Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Industries Assistance Commission - Reports - Government (Non-tax) charges, 29 September 1989 (No. 422) - Volume 3


Download PDF Download PDF

Government (Non-Tax) Charges

Volume 3 Efficiency Issues and Public Enterprises: Appendixes G to N

29 September 1989

Industries Assistance Commission

Government (Non-Tax) Charges

Volume 3

Efficiency Issues and Public Enterprises : Appendixes G to N

Report No. 422

29 September 1989

Australian Government Publishing Service Canberra

© Commonwealth of Australia 1989

ISBN 0 644 10780 4

This work is copyright. Apart from any use as permitted under the Copyright Act 1968, part may be reproduced by any process without written permission from the Director Publishing and Marketing, AGPS. Inquiries should be directed to the Manager, AGPS Press, Australian Government Publishing Service, GPO Box 84, Canberra ACT 2601. A

THE PARLIAMENT OF THE c o m m o n w e a l t h o f Au s t r a l ia PARLIAMENTARY PAPER

No. 30 4 O F 1989

Ordered to be printed by authority ISSN 0727-41a 1

P rin te r! in AnctralSo W., n rx n ,

CONTENTS

Abbreviations v

Appendixes

G Pricing l

H Impact of ownership on economic performance 17 I Community service obligations 39 J Natural monopoly 79

K Contracting out of public goods and services 107 L Setting performance targets for public enterprises 115 M Valuing the assets of public enterprises 141 N Overseas developments in relation to public enterprises 163

Literature references appear at the end o f each appendix

Appendix figure

J 1 Policy implications of alternative natural monopoly situations 89

Appendix tables

HI International studies of comparative performance 21 H2 Australian empirical studies of comparative performance 23 II Allocation of New South Wales special funds, 1988-89 48 12 Government rail authority community service obligations 51 13 Telecom’s estimated cross-subsidisation, 1984-85 56 K1 Municipal use of tendering, South Australia and Tasmania 110 K2 Percentage cost savings resulting from tendering of services in the United

Kingdom 111 K3 Percentage cost savings resulting from tendering of services in Australia 112 Ml Financial aggregates under alternative accounting systems 145

iii

CONTENTS OF OTHER VOLUMES

VOLUME 1 : REPORT

Preface Summary Findings Report chapters

Appendixes A Terms of reference and conduct of the inquiry B Government policies for public enterprises

VOLUME 2 : APPENDIXES ABOUT THE IMPORTANCE OF GOVERNMENT (NON-TAX) CHARGES : APPENDIXES C TO F

Appendixes C The government goods and services sector D Extent and significance of government charges E Survey of government charges on business, 1987-88 F Surveys by industry associations of the impact of government charges on

industry competitiveness

VOLUME 4 : STUDIES - PUBLIC RAIL FREIGHT, ELECTRICITY AND WORKERS COMPENSATION ARRNAGEMENTS

Public rail freight services The electricity supply industry in Australia Workers compensation arrangements in Australia

iv

ABBREVIATIONS

Government departments/agencies

AGPS AN ANL AUSSAT

AUSTEL Australia Post Australian Airlines BIE

BTCE CBC CDB CSB DOE ECNSW HECT LAC

OTC PSA Qantas QEC

QR SECV SECWA SMHEA

SRA SRWSC Telecom TPC UTA V/Line

Westrail

Australian Government Publishing Service Australian National Railways Commission Australian National Line Aussat Pty Ltd

Australian Telecommunications Authority Australian Postal Commission Australian Airlines Ltd Bureau of Industry Economics

Bureau of Transport and Communication Economics Commonwealth Banking Corporation Commonwealth Development Bank Commonwealth Savings Bank

NSW Department of Energy Electricity Commission of New South Wales Hydro-Electricity Commission of Tasmania Industries Assistance Commission

Overseas Telecommunication Commission Prices Surveillance Authority Qantas Airlines Ltd Queensland Electricity Commission Queensland Rail

State Electricity Commission of Victoria State Energy Commission of Western Australia Snowy Mountains Hydro-Electric Authority State Rail Authority of New South Wales

Victorian State Rivers and Water Supply Commission Australian Telecommunication Commission Trade Practices Commission Urban Transit Authority of New South Wales

State Transport Authority of Victoria Western Australian Railways Commission

Other

ACM ACTU ANU BCA

CAPM CARES CCA COF

CPI CPI-X CRA CRES

CSOs Esso et al

Australian Chamber of Manufactures Australian Council of Trade Unions Australian National University Business Council of Australia

capital asset pricing model community advisory and retrofitting energy service current cost accounting cost of funds Consumer Price Index Consumer Price Index less an allowance for productivity gains CRA Ltd Centre for Resource and Environment Studies community service obligations Esso Australia Ltd and others g gram

GBEs government business enterprises

GDP gross domestic product

LCL LRMC NZ OFGAS

OFTEL OPEC OPPAS OPRA PI-X PPMFA R&D RAP AS RCCA RRR SAP SERDF SOC SRMC SRTP STD TRESS

UK USA/US

less than a full container load long ran marginal cost New Zealand United Kingdom Office of Gas Supply United Kingdom Office of Telecommunications

Organisation of Petroleum Exporting Countries off-peak hot water purchase assistance Canadian Office of Privatisation and Regulatory Affairs Price Index minus an allowance for productivity gains Pulp & Paper Manufacturers Federation of Australia Ltd research and development Remote Area Power Assistance Scheme real current cost accounting rate of return reporting statement of accounting practice a research and development program social opportunity cost short ran marginal cost social rate of time preference Subscriber Trank Dialling Traffic Route Lighting Subsidy

United Kingdom United States of America

APPENDIX G : PRICING

In a competitive environment prices charged for goods and services signal the cost o f their supply to users while also indicating to producers the values which consumers place on their products. The incentives associated with such an environment encourage least-cost production and prices which mirror the cost o f producing an additional (or marginal) unit of output. In an economic context, such prices are viewed as being ‘efficient’.

While production o f many government-provided goods and services occurs in a non-competitive environment, it is important for prices to reflect efficient supply costs as far as practicable. Pricing on this basis will promote appropriate consumption, production and investment decisions

and in the process enhance the competitiveness of the economy.

G1 Introduction

The principal aims of this appendix are to identify the requirements of efficient pricing and to discuss the main factors which influence the level and structure of prices charged by public enterprises.

Questions of what constitutes an efficient price and why efficient pricing is important are addressed in section G2.1 The level and structure of prices charged by public enterprises are significantly affected by a number of external influences which shape the incentives which enterprise managers face to price and produce efficiently. These

influences and their implications for the pricing practices of public enterprises are outlined in section G3. Some broad features of the current pricing practices of public enterprises in Australia are dealt with in section G4. A summary is presented in section G5.

G2 What are efficient prices and why are they important?

The first requirement of an efficient price is that it reflect the opportunity costs of supply (ie the value of production foregone through the use of resources employed in their best alternative). A concept closely related to opportunity costs is the concept of marginal costs. The marginal costs of supply reflect the change in total costs of supply

as output is varied. Increasing the supply of a good or service implies that additional inputs will be needed to facilitate this increase in supply. Since these inputs could be used to produce other goods and services, the marginal cost of producing a good or service is the value of alternative production foregone to produce that output. 1

1 There is an extensive literature on the pricing of government-provided goods and services both in general (eg Coase 1946; Joskow 1976; Kay 1971; Rees 1984; Turvey 1969, 1971; Williamson 1966) as well as for particular goods and services such as electricity (eg Houthakker 1951; Kolsen 1966; McColl 1976), water (eg Dragun 1983; Gallagher & Robinson 1977; Hanke 1981; Hunter District Water

Board 1982; Ng 1987; Randall 1981) and telecommunications (eg Albon 1988; Hazlewood 1950).

Provided there are no externalities2 or other distortions, the marginal costs of supply are equivalent to the opportunity costs of supply.

Basing prices on the marginal costs of supply will, in principle, result in the community obtaining the greatest value from its resources. Consumers will only purchase a good or service if the marginal benefit to them (as represented by the price paid for the good or service) at least equals the costs of supply. Charging each user a price based on marginal cost means that the good or service will only be purchased if the benefit to the consumer reflects the costs to society of providing it.

Pricing above marginal cost would result in a loss of efficiency as consumers would fail to purchase additional units of a good or service which would have yielded benefits to them in excess of costs. Similarly, pricing below the marginal costs of supply would lead to consumption beyond an efficient level, with marginal units

having costs in excess of their perceived benefits. As a result, potentially valuable resources would be wasted, since resources could be used to produce goods and services which are valued in excess of their opportunity costs.

In competitive markets, rivalry among firms encourages production at least cost (another requirement of efficient pricing) and firms face pressures to set prices close to or equal to marginal cost. Because of these pressures, prices which prevail in competitive markets are commonly regarded as ‘efficient’. In the non-competitive markets in which many public enterprises operate, efficient prices will not emerge automatically. This is because, without rivalry, neither requirement for pricing efficiency (ie least-cost production combined with prices based on marginal cost) will necessarily prevail.

Apart from being important economic activities in their own right, government- provided goods and services are an important input for many businesses (see Chapter 3 and Appendix D). To give the correct signals to purchasers of the goods and services they produce, it is essential for public enterprises to charge efficient prices or to seek to approximate them as closely as is practicable.

If prices exceed the marginal costs of efficient supply, they act like a tax on business, raising production costs and undermining competitiveness. Conversely, prices which do not recover efficient supply costs act as a subsidy to user industries. Businesses benefiting from such implicit subsidies will improve their competitiveness, but at the expense of others. This is because other users pay more than they should, or taxes are higher than they need be to fund loss-making public enterprises which underprice their outputs. Thus inefficient pricing practices by public enterprises distort the structure of production generally.

The level and structure of prices for government-provided goods and services also play an important role in investment decisions. This is because prices charged for government-supplied outputs affect the demand for these outputs, and thus the investment undertaken to satisfy that demand. Forms of pricing which fail to recognise capacity constraints, the need to replace assets, or provide an appropriate return on capital employed, will generally encourage excessive investment by public

2 Externalties arise where the activities of one economic agent impose benefits or costs on other parties which are not reflected in market prices. In consequence, particular activities may be smaller or larger than would be desirable from society’s point of view, compared with the outcome if these benefits or costs were to be fully captured in market prices.

2

enterprises. Inappropriate government charges can also affect the profitability of user industries and therefore the investment decisions they make.

Calculating the appropriate marginal cost on which to base prices is often not a straightforward exercise and, in consequence, the theoretical benchmark of marginal cost pricing may be difficult to apply in practice. For example, when enterprises supply more than one good or service, it can be difficult to apportion joint or common

costs to particular outputs and users. In addition, cost data may not be available in enough detail or be too expensive to obtain relative to the benefits, to provide a sound basis for setting prices based on marginal cost.

