3 Price Cap Regulation

4.3.1 Overview

This Section provides an overview of price capregulation, which is the preferred form of rules-based price regulation around the world today.

Price cap regulation uses a formula to determine themaximum allowable price increases for a regulatedoperator’s services for a specified number of years.The formula is designed to permit an operator torecover its unavoidable cost increases (e.g. inflation,tax increases, etc.) through price increases.However, unlike ROR regulation, the formula doesnot permit the operator to increase rates to recoverall costs. The formula also requires the operator tolower its prices regularly to reflect productivity


4 - 10

increases that an efficient operator would beexpected to experience.

Price cap regulation has several advantages overROR regulation:

➢ It provides incentives for greater efficiency;

➢ It streamlines the regulatory process;

➢ It provides greater pricing flexibility;

➢ It reduces the possibility of regulatory interven-tion and micro-management;

➢ It allows consumers and operators to share inexpected productivity gains;

➢ It protects consumers and competitors by limit-ing price increases; and

➢ It limits the opportunity for cross-subsidization.

For these advantages to materialize, price capregulation must be implemented in an effective andinternally consistent manner. We discuss some ofthese implementation challenges in the Sectionsbelow.

Price cap regulation is meant to provide incentivesthat are similar to competitive market forces. Com-petitive forces require operators to improveproductivity and, after accounting for unavoidableincreases in their input costs, pass these gains on totheir customers in the form of lower prices. The pricecap formula has a similar effect.

Price cap regulation is a means to regulate pricesover time. The price cap formula determines the rateof change in prices from an initial level. The initiallevel of prices may be set by the regulator (see Sec-tion 4.1.2). Alternatively, the regulator may establisha transition period at the end of which the regulatedoperator must reach target price levels or ranges(see Section 4.4.5). Future financial performance fora price cap regulated operator formulae is highlydependent on the initial price levels. Therefore, it iscritical for the regulator to ensure that the initial levelof prices are consistent with the operator’s revenuerequirement.

4.3.2 The Basic Price Cap Formula

There are a number of ways to express the pricecap formula. In its simplest form, a price cap formulaallows an operator to increase its rates annually byan amount equal to an inflation measure, less anamount equal to the assumed rate of productivityincrease. A simplified very basic price cap formula isset out in Box 4-2.

It can be seen from this simple example that opera-tors may increase their prices to include the effectsof inflation, but no more. Inflationary cost increasesof 5% may be passed on because it is assumed thatthe operator cannot control them. However, theexample also assumes that telecommunicationsindustry productivity will increase by 3%. Such pro-ductivity increases result from technologicalimprovements, lower switching and transmissioncosts, and many other factors. Therefore, in theabove example, the operator must pass on aproductivity benefit to its customers by lowering itsyear 2001 prices by 3%.

In this example, the operator may reap the benefitsof any measures it takes to reduce its costs below3%. If the operator has been very efficient, it may

Box 4-2: Simplified Basic Price CapFormula

Allowable price increase for a year = StartingPrice + I – X

Notes:

(1) I = Inflation Factor for the year

(2) X = Productivity Factor

(3) These factors are discussed in greater detail in later Sections of this Module

Example:

In year 2000, the price is 100

I = 5

X = 3

Therefore, the allowable price increase for2001 equals 100 + 5 – 3 = 102


4 - 11

Price Regulation

have reduced its actual costs by 10%. In such acase, the operator may retain the benefits oflowering its costs from the assumed 3% productivityfactor to 10%. The additional earnings which resultfrom such efficient operations may be retained asprofits to shareholders or used for other purposes,such as new investment. The earnings could also beused to reduce prices further, for example to meetcompetition. However, such additional reductionswill not be required by the regulator. The price capformula determines the maximum required pricedecreases.

4.3.2.1 Price Indices and Weights

The sample price cap formula in Box 4-2 is highlysimplified. In practice, telecommunications operatorsdo not offer a single service at a single price. Theyoffer a range of different services at different prices.A typical price cap formula will, therefore, generallyuse an index of the prices charged by an operatorand not a single price. In such cases, the operatorwill be required to keep its actual prices below aPrice Cap Index (PCI).

In developing indices for a price cap formula, pricesof different services are weighted so that the pricesfor major services receive a proportionately greaterweight. Consider a simple example, where anoperator provides only two services, local serviceand international service. An index of the operator’sactual prices (Actual Price Index or API) can bedeveloped for this operator using service revenuesas weights. For example, assume that local serviceaccounts for 75% of the operator’s revenues, andinternational service accounts for 25%. The sameproportions (“weights”) will be used to determinewhether the operator’s API exceeded the price cap,or PCI.

Let us use the same price increase assumptions asdescribed in Box 4-2. In the year 2001, prices will beallowed to increase from 100 to 102. Therefore, letus assume that 102 is the PCI. To determinewhether the operator’s actual prices in 2001 exceedthe PCI of 102, we must compare that PCI to theAPI. Box 4-3 contains examples comparing theoperator’s API for 2001 to its PCI of 102.

These simple examples illustrate the following basicfeatures of price cap formulae that are based onindices:

➢ The actual prices of the operator (as measuredby the API) may not exceed the price cap for theyear (as measured by the PCI).

➢ The operator has pricing flexibility; some pricesmay be increased above the weighted averageof the change in prices, as long as others arenot.

➢ Prices for services with heavier weightings in anindex will affect the index more. Therefore,prices for major services (measured byrevenues) may not be increased as much asprices for less significant services.

4.3.2.2 Basic Indexed Price Cap Formula

Box 4-4 restates the basic price cap formula usingthe concept of price indices described above. Theformula assumes that prices will be calculated foreach year. The symbol “t” is used in the formula torepresent the appropriate time period (e.g. a year).In practice, different time periods can be used in-stead of years.

The factors I and X which are used in the formulaset out in Box 4-4 are discussed in greater detail inlater Sections of this Module.

4.3.2.3 Service Baskets

Under price cap regulation, services are usuallygrouped into one or more service baskets. Differentservice baskets may be subject to different price capindices.

For example, a residential service basket might bedeveloped to limit price increases affecting residen-tial consumers. This basket might include localresidential connection charges, monthly subscriptionfees, and local and international usage charges. Aseparate basket might include services used by typi-cal business customers.


4 - 12

Box 4-3: Using Price Indices - Simplified Calculation of API

Basic Price Cap Rule: API @ PCIi.e. the Actual Price Index (API) for the year 2001 must be equal to or less than the Price Cap Index (PCI)for 2001. The objective of this example is to calculate the API for year 2001 and determine whether theproposed price changes comply with the Basic Price Cap Rule. The API for year 2001 is the product of theAPI for year 2000 and the weighted average of the change in prices from 2000 to 2001.Notes:(1) Set the API, the PCI and all prices equal to 100 in year 2000(2) Year 2001 PCI = 102 (i.e. a 2% increase over year 2000)(3) Indices are weighted by revenues(4) The operator provides only 2 services:

(a) Local Services = 75% of revenues(b) International Services = 25% of revenues

(5) The weighted average of the change in prices is the sum of the following calculation for each service:the change in prices (expressed as the division of year 2001 price by year 2000 price) multiplied by therespective revenue weight (expressed as the division of service revenue by total revenue).

Example A:Proposed price changes: Local price increases by 1% from year 2000 to 2001 (100 to 101)

International price increases by 4% from year 2000 to 2001 (100 to 104)

Weighted average of thechange in prices:

Local service: 1.01 x 0.75International service: 1.04 x 0.25Total:

= 0.7575= 0.2600= 1.0175

Since the API for 2000 was 100, the API for 2001 is the product of 100 and the weighted average of thechange in prices, i.e. 100 x 1.0175 = 101.75. Therefore API < PCI (i.e. 101.75 is less than 102). Since theproposed year 2001 prices are less than the PCI, no additional price reductions would be required by theregulator.Example B:Proposed price changes: Local price increases by 4% from year 2000 to 2001 (100 to 104)

International price increases by 1% from year 2000 to 2001 (100 to 101)

Weighted average of thechange in prices:

Local service: 1.04 x 0.75International service: 1.01 x 0.25Total:

= 0.7800= 0.2525= 1.0325

Since the API for 2000 was 100, the API for 2001 is the product of 100 and the weighted average of thechange in prices, i.e. 100 x 1.0325 = 103.25. Therefore API > PCI (i.e. 103.25 is greater than 102). Sincethe proposed year 2001 prices are higher than the PCI, the proposed prices would not be approved by theregulator. The regulator would require that prices must be reduced further.


4 - 13

Price Regulation

Box 4-4: Basic Price Cap Formula Using Indices

Price cap regulation requires:

APIt @ PCIt for all t

That is, the API for a particular time period must always be less than or equal to the PCI for that period. Fromyear to year, the PCI is adjusted according to the following formula:

PCIt = PCIt-1 x (1 + It - X)

i.e. the PCI for a given year (t) will be equal to the PCI for the previous year (t-1) multiplied by 1 plus the Infla-tion Factor for year t (It ) minus the Productivity Factor (X).

Notes:

(1) APIt means the Actual Price Index at year t. The API is a weighted average of the change in pricesactually charged by the operator.

(2) PCIt means the price cap index at year t. The PCI is a weighted average of the change in themaximum allowable prices of the operator.

(3) It is the inflation factor at time t.

(4) X is the productivity factor.

(5) It is common to express It and X in percentage terms especially when referring to them outside thecontext of actual price cap calculations. Note, however, that in the price cap formulae these variables areexpressed in decimal, not percentage terms.