Apart from difficulties in measuring marginal costs, a number of other issues arise:

. whether the basis for setting charges should be short run or long run marginal cost;

. what to do when marginal cost pricing would lead to operating losses; and

. the relevance of so-called ‘second-best’ considerations, which seek to take account of non-competitive conditions existing elsewhere in the economy.

These are discussed separately below.

Short run versus long run marginal costs

An important issue for marginal cost pricing is the time frame in which prices and costs are to be related. Short run marginal cost (SRMC) corresponds to the incremental costs of supplying a good or service in situations where the ability of a supplier to vary output is constrained by the existing capacity of a supply network (eg

electricity generation system). In cases where demand exceeds available production capacity, pricing in accordance with SRMC would include an increment to ensure that available supplies were allocated to their highest valued uses; in other words, prices would be increased to ration available supplies among users.

Long run marginal cost (LRMC) corresponds to the costs of supplying an additional unit of a good or service when capacity can be varied. Thus, LRMC comprises not only the operating costs (such as material, fuel and maintenance) which comprise SRMC, but also the capital costs involved in installing additional productive capacity.

As such, the time frame applicable to LRMC is an extended and variable one, possibly of several years’ duration. In practice, the measurement of LRMC may require complex calculations incorporating the impact of new capacity on the production system already in existence (Turvey 1969).

In a strict efficiency sense, it is often argued that SRMC is the relevant basis for setting prices, since users should be charged a price which reflects the incremental costs of supplying them. Capacity (eg congestion) costs are not usually part of marginal costs unless capacity constraints are being reached. The practical

implications of this are that the price per unit would be low when capacity is large relative to demand, and increase as demand increases. When demand varies over time (as occurs with activities such as electricity, telecommunications and water) an

3

efficient pricing regime will need to set varying prices to ensure that available capacity is efficiently used.

LRMC is relevant to investment decisions. For example, when demand is such that price in a given period based on SRMC exceeds the assessed LRMC of supply, it will indicate that investment in new capacity may be warranted. However, given that capacity can often be expanded only in relatively large units, the timing of such expansions is important. LRMC (or some approximation of LRMC) is often preferred in practice as the basis for setting prices because it is seen as producing less fluctuations in prices and operating revenues compared with prices based on SRMC.

Marginal cost pricing and operating losses

Adherence to marginal cost pricing will sometimes result in total sales revenue falling short of the total costs of supplying a government-provided good or service. Examples include situations characterised by excess capacity,3 the production of ‘public goods’4 and goods characterised by increasing returns to scale.5 Adherence to a pricing benchmark which produces such an outcome is unlikely to be acceptable in circumstances in which public enterprises are expected to cover all the costs they incur.

In these situations two questions arise: how is the operating deficit to be funded? and, what implications do different approaches to funding have for efficient pricing based on marginal cost?

One approach is to retain marginal cost pricing and provide the public enterprise concerned with an operating subsidy to meet any deficit which arises. However, as many studies show, raising the necessary revenue through additional taxes imposes collection costs and also distorts individual choices between work and leisure, or saving and spending (eg Findlay & Jones 1982; Piggott & Whalley 1984). The use of alternative pricing arrangements may, in these circumstances, permit cost recovery with smaller efficiency costs.

3 When capacity is in excess of current demand, SRMC pricing will result in an enterprise generating a loss since SRMC will be less than LRMC. In practice, it is unlikely that production capacity will be optimal at any particular point in time due to uncertainties in demand and supply, the lumpy nature of investments for capacity increases, and the lead times associated with installing new capacity. Given these considerations, public enterprises could make either losses or profits for considerable periods of time if they set prices on a SRMC basis. If these enterprises are required to be self-financing this will require prices to deviate from SRMC so as to cover total supply costs. 4 Pure public goods have the attribute that an individual’s use of the good does not reduce the amount

available to others and it is too costly to exclude those who do not want to pay to use the good. Goods exhibiting these characteristics generally have a zero price and are funded out of taxation revenue (eg defence). Of more significance to this inquiry are goods for which marginal costs of provision are close to zero, but non-payers can be excluded at reasonable cost. These goods are sometimes called ‘impure public goods’. While the marginal cost of supplying these goods to additional users may be close to zero (eg broadcasting), the total cost of providing these goods is not, so setting prices on the basis of marginal cost would result in losses for the enterprise concerned. 5 Increasing returns to scale can arise from economies of scale and scope (refer to Appendix J). Over

the levels of production for which increasing returns to scale apply, long run average costs will exceed long run marginal costs. A price based on marginal cost would not therefore enable total costs of production to be covered and would therefore result in the public enterprise running at a loss.

Hence, where direct charging for goods and services is feasible, it can be argued that prices should be set to recover the total costs of production. Arguments in support of this approach include the user pays principle6 and the greater incentive that managers of public enterprises may have to minimise costs when they are required to avoid

losses. Modified marginal cost pricing regimes can thus be used to provide alternatives to the use of subsidies. The issue is how to set prices to minimise potential losses which might be associated with departures from marginal cost pricing.

Two pricing regimes which are consistent with cost recovery and which attempt to minimise the efficiency costs of deviating from marginal cost pricing are multi-part tariffs (frequently two-part tariffs) and Ramsey prices.

Under a two-part tariff, users pay an access fee as well as a usage fee. The access fee is designed to cover fixed costs (ie those costs which do not vary with usage), while the usage fee is based on the marginal costs of satisfying the demands of individual users. Considerable care needs to be exercised in setting the components of a two-part tariff, j This is because charging an access fee in excess of the cost of entry/connection may

deter potential users, while charging usage fees which exceed the marginal cost of | supply may limit use. Both components could involve an economic cost and, in raising revenue to cover production costs, this possibility needs to be taken into account.

Sometimes, it may be more efficient to charge only an access fee. For example, the marginal damage cost imposed by a car on an uncongested highway could be close to zero. Given the costs of excluding individuals from using roads, it may be more efficient to include any marginal usage costs for cars in an access fee.

Where there are major difficulties in applying multi-part tariffs, it may be more appropriate to consider so-called Ramsey pricing7 as an alternative.

With Ramsey-based usage fees, the price to each user (or user category) would include a mark-up above the marginal cost of supply designed to cover the overall costs of production. Mark-ups would vary among users according to their sensitivity to price changes. Thus, those users who are very responsive to price changes (ie have

a high elasticity of demand) would be charged a lower mark-up compared with those users who are less sensitive to price changes. This approach seeks to minimise any undesirable effects on resource allocation which would arise from applying a uniform mark-up to all users.

'

The choice between charging for units used (through either a two-part tariff or Ramsey-based fees) depends on several factors, including the ease of monitoring use of the goods and services and on their demand characteristics. Charging users different prices requires, amongst other things, that it is not possible for them to resell

the good or service.

6 The user pays principle refers to situations where the total costs of producing a good or service are recouped from charges to individual users directly related to their consumption or valuation of the good or service. Resort to this principle avoids the redistribution of income from taxpayers (some of whom may not be users of a particular good or service) to users arising from relying on taxes to fund

subsidies to cover operating losses. 7 Ramsey pricing is based on the taxation principles first advocated by Ramsey (1927). For a discussion of Ramsey pricing see Baumol and Bradford (1970).

Second-best considerations

A further complication with marginal cost pricing arises when prices elsewhere in the economy, notably those in markets of particular relevance to the industry under consideration, are not based on this principle. In such cases, departures from marginal cost pricing may be justified (Lipsey & Lancaster 1956). For example, if gas (which is substitutable for electricity in some uses) is priced below marginal cost, there will be excessive use of gas relative to electricity. In these circumstances, the application of second-best pricing rules would require the pricing of electricity below marginal cost to offset the distortion created by the incorrect pricing of gas.

Adopting second-best pricing regimes is likely to be complex (Mishan 1962). In order to calculate a set of second-best prices, a large amount of information on existing distortions in the economy may have to be collected, and inter-relationships between

a multitude of goods and services taken into account (Turvey 1971). In practice, it is unlikely that attempting to take account of distortions elsewhere in the economy will result in lasting improvements in the efficiency of resource use. Indeed, adherence to pricing rules which approximate marginal cost pricing as closely as possible may help to create pressures for distortions elsewhere to be dealt with directly (such as resetting the price of gas to correspond more closely to the marginal cost of its supply). In general, correcting distortions at their source is preferable to less direct methods of adjustment.

G3 External influences on the pricing practices of public enterprises

The pricing practices of public enterprises are largely shaped by the incentives environment in which they operate. A limited number of public enterprises in Australia (eg those involved in banking and some areas of insurance) operate in direct competition with other suppliers. These enterprises experience competitive pressures to price and produce efficiently. However, public enterprises are often the sole suppliers of goods or services either at a local, State or national level. Given their monopoly position and the often limited extent of competition from substitutes, these enterprises have considerable discretion in determining prices. Nonetheless, this discretion is restricted by various institutional provisions which affect their capacity to set efficient prices, including:

. the existence of unclear and conflicting objectives;

. a requirement to subsidise the supply of certain goods and services (commonly referred to as CSOs);

. the requirement to comply with pricing controls or guidelines; and

. the requirement to meet a financial target (or targets).

The influence of these provisions on the pricing practices of public enterprises is discussed below. A broader treatment relating to their effects on production, as well as pricing practices, is set out in Chapter 6. An examination of their effects on rail and electricity authorities is contained in Volume 4 of this report.

6

Unclear and conflicting objectives

Public enterprises are often required to meet a variety of loosely specified objectives which can and do conflict, necessitating compromises in the design of their pricing structures. For example, it is not uncommon for public enterprises to be required to ensure the efficient supply of goods and services while also ensuring an equitable

allocation of supply costs among consumers. However, equity can be interpreted in different ways. For example, it can be taken to require that prices reflect the costs of supplying a given user (or more often a group of users with common characteristics), often referred to as the user pays principle. Or it can mean that costs should be

allocated to users having regard to their capacity to pay in support of the income distribution or welfare goals of governments. The former requirement is consistent with setting efficient prices (provided production is undertaken at least cost), while the latter is likely to involve users facing prices which are unrelated to the costs of

supply. In such cases, the capacity of prices to promote efficient production and consumption decisions is likely to be impaired. The tax and social security systems often represent more effective ways of meeting social welfare or income distribution policy objectives.

The merits of this approach seem to have been recognised by some governments in recent years following reviews of the pricing structures of public enterprises in areas such as electricity and water. The pricing policies of some of these enterprises have been modified (or are being modified) to require the progressive elimination of cross­

subsidies between users of commercially supplied goods and services, with direct budget funding being substituted to finance price concessions to certain users.

The supply of non-commercial goods and services

Many public enterprises are directed to provide goods and services to some users on a non-commercial basis to achieve the social and development objectives of governments (see Appendix I). Well known examples of such directives (commonly

known as CSOs) include the requirement that Telecom and Australia Post provide a ‘standard service at a uniform price’.

While some CSOs are directly funded by governments through revenue supplements to the enterprises concerned (eg some of those applying to telecommunications, electricity, rail and water authorities), most are financed internally from charges applied by public enterprises on their goods and services, a practice known as cross­

subsidisation. This method of financing CSOs precludes the setting of efficient prices, since some users are required to pay more than their marginal supply costs while others pay less. As discussed in Appendix I, there are alternative and potentially more cost-effective ways of financing CSOs.

Pricing controls/guidelines

Another way in which governments influence the level and structure of prices set by public enterprises is through price controls and guidelines. Common rationales for

7

these controls/guidelines include:

. concerns by government to constrain the market power bestowed on some public enterprises by their sole supplier status;

. attempts to meet welfare or development goals through price concessions to various users;

. attempts to constrain inflationary pressures; and

. efforts to limit the size of price increases between years.

Various control mechanisms have been adopted, including the need for enterprise managers to gain ministerial approval to change their prices, the use of formal price review mechanisms such as the PSA, and the use of formal or informal price capping arrangements. While such measures can promote accountability and encourage greater operational efficiency, they can also act to impede the adoption of efficient pricing strategies. For example, while the requirement for ministerial approval of proposed price changes enables governments to exert some control over the pricing policies of their enterprises, it can also politicise the pricing decisions of these

enterprises and constrain their ability to adapt their pricing structures to changing market conditions. Since not all ministerial intervention in pricing is obvious, it is correspondingly difficult to determine the extent to which enterprise managers are responsible for the current level and structure of prices.

Another form of price control has been that annual price increases are limited to less than the general rate of inflation (as measured by the CPI). Such a price capping requirement is usually referred to in the annual reports of the affected enterprises or in ministerial statements relating to the enterprises concerned (eg electricity, gas and water authorities).

While such action by governments may be seen as a way of containing inflationary pressures in the economy, it may also limit the capacity of enterprises to restructure their prices in response to changing economic conditions. A general price index such as the CPI is unlikely to accurately reflect changes in the cost conditions of a public enterprise. If these costs have legitimately increased by more than the CPI, then it is appropriate for the price of goods and services produced by the enterprise to rise relative to the general movement in prices. Such a shift in the relative level of prices needs to be communicated to users in order to avoid inefficient consumption and investment decisions.

An alternative, and potentially more effective price capping measure, is a general or industry-based price index which allows for potential productivity gains (ie a CPI-X mechanism). The main features of this form of price regulation are discussed in Appendix L. While such an arrangement avoids the deficiencies of a single general price index like the CPI, it would do little to promote greater efficiency in the structure of pricing - if charges are excessive for one product but too low for another, they could remain so.

8

Financial targets

Financial pressures on governments and a growing recognition of the adverse effects of under recovery of costs by public enterprises (such as higher levels of taxes and/or large net borrowings by the public sector) have prompted governments to introduce financial targets to promote greater cost recovery by their enterprises. Such targets

have implications for the prices established by public enterprises for their goods and services, particularly where strict adherence to marginal cost pricing would result in enterprises making operating losses. In such cases, self-financing targets will require the development of alternative pricing strategies in the absence of an operating

subsidy from the government. As discussed in section G2, there are a number of alternative pricing strategies which are consistent with cost recovery, such as multi­ part tariffs and Ramsey-based charges. Their aim is to minimise the efficiency costs of applying charges to ensure cost recovery.

While the adoption of financial targets can promote better resource use, their capacity to achieve this outcome depends crucially on the manner in which they are specified, and whether appropriate adjustments are made for the specific circumstances of particular enterprises (see Appendix L). As normally specified, cost

recovery targets require that the total of all charges imposed on all users of the good or service cover the accounting costs of their provision. As such, the charges need not be linked closely to the economic costs of supply. In many enterprises, charges seem to be based on costs which exclude a requirement to recover adequate funds for asset replacement or earn an appropriate return on capital. Costs may be recovered

through an inefficient or efficient structure of prices with very different impacts on users and on the production and investment decisions of public enterprises.

G4 Public enterprise pricing in practice

The preceding discussion suggests that many external influences combine to limit the ability of and incentive for public enterprises to adopt efficient pricing practices. Indeed, the specification of unclear and conflicting objectives, together with the requirement to deliver CSOs, can preclude the setting of efficient prices and weaken

disciplines on enterprises to undertake production at least cost. While price controls/guidelines and financial targets have the potential to promote more efficient pricing practices, those currently in operation are characterised by important weaknesses which undermine this potential. Given these circumstances, the question

arises as to whether opportunities exist for public enterprises to price more efficiently.

Some of the main features of the current pricing practices of the more significant public enterprises are discussed in Chapter 4. In addition, the Commission has undertaken a more detailed examination of the pricing practices of rail and electricity authorities (refer to the studies in Volume 4 of this report). There is evidence of

improvements to the pricing practices of these enterprises in response to initiatives by the enterprises themselves, changes in policy settings by government, or a combination of the two. Nevertheless, the Commission’s studies, together with the comments of user industries, suggest widespread inefficiencies in pricing practices.

These arise, at least in part, from the operations of public enterprises themselves rather than being fully attributable to the impact of the external influences discussed in section G3.

9

Internal factors governing the operation of public enterprises which impact adversely on the efficiency of their prices can be grouped into three main areas, namely:

. inefficiencies in the production of goods and services which inflate their charges;

. incomplete accounting for the full economic costs of service provision when setting charges; and

. insufficient regard for reflecting variations in the costs of supply to different users in the structure of prices.

These factors are discussed below.

Inefficiencies in production

As noted in Chapter 4, inefficiencies in production will make costs and consequently prices higher than they would otherwise be, thus impairing the application of efficient pricing. Indeed, a number of industries participating in the inquiry indicated that their support for the principle of ‘user pays’ was dependent on public enterprises improving the efficiency of their operations.

While the external influences discussed in section G3, (and more generally in Chapter 6), influence the extent of inefficiencies in production, differences in the performance of key public enterprises between States suggest that other factors, within the control of enterprise managers, are also relevant. This, in turn, implies that there are opportunities for managers to reduce the level of their charges by raising the productive efficiency of their enterprises.

The efficiency of the production process is also affected by the extent to which charges adequately signal the actual costs which a user’s demand imposes on a supply system. For example, a structure of charges which fails to reflect the differing costs of

servicing user demands in peak as opposed to off-peak periods is unlikely to encourage the effective use of an existing supply network. Indeed, if the charge facing users is artificially low it will tend to stimulate demand and encourage increased investment to expand the capacity of a system at greater overall cost to meet a given level of demand. Examples of this and other production inefficiencies in the supply of government-provided goods and services arising from inappropriate price signals to users are discussed in Chapter 4.

Incomplete accounting for the costs of service provision

If charges are to reflect the full economic costs of supplying goods and services on an efficient basis, it is necessary to cost all the inputs used to supply a good or service at their opportunity cost (ie their value in next best alternative use). This is relatively straightforward for labour and material inputs (ie simply use the prices paid for these inputs), but less so for capital inputs employed by a monopoly supplier. In the case of capital, an appropriate allowance for depreciation/asset replacement and a ‘commercial’ rate of return need to be derived. While the importance of the concept

10

of opportunity cost is often accepted in principle, it is frequently not applied in practice.

The cost information systems of most public enterprises are based on historical cost accounting systems; revenues are valued in terms of current prices, and so are outlays on non-capital inputs, but costs for capital inputs (eg depreciation) are valued in terms of the prices paid for these inputs at the time of their acquisition.

Consequently, depreciation provisions often fail to provide adequately for the maintenance of the enterprise’s capital through time (see Appendix M). Moreover, the financial costs implicit in any accounting system ignore the need for an enterprise to provide for a return on its capital. A charge based on such cost information will

tend to significantly understate the full costs of resources used to supply users.

When charges are set ‘too low’ for these reasons, demand by users for goods and services will be stimulated. In order to meet current and future demand, public enterprises may expand their productive capacity prematurely. As discussed in

Appendix L, valuing assets on a replacement cost basis and establishing appropriate allowances for depreciation and a rate of return on capital, represent ways of ensuring that the costs of delivering goods and services are fully reflected in the prices charged to users. While there may be some difficulties in making the required calculations, they are not insurmountable. Indeed, a number of public enterprises have already adopted current cost accounting in order to improve their pricing practices.

One objection often raised in response to proposals for full economic costing of government-provided goods and services is that lack of discipline on costs (notably for monopolistic suppliers) may result in higher charges. Hence, a necessary pre­ condition for the unqualified acceptance of cost-based charges by users is the efficient

supply and delivery of goods and services.

Deficiencies in the structure of prices

Another attribute of an efficient pricing system is that charges reflect, as closely as practicable, variations in the cost of supplying different users. Since marginal costs of supply may vary between users, for example by time-of-use and depending on their location, an efficient pricing strategy will require that prices reflect these sources of

variations in cost as far as practicable, having regard to the costs and benefits of securing the necessary information and administering the pricing system. Related to this requirement is the need for pricing structures to be readily understood by users so that price signals actually affect their decisions in the desired way. In general, a

simple pricing system will be preferable to a complex system.

The capacity of public enterprises to develop such a pricing strategy requires information systems which are appropriately designed and sufficiently disaggregated to facilitate the collection of the relevant data. However, the information systems of many public enterprises are geared predominantly to the collection of data for the preparation of their financial statements, so that the identification of marginal costs

or variations in the cost of supplying different users is hampered by the absence of appropriate data. Many public enterprises do not possess the relevant information to set efficient prices. For example, in commenting on the accounting practices of the

11

SRA, the NSW Commission of Audit (1988, p. A.2.8) stated:

...the SRA has no acceptable cost allocation system. Hence, it is difficult to set appropriate prices, and the full extent of the implicit subsidies is not known.

Examples of how information has been or could be used more widely to develop more efficient pricing structures are briefly referred to below.

End Use

For many public utilities, costs are generally allocated to broad categories or classes of consumers, usually based on end use (eg domestic, farm, commercial). Such an approach is aimed at developing user classes which are relatively homogeneous (ie each consumer costs much the same to supply) for cost allocation purposes. In the case of electricity, evidence suggests that many customers within each end-use category have a wide range of usage and supply cost patterns. This implies that gains may be available from seeking to reflect these differences more fully in the prices charged to users. Failure to do so limits the capacity of prices to influence the behaviour of users.