Example

Using the same PCI assumptions as described in Box 4-2 and Box 4-3, in a period where the inflationfactor is 5% and the productivity factor is 3%, the maximum amount that the weighted average of thechange in prices would be permitted to increase would be by 2%.

i.e. The formula PCIt = PCIt-1 x (1 + It - X) produces the following result: 102 = 100 x (1+.05 -.03)

There may also be restrictions on the absolute orrelative movement of prices for services subject toprice cap regulation. Operators may change pricesfor individual services within the baskets as long asthe API for the services in the basket complies withthe price cap formula, and as long as no individualservice pricing restrictions are breached.

An example of an individual service pricing restric-tion is a rule that no price for an individual servicemay increase by more than 10% per year. Suchrestrictions may be applied, for example, to limit theimpact on residential consumers of rate rebalancing.The concepts of service baskets and individualrestrictions are discussed further in Sections 4.3.7and 4.3.8 of this Module. Service baskets may alsobe used to restrict or prevent the cross-subsidization

of service open to competition (e.g. domestic andinternational long distance) by monopoly services(e.g. access and local calling).

4.3.3 Calculating Price Cap Variables:Looking Ahead or Back

The basic price cap formula contains a number ofvariables that must be calculated. To mimic theworkings of competitive markets, the price capformula should ideally be forward-looking. Variablessuch as the inflation (I) and productivity (X) factors,and the weights used to calculate the indices shouldideally be determined based on expected futurevalues.


4 - 14

In practice, however, the majority of regulators onlyset the productivity factor based on future values.The inflation factor and index weights aredetermined based on the most recent availablehistorical data.

There are a number of practical reasons for settingthe inflation factor and index weights based onhistorical data:

➢ In many economies, past inflation performanceis a good predictor of future inflation.

➢ The process of forecasting inflation and the de-mand and revenue variables needed to forecastweights is complex, time consuming, andsubject to controversy and possiblymanipulation.

➢ A forecasting approach may necessitatecorrections to offset the effect of forecastingerrors, thus adding complexity and regulatoryuncertainty.

Basing the inflation factor and weights on historicaldata also has disadvantages. For instance, futureinflation may vary significantly from past inflation.This disadvantage may be mitigated by increasingthe frequency of adjustments to the inflation factor,or by establishing trigger mechanisms as discussedbelow.

In principle, index weights may be based on costs orrevenues. Cost weights are generally considered tobe the more theoretically correct choice, but reliableforward-looking costing data is often not available. Inpractice, therefore, most regulators have chosenrevenue weights to calculate the aggregate indicesin the price cap formula. Regulators should be espe-cially vigilant in the choice of weights when pricesare not balanced and heavy cross-subsidizationexists. In this type of scenario there may besignificant differences in the cost and revenueweights and use of the latter may bias thecalculation of the API.

Another approach is to set fixed weights that do notvary from period to period. This approach isadministratively simpler and limits any possibility forthe operator to manipulate the price cap formula bysetting prices strategically. Setting weights based onforward-looking cost benchmarks is one possiblealternative under this approach.

4.3.4 The Inflation Factor

The price cap formula includes an inflation factor toaccount for changes in input costs of the operator.For example, holding all the other variablesconstant, a 5% inflation factor would allow aregulated operator to increase its average prices by5%.

4.3.4.1 Selection Criteria

In most economies, a number of different indices areused to measure inflation. For example, a consumerprice index or retail price index (CPI or RPI) meas-ures changes in the prices of goods and servicespurchased by typical consumers (e.g. food,passenger transportation, residential electricalpower, etc.). A Producer Price Index (PPI) measureschanges in the prices of goods and servicespurchased by different types of production industries(e.g. prices for labour, freight transport, industrialelectrical power, etc.).

In developing a price cap formula, regulators mustselect an appropriate inflation factor (I). A choicemay be made from among existing inflation indices,or a new inflation factor may be calculated.Regulators that have implemented price cap regula-tion have identified a number of criteria for selectingan inflation index to be used as the inflation factor.Frequently used criteria are set out in Box 4-5.

Particular national circumstances may dictate thatother criteria should be considered. It is unlikely thatany one potential inflation measure will rank highestin all of the selection criteria. Ultimately, the selectionmust be based on the informed judgment of theregulator.


4 - 15

Price Regulation

Box 4-5: Selection Criteria for an Inflation Factor

Reflective of changes in the operator’s costs

➢ For the inflation factor to be a useful variable, it must reflect changes in the operator’s input costs.This is particularly critical in situations of economic instability, when the inflation factor will have tocapture sudden and large changes in the country’s exchange rate. This is particularly important foroperators that typically purchase a large proportion of their equipment in foreign currency.

Availability from a credible, published, independent source

➢ This is important if price cap regulation is to have credibility with all parties involved. Private sectorparticipants as well as international investors in the sector must be able to trust the source of thedata.

Availability on a timely basis

➢ In order for the price cap formula to respond quickly to any changes in input costs, the inflationfactor should ideally be available with a lag of less than 6 months and preferably 2 to 4 months.

Understandability

➢ There is significant benefit in including an inflation factor that is easily understood not only by allthe players in the telecommunications sector, but by the public at large.

Stability

➢ The values of some statistical indices are subject to revision after their initial release. For example,in March 2001, the January 2001 CPI may be announced at 123.47; however, that value may berevised to 123.58 in June 2001. If possible, an inflation factor should be chosen that is not subjectto large frequent revisions.

Consistency with total factor productivity of the economy

➢ The choice of price index will have a direct impact on the manner of calculating the productivityfactor (X) because efficiency gains in the rest of the economy affect the operator through thisindex. As we discuss below, the inclusion of specific variables in the price cap formula will dependon whether an economy-wide price index or a price index for the operator's principal inputs isused. This aspect is discussed in greater detail in Section 4.3.5.

4.3.4.2 Potentially Useful Inflation Indices

With these selection criteria in mind, the next step isto examine existing inflation measures available inthe country. A number of indices are normally pub-lished by or available from the government statisticaloffice (if one exists), and/or the country’s centralbank. In some countries, these statistics areproduced by government ministries, such as theMinistries of Finance, Statistics, Planning orEconomic Development.

Potentially useful inflation measures may be classi-fied as either economy-wide indices or non-economy wide price indices. Some inflationmeasures are designed to reflect national ordomestic output price changes. For example, theGross Domestic Product (GDP) price indexmeasures the cost of a fixed basket of goods andservices that make up the GDP in a particular baseyear. This is updated at periodic intervals. Similarly,the price index for the Gross National Product (GNP)gives economy-wide coverage.


4 - 16

A related index is the GDP or GNP deflator. Tradi-tionally, the deflator is determined by dividing thecost of the basket of goods and services that makeup the GDP (or GNP) at current prices by the cost ofthe same basket at constant prices. Hence, thedeflator reflects not only pure price changes, butalso changes, if any, in the weights attached to theGDP (or GNP) components.

The GDP (and GNP) indices and deflators arebroadly based. They reflect changes in the pricesaffecting a large basket of goods and services. Manyregulators in the U.S. and Canada have chosen oneof these economy-wide indices as the inflation factorto be included in their price cap formula.

Other indices are narrower in scope. For example,the Consumer Price Index (CPI) or the Retail PriceIndex (RPI) measures the changes in prices paid byconsumers. They typically measure the cost of afixed basket of goods and services that are boughtby consumers in a particular base year, and updateit at periodic intervals. This narrow scope is theirgreatest disadvantage because telecommunicationsoperators incur only a portion of their costs in retailconsumer markets. Hence, the CPI or RPI may berelatively poor indicators of inflation affecting theoperator’s cost structure.

Another set of inflation measures that is narrower inscope are the producer, industrial or wholesale priceindices. Generally, they measure changes in pricespaid by companies economy-wide, or in particularsectors of the economy.

A number of regulators in the United Kingdom andEurope have selected retail price indices as theinflation factor to be included in their price capformula. In fact, price cap regulation is sometimesreferred to as “RPI-X” regulation, referring to theinitiative of the United Kingdom in first implementingthis type of regulation in the early 1980s, whenBritish Telecom was privatized.

4.3.4.3 Other Inflation Factors

Based on the general criteria set out above and on asurvey of existing indices, the regulator should con-

sider the advantages and disadvantages of eachavailable index as a potential inflation factor. It ispossible that the regulator will decide that none ofthe existing national indices is appropriate. Box 4-6presents some possible alternative inflation factors.

4.3.4.4 Period of Adjustment

The regulator must decide how often changes in thechosen inflation index will be used to adjust the pricecap formula, and how often the operator will beallowed to adjust its rates. This is referred to as theperiodicity of adjustment to the price cap formula. Inindustrialized countries, the period of adjustment isusually once a year. This is a feasible option be-cause inflation rates tend to be relatively low andstable in such countries.

Many developing countries, however, are subject togreater economic instability. Hence, the ideal perio-dicity may be less than a year, say 3 or 6 months. Arelatively short period between updates lessens theimpact that an acceleration or deceleration ofinflation can have on the operator’s expenses. Theregulator should weigh the benefits of frequentadjustment against the administrative costs ofchanging and publishing new prices on a regular,short-term basis.

4.3.4.5 I-Factor Adjustment Mechanism

One approach developed to deal with economicinstability is to include a trigger mechanism in theadjustment of the price cap formula. Under thisapproach, the regulator may select a standardnational inflation index as its inflation factor with arelatively long period of adjustment. However, as a“fall back”, an immediate adjustment may be madeto the inflation factor in the event of certain large andunexpected economic developments.