Time o f use

The outputs of some public utilities (eg electricity, water, telecommunications) are characterised by significant variations in costs according to time of use. However, many of the charges for these goods do not reflect these cost variations, or provide

only very crude indicators of such variations. In the case of electricity, for example, many time-of-use tariffs are differentiated into two rating periods, peak and off-peak, although the costs of supply vary over several parts of the day. Consequently, consumers have limited incentive to adjust their consumption patterns towards off- peak periods. As a result, the supply networks have lower load factors and use the

higher cost segments of the network more frequently to meet demand than would be the case if time-of-use tariffs were more widely applied or better designed.

These deficiencies in electricity tariffs have been identified by price reviews held in several States. The old Bulk Supply Tariff offered by the Electricity Commission of New South Wales to county councils and large industrial customers, has recently been

restructured to more accurately reflect supply costs. The main features of the revised tariff include three rating periods (eg shoulder, peak and off-peak), differentiation of charges to reflect the costs of supplying different voltages, and energy charges based on marginal rather than average costs of supply. The off-peak rate reflects the average of the short run marginal costs of supply, whereas the peak and shoulder rates reflect the average of the long run marginal costs of generation and transmission in these periods. In view of the pricing principles outlined earlier, the new Bulk Supply Tariff is a major improvement over the earlier structure of electricity tariffs in NSW. The SECV has also moved strongly in the direction of relating tariffs more closely to supply costs. In contrast, despite the findings of a 1985 inquiry into gas and electricity tariffs in Western Australia, which identified opportunities for developing more efficient tariffs, few changes have been made to that State’s tariffs.

Location

Some element of price equalisation between supply points is common in the private sector. However, the extent of this practice appears to be excessive for some goods and services provided by public enterprises. Indeed, many public utilities set uniform I prices for their goods and services on a state wide or national basis. For example, a basic tenet of pricing by Telecom and Australia Post is the concept of a ‘uniform price for a standard service’. This has led to the setting of uniform usage and access fees

throughout Australia for some goods and services. For instance, the price of a local call is the same for all city and country centres. Charges for STD calls vary in accordance with discrete intervals of distance, but do not appear to vary with the likely extent of usage (ie density) of a particular route. The price for the delivery of a 1 standard letter is uniform throughout Australia despite significant variations in the

cost of delivery.

Such pricing policies have facilitated the development of more extensive public utility networks to the benefit of non-metropolitan users, but at a cost to the rest of the I community. If the price that one group pays for a good or service does not cover its costs, it will encourage excessive consumption by these subsidised users and

discourage consumption by those users charged above their full costs of supply. This may inhibit the use of alternatives or the development of new products to meet user needs at a lower overall cost to the community.

G5 Summary

Important requirements for efficient pricing are that charges reflect the opportunity cost of resources used and that they are based on efficient production (ie production at least cost). In a competitive market environment (and in the absence of externalties and other distortions), these requirements are satisfied by marginal cost pricing. However, most public enterprises do not face direct competition and, in

consequence, experience little pressure to adopt efficient pricing regimes.

Further complications arise in attempting to apply marginal cost pricing in the case of impure public goods, decreasing cost industries and where conditions of excess capacity arise, since charges may not recover the overall costs of supply. Such situations require the development of alternative pricing arrangements. These need to be structured to achieve a reasonable approximation to marginal cost pricing; they

will otherwise impose avoidable efficiency costs on the community. i

Against this background, it is not possible to state unequivocally that a particular form of pricing will be the most efficient in all cases. The form of pricing consistent with pricing efficiency will be influenced by the cost and demand conditions associated with a particular good or service and the costs associated with administering particular pricing regimes relative to their benefits. For example, a single access-

based charge may be appropriate in some cases, while in others a usage fee or some form of price discrimination (eg multi-part tariff or Ramsey pricing) may be appropriate.

The incentives environment in which public enterprises operate affects the extent to which enterprise managers are able to apply efficient prices. In this context,

13

government imposed constraints on their enterprises often inhibit the development of efficient pricing structures. Notwithstanding these constraints, it is also apparent that a number of inefficiencies (attributable to internal rather than external influences alone) characterise the pricing practices of public enterprises. These pricing inefficiencies appear to arise from inefficiencies in production, incomplete accounting for the full economic costs of producing a good or service, and insufficient regard for

reflecting variations in the cost of supply to different users. It is clear that while some governments and enterprise managers have taken initiatives in recent years to improve the pricing efficiency of public enterprises, much remains to be done. Many of the potential reforms identified in Chapter 7 offer opportunities for realising further significant improvements in the pricing efficiency of public enterprises.

REFERENCES

Albon, R. 1988, ‘The Welfare Costs of the Australian Telecommunications Pricing Structure’, Economic Record, vol. 64, no. 185, June, pp. 102-12.

Baumol, W.J. & Bradford, D.F., 1970, Optimal Departures From Marginal Cost Pricing’, American Economic Review, vol. 60, no.3, June, pp. 265-83.

Coase, R.H. 1946, ‘The Marginal Cost Controversy’, Economica, vol. 13, no. 5, August, pp. 169-83.

Committee of Inquiry into Gas and Electricity Tariffs in Western Australia 1985, Final Report, Perth, May.

Committee of Inquiry into Electricity Tariffs and Related Matters 1989, Report, (J.W. Butters, Chairman), Sydney, March.

Dragun, A.K. 1983, ‘The Marginal Cost Pricing of Water : The Hunter District Water Board Application, CRES Working Paper, ANU.

Findlay, C.C. & Jones, R.L. 1982, ‘The Marginal Cost of Australian Income Taxation’, Economic Record, vol. 58, no. 162, September, pp. 253-62.

Gallagher, D & Robinson, R. 1977, Influence of Metering, Pricing Policies and Incentives on Water Use Efficiency, Technical Paper no. 19, Australian Water Resources Council, Canberra, AGPS.

Hanke, S. H. 1981, O n the Marginal Cost of Water Supply’, Water Engineering and Management, February, pp. 60-3.

Hazlewood, A. 1950, Optimum Pricing As Applied to Telephone Service’, Review of Economic Studies, vol. 18, pp. 67-78.

Houthakker, H.S. 1951, ‘Electricity Tariffs in Theory and Practice’, Economic Journal, vol. 51, March, pp. 1-25.

Hunter District Water Board, 1982, ‘An Equity Based Water and Sewer User-Pays Tariff, Newcastle, July.

Joskow, P. 1976, ‘Contributions to the Theory of Marginal Cost Pricing’, Bell Journal o f Economics, vol. 7, no. 1, pp. 197-206.

Kay, J.A. 1971, ‘Recent Contributions to the Theory of Marginal Cost Pricing : Some Comments’, Economic Journal, vol. 81, June, pp. 366-71.

15

Kolsen, H.M. 1966, ‘The Economics of Electricity Pricing in NSW’, Economic Record, vol. 42, December, pp. 555-71.

Lipsey, R.G. & Lancaster, K. 1956, ‘The General Theory of Second Best’, Review of Economic Studies, vol. 24, no. 1, pp. 11-32.

McColl, G.D. 1976, ‘The Economics o f Electricity Supply in Australia’, Melbourne University Press, Melbourne.

Mishan, E.J. 1962, ‘Second Thoughts on Second Best’, Oxford Economic Papers, vol. 14, no. 3, October, pp. 205-17.

New South Wales Commission of Audit, 1988, Focus on Reform : Report on the State’s Finances, (C.P. Curran, Chairman), Government Printer, New South Wales, July.

Ng, Y.K. 1987, ‘Equity, Efficiency and Financial Viability : Public-Utility Pricing with Special Reference to Water Supply’, Australian Economic Review, 3rd Quarter, pp. 21-35.

Piggot, J.R. & Whalley, J. 1984, ‘Public Goods and Welfare with Distortionary Taxation : An Applied General Equilibrium Approach’, Australian National University, Mimeo.

Ramsey, F.P. 1927, ‘A Contribution To The Theory of Taxation’, Economic Journal, vol. 37, March, pp. 47-61.

Randall, A. 1981, ‘Property Entitlements and Pricing Policies for a Maturing Water Economy’, Australian Journal o f Agricultural Economics, vol. 25, no. 3, pp. 195-221.

Rees, R. 1984, ‘Public Enterprise Economics’, Weidenfeld & Nicolson, London .

Turvey, R. 1969, ‘Marginal Cost’, Economic Journal, vol. 79, June, pp. 282-99.

____ 1971, ‘Economic Analysis and Public Enterprises’, Allen & Unwin, London.

Williamson, O.E. 1966, ‘Peak Load Pricing and Optimal Capacity Under Indivisibility Constraints’, American Economic Review, vol. 56, no. 4, September, pp. 810-27.

APPENDIX Η : IMPACT OF OWNERSHIP ON ECONOMIC PERFORMANCE

Different forms of ownership provide different incentive structures for owners, managers and employees. Private firms have incentives and disciplines for better performance - the profit motive, the sharemarket and the threats o f takeover and insolvency - which are not available to the public sector. Empirical studies o f comparative performance of public and private firms indicate that these incentives operate sufficiently well in

competitive markets to cause private firms to out-perform public firms. Where competition is restricted or prohibited by regulation, the evidence as to relative performance is inconclusive; it appears that the effectiveness of incentives in private firms is impaired by regulation and the absence of

competition.

HI Introduction

The extent to which the nature of ownership influences the economic performance of an enterprise has been the subject of considerable debate. One view is that ownership does matter; private ownership promotes the internal efficiency of firms and overall efficiency in resource use.

In considering the performance of public enterprises, the Commission has sought to evaluate the relationship between ownership and economic performance. Its approach has been twofold. One has been to consider how the different forms of ownership may influence the structure of incentives facing managers and hence managerial behaviour and firm performance. The other approach has been to determine whether differences in the structure of incentives under different

ownership regimes are reflected in differences in the comparative performance of public and private firms.

Attachment H I to this appendix reports available theoretical and empirical research on the divergence of managerial and market incentives arising from the different forms of ownership. It considers these incentives in terms of the objectives of private and public firms, the mechanisms available for internal monitoring of performance,

and the role of takeovers and insolvency in disciplining potentially poor performance. The conclusion that can be drawn from this review is that theoretical and empirical research does not provide unambiguous conclusions about the strength of managerial

incentives in promoting the efficiency of private firms relative to public firms. It indicates that the internal mechanisms for monitoring private sector performance and the external disciplines provided by the threat of takeover and insolvency do not operate perfectly.

These external disciplines are not available in any significant sense to govern the behaviour of public sector managers. Various alternative mechanisms may be adopted in public enterprises, but they have important limitations with the result that ‘... it is likely that public monitoring systems are generally less effective than their

private counterparts’ (Vickers & Yarrow 1988, p. 44).

These conclusions can be tested further by reviewing the available empirical studies of the comparative performance of public and private firms.

H2 Empirical evidence: a brief review

Studies of the performance of public and private firms fall into two broad categories. One large group comprises studies which are restricted to the analysis of the performance of public enterprises alone, without comparable data on private firms. The focus of the other group is a direct comparison of the performance of public and private firms, between countries or within one country, within the same industry, or over time or at a particular time.

In the case of those studies which evaluate the performance of public enterprises alone, there are sometimes explicit or implied conclusions that any observed inefficiency is due to their public ownership. However, inefficiencies can arise from a variety of factors which have little to do with ownership. Of these, perhaps the most important is that public enterprises are commonly not subject to direct competition. The role of these studies is more important as a means of calling governments and managers to better account for their policies and actions and as a basis for considering proposals for reform rather than indicating anything on the relative merits of public and private ownership.1

In evaluating the influence of ownership on performance, this section surveys only the results of those studies which focus on the comparative performance of public and private enterprises. A number of such studies has been conducted overseas employing a variety of technical methods to measure different indicators of comparative performance and a variety of ways of allowing for other influences on performance. Relatively few studies have been conducted in Australia.

In their survey of some of these studies, Domberger and Piggott (1988, p. 150) state that there is no single study which is entirely compelling in its findings ‘... because cost comparisons are rarely straightforward in the absence of a controlled experiment’. As Marchand et al. (1984, p. 26) noted: ‘The hard problem is in fact not so much to obtain the evidence as to make good economic sense of it’.

One of the main difficulties in the evaluation of comparative performance is in finding suitable cases for comparison. In Australia and Western Europe many of the activities carried out in the public sector are not undertaken in the private sector and the number of cases which can be reviewed is limited. It is difficult to isolate the influence on performance of ownership from the influence of other factors such as the regulatory environment in which the firms being compared operate. Public enterprises frequently possess a number of advantages in competing with private enterprises, as well as being subject to certain restrictions on their activities. In some industries, the operation of both private and public firms is highly regulated and barriers may exist to the entry of other potential competitors. Hence, it is necessary

to understand the nature of the comparisons in order to evaluate the reliability of the results of these studies.

1 Such studies have been carried out in a number of countries. Some of them are summarised in Marchand et al. (1984), Parris et al. (1987) and Vickers and Yarrow (1988).

18

In addition, there are other methodological and technical difficulties involved in the conduct of these studies which influence their reliability and the interpretation of results. In particular, the various types of quantitative techniques employed to measure relative efficiency of private and public firms are susceptible to

methodological and data error (Forsund et al 1981).

International evidence

The inconclusive nature of the literature is reflected in the two major surveys. Borcherding et al. (1982) cited more than 50 studies from five countries and concluded that findings in most of the studies ‘...are consistent with the notion that public firms have higher unit cost structures’ (p. 134). This finding, however, is

contradicted by the other major surveys of the literature on this subject (Millward 1982, p. 83; Millward & Parker 1983, p. 258), which found no broad support for private enterprise superiority’.

Domberger and Piggott (1986), after reviewing the available evidence, favoured Borcherding’s finding. They pointed out that, of the more than 50 studies surveyed, in only nine did private enterprise fail to outperform public enterprise. In a further four studies examined by Millward, but not by Borcherding, this was also the outcome.

Similarly, Yarrow (1986, p. 375), in a review of some studies of comparative performance (see Table H I) concluded that:

... taken as a whole, the results do point to a presumption in favour of private ownership, provided that other product market failures are insignificant or can be adequately corrected by means of alternative

policy instruments.

He also noted:

Where product markets are less monopolised, the comparative performance studies suggest a more favourable verdict on private enterprise, implying that incentive failures associated with government monitoring are empirically significant.

The observation can be confirmed by considering the results of some of the studies summarised in Table HI, with particular attention to the market conditions in which the comparisons are made.

The industries that have been most intensively studied include electricity generation and distribution, water, and refuse collection. Most of the studies have been conducted in the United States where the two types of ownership frequently co-exist in similar market conditions, although they generally operate in separate regions. In both the electricity and water industries, direct competition is generally absent and

both public and private firms tend to be highly regulated.2 In contrast, refuse collection appears less regulated in the sense of being more open to entry for new suppliers tendering for contracts.

2 As a result, comparisons of performance have to be made between the two types of ownership across regions, rather than in the one geographical location.

19

Table HI indicates that public firms more often do as well as, or perform better than private firms in the more highly regulated industries of electricity and water supply. Vickers and Yarrow (1988, p. 41) made the following conclusion from their analysis of the comparative studies in the US electricity and water industries:

... we are therefore led to the conclusion that, where firms face little product market competition and are extensively regulated, there is not generally decisive evidence in favour of one or other type of ownership.

The conclusion seems also valid in circumstances where public and private firms directly compete with each other under a regime of extensive regulation. A study by Caves and Christensen (1980) of productivity growth in Canadian railroads found there was little difference in the performance of the public and private railway enterprises. Moreover, in this case significant competition was present from other modes of transport.

In less regulated environments where the potential for competition is greater, the results in Table H I seem to indicate that private firms exhibit greater internal efficiency than public firms. The research on refuse collection in the United States highlights the point. Other studies included in Table H I support this. For example, a study carried out in the UK by Pryke (1982) compares three activities where services were provided by both the public and private sector: airlines; ferries and hovercraft;

and the sale of electricity and gas appliances. With the possible exception of airlines, there were no major barriers to the entry of new competitors and the regulatory environment was not extensive. In each of these activities Pryke’s analysis shows a picture of a more profitable private enterprise increasing its market share at the expense of the public sector. In cases where comparisons of cost levels and efficiency were feasible, Pryke’s analysis generally shows the private enterprises performing better.

Where government regulatory barriers to competition are not significant, the evidence suggests that private firms are generally more efficient than public firms. Where natural monopoly elements are present and extensive government regulations are in place, the results are inconclusive. In these market conditions the internal incentives to efficiency of private firms appear to be impaired, as in a greater number of cases their performance appears to be equalled or bettered by public firms.3

3 There are, of course, exceptions to these generalisations. For example, a study of German automobile and life insurance by Finsinger (1984) shows on the basis of a sample of about 100 companies that there are no obvious differences in cost efficiency between private and public companies.

20

Table H I : International studies of comparative performance

Industry A u th ors3

C on clu sion on pu blic ow n ersh ip

B a sis o f com p arison between p u blic an d private

Airlines Pryke (1982) Anti daily flying hours, capacity

tonne km, profitability

Appliance showrooms

Pryke (1982) Anti expenditures/staff, mark-up,

market share, profitability

Electric utilities

Atkinson and Halvorsen (1986)

Neutral costs

De Alessi (1977) Anti pricing policy

Di Lorenzo and Robinson (1982)

Neutral costs

Fare et al. (1985)

Neutral costs

Meyer (1975) Pro costs and prices

Moore (1970) Anti costs

Neuberg (1977) Neutral cost-efficiency

Peltzman (1971) Anti pricing policy

Spann (1974) Neutral operating costs, investment

costs

Yunker (1975) P ro/

neutral

unit costs, customer costs

Pescatrice et al. (1980)

Pro costs

Fire services

Ahlbrandt (1973)

Anti costs

Ferries Pryke (1982) Anti turnover per employee

profitability, growth in traffic

Insurance Finsinger

(1984) Freeh (1979)

Pro premiums, dividends, complaints

Anti costs, service, number of errors

Medicine Wilson et al.

(1982) Caves et al. (1980)

Anti productivity

Railways Neutral total factor productivity

Refuse Hirsch (1965) Anti costs

Kitchen (1976) Anti costs

Pier et al. (1974)

Pro costs

Pommerehne (1977) Savas (1974)

Anti costs

Anti truck utilisation

Savas (1977) Anti overall costs

Steel Rowley et al.

(1981)

Anti total factor productivity,

technological diffusion

Urban Pashigian Anti profitability, revenue, mile,

transit (1976) demand

Water Bruggink (1982) Pro costs

Crain et al. (1978)

Anti costs

Feigenbaum et al. (1983)

Neutral costs

Mann et al. (1976)

Pro costs

a The following studies are sourced in Yarrow (1986) and Atkinson and Halvorsen (1986).

The Australian evidence, although not as wide ranging as the international literature, on balance appears to provide qualitative support to the case that private enterprise is relatively more efficient than public enterprise where competition occurs. Table H2

provides a summary of the results of some major Australian studies.

Domestic airlines

The case of the airlines, in particular, has been widely studied. The provision of air services in Australia is highly regulated through the two airline policy. Scheduled services on the major air routes are allocated in almost equal proportions between the two operators, Australian Airlines (publicly-owned) and Ansett Airlines of Australia (private). Under this arrangement competition is extremely limited and new entrants

are effectively excluded.

Davies (1971) concluded, on the basis of labour productivity measures, that the private airline was more efficient than the public airline. Subsequent studies by Davies (1977), Mackay (1979), Kirby (1984), and Kirby and Albon (1985) reached similar conclusions. Other studies - Hocking (1977); Forsyth and Hocking (1980);

Millward & Parker (1983) - concluded that little distinction could be made between the efficiency of the two firms.

Kirby and Albon (1985, p. 539) suggest that the evidence as to the comparative performance of public and private airlines should be kept in perspective. While there is some evidence that the public firm is less efficient than the private firm, the difference is likely to be small in comparison with the inefficiencies of both operators due to the regulatory environment in which they have been operating. Some of the studies (Hocking 1977; Mackay 1979) of the cost efficiency of the major airlines have suggested that costs are higher than necessary, especially in comparison with North American airlines.

The studies of the airline industry appear to provide support to the conclusion that where firms are extensively regulated and protected from competition, the nature of ownership may not be the most significant influence on performance. It is the

regulatory environment and the lack of competition which seems to be the major impediment to more efficient operations.

Banking industry

The banking industry, unlike the domestic aviation industry, is not nearly as regulated. The publicly-owned Commonwealth Bank competes directly with private banks. While there are minor differences in their service quality, interest rate charges, borrowing limits, and the risk profiles of their investment portfolios, there are only minimal differences in pricing policies.