As an example, an adjustment might take placewhen the selected national inflation index increasesor decreases by a significant amount. In countrieswith a history of relatively low and stable inflation,this amount could be in the order of 10% to 20%.


4 - 17

Price Regulation

Box 4-6: Alternative Inflation Factors

➢ One option is to use an inflation index from another country (or inflation measures produced byUnited Nations organizations and/or international financial institutions, regional development banks,The World Bank, the IMF, etc.).

➢ In Argentina, for instance, some regulated utilities use the producer price index for the UnitedStates. This is then converted into the national currency. This choice was designed to reassureforeign investors by relating their revenues to a hard currency.

➢ Another option is to construct a new measure of inflation that more accurately reflects the coststructure of the operators. This new “composite” index may be a weighted combination of severalexisting indices.

➢ In Colombia, for instance, the interconnection access rates paid by wireless and long distance op-erators to local telephone operators is indexed on a monthly basis to a composite index made upof the following:

➢ An index of the US/Colombia exchange rate and the average customs duty; weight: 0.38

➢ An index of the minimum industrial wage in Colombia; weight: 0.29

➢ The Producer Price Index of Colombia; weight: 0.33

➢ Similarly, in Chile, the access rates paid by mobile operators to terminate calls on the networks ofPSTN operators is indexed on a monthly basis to a weighted aggregate index made up of thefollowing:

➢ An index of the imported goods and services component of the Chilean wholesale priceindex; weight 0.263

➢ The Chilean wholesale price index; weight: 0.542

➢ The Chilean consumer price index; weight: 0.195

Source for Argentina example: Green and Pardina (1999)

An I-Factor adjustment mechanism can also be tiedto other key changes that would seriously impact onthe cost of operating a telecommunications system.In many countries, the most serious potentialchange is a devaluation of the national currency.While this may reduce labour costs, it can signifi-cantly increase the costs of equipment, foreignconsulting services, financing charges, etc. Anadjustment mechanism to deal with this type ofchange is presented in Box 4-7.

4.3.5 The Productivity Factor

The price cap formula includes a productivity factor,which is based on an estimate of the operator’s ex-pected productivity increases over the relevantperiod. This variable, commonly referred to as the“X-factor” or the “productivity offset”, ensures that

consumers receive partly or fully the benefits of theoperator’s expected productivity gains in the form oflower prices. For example, if all other variables areheld constant, a 3% X-factor will result in annualreductions of 3% in average consumer prices.

The proper choice of an X-factor is critical for thelong-term viability of any price cap plan. Selectingthe X-factor is often the most contentious aspect ofimplementing price cap regulation. The X-factorshould be set so that it poses a challenge to theoperator. It should promise consumers higher gainsrelative to alternative regulatory regimes. If the X-factor is set too low, the operator will earn excessiveprofits and the regulatory regime could fall intodisrepute. If too large an X-factor is selected, theoperator may not be permitted to meet its revenuerequirement.


4 - 18

Box 4-7: Inflation Factor for Foreign Exchange Rate Changes

A mechanism to adjust the inflation factor in a price cap formula can be triggered by large foreignexchange (FX) rate changes. It is possible that national inflation measures will not adjust rapidly enough toreflect the real impact of large FX changes. For example, this occurred in Indonesia in 1997, when theAsian economic crises caused the Indonesian rupiah to drop rapidly from approximately 2400 rupiah perUS dollar to 14,000 per dollar. In comparison, the Indonesian inflation indices remained relatively stable.Since telecommunications operators paid for equipment purchase, financing charges, etc. in foreigncurrencies, the drop in the rupiah translated into a massive increase in operating costs, which was notreflected in the national inflation indices.

To account for such large FX changes, a pre-established mechanism could provide for an adjustment tothe inflation factor – for example, if the percentage change in the average monthly exchange rate is higherthan the corresponding percentage change in the inflation factor by a specified amount (perhaps 20 to30%) within any specified period.

By way of illustration, let us assume a 25% threshold. If the Indonesian rupiah depreciated by 35% duringthe relevant period (hence, increasing by 35% the number of rupiah required to purchase a US dollar), butthe national inflation measure increased by 30%, the trigger mechanism would not apply. However, if theinflation measure only increased by 5%, an adjustment would be triggered.

4.3.5.1 X-Factor Determination

The X-factor may be divided into the “basic offset”and adjustment factors. The basic offset shouldreflect the regulated operator’s historical achieve-ment of productivity growth. If the operator has hada history of lower input price inflation than other firmsin the economy, that should be reflected in the basicoffset. Adjustment factors are included to take intoaccount changes in the operating environment of theregulated operator. For example, an adjustmentfactor might reflect the introduction of price capregulation, the introduction of competition or the pri-vatization of the operator.

There are two major approaches to determination ofthe X-factor. One approach, which we will refer to asthe historical productivity method, relies on historicinformation about the productivity performance ofthe regulated firm to set the basic offset. Once thebasic offset is calculated, certain adjustment factorsmay be added or subtracted to take into accountchanges in the operating environment of theoperator. These adjustment factors are based onregulatory benchmarking or other predictivemethodologies. This approach is based on theunderstanding that past productivity, withadjustments, is a good indicator of future productiv-

ity. The implementation of this approach is subject tothe availability of specific data. The calculation maybe very data-intensive and requires reliable andconsistent data of a very specific nature at anadequate level of detail for an adequate period oftime.

The other approach, which we will refer to as theregulatory benchmarking method, recognizes that insome instances, past productivity performance maynot be a good indicator of future expected perform-ance. This may be the case where the sector waspreviously regulated by discretionary price setting (ornot regulated at all). It may also be the case wherethe sector has been inefficiently operated underpublic ownership or is subject to very significantstructural change, for instance, divestiture. In thesecases, the adjustment factors may be much moresignificant than the calculated basic offset. Abenchmarked productivity factor is likely the onlypractical alternative in many developing countries.There the regulator is not likely have access toreliable and consistent historical productivity data todetermine the historical productivity factor.


4 - 19

Price Regulation

4.3.5.2 Historical Productivity Method

A number of empirical methods can be used to helpthe regulator set the X-factor. Most of these methodswere developed in the countries that first imple-mented price cap regulation (United Kingdom,United States, Canada, etc.). The preferred methodto determine an X-factor is to carry out a total factorproductivity (TFP) study using historical data on theregulated operator and/or on the sector. Box 4-8

provides an overview of TFP and how it may beapplied to the telecommunications sector.

Historical Productivity - Basic Offset

Price cap regulation is intended to replicate thediscipline of competitive market forces. These forcesrequire operators to improve productivity and passtheir gains on to their customers in the form of lowerprices, after accounting for increases in input prices.If all sectors in the economy were fully competitive,

Box 4-8: Total Factor Productivity

Productivity is the measure of how effectively an entity employs inputs to produce outputs. It is a measure ofoperational efficiency. A typical, although partial, measure of productivity in the telecommunications industryis lines (one output) per employee (one input). Lines per employee is obviously only a partial measure,given that one could increase the number of lines by increasing capital investment or materials.Simultaneously, a telecommunications operator produces many more outputs than just the number of lines.

TFP (also known as multi-factor productivity) measures how effectively an operator, an industry or aneconomy employs all inputs to produce all outputs. TFP can be said to have increased if the operatorproduces more outputs with the same amount of inputs, or if it produces the same outputs with fewer inputs.TFP is equal to the ratio of output volume to input volume. Algebraically the TFP index may be expressedas:

TFP = Q/Z

Where Q is an index of aggregate output volume and Z is an index of aggregate input volume. Note that forprice cap regulation we are primarily interested in the changes in the TFP index, rather than its level. If werefer to changes by the symbol ∆, the change in the TFP index may be expressed in the following manner:

∆ TFP = ∆Q/∆Z

Example:

If the output volume index has increased by 5% (i.e. ∆ Q=1.05) and the input volume index has increasedby 2% (i.e. ∆ Z= 1.02), the change in the TFP index is 2.94%:

∆ TFP = 1.05/1.02 = 1.0294

Note that for the sake of simplification regulators and analysts often approximate the multiplicativerelationship between TFP, Q and Z by an additive relationship. In this instance, if output has increased by5% and inputs by 2%, it can be said that TFP has approximately increased by 3%:

Approximation: ∆ TFP ´ ∆Q – ∆Z´ 5% - 2%´ 3%

It should be stressed that while this type of approximation is fairly common, it is not always accurate. Whilein the example above the approximation (2.94%) was quite close to the actual number (3.00%), this will notalways be the case. Generally, the larger the change in TFP the larger the inaccuracy of the approximation.


4 - 20

output prices in the economy would grow at a rateequal to the difference between the growth rate ofinput prices and the rate of productivity growth.

As described in Bernstein and Sappington (1998), ifregulated telecommunications operators were like atypical company, the telecommunications regulatorcould replicate market discipline by restricting in-creases in the operator’s prices to the economy-wide rate of price inflation. This restriction wouldrequire the regulated operator to achieve the sameproductivity gains as that of the typical company,and to pass these gains on to its customers, afteradjusting for the typical input price inflation rate. Ifthe regulated operator faces the same input priceinflation rate as other companies in the economy,the X-factor should be set at zero.

Generally, therefore, the X-factor should reflect theextent to which:

➢ the regulated operator is capable of increasingits productivity more rapidly than othercompanies in the economy; and

➢ the prices of inputs used by the regulatedoperator grow more slowly than the input pricesfaced by other companies in the economy (thisis often referred to as the input price differentialor IPD).