Australian studies of comparative performance

22

Table H2: Australian empirical studies of comparative performance

In du stry A u thors

C on clu sion on p u blic ow n ersh ip

B a sis o f com p arison between p u blic an d private

Airline Davies (1971,1977) Anti labour productivity

Hocking (1977) Neutral physical units of

inputs and outputs

Forsyth and Hocking (1980)

Neutral labour productivity

Kirby (1984) Anti index of cost

efficiency

Kirby and Albon (1985)

Anti index of cost

efficiency

Mackay (1979) Anti operating costs

Millward and Parker (1982)

Neutral productivity

Bus and Bus and Coach Association Anti unit costs of

Coach (1985) operation

Hensher (1987) Anti operating costs

Banking Davies (1981) Anti various measures

Coughlin (1987) Anti various measures

23

There have been two studies comparing the relative efficiency of the Commonwealth Bank and the private banks; one was conducted prior to deregulation of the financial system and one after. Davies (1981) found that the managers of the government- owned bank held a higher proportion of their assets in low risk, low return investments than did private banks. The Government bank was found to have larger

staffs and less efficient monitoring and work practices resulting in a lower overall profit rate.

Coughlin (1987) analysed the relative efficiency of the Commonwealth Bank by comparing its returns on equity and on assets with the returns of the three major private banks. He concluded that, in terms of profitability, efficiency and growth, the Commonwealth Bank has lagged behind the three major private banks.

These studies appear to support the general conclusion that in a competitive, unregulated market, the private firm will out-perform the public firm in terms of efficiency. However, because of the limited number of studies in this area and the fact that neither of the studies cited appear comprehensive, this conclusion is tentative.

Bus and coach services

Studies of bus and coach services have focussed on comparing the operating efficiency of the government-owned Urban Transit Authority (UTA) with private bus operators in the Sydney Metropolitan area. A study undertaken for the Bus and Coach Association of NSW (1985) compared the unit costs of operation of the UTA and nine private firms, and found that the private bus services could be provided at half the cost of the municipal bus system.

A critical issue in any comparison is the level of service provided. Public bus operators are required, as a community service obligation, to service geographically distant or low traffic routes. Such services may not have to be provided by private operators. The above studies took no account of this, and may have overstated the likely unit cost differentials between the public and private operators.

Even when account is taken of the service obligation of public bus operators, they still appear to be less efficient than their private counterparts. A recent study by Hensher (1987) which accounted for the ‘social responsibility’ of public operations found significant differences in operating costs, with the private operator much less costly than the public operator. Hensher concluded:

Even if the private operators were required to supply the range of services required to maximise net social surplus (rather than maximum profits), their approach to the supply of labour and other variable costs items explains much of the cost efficiency difference between the provision of services by private and public operators.

Competition and ownership

The previous sections suggest that evidence of superior private sector performance is observed largely in competitive markets. Where market disciplines are blunted, as is in the case of regulated monopolies, or where public and private firms are protected

from competition by regulation, the evidence on comparative performance is largely inconclusive. However, public firms have been found to be more efficient in a larger number of cases in regulated markets than occurs in competitive markets. It may be concluded that the margin between the relative performance of private and public

firms in markets appears to be eroded where restrictions on competition exist.

Mechanisms reviewed in Attachment H I promoting the relative internal efficiency of private firms depend on the operation of competitive market forces. If product markets are competitive, lower efficiency tends to result in falling market share, low

profits, and takeover of unsuccessful firms. Where regulation restricts competition, for example by guaranteeing market share and restricting entry to the industry, the market discipline associated with private ownership appears to work little better, and may even do worse, than those monitoring devices available to the public sector. The

regulation of private firms to restrict competition blunts market disciplines and may distort incentives in ways which lead to cost levels higher than in competitive markets and sometimes higher than in regulated public enterprises.

The major surveys cited above both agree that the presence or absence of competition are important determinants of a firm’s economic performance. As Kay and Thompson (1988, p. 25) have stated in their consideration of the research on comparative performance: ‘it is not ownership as such, but the interaction of

ownership and competition that promotes efficiency.’

H3 Assessment

The nature of ownership clearly does have an influence on the performance of an enterprise. Whether this is a decisive influence in all market conditions is questionable.

In competitive markets there are incentives and disciplines which can usually be expected to promote greater efficiency on the part of private firms. The evidence for superior private performance comes largely from studies of competitive markets. The incentive to produce and price efficiently comes from the requirement on private

firms to make profits — to stay in business in a competitive market. Low efficiency results in falling market share (share price) and profits, even bankruptcy. The possibility of a takeover or merger poses a threat to incumbent managers and firms, and hence provides some control over managerial behaviour.

These incentives and disciplines may not operate perfectly; firms may have objectives other than profit maximisation, managers may not always act in the interests of shareholders, share markets may not always serve as adequate performance monitoring mechanisms, and the threat of takeover or insolvency may not always lead

to improved management performance. As a result, there may be some limitations to the impact of these incentives and disciplines on the internal efficiency of private firms.

Nevertheless, even these imperfect mechanisms are not available in the case of public firms. The inability to transfer ownership or property rights under public ownership weakens the mechanisms that would otherwise act to align the actions of management with the interests of owners. Public enterprise managers are not

normally threatened by takeover. Their performance is difficult to assess, they are

less subject to scrutiny and external monitoring by the capital market and they are not exposed to bankruptcy. Public firms are also susceptible to political interference and are often required to pursue non-commercial objectives.

The weight of empirical evidence on the relative performance of public and private firms generally supports the view that, under competitive conditions, a private firm will be more efficient than a public firm. Although many of the studies suffer from methodological problems, a reasonably strong case for private ownership of firms in a

competitive market can be made. The incentives facing private firms work sufficiently well in competitive market conditions to ensure that private firms operate more efficiently than public firms. The effectiveness of these incentives appears to be impaired where regulation restricts or prohibits competition. The empirical evidence on the comparative performance of private and public firms where competition is limited by regulation remains inconclusive.

ATTACHMENT HI: OBJECTIVES AND MANAGERIAL INCENTIVES

The expectation that ownership does matter for the efficiency of firms relies on the judgement that differences in the structure of incentives facing managers in public and private firms results in superior economic performance by private firms. These differences emerge in relation to: the objectives of firms, the mechanisms available

for internal monitoring of performance, and external discipline on performance from takeovers and insolvency.

This Attachment draws on available theoretical and empirical research to consider the extent to which such differences occur in the incentive environment facing public and private sector managers.

Objectives of public and private firms

Economic analysis of the behaviour of privately owned firms traditionally rests upon the assumption that the aim of decision makers is profit maximisation by producing its chosen level of output in the most efficient way. Private firms appear to have a clear cut objective - profit maximisation through cost minimisation and efficient production.

Critics of the profit maximisation assumption are legion. A number of alternative objectives of firms have been proposed, including sales maximisation, the asset value of the firm, and achieving satisfactory profit and sales standards rather than maximising them. Whatever their relative merits, and these have been debated intensively over decades, the central point remains that in the presence of actual or potential competition, some firms are likely to be driven from the market or taken

over by other firms that can produce at lower cost. Thus, profit generally remains a powerful motivation for firms to reduce costs. As Robinson and Eatwell (1973, p. 236) observe:

... it remains true that all business decisions are taken under pressure of the need to maintain profits and avoid losses.

Irrespective of whether or not private firms actually maximise profits, they do prefer more profit to less.

In contrast, the Commission has found in this inquiry that the objectives of the public firm are often imprecise and in some cases inconsistent with that of profit maximisation. Public firms may, for example, be directed by government to pursue non-commercial objectives, such as providing services to a particular community

group at less than cost. Public firms may also be vulnerable to political interference. Such interference and the pursuit of non-commercial objectives may result in outcomes not consistent with efficient production.

The possibility that firms and their managers may pursue other objectives raises the question of the relationship between owners and managers of firms in the pursuit of

27

more profits rather than less. In the economics literature this is known as the principal-agent problem .

An incentive problem arises from the separation of the principal or owner (a shareholder in a private firm, the community in the case of a public firm) from the agent (management). The problem is that owners and managers may not have the same objectives. The owner naturally wants to induce management to act in the owner’s interests, but he does not have full information about the circumstances and behaviour of management, and so he has a monitoring problem. This reduces the owner’s ability to tell management what to do, because he cannot fully observe what is happening.

In large public or private corporations, decision makers are usually professional managers whose rewards will not exclusively be dependent on profit flows or salaries. Managers may derive non-pecuniary rewards by pursuing objectives other than those of the owner, such as corporate sales and growth, or a discretionary level of managerial expenditures. In other words, they may seek greater power and prestige for the manager.1

Although both private and public enterprises are vulnerable to principal-agent type problems, it can be argued that, for private firms, exposure to competition through the sharemarket and the market for corporate control enables such problems to be more effectively controlled than in publicly owned firms, leading to better performance. The effectiveness of those mechanisms is considered in the following sections.

Sharemarket monitoring

The typical large private firm has many shareholders, each of whom possesses a relatively small fraction of the total equity of the firm and can transfer property rights in the firm to another investor.

Shareholders have a direct interest in the performance of the firm because their returns can only be maximised when the firm maximises its profits. In this sense the shareholder is assumed to desire to maximise expected profits.

Information about the relative performance of firms and their management is reflected in the sharemarket. The share price reflects publicly available data on the current performance and future earnings of firms, and this in turn reflects the value of decisions made by its managers. Because shareholders and potential owners can trade inexpensively in ownership rights, they can individually and collectively signal their views of corporate and managerial performance. Thus, if dissatisfied with the performance of management, the shareholder can remove funds invested and, because this affects the firm’s access to finance, it acts as a signal to management.

However, the effectiveness of the sharemarket as a monitoring mechanism may be limited. The information generated by the sharemarket is not uniformly distributed

1 Niskanen (1971) illustrated the effect of the incentive problem in public firms where the bureaucracy dominates the government because of its informational superiority and, hence, pursues the objectives of more pay, power and prestige. He suggested that managers of public enterprises wanting to maximise their utility will adopt pricing rules and investment patterns which maximise the number of their subordinates and the amount of money they decide on.

among managers, and different managers do not respond to a given set of information in the same manner. Moreover, the interests of shareholders may vary in that they will have different assessments of alternative managerial policies. There is uncertainty in the market and, as a result, profit maximisation cannot be attributed to

a particular managerial action/strategy. Consequently, particularly in the short term, share prices may not accurately reflect corporate performance.

The dispersed ownership of a firm’s shares also may limit the effectiveness of shareholder monitoring of management because the activity of specifying and enforcing managerial contracts confers external benefits to others. To illustrate, if only one shareholder is engaged in this task, he bears the full cost of the activity but

receives only a fraction of the total gain. Vickers and Yarrow (1988, p. 13) argue that:

In such circumstances there is a danger that, from the perspective of shareholders as a whole, the intensity with which managers are monitored will be suboptimally low. If that is the case managers can be said to have discretion to pursue their own objectives and it may not be appropriate to base analysis of company behaviour on the expected- profit-maximization assumption.

There are some equally strong arguments in defence of the traditional line. In cases where a particular group holds a significant proportion of the share capital, the effects of share dispersion are unlikely to be severe. Where shareholders maintain a large diversified portfolio of shares active monitoring of one company may have an impact

on the managers of other companies in which the investor has an interest.