Telecommunications operators should normallyenjoy faster productivity growth than othercompanies due to the more rapid rate oftechnological change in the telecommunicationsindustry. Telecommunications operators may alsohave lower input price inflation due to the decreasingunit costs of processing, switching and transmission.

If the regulated operator can achieve faster produc-tivity growth or enjoy lower input price inflation thanother companies in the economy, then the regulatedoperator should be required to pass the associatedbenefits on to customers in the form of lower prices.

For example, assume the expected annual rate ofproductivity growth of the regulated operator is 3%.The corresponding growth rate elsewhere in theeconomy is 1%. Input prices in the regulated

industry are expected to increase 0.5% annually,and the corresponding growth rate of input priceselsewhere in the economy is 2.5%. In this setting,the X-factor should be set at approximately 4% (= [3- 1] + [2.5 - 0.5]). Note that for simplicity we haveapproximated the X-factor by adding and subtractingthe different variables. As pointed out in Box 4-8, forsmall numbers this is generally a fair approximationto the mathematically-correct multiplicativecalculation.

Table 4-2 presents the results of some studies ofTFP for the US communications industry andcorresponding TFP performance of the US economyas a whole. Based on Table 4-2 and other studies(including those of the Canadian telecommunica-tions industry), it appears that in the long-termproductivity growth of the communications industryin North America has been about 2% to 2.5% higherthan productivity growth of the respectiveeconomies. Some of these studies are dated andthe productivity differential may have changedrecently.

The choice of the inflation factor will have an impacton the choice of variables to calculate the basicoffset. If a general inflation index is selected for the I-factor (e.g. GDP-PI or CPI or RPI, etc.), the basicproductivity offset should be calculated as in theexample presented two paragraphs above. This isreferred to as the differential approach. Based onthis approach, the figures in Table 4-2 suggest abasic offset between 2.0% and 2.5%. If a sector oroperator-specific index is constructed, however, theappropriate basic offset is simply thetelecommunications TFP estimate. This is referredto as the direct approach. Based on his approach,the figures in Table 4-2 suggest a basic offsetbetween 3.0% and 3.5%.

Historical Productivity Adjustments

Many regulators have adjusted the basic offset byother factors to take into account significant changesin the operating environment of the regulatedoperator. We review some of the key adjustmentfactors below. These adjustment factors are oftendetermined based on benchmarking or predictivemethods, such as time-series, cross-sectionaleconometric studies.


4 - 21

Price Regulation

Table 4-2: Selected Estimates of TFP for the US.

Study Period COM US DIFF

Nadiri-Schankerman

1947-76 4.1 2.0 2.1

Jorgenson 1948-79 2.9 0.8 2.1

Christensen 1947-79 3.2 1.9 1.4

AT&T 1948-79 3.8 1.8 2.0

A.P.C. 1948-87 4.0 1.7 2.3

Christensen 1951-87 3.2 1.2 1.9

Crandall 1960-87 3.4 1.3 2.1

DRI 1963-91 3.0 0.2 2.8

Christensen 1984-93 2.4 0.3 2.1

Note: US Communications Industry (COM); US Economy and Differential (DIFF %)

Source: Taylor (1997)

Incentive Regulation Factor

After price cap regulation replaces ROR regulationor, more likely, when it becomes the first form ofrules-based price regulation adopted, operators inthe industry can be expected to achieve a higherproductivity growth rate than they have in the past.

In such circumstances, some regulators have sup-plemented the basic offset with what is sometimescalled a customer productivity dividend (CPD). Anumber of econometric studies have examined theimpact of incentive regulation plans on productivityof telecommunications operators. On the whole,these studies have concluded that incentive regula-tion has a positive impact on productivity growth.

In principle, the CPD should reflect the best estimateof the increase in the productivity growth rate in theregulated sector that will be brought about by theimproved incentives inherent in the new regulatoryregime. This variable, also referred to as the stretchfactor, could be allowed to vary over the life of theprice cap plan. For example, the variable may behigher at the beginning of the plan and reduced nearits end. CPDs adopted in the US and Canada havegenerally been below 1% per year.

Competition Adjustment

The rise of strong competition is another structuralchange that can affect the value of the X-factorunder price cap regulation. The effect of increasedcompetition, however, is unclear.

On the one hand, increased competition, like achange in regulatory regime, can force the regulatedoperator to operate more efficiently, and therebyachieve a higher productivity growth rate. This wouldseem to favour a higher X-factor, particularly if aCPD has not been imposed.

On the other hand, increased competitive forces canshift market share from incumbent operators to newentrants. The result can be an unavoidable reductionin the growth rate of the incumbent’s outputs. Par-ticularly in the short run, this lowering of the growthrate of the incumbent’s outputs can be higher thanany associated lowering of the growth rate of itsinputs. This leads to a lower productivity growth ratefor the incumbent, arguing for a lower X-factor. Theempirical evidence on the effect of competition onproductivity growth is mixed. A number of recenttime-series, cross-sectional econometric studieshave found no relationship between competition and


4 - 22

productivity growth, after taking into account otherfactors.

Privatization Factor

The theoretical literature suggests that privatizationshould increase productivity growth. The theory issubstantiated by recent econometric studies thathave found privatization to increase productivity byat least 0.5% to 1.0% per year.

4.3.5.3 Regulatory Benchmarking Method

In some instances, past productivity performancemay not be a good indicator of future expectedperformance. This may be the case where thesector was not price regulated, was not operatedefficiently or is the subject to very significant struc-tural change.

In these circumstances, or when the operator and/orits operating environment are undergoing drasticchange, the X-factor may have to be developedbased on the informed judgement of the regulatorand its advisors. International experience with pricecap regulation can provide a useful benchmark insuch cases. This is why we refer to this method asregulatory benchmarking.

Furthermore, this approach may be the only practi-cal alternative in many developing countriesbecause of lack of the very specific detailedhistorical data over an adequate period of time tocalculate TFP. More generally, the historical methodmay be less applicable to developing economies forthe following reasons:

➢ Low teledensity levels, and privatization offormer government telecommunications opera-tors, can be expected to lead to significantproductivity improvements;

➢ Significant political and economic instability, andthe lack of a clear legal and regulatoryframework may affect productivity levels, and;

➢ Recent evidence suggests that technologicalcatch-up and the possibilities for greater sectorgrowth in developing countries mean thatproductivity growth should be higher than inindustrialized economies. This would suggest

that the X-factor should be set at relatively highlevels. On the other hand, there are some veryefficient telecommunications sectors in thedeveloping world that may not be subject tosuch “catch-up” phenomenon.

Countries that have had a number of price cap planshave generally increased the X-factor over time.One example is that of British Telecom (BT) in theUnited Kingdom (UK), where the X-factor has beenincreased from 3% in the 1984-1989 period to muchhigher factors in recent years (see Table 4-3). Thisregulatory “tightening” has been a result of better-than-expected performance by the regulatedoperator. The major increases in the X-factor from amodest initial figure also reflects a degree ofregulatory caution, with an initial bias towardsensuring the regulated operator's revenuerequirement is met.

No price cap plan, no matter how carefully designed,will be perfect or permanent. It is in the nature ofgood regulation to evolve with market and policydevelopments. The evolving nature of price capregulation is perhaps best illustrated by the changesin the various price cap plans that have been appliedto British Telecom. BT was the first telecommunica-tions operator subject to price cap regulation. Itremains subject to this form of regulation, but asillustrated in Table 4-3, there have been significantchanges over the years. Regulators that areconsidering introducing price cap regulation shouldtake comfort from the British experience. The mostsignificant decision at the time of the privatization ofBT was to adopt price cap regulation – not to specifyits particular X-factor, and other details. The actualform of regulation was not cast in stone. As in othercountries where price cap regulation has since beenadopted, adjustments continue to be made as theregulator’s experience with this form of regulationincreases, particularly with respect to the determina-tion of the X-factor.

Table 4-4 and Table 4-5 provide examples of currentX-factors adopted by regulators around the world.Although there is some variation in the actual X-factors set by regulators, based on this selectedsample, when a majority of the operator’s servicesare included in price cap regulation, many regulatorshave selected an X-factor in the range of 3.5% to4.5% as the initial X-factor. This range is generally


4 - 23

Price Regulation

Table 4-3: A Summary of British Telecom' s Price Cap Plans

Duration X Services Subject to PriceCaps

Other Main PricingConstraints

Main Services NotSubject to Price Caps

1984-89 3.0 Subscription; local andnational calling

Residential subscription(RPI+2)

Rental, international calls,operator services , con-nection charges, publictelephone calls

1989-91 4.5 Subscription; local andnational calling

Subscription (RPI+2);connections (RPI+2);private circuits (RPI+0).

Rental; international callsand public phone calls.

1991-93 6.25 Subscription; local andnational call charges;international calls; volumediscounts.

Residential and single linesubscription (RPI+2):multi-line subs. (RPI+5);connection (RPI+2);private circuits (RPI);median res.bill (RPI)

Telephone rental; publictelephone calling.

1993-97 7.5 Subscription; local andnational calling; interna-tional calls; connectioncharges.

All subscriptions (RPI+2);all individual prices inbasket limited to RPIincluding connectioncharges; private circuitbasket (RPI).

Public telephone calling.

1998-2001 4.5 Retail charges: residentialconnection subscription;local, national and inter-national calls. Based onexpenditure patterns oflowest spending 80% ofresidential customers.

Business assurancepackage, includingsubscription (RPI),analogue private circuits(RPI).

Public telephone calling.

8.0 Network charges: non-competitive accessservices (call originationand termination, singletransit, local conveyance)and interconnectionspecific service.

Services divided into threebaskets, each basketsubject to RPI-8 cap.

Source: Adapted from OECD (1995) and Oftel (2000a)

consistent with the differential approach to calculat-ing the X-factor. The more detailed guidelines

discussed below may assist in regulatoryjudgements to set the X-factor.


4 - 24

Table 4-4: X-Factors of Selected NationalPrice Cap Regulation Plans

Table 4-5: X-Factors of Selected State PriceCap Regulation Plans in US

Country X-Factor State X-Factor

Argentina 5.5 Connecticut 5.0

Australia 7.5 Delaware 3.0

Canada 4.5 Georgia 3.0

Chile 1.1 Illinois 4.3

Colombia 2.0 Maine 4.5

Denmark 4.0 Massachusetts 4.1

France 4.5 Michigan 1.0

Ireland 6.0 New York 4.0

Mexico 3.0 North Carolina 2.0

Portugal 4.0 Ohio 3.0

UK 4.5 Rhode Island 4.0

US 6.5 Wisconsin 3.0

Regulatory Benchmarking

➢ Differential Approach

The long-term historical productivity differentialbetween the telecommunications sector and theeconomy is generally accepted to be 2% to 2.5%or higher. We discussed this range in theprevious section. This benchmark can be higherwhere the telecommunications sector isexpected to grow at a rate significantly higherthan that of the economy.

The long-term historical input price differential(IPD) between the telecommunications sectorand the economy is generally accepted to bepositive, but smaller than 1%. The IPD could belowered if, for example, a telecommunicationsworker’s wages grew faster than that of theaverage worker. Conversely, the IPD should beraised if the rate of productivity-improving tech-nological development in the telecommunicationsindustry increases.

➢ Direct Approach

The long-term historical productivity performanceof the telecommunications sector is generallyaccepted to be 3% to 3.5% or higher. Wediscussed this range in the previous section. Thisbenchmark could be higher where productivity inthe telecommunications sector is expected togrow at a rate significantly higher than that of theeconomy.

Regulatory Benchmarking – Adjustments

Adjustments can be made for the effects of theintroduction of incentive price regulation, competi-tion, and privatization, where these conditions apply.These factors and their effects are discussed above.Table 4-6 provides a numerical summary of thebenchmark estimates discussed in this Section.These are of a general nature. It is recommendedthat each country carry out an appropriate TFP orbenchmarking study based on specific nationalconditions.


4 - 25

Price Regulation

Table 4-6: Summary of Benchmark Estimates for Setting X-Factor (%)

Differential Approach Direct Approach

Basic Offset 2.0 to 2.5 3.0 to 3.5

Adjustment Factors

Incentive Regulation 0.5 to 1.0

Competition 0.0*

Privatization 0.5 to 1.0

Note: * Could be increased to up to 0.5 if competition is combined with privatization.

Source: Based on McCarthy Tétrault review of the literature and experience with price cap regulation inindustrialized countries. Estimates may be less applicable to developing countries. These estimates are of ageneral nature. It is recommended that each country carry out an appropriate TFP or benchmarking studybased on specific national conditions.

4.3.6 Capped and Non-Capped Services

A basic decisions to be taken in price cap regulationis the selection of which services to regulate. Ingeneral, regulators apply price cap regulation toservices that are provided on a monopoly or domi-nant provider basis. The rationale for price regulationis discussed in Appendix B of the Handbook.

In many markets, the distinction is made between“basic services” which are price-capped, and otherservices which are not. Services provided in fullycompetitive markets are normally excluded fromprice cap plans. There is sometimes a grey linebetween the categories, and regulators have treatedthe same types of services differently. Table 4-7 andTable 4-8 describe the types of services covered byprice cap plans in the same jurisdictions as in Table4-4 and Table 4-5 around the world.

Services are sometimes included in price capbaskets to promote competition and to protect con-sumers. An example is the case of interconnectioncharges. Interconnection access charges can beincluded under a global price cap that couldincorporate consumer “retail” and access“wholesale” services. This would make it possible forexpected productivity gains in the provision of ac-cess services to be passed on to competitors and be

ultimately reflected in retail prices. Access servicescan also be placed in a separate basket from retailservices to prevent the dominant supplier from “pricesqueezing” its competitors through its control of bothretail and wholesale pricing.

4.3.7 Service Baskets

Having selected the services to be included in pricecap regulation, the structure of the price cap planshould be determined. One of the features of pricecap regulation is that the regulated operatormaintains some pricing flexibility. This flexibility isparticularly important when significant rate rebal-ancing is required within the price cap plan. It is alsoimportant when the operator is facing competitionand must respond quickly to competitive pricechallenges. Nevertheless, there are a number ofreasons for the regulator to restrict pricing flexibility.

One reason is to restrict the operator’s ability toengage in inappropriate cross-subsidization. Such arestriction can be implemented through the creationof groups of services, or service baskets, within theprice cap plan. An example of how service basketsconstrain flexibility is provided in Box 4-9.

It is common practice to place capped services inmore than one basket. For example, Figure 4-3 and


4 - 26

Table 4-7: Service Coverage of SelectedNational Price Cap Regulation Plans

Table 4-8: Service Coverage of SelectedPrice Cap Regulation Schemes in US

Country Service Coverage State Service Coverage

Argentina Basic services Connecticut Basic and non-competitiveservices

Australia Basic and mobile services Delaware Basic services

Canada Basic local services Georgia Basic and other services

Chile Local and access services Illinois Non-competitive services*

Colombia Local services Maine All services

Denmark Basic and ISDN services Massachusetts Non-competitive services*

France Basic services Michigan Non-competitive services

Ireland Basic and ISDN Services New York Basic services

Mexico Basic services North Carolina Basic services

Portugal Basic and leased line services Ohio Basic Services*

UK Basic residential services Rhode Island Basic services

US Interstate access services Wisconsin Basic and other services

Note: * excludes basic residential services

Figure 4-4 illustrate the service baskets for theTelecom Australia price cap plan. Different types ofservices are grouped in different baskets, and serv-ices with common characteristics are grouped withina single basket.

Many regulators have established different “sub-caps” on different baskets. In effect, these basket-pricing restrictions are used by regulators to furtherconstrain the pricing flexibility of the operator. Forinstance, in Figure 4-4, subscription services wouldbe subject to the CPI-2% sub-cap of its servicebasket and also to the overall price cap of CPI-5.5%.

The assignment to baskets is intended to replicatethe effects of competition. The following are generalcriteria for assigning capped services into servicebaskets:

➢ degree of competition in each service basket;

➢ homogeneity of services (including similarity indemand price elasticities); and

➢ degree of substitutability of each service.

4.3.8 Individual Service Pricing Restrictions

Restrictions can be placed on the relative and/orabsolute movement of prices of individual services,as well as on service baskets. This may be done, forexample, if the regulator is concerned that the resi-dential subscription rate may rise too quickly as aresult of rate rebalancing.

The maximum allowable price increase for individualservices will be inversely proportional to the weightof the individual service within the services basket.As a result, the price for services with relatively smallweights could increase significantly if the allowedincrease were channelled towards one of theseservices and there were smaller compensatingdecreases in charges for services with relativelylarger weights. Conversely, a service with arelatively heavy weight within the proposed servicesbasket would be subject to only moderate priceincreases if the allowed increase were channelled to


4 - 27

Price Regulation

this service, even though compensating decreaseswere made in services with relatively smallerweights.

There are two alternatives to individual restrictions,each of which may be applied to restrict thedecreases and/or increases in prices.

One of these methods, commonly referred to as“banding”, limits the price movement of specificservices relative to another variable, usually theinflation factor. For instance, if the regulator isconcerned about increases in the residentialsubscription price, an upper restriction could providethat the price could not increase at a rate greaterthan the inflation factor plus 5% (CPI + 5). If the X-factor has been set at 4% and the corresponding I-factor was 7%, the weighted average of prices couldincrease by approximately 3% (7% - 4%). Theresidential subscription price, however, couldincrease up to a maximum of approximately 12%(7% + 5%). This is an example of a relative restric-tion. In the case of Telecom Australia provided inFigure 4-3, the residential subscription and localcalls are subject to an individual restriction of CPI.

The other type of restriction is absolute. Forexample, if the regulator is concerned that nationallong-distance rates may decline too quickly, adownwards restriction could provide that the aver-age price of these calls should not decrease bymore than 20% per year.

It is generally considered that relative restrictions arepreferable to absolute ones because they providethe regulator with greater certainty as to themovement of the real (inflation-adjusted) prices ofthe services.

Restrictions may be upwards or downwards, but donot necessarily have to be symmetrical. Forexample, if access prices are included in the pricecap, they could be subject to relative restrictions withupward and downward bounds (e.g. inflation factor±5%).

Restrictions on the regulated operator’s pricing flexi-bility have been implemented by most regulators.Care must be taken to design restrictions that areinternally consistent and do not unduly constrain theoperator. Judgment is required to set restrictions that

provide sufficient flexibility to permit necessary raterebalancing, while protecting consumers fromexcessive rate increases and competitors from anti-competitive subsidization. Too many restrictions onprices will eliminate pricing flexibility, one of the mainbenefits of price cap regulations.

4.3.9 Duration and Review of Price CapPlans

The longer the term of a price cap plan, the strongerthe incentive for the operator to improve itsperformance. In theory, the duration of a price capplan should be indefinite, so that the regulator wouldnot intervene in the setting of future prices.