Shareholders may appoint a board of directors to serve as an agent in specifying and enforcing managerial contracts. While such appointments also involve principal/agent problems, they may at least partially limit the discretion of managers. Shareholders may also provide senior managers with incentive schemes which are related to the firm’s profitability, hence increasing the competition for these positions

and promoting managerial and firm performance.

Thus, while it cannot be expected that managers of privately owned firms will always act in the best interests of shareholders as profit maximisers, the exposure of the firm to the market for share trading, together with certain institutional devices, generates some degree of control over managerial incentive problems.

In contrast, similar monitoring controls do not exist for the public firm because ownership is non-transferable and hence shares are not traded. Alchian (1965) noted that the non-transferability of ownership in publicly-owned firms decreases the concentration of rewards and costs on the persons responsible for them and also inhibits specialisation of ownership according to comparative advantages in knowledge or attitudes towards risk. He concluded that public ownership weakens the relationship between managerial utility and firm profit, and therefore adversely

affects efficiency.

Furthermore, in the case of the public firm, the monitoring mechanisms that are available to the citizen appear particularly weak. The main arguments were summarised by Clarke and Porter (1982, p. 10) as follows:

oo

First, each citizen is in a free rider situation: the benefits of any individual action intended to increase the value of the firm accrue to all. Unlike private ownership, an individual cannot appropriate the returns to his superior ability as an owner by increasing his equity in a state enterprise. His equity, and consequently his gain as an owner from a change in corporate policy, must remain small. Second, in a majority rule system, in which governments are elected on the basis of a package of policies, the voices of individual voters on particular issues carry little weight. Third, it is very costly for an individual to obtain information about how the enterprise might be better run. there are not always market incentives for the provision of information as there are in private enterprise. Fourth, because the individual cannot sell his equity in the enterprise, management will not receive information on the market’s grading of its (relative) performance, which is in the private sector concisely expressed in (relative) share prices. For these reasons, incentives are few for individual shareholders in publicly- owned enterprises to attempt to increase the value of the enterprise.

The controls on incentive problems in public firms appear much weaker because the property rights cannot be transferred, and this precludes the operation of the sharemarket as a monitoring mechanism.

There are alternative mechanisms available to monitor the performance of public sector managers. The public reports of Auditors-General and of parliamentary sub­ committees provide some discipline on the operation of public firms, although the extent does vary between the various governments of the Commonwealth. Moreover, public sector managers can be fired in the same way as managers of private firms.

Nevertheless, these mechanisms do not seem to be as effective as those applying to private enterprise. The absence of an exclusive right to any income accruing from an individual’s efforts to improve the use of public assets impairs the mechanism for controlling incentive type problems and significantly reduces incentives for good economic performance. This absence is compounded by the lack of a market for corporate control (takeovers and mergers) and lack of insolvency risks for public firms.

One final consideration which is more specific to the environment in which firms operate in Australia is the efficiency of sharemarket monitoring in the case of transnational corporations. The presence of a number of firms whose head offices and major shareholders are located in other countries may be said to inhibit the potential for monitoring and influence by Australian investors. This may be true, but there is no reason to expect the objectives of overseas shareholders in these firms to differ markedly from those of Australian shareholders - they, too, will prefer more profit to less.

It might be argued that shareholders in the parent company will be more concerned with the overall performance of a corporation than with the performance of each subsidiary separately. As a result, poor performance of a subsidiary in Australia may be neglected at times as long as overall performance of the corporation is satisfactory.

Poor performance may also be sustained by the head office in order to maintain a market presence as part of its long-term strategy. But it cannot be expected that poor

30

performance will be neglected over the long term if it becomes a drain on potential overall profits and returns to shareholders. Managements do review the performance of the various arms of a corporation in considering future investment strategies and firm restructuring to maintain overall performance to the satisfaction of shareholders.

Takeovers

The operation of a market for shares (equity) in private firms provides a mechanism for transferring ownership rights. It allows an entity (an individual or an institution for example) to bid for a stake in a company - a takeover bid. It is argued that the marketability of ownership reinforces incentives for management in private firms to

maximise profits.

If a management team were to depart from profit maximising policies, profits would fall below those expected by the market and this would see the value of the company’s shares fall. Outside interests which saw that profits would be higher if different strategic choices were made could bid on the stock market for a controlling stake in the firm. A successful bid would see the old management team displaced and new profit maximising policies enforced. As a result of this threat, management is likely to pursue various strategies to minimise falling performance, such as selling off unprofitable divisions before a takeover threat becomes imminent. The existence of a perceived threat of a takeover acts as an incentive mechanism that deters

management in private firms from pursuing policies that are substantially removed from the interests of shareholders.2 By contrast, in the case of the public firm no such deterrent exists as ownership cannot be transferred through the market for shares.

There has been a considerable amount of economic research, both theoretical and empirical, aimed at testing the extent to which takeovers act as an incentive mechanism which promotes efficiency in private firms. The strength of the relationship has been questioned on several grounds.

First, it is alleged that takeovers bias decision-making and profit maximisation toward the short-run, to the detriment of research and development and long-term investment planning. To date, this allegation has not been tested adequately. Scherer (1988, p. 80) reports that: ‘In the present state of knowledge, the question of how

takeovers affect R&D spending and other far-sighted investment remains unresolved.’

Grossman and Hart (1980) argue that when shareholdings are dispersed, it is in each shareholder’s interest to decline a takeover offer and ‘free ride’ on the performance improvement that it is anticipated the takeover will generate. If enough shareholders behave in this way, the takeover bid will fail. Consequently, the link between

efficiency and a takeover is weakened. However, Schleifer and Vishney (1986) find that the presence of a large minority shareholder is a partial solution to this ‘free­ rider’ problem and report that in a large proportion of enterprises at least one large minority shareholder exists.

2 This argument does not imply that any deviation from profit maximisation will result in a takeover: some degree of managerial discretion will exist because of factors such as transaction costs associated with takeovers or imperfect markets.

31

Takeovers have also been questioned as an incentive for better performance on the grounds that the sharemarket is an inadequate means of performance evaluation. In his examination of the efficiency arguments for takeovers Scherer (1988, p. 72) stated:

The market processes that put a company in play are at best noisy, creating incentives for takeover when management errs and when the market errs.

and (p. 80) that:

... stock prices move over time in something approximating a random walk. Companies can therefore become targets not only because their managers have erred in failing to maximise profits, but because the stock market has erred, randomly setting share prices so low as to make

their issues a bargain worth snapping up.

As Summers (1986) has noted, empirical proof or refutation of the efficiency of the stock market is extremely difficult. Yet, it is this question which lies at the heart of much disputation about the efficiency-enhancing effects or otherwise of takeovers.

Finally, there is the question as to whether the objective of the acquiring firm is expected profit maximisation. And even if the acquiring company is a profit seeker, takeovers may be motivated by factors such as the gains from increased market power or from reductions in tax liabilities. For example, in the wave of takeover activity in

recent times in many countries, the charge is often raised that the merger activity is closely correlated with stock market prices as a result of tax-induced distortions in the pricing of financial assets. As in the argument above that the sharemarket may err in

its judgement, the implication is that even an efficient management may be vulnerable to takeover bids and so the link between current performance and the takeover threat is weakened.

The empirical studies which consider the efficiency promoting qualities of takeovers are broadly grouped according to those which attempt to estimate the empirical relationship between company performance and the likelihood of takeover, and those that consider the effect of takeovers on subsequent performance.

The relationship between company performance and the likelihood of takeover is the basis of the assumption that takeovers generate incentives for managers to act in the interests of owners. The extent of research on this relationship appears limited. One UK study (Singh, 1975) has found only small differences in profitability and other measures of financial performance between companies that became takeover victims and those that did not, casting some doubt on the notion that relatively poor performance leads to a sharp increase in the threat of a takeover.

The other side of the argument concerns the post-takeover performance. If takeovers are motivated by the gains that can be realised from improving the performance of a company, it is to be expected that the financial performance of the acquiring company will itself improve following the takeover. Here, also, the empirical studies are

inconclusive.

Most post-takeover studies focus on stock movements during only a few weeks around the takeover date. The results indicate large positive returns to shareholders of target firms and small to zero returns to acquiring firms (Jarrell, Brickley & Netter, 1988). When the period covered by studies is extended to one to three years after the event,

acquiring firms are found to experience negative abnormal returns. A survey of such studies by Jensen and Ruback (1983, p. 20) concludes:

These post-outcome negative abnormal returns are unsettling because they are inconsistent with market efficiency and suggest that changes in stock price during takeovers over-estimate the future efficiency gains from merger.

Some extensive studies of the relationship between takeovers and subsequent performance over prolonged periods have been conducted utilising data from several countries. One example is an extensive study by Mueller (1980) of some 800 takeovers in seven countries which attempted to test the empirical relationship

between performance and takeover. The results were inconclusive. Mueller found that takeovers lead to only modest changes in profitability, both up and down in the post-merger period, and that if efficiency gains were generated, they were only small. A recent study by Franks, Harris and Mayer (1988) of some 2500 acquisitions in

Britain and the US between 1955 and 1985 finds that after two years the shares of cash bidders have appreciated as fast as they did before the merger, while the shares of equity bidders do substantially worse than before the bid. Thus, the stockmarket’s initial reaction to a bid was often belied by performance in the longer term, the

results being ‘... inconsistent with theories of efficient capital markets’ (p. 254).

A recent study by McDougall and Round (1986) of the causes and consequences of takeovers in Australia used a very similar approach to that of Mueller (1980) and made similar conclusions. In contrast, Bishop, Dodd and Officer (1987, p. 80) in their review of takeovers in Australia found:

Large increases in shareholders’ wealth are generally associated with takeovers. This evidence is consistent with the view that takeovers on average lead to more profitable uses of company assets, and as such they play a vital role in the capital allocation process.

One other approach to testing the effect of takeovers on the subsequent performance of firms is that employed by Ravenscraft and Scherer (1987a, 1987b). This study investigated the internal financial health of companies at the time they became takeover targets to ascertain how their profit performance changed over a substantial period of time following the takeover. They found that the acquired companies had

slightly inferior profit performance before takeover, but this deteriorated over a prolonged period after takeover:

An important reason for their deteriorated post-takeover returns was the write-up of asset values stemming from the payment of acquisition premiums. But those premiums were supposedly paid in anticipation of enhanced profitability, which is not evident in our post-takeover

operating income and cash flow regressions. This is an anomaly for the theory of takeovers as an efficiency-increasing mechanism. (Ravenscraft & Scherer 1987a, pp. 154-5)

33

On balance, the empirical research provides no firm conclusions on the relationship between takeovers and efficiency. With the important exception of the above study, which utilises an innovative methodology, many studies demonstrate substantial gains in the returns to shareholders of acquired companies. An unresolved debate continues over the significance of these gains. It revolves around the question as to whether the gains represent an increase in wealth and economic efficiency or a re­ distribution of wealth.3 In contrast, most commentators concede that the empirical evidence on post-takeover performance is inconsistent with expected efficiency­ enhancing qualities of the takeover mechanism.