In practice, however, this type of price cap regime isneither feasible nor desirable. A regulator cannotestimate future productivity growth with certainty; norcan it set the X-factor at the right level for anindefinite period. With the X-factor set imperfectly,

Box 4-9: How Service Baskets ConstrainPrice Flexibility - Example

Assume a one-basket price cap plan. Itincludes international services and residentialsubscription services. Assume both have thesame weight in the price cap index. Holdingall other prices constant, a decrease indomestic long distance prices (say 30%) canbe offset by a significant increase (also 30%,assuming the same revenue weights) in theresidential subscription rates.

To constrain this type of offsetting raterebalancing, residential subscription servicesand international services can be placed inseparate baskets. If this were done, pricedecreases in one service cannot be offset byequivalent increases in the price of otherservices.

In practice, of course, regulators should notconstrain operators from implementingnecessary rate rebalancing. Individual priceconstraints or bands can be used to limitprice increases to particularly sensitiveservices, with less impact on the overallpricing flexibility of operators.


4 - 28

Figure 4-3: Price Cap Plan for Telecom Australiafrom 1989 to 1992

Figure 4-4: Price Cap Plan for TelecomAustralia from 1992 to 1995

• International Calls

International Cap:CPI-4 %

• Subscription• Local Calls• Domestic LD Calls

Domestic Cap:CPI-4 %

• Subscription• Residential• Local Calls

Individual Restrictions: CPI

• Connections• Subscription• Local calls

Sub Cap:CPI-2 %

• International Calls

Sub Cap:CPI-5.5 %

• Domestic LDCalls

Sub Cap:CPI-5.5 %

• Domestic DedicatedLines

• International DedicatedLines

• Mobile TelephoneService

Charges Subject toOverall Cap Only

• Connections• Rentals• Local Calls• Trunk Calls

Individual Restrictions: CPI(in any one year)

Overall Price Cap: CPI-5.5 %

the operator would either earn insufficient revenuesor unacceptably high profits. Both outcomes areinefficient and unsustainable. As a result, in a real-world price cap plan, the regulator generally sets aminimum period during which the X-factor will not berevised. At the end of the duration period, a review isundertaken. Regulators have typically chosenperiods of three to five years.

The duration of the plan should be sufficiently long toallow efficiency incentives to be acted on. However,it should not be so long that market developmentsundermine the regime. In the Regulatory Bench-marking Section, above, it was suggested that aprudent approach would be to set an initial X-factorconservatively. In such as case, the plan should bereviewed reasonably soon, to minimize the negativeimpact of miscalculations or errors of judgement insetting the X-factor.

The price cap plan review process should becarefully designed. The key variable that will have tobe reviewed, and perhaps reset, is the X-factor. Thiswill be a primary focus of the review. It involvessome complex incentive issues for the regulator tograpple with.

If productivity improvements achieved by theoperator exceed the X-factor by a substantialamount, the operator will make significant profits.There may be pressure on the regulator to adjust thevalue of the X-factor upwards. Rate of return orother profit indicators are generally used to reset theX-factor. This review mechanism will reduce theregulated operator’s incentive to continue toincrease its productivity. The incentives for furtherefficiency improvements will tend to fall as thereview approaches, particularly if the operator knowsthat any additional cost savings will result in a higherX-factor resulting from the review process. In this


4 - 29

Price Regulation

instance, price cap regulation will approximate RORregulation during the review period. The optimalselection of incentives/disincentives will ultimately bebased on regulatory judgement.

The approach to resetting the X-factor will dependon how the regulator evaluates the need for the op-erator to earn higher profits to increase its ability toattract investment, compared to the consumerbenefits of lower prices (operator-consumer equityobjective). It will also depend on the relativeimportance placed on productive and dynamicefficiency. The higher the weight placed onconsumer benefits relative to profits in the shortterm, the more the regulator will tend to reducefuture profits by setting a higher X-factor at the timeof the review.

4.4 Price Cap Variations

4.4.1 Introduction

This Section considers some of the variations thathave been applied to the basic price cap formula.Depending on national telecommunications marketconditions, regulators may include some of thesevariations in their price cap plan.

4.4.2 The Exogenous Factor

As discussed above, the inflation factor is a proxy forthe changes in the regulated operator’s input prices.There may be instances, however, when theoperator faces a significant change in input pricesthat are outside its control and not captured by theinflation factor. An example is provided in Box 4-10.Regulators must decide whether to include in theprice cap formula a cost pass-through variable (alsoreferred to as an “exogenous” variable or “Z-factor”)to address this possibility.

The inclusion of a Z-Factor in a price cap formula isnot always warranted. Many US state regulatorshave not included a Z-factor in their price cap plans.They may consider that there are very few trulyexogenous events that are not captured in the

Box 4-10: Examples of Unexpected InputPrice Change Outside Control ofRegulated Operator

An increase in customs duty from 20% to40% is imposed on imported capitalequipment, including telecommunicationsequipment. The telecommunications operatorfaces a significant input price increase.Assuming new equipment purchases accountfor 20% of the annual costs of the operator,the customs increase could increase totalcosts by 4%. This change may not be fullyreflected in the inflation factor. Assuming thatnew foreign equipment purchases account for5% of economy-wide costs, an economy-wideinflation index would increase by only 1%. Asa result, the operator would have to absorbthe remaining 3% increase in its costs.

inflation factor. Most developing country price capplans do not include a Z-factor. However, the situa-tion may be quite different in emerging marketswhere significant exogenous events are morecommon.

If the regulator decides to include a Z-factor adjust-ment, the amended price cap formula would be asfollows:

PCIt = PCIt-1 x (1 + It – X ± Zt)

It should be recognized that the practical applicationof a Z-factor adjustment could be administrativelychallenging and a source of controversy. Some ofthe uncertainty can be eliminated by carefullydefining the types of cost changes covered by the Z-factor.

Based on the considerations outlined in Box 4-11,the regulator should define the exogenous factor topromote certainty in price regulation of the sector. Inpreparing this definition, the criteria should begeneral enough to capture the impact of certainevents without diminishing the operator’s incentiveto control its costs.


4 - 30

Box 4-11: What are Exogenous Cost Changes?

The following considerations are relevant in determining which cost increases may be covered by a Z-factor:

➢ Generally, legislative, judicial or administrative actions that have a significant impact on the regu-lated operator should be considered. Such actions are usually beyond the control of theoperator. With respect to “significant”, there may be an advantage to setting a threshold belowwhich adjustments would not be considered. A threshold in the range of 1% to 2% of revenuesmay be reasonable.

➢ Regulators should only consider events that do not represent normal business risk. In assessingwhether costs should be included in a Z-factor, the regulator should consider whether theoperator can take reasonable measures to mitigate the consequences of the cost-producingevents.

➢ The Z-factor costs should not otherwise be reflected in the price cap formula and must be suchthat they have specific or disproportionate impact on the operator. The burden of proof should beon the operator to show that the proposed event is not already accounted for in the inflationfactor and would be reflected in prices charged by operators operating in competitive markets.

➢ Events, such as an economic downturn, that affect the whole economy would generally not beconsidered to produce exogenous cost increases eligible for Z-factor treatment. While suchevents may have a negative impact on the demand for the operator’s services, and decrease itsability to recover costs, the purpose of the Z-factor is not to guarantee a rate of return for theoperator. Such a guarantee would not be consistent with the objective of using price capregulation as a proxy for competitive market conditions.

➢ Z-factor costs should be quantifiable and known. The operator must be able to estimate the spe-cific costs in monetary terms.

In theory, the impact of the exogenous event shouldbe allocated across capped and uncapped services,with only the impact allocated to capped servicesbeing included in the price cap formula. In practice,the regulator could use revenue shares or otherweights to allocate the impact to capped servicesonly.

Generally, the regulator will design the exogenousfactor so that the regulated operator must requestthe inclusion of exogenous cost changes in the pricecap formula. The onus is placed on the operator totake action in the matter. The regulator only has todecide on the issue if and when there is an applica-tion before it.

An exogenous event may increase or decrease thecosts of the operator. It will be in the interest of the

operator to request the consideration of an eventthat has increased its costs. Where an event hasdecreased the operator’s costs, however, theoperator has no incentive to request consideration. Ifa Z-factor exists, the regulator will likely have toensure that savings are passed on to consumers.

4.4.3 Quality of Service

Like other services, telecommunications serviceshave a quality component and a price component. Intheory, a telecommunications operator subject toprice cap regulation could increase profit by loweringthe quality of its service. This prospect is of mostconcern when the operator is a monopolist, or isdominant so that it’s service levels are not subject toeffective competitive pressure from other operators.


4 - 31

Price Regulation

Table 4-9: Q-Factor Example –Rhode Island Scheme

The price regulation plan for the incumbent operator in Rhode Island, NYNEX, includes a Service QualityAdjustment Factor "SQAF". It was added to the basic price cap formula in the following manner

PCIt = PCIt-1 x (1 + It – X ? SQAFt)

Each month, NYNEX provides reports to the regulator on QoS performance. As illustrated in the tablebelow, the maximum value for the Service Quality Index (SQI) is 42. The regulator has determined that apassing monthly score is 25. The price cap formula is adjusted once a year. At that time, for each of the 12most recently measured months that NYNEX has not achieved a passing score in the SQI, the SQAF willbe increased by .0417%. Hence, if NYNEX does not receive a passing score in 6 months, the SQAF willtake the value of 0.25%, and prices must be decreased by that amount in the next period to compensatefor poor QoS performance.

Nynex Performance Points

New Installations orders not completedwithin 5 working days (%)

<1212.0-13.99

A14.0

210

Installation appointments missed (%) < 2.52.5 – 3.49

A3.5

210

Line out of service > 24 hours (%) < 4040 – 44.99

A45

420

Repeat repair reports (%) < 1111.0 – 13.99

A14

210

Repair service answer time (sec.) < 14.014.0 – 16.99

A17

420

Directory assistance answer time (sec.) < 4.04.0 – 5.99

A6.0

210

Average duration time – special access1.5 Mbps Circuits (hours)

< 2.52.5 – 4.49A4.49

210

Sub-total (maximum available) 22

Customer trouble reports per 100 linesper central office (CO)

< 4.04.0 – 4.99

A5

210

Sub-total (maximum available assuming10 CO’s reviewed)

20

TOTAL POSSIBLE POINTS/MONTH 42


4 - 32

Telecommunications Quality of Service (QoS) hasmany aspects. It has traditionally been measured bya number of QoS indicators, such as:

➢ Call Completion Rate

➢ Dial Tone Delay

➢ Delivery Precision

➢ Call Failure Rate

➢ Fault Clearance

➢ Complaints

➢ Billing Accuracy.

If a regulator decides to regulate QoS of an operatorsubject to price cap regulation, it can adopt severalapproaches. The traditional approach is to set a se-ries of QoS targets or standards for each indicator.Sub-standard performance can be dealt with on acase-by-case basis, or by pre-set sanctions (e.g.monetary fines or penalties payable by theoperator).

An innovative approach is to integrate a QoSvariable, often referred to as a Q-factor, in the pricecap formula. This is a relatively new approach. It isbeing implemented in a few states of the US. Table4-9 provides a summary of such an approach in theUS State of Rhode Island. A similar approach hasalso been recently adopted in Colombia at thenational level. This approach is consistent with theobjectives of incentive regulation. In addition, it hasthe advantage of directly linking QoS with the pricemechanism, thus mimicking the quality/price trade-off in competitive markets. The following formulaillustrates how a Q-factor fits into the basic price capformula:

PCIt = PCIt-1 x (1 + It – X � Qt)

The objective of including a Q-factor is that a reduc-tion in quality should result in lower prices forconsumers. Conversely, increased quality may leadto higher prices. If there is concern that quality maydrop to unacceptable levels, the regulator may set

minimum quality standards similar to minimum pricefloors. It should be recognized that the incorporationof a Q-factor can be complex and administrativelychallenging. Few regulators have integrated QoSand price cap regulation in this manner.

4.4.4 New Services

A key objective of telecommunications sector reformis to promote innovation, particularly in the introduc-tion of new services. The regulator must determinewhether or not to subject new services to priceregulation. If the decision is affirmative, price capregulation is sufficiently flexible to accommodatemost new services.

In markets that are subject to competition, manyregulators have concluded that it is not in the publicinterest to regulate most new services. Suchdecisions provide an additional incentive for opera-tors to introduce innovative services; mobile servicesare a common example.

Where such a decision is taken, it is important forthe regulator to ensure that a regulated operator’s“new” service is truly new. Operators will have anincentive to try to repackage existing services as“new” services in order to avoid price regulation. Toavoid confusion in the industry, the regulator maywant to consider publishing a definition of a newservice based on the criteria, such as the following:

➢ Does the new service include a new technologyor functional capability?

➢ Does the new service replace an existingservice and consequently not expand the rangeof services available?

4.4.5 Rate Rebalancing and Price Caps

Rate or price rebalancing is discussed in Section4.1.2 and in Appendix 4.2. It refers to the adjustmentof price levels for different services to more closelyreflect the costs of providing each service. Rebal-ancing can be achieved under most forms of priceregulation.


4 - 33

Price Regulation

A regulator that implements a price cap plan shouldconsider including a transitional period for raterebalancing, either before the plan comes into effector as part of the plan. The transition period shouldbe kept as short as possible and, depending on thelevel of price-cost imbalances, should not last morethan 5 to 7 years. This will ensure that prices at thebeginning of a price cap plan are more in line withcosts than they would be without a transition period.In a number of countries, regulators have allowedthe regulated operators a period of several years to

achieve limited rebalancing. This decision is basedon the conviction that the benefits of price capregulation are greater when prices are balanced.Rebalanced prices are clearly closer to those foundin a competitive market.

A new regulator will likely be confronted with thenecessity to rebalance prices and to introduce aform of price regulation for the first time. Given thebenefits of rebalancing and price cap regulation,neither should be delayed. Hence, there may be noopportunity to attain any significant rebalancing priorto the implementation of price cap regulation. It willhave to be done as part of a price cap plan. Theregulator may prescribe the specific targets or targetranges for some or all prices for regulated services.Some regulators have only specified end-of-periodtargets while others have also specified intermediatetargets. In this manner the regulator may be surethat over the transition period the operator will moveprices in the desired direction. If this is done, theoperator should be given sufficient pricing flexibilityto achieve rebalancing.

Establishing a transition period for rebalancingbefore a price cap plan is implemented is only apractical option when there is another form of priceregulation in place. In Canada, the incumbentoperators were subject to ROR regulation during thetransition period. Clearly, where the operator is aprivately-owned monopoly or dominant operator,some form of price regulation is preferable to noneat all. In countries where price cap regulation is thefirst form of price regulation to be introduced, themore desirable option will be to undertake raterebalancing within a price regulation regime.

4.4.6 International Accounting Rates

Technological developments and the liberalization oftelecommunications markets have put downwardpressure on international accounting rates.Accounting rates are the charges payable to inter-connecting international operators under traditionalsettlement arrangements for mutual termination oftraffic between their networks. In most countries,during the last few decades of the 20th Century, thelevel of accounting rates was well above the cost ofproviding international service termination.

Profits from high accounting rates provided a signifi-cant source of cross-subsidies, particularly todeveloping countries. It also led to a majorimbalance in payment of accounting rates fromcountries that originated more calls than theyterminated. There has been strong pressure fromthe US and other countries with outbound account-ing rate imbalances to reduce accounting rates. Thispressure, the ITU response, international servicecompetition and technological developments haveall led to significant decreases in accounting rates.

One recent technological development that under-mines the accounting rate regime is Internettelephony, also referred to as “Voice over theInternet”, or “voice over IP” (VoIP) technology.Internet telephony generally bypasses the account-ing rate regime, and hence allows VoIP providers toprice their services below those of operators ofconventional PSTN networks.

The downward trend in international accountingrates can be seen as an international form of raterebalancing - between international and nationalservice. Operators in a large number of countries willneed to increase revenues from national services tooffset potential losses from international settlements.

The demand for international calling is generallyconsidered to be price elastic, especially at higherprices. Reductions in international rates will there-fore usually lead to increases in international calling.Local access and calling, on the other hand, aregenerally less price elastic. The result of thisrebalancing could therefore be higher overallrevenues for operators providing both services.


4 - 34

The need to rebalance international and nationalrates has important implications for price cap regula-tion. For many countries, a significant amount of raterebalancing may be both desirable and necessary.Accordingly, pricing restrictions should not deprivethe operator of sufficient pricing flexibility to

implement rebalancing. The potential volatility ofinternational prices and uncertainty of customerresponse, may make it beneficial for the regulator toimplement a fixed-weights scheme for the price capformula, at least until the majority of the rebalancinghas occurred.


4 - 35

Price Regulation

Appendix 4-1: OECD Rate Rebalancing

This appendix provides an overview of the OECDtariff comparison methodology and recent analysisof rate rebalancing trends in the OECD membercountries.

Figure 4-5 and Figure 4-6 show the most recentBusiness and Residential Tariff basket comparisonsfor OECD member countries. Note that thesebaskets are based on standard listed prices ratherthan the myriad of discount schemes generallyavailable in competitive markets.

In its 1999 publication, Communications Outlook, theOECD noted significant rate rebalancing in its 29member countries. For instance, it noted a majortrend towards postalized rates at the national level.Postalization is the term given for the trend towardsflat rates for long distance services regardless of thedistance. In other words, long distance service isheading for a world where, like postal services, it isusually priced the same irrespective of distance.This has been referred to in the industry as “thedeath of distance.”

For instance, Figure 4-7 shows the differencebetween the cost of long distance calls and a localcall (3 km) between 1990 and 1998. In 1990, theaverage price of a call at 490 km was 20 times

greater than a local call at 3 km. In 1998, the marginwas reduced to about seven times.

There are a number of reasons for postalized rates.Incumbent operators typically tend to reduce thenumber of long-distance bands in response to com-petitive entry. Another reason is the prevalence ofdiscount plans that require consumers to sign up toa specific operator, often having to pay a fee. Inreturn, loyal consumers receive significant savingsover standard listed prices. Figure 4-1 in the maintext of this Module indicates that rate rebalancing isalso evident in the price trends of calls at differentdistances.

Rebalancing has been slower for the residentialbasket, as shown in Figure 4-8, than the businessbasket, which is illustrated in Figure 4-2 in the maintext of this Module. When rate rebalancing occurredin 1994, however, significant cost savings for bothresidential and business consumers were realized.Fixed charges are now slightly lower than in 1990.Combined with usage charge reductions in the orderof 25%, the overall price of the basket has beenreduced by nearly 15%. Note that overall teledensityin the OECD countries has increased steadily,despite rebalancing.

Box 4-12: OECD Tariff Comparison Methodology

In 1990, the OECD established a harmonized methodology that enables international comparisons ofnational telecommunications prices, using a basket of the different elements for a particular service. Thiscomparison can be made across countries and across time.

In recognition of the calling patterns and different prices faced by residential and business consumers, theOECD constructed a Residential basket and a Business basket. Each basket is made up of two elements,a fixed charge and a usage charge. The fixed charge covers a one year’s subscription with the installationcharge discounted over 5 years.

Once the fixed charge is calculated, the usage charge is based on OECD averages for the overallallocation of fixed charges to usage charges. Based on telephone usage patterns, the usage charge canbe allocated to national calls. These calls are then priced out for each of the countries to arrive at themonetary amount, which is either expressed in US dollars at prevailing exchange rates or based onpurchasing power parity (PPP).


4 - 36

Figure 4-5: OECD Business Tariff Basket

0

500

1000

1500

2000

2500

Luxem

bourg

Denmark

Swede

nNorw

ayKo

rea

Finlan

d

Netherl

ands

Canada

Japan

United K

ingdo

nBe

lgium

Turke

y

Irelan

d

Switze

rland

OECD av

erage

France

Germany

United S

tates Ital

y

Greece

Austri

a

New Ze

aland

Austra

lia

Hungary

Czech R

epub

licPo

rtugal Sp

ainPo

land

Mexico

Fixed Charges Usage Charges

Source: OECD (1999)

Note: Calculated based on PPPs, expressed in USD

Figure 4-6: OECD Residential Tariff Basket

0

100

200

300

400

500

600

700

800

900

Icelan

dKo

rea

Luxem

bourgSw

eden

Denmark

Norway

Turke

y

Canada

United S

tates

Japan

United K

ingdo

n

Netherl

andsFin

land

Germany

France

New Ze

aland

Austra

lia

OECD avera

geIre

land Ital

y

Switzerla

nd

Belgiu

m

Czech R

epublicAu

striaGree

ce

Hungary

Portu

gal Spain

Poland

Fixed Charges Usage Charges

Source: OECD (1999)

Note: Calculated Based on PPPs, expressed in USD


4 - 37

Price Regulation

Figure 4-7: Index of OECD Tariff Rebalancing, by year- The "Death of Distance"

0

500

1000

1500

2000

2500

3 km 7 km 12 km 17 km 22 km 27 km 40 km 75 km 110 km 135 km 175 km 250 km 350 km 490 km0

500

1000

1500

2000

2500

1998 1990

Note: 3km call charge = 100

Source: OECD (1999)

Figure 4-8: Index of OECD Residential Charges and Teledensity

6 0

7 0

8 0

9 0

1 0 0

1 1 0

1 2 0

1 3 0

1 9 9 0 1 9 9 1 1 9 9 2 1 9 9 3 1 9 9 4 1 9 9 5 1 9 9 6 1 9 9 7 1 9 9 8

T o ta l C h a r g e s U s a g e C h a r g e s F ixe d C h a r g e s T e le d e n s it y

Notes: All indices set to 100 in 1990 Average weighted by number of access lines. Calculation based on PPPs expressed in US$Source: OECD (1999)


4 - 38

Appendix 4-2: Welfare Benefits of Rate Rebalancing

This Appendix provides an overview of the potentialbenefits to public welfare that may be expected fromrate rebalancing.

The OECD study of telecommunications price trends(Appendix 4-1) indicates that rate rebalancing pro-vided most consumers with lower prices in a majorityof the countries surveyed. This is not the only benefitof rebalancing. Rate rebalancing will also increasesocial welfare by moving prices closer to costs. Thiswill provide benefits to the economy in addition tothose that result in lower overall prices. Rate rebal-ancing, therefore, should be undertaken whethercompetition is being considered or not.

Regulators may be requested to justify raterebalancing. The recent modelling of raterebalancing that carried out in Australia may beuseful in this regard. The model was prepared forAustralia’s incumbent operator, Telstra, to estimatethe potential efficiency gains from differentrebalancing scenarios. Similar analyses have beencarried out in other countries. This example uses anumber of concepts, including long run incrementalcosts (LRIC), demand elasticities, revenuerequirement, and Ramsey Pricing, which arediscussed in Appendix B of the Handbook.

Table 4-10 provides a summary of the main esti-mates used in the model. By way of explanation, theunit of measurement for access is connections. Forlocal calling, it is number of calls, for long distanceand international calling, it is minutes. Net revenue isthe difference between price and cost (LRIC) timesthe quantity. For instance, the net revenue loss forresidential access of $614 million is equal to thedifference between price $139.80 and LRIC$235.00, times 6.45 million connections. Note thatthe sum of the net revenue is $2,909 million, whichwe later assume to be its net revenue requirement.

The price elasticity of demand was based on areview of available estimates that were consideredappropriate for national conditions. With theexception of local calling elasticity, the estimates arewithin the intervals discussed in the Appendix B ofthe Handbook.

The concept of efficiency loss requires some expla-nation. It is based on the theory that marginal costpricing is optimal, that is, it maximizes the sum ofconsumer and producer surplus. (This concept isdiscussed in Appendix B of the Handbook.) Whenprices do not equal marginal costs, there are effi-ciency losses because either consumer or producer

Table 4-10: Estimates Used in the Telstra Rate Rebalancing Model – Base Scenario

Markets Price($ per unit)

LRIC($ per unit) Quantity

NetRevenue($m)

EfficiencyLoss ($m)

PriceElasticity ofDemand

Res. Access 139.80 235.00 6.45 m -614 8 -0.04

Bus. Access 240.00 235.00 2.76 m 14 0 -0.00

Local calls 0.232 0.099 11.20 b 1492 26 -0.006

Dom. LD calls 0.311 0.124 9.51 b 1782 322 -0.60

Intntnl. Calls 1.129 0.759 638.00 m 236 46 -1.20

Total 2909 402

Source: Australia Productivity Centre (1997)


4 - 39

Price Regulation

surplus are reduced. Figure 4-9 provides a graphicalpresentation of the main estimates used in theanalysis. The black shaded areas are the efficiencylosses associated with each instance of non-marginal cost pricing. Note that efficiency lossesincrease when the price-cost disparity is greater andwhen demand is more elastic. (Note the light shadedareas represent net revenues for each service).

The price-cost gap at initial conditions (the basescenario) imposes a loss in economic efficiency of

$402 million, or nearly 15% of total operatorrevenues. Note, however, that while marginal costpricing will eliminate this loss in economic efficiency,it will not meet Telstra’s net revenue requirement,which is assumed to be equal to $2,909 million. Asdiscussed in Appendix B of the Handbook, the solu-tion to this dilemma is to calculate the correspondingRamsey prices, that is, the set of prices thatminimize efficiency losses and meet the revenuerequirement.

Figure 4-9: Rate Rebalancing - Base Scenario

$

P = 0.232

LRIC = 0.099

11.2b

Local Calling (calls)

$

LRIC = 235

2.76m

Business access (connections)

P = 240

$

LRIC = 0.124

9.51m

Domestic LD Calling (minutes)

P = 0.311

$

P = 1.129

LRIC = 0.759

638m

International Calling (minutes)

$

LRIC = 235

P = 139.80

6.45m

Residential access (connections)

Note: Figures are not to scale.

Source: Australia Productivity Centre (1997)


4 - 40

Table 4-11 presents the results of five rebalancingscenarios, with the corresponding net revenue con-tribution and efficiency losses, and contrasts theseto the base scenario.

Scenario #1, which totally eliminates the efficiencyloss, would appear to be an extreme case. Ramseyprices call for the highest price over cost mark-up forthe least price-sensitive service. In this instance,business access price is raised to $1,287 andcontributes the entire net revenue of $2,909 millionbecause it has an estimated zero price elasticity.

Scenario #2 holds the business access price to$350 and calculates the constrained Ramsey prices.Scenarios #3, #4, and #5 are other permutationsthat put further constraints on prices. Note that themore constraints that are placed on Ramsey prices,the smaller the efficiency gains. Note also, however,that even modest price movements towards LRICcan result in significant efficiency gains. These gainsare likely to be greater in developing countriesbecause of the generally greater disparity betweenprices and costs.

Table 4-11: Results of Rate Rebalancing Scenarios

Variable ResidentialAccess

BusinessAccess

LocalCalls

DomesticLD Calls

InternationalCalls

Total

Scenarios LRIC $ 235.00 235.00 0.099 0.124 0.759

Price ($) 139.80 240.00 0.232 0.311 1.129

Net. rev ($m) -614 14 1492 1782 236 2909

Base Scenario

Eff. Loss ($m) 8 0 26 322 46 402

Price ($) 235.00 1287.00 0.099 0.124 0.759

Net. rev ($m) 0 2909 0 0 0 2909

Scenario 1:UnconstrainedRamsey Pricing

Eff. Loss ($m) 0 0 0 0 0 0

Price ($) 354.00 350.00 0.235 0.148 0.804

Net. rev ($m) 723 318 1529 301 38 2909

Scenario 2:ConstrainedRamsey Prices

Eff. Loss ($m) 13 0 27 5 0 45

Price ($) 235.00 350.00 0.291 0.158 0.822

Net. rev ($m) 0 318 2120 418 53 2909


Eff. Loss ($m) 0 0 54 10 1 65

Price ($) 235.00 350.00 0.232 0.209 0.919

Net. rev ($m) 0 318 1492 975 124 2909


Eff. Loss ($m) 0 0 26 67 8 101

Price ($) 140.00 350.00 0.232 0.272 1.036

Net. rev ($m) -614 318 1492 1520 194 2909

Scenario 5:ConstrainedRamsey prices

Eff. Loss ($m) 8 0 26 203 25 262

Source: Australian Productivity Centre (1997)

Documents

· cost pricing of international services discourages ... to make trade-offs between these objectives in the ... that prices are more closely aligned with costs