It can be argued that the inconclusive nature of the studies reflects inadequacies in the empirics. For example, they do not adequately separate the effects of takeovers from changes in political, social and economic environments. Nor do they take proper account of takeovers which result from factors other than the perceived failure of management teams to profit maximise, ie. economies of scale or scope, tax advantages, decreased risk through diversification and greater prestige, power or convenience for managers.

The effectiveness of the takeover mechanism in establishing incentives for good performance is crucial in the analysis of public versus private ownership. Together, the theoretical and empirical evidence suggest that in reality the link between the threat of takeover and efficiency may be weakened by market imperfections and characteristics. That is not to say that takeover threats have no role to play in promoting efficiency. Rather, it suggests that while the threat of takeover generates strong incentives to profit maximise, its effectiveness in this regard is influenced by

the accuracy of share market valuations and the dispersion of shareholdings, as well as other factors including the extent of shareholder protection afforded by the regulatory and legal frameworks, and the constraints imposed by competition law and the relevant fiscal system.

Insolvency

Insolvency is another mechanism which removes the control of a private firm from management and can therefore be considered to act in a similar way to takeovers in promoting efficiency. The threat of insolvency serves as the ultimate market

discipline on the performance of managers on behalf of shareholders. This threat may at times be circumvented by government intervention, particularly in the case of large financial companies. Nevertheless, the threat of insolvency does provide a strong incentive to profit rather than losses. Such a mechanism is absent in the case of the public firm, where bankruptcy is virtually impossible because losses are underwritten by governments.

3 Examples of the polar positions are Jensen (1984, 1988) and Drucker (1986). See also Shleifer and Summers (1988).

34

REFERENCES

Alchian, A.A. 1965, ‘Some Economics of Property Rights’, II Politico, vol. 30, pp. 816­ 829.

Atkinson, S.E. & Halvorsen, E. 1986, ‘The Relative Efficiency of Public and Private Firms in a Regulated Environment: The Case of US Electricity Untilities’, Journal o f Public Economics, vol. 28, no. 3, April, pp. 281-94.

Bishop, D., Dodd, P. & Officer, R.R. 1987, Australian Takeovers: The Evidence 1972­ 1985, The Centre for Independent Studies, Policy Monograph 12, Sydney.

Borcherding, T.E., Pommerehne, W. & Schneider, F. 1982, ‘Comparing the Efficiency of Private Production: The Evidence From Five Countries’, Zeitschrift fur Nationalokonomie, Supplementum Z, pp. 127-56.

Bus and Coach Association, NSW 1985, The $90m Case for Private Buses, Bus and Coach Association, Sydney.

Caves, D.W. & Christensen, L.R. 1980, ‘The Relative Efficiency of Public and Private Firms in a Competitive Environment: The Case of Canadian Railroads’, Journal of Political Economy, vol. 88, no. 5, pp. 958-76.

Clark, P. & Porter, M.G. 1982, ‘State Enterprise Accountability: A Contradiction in Terms?’ Paper presented at conference on ‘The Dilemma of Accountability in a Quango State’, Monash Uni., Melbourne, July.

Coughlin, P. 1987, ‘The Commonwealth Banking Corporation: A Case for Privatisation’, in Privatisation: An Australian Perspective, ed. P. Abelson, Australian Profession Publications, Sydney, pp. 204-25.

Davies, D.G. 1971, ‘The Efficiency of Public versus Private Firms: The Case of Australia’s Two Airlines’, Journal of Law and Economics, vol. 14, no. 1, pp. 149-65.

____ 1977, ‘Property Rights and Economic Efficiency: The Australian Airlines Revisited’, Journal o f Law and Economics, vol. 20, no. 1, p. 223-26.

____ 1981, ‘Property Rights and Economic Behaviour in Private and

Government Enterprise: The Case of Australia’s Banking System’, Research in Law and Economics, vol. 3, pp. 111-42.

Domberger, S. & Piggot, J. 1986, ‘Privatisation Policies and Public Enterprises: A Survey’, Economic Record, vol. 62, no. 173, June, pp. 145-62.

Drucker, P. 1986, ‘Corporate Takeovers: What is to be Done?’, The Public Interest, vol. 82, Winter, pp. 3-24.

35

Finsinger, J. 1984, ‘The Performance of Public Enterprises in Insurance Markets’, in The Performance o f Public Enterprises, eds. M. Marchand et al, Elsevier, Amsterdam, pp. 223-41.

Forsund, F.R., Lowel, C. & Schmidt, P. 1981, ‘A Survey of Frontier Production Functions and their Relationship to Efficiency Measurement’, Journal of Econometrics, vol. 13, no. 1, pp. 5-25.

Forsyth, P.J. & Hocking, R.D. 1980, ‘Property Rights and Efficiency in a Regulated Environment: The Case of Australian Airlines’, Economic Record, vol. 56, no. 153, pp. 182-5.

Franks, J.R., Harris, R.S. & Mayer, S. 1988, ‘Means of Payment in Takeover: Results for the United Kingdom and United States’, in Corporate Takeovers: Causes and Consequences, ed A. Auerbach, University of Chicago Press, pp. 221-58.

Grossman, S. & Hart, A.D. 1980, ‘Takeover Bids, the Free-Rider Problem and the Theory of the Corporation’, Bell Journal o f Economics, vol. 11, no. 1, pp. 42-64.

Hensher, D.A. 1987, Productivity in Privately Owned and Operated Local Bus Firms in Australia, Working Paper no. 1, vol. 10, Research Grant, Shcool of Economic and Financial Studies, Macquarie Uni., NSW.

Hocking, R.D. 1977, ‘The Economic Efficiency of the Australian Airlines’, in Domestic Air Transport Review, vol. 2, Department of Transport, AGPS, Canberra.

Jarrell, G.A., Brickley, J.A. & Netter, J.M. 1988, ‘The Market for Corporate Control: The Empirical Evidence since 1980’, Journal of Economic Perspectives, vol. 2, no. 1, pp. 49-68.

Jensen, M.C. 1984, ‘Takeovers: Folkforce and Science’, Harvard Business Review, no. 6,November-December, pp. 109-21.

____ 1988, ‘Takeovers: Their Causes and Consequences’, Journal o f Economic Perspectives, vol. 2, no. 1, pp. 21-48.

Jensen, M.C., & Ruback, R.S. 1983, ‘The Market for Corporate Control: The Scientific Evidience’, Journal o f Financial Economics, vol. 11, pp. 5-50.

Kay, J.A. & Thompson, D.J. 1986, ‘Privatisation: A Policy in Search of a Rationale’. Economic Journal, vol. 96, no. 381, March, pp. 18-32.

36

Kirby, M.G. 1984, Airline Economies of ‘Scale’ and Australian Domestic Transport Policy, in Discussion Paper no. 112, Centre for Economic Policy Research, ANU, November.

Kirby, M.G. & Albon, R.P. 1985, ‘Property Rights, Regulation and Efficiency: A Further Comment on Australia’s Two Airline Policy’, Economic Record, vol. 61, no. 173, pp. 535-9.

Mackay, K.R. 1979, ‘A Comparison of the Relative Efficiency of Australian Domestic Airline and Foreigh Airlines’, in Domestic Air Transport Review, vol. 2, Department of Transport, AGPS, Canberra.

Marchand, M., Pestieau, P. & Tulkens, H. 1984, ‘The Performance of Public Enterprises: Normative Postitive and Empirical Issues’, in The Performance of Public Enterprises, eds. Marchand, Pestieau and Tulkens, Elsevier, Amsterdam, pp. 3-41.

McDougall, F.M. & Round, R. 1986, The Determinants and Effects o f Corporate Takeovers in Australia 1970-81. A study commissioned by the National Companies and Securities Commission and the Australian Institute of Management.

Millward, R. 1982, ‘The Comparative Performance of Public and Private Ownership’, in The Mixed Economy, ed. E. Roll, pp. 58-93.

Millward, R.W. & Parker, D.M. 1983, ‘Public and Private Enterprise: Comparative Behaviour and Relative Efficiency’, in Public Sector Economics, eds. R.W. Millward et al., Longman, London, pp. 199-274.

Mueller, Q.D. 1980, The Determinants and Effects o f Mergers: An International Comparison, Cambridge.

Niskanen, W.A. 1971, Bureacracy and Representative Government, Aldine-Atherton, Chicago.

Parris, H., Pestieau, P & Saynor, P. 1987, Public Enterprises in Western Europe, Croom Helm, London.

Pryke, R. 1982, ‘The Comparative Performance of Public and Private Enterprise’, Fiscal Studies, vol. 3, no. 2, pp. 68-81.

Ravenscraft, D. & Scherer, F.M. 1987a, ‘Life After Takeover’, Journal of Industrial Economics, vol. 36, no. 2, December, pp. 147-56.

1987b, Mergers, Sell-Offs and Economic Efficiency, Brookings, Washington.

Robinson, J. & Eatwell, J. 1973, An Introduction to Modem Economics, McGraw-Hill, London.

Scherer, F.M. 1988, ’Corporate Takeovers: The Efficiency Argument’, Journal of Economic Prospects, vol. 2, no. 1, pp. 69-82.

Shleifer, A. & Summers, L.H 1988, ‘Breach of Trust in Hostile Takeovers’, in Corporate Takeovers: Causes and Consequences, ed. A. Auerbach, University of Chicago Press, pp. 33-56.

Shleifer, A. & Vishney, R.W. 1986, ‘Large Shareholders and Corporate Control’, Journal o f Political Economics, vol. 94, no. 3, pt. 1, pp. 461-88.

Singh, A. 1975, ‘Takeovers, Economic Natural Selection and the Theory of the Firm’, Economic Journal, vol. 85, no. 339, pp. 497-515.

Summers, L.H. 1986, ‘Does the Stock Market Rationally Reflect Fundamental V a lu e sJournal o f Finance, vol. 41, pp. 591-601.

Vickers, J. & Yarrow, G. 1988, Privatisation: An Economic Analysis, The MIT Press, Cambridge US.

Yarrow, G. 1986, ‘Privatisation in Theory and Practice’, Economic Policy, vol. 2, pp. 321-77.

38

APPENDIX I : COMMUNITY SERVICE OBLIGATIONS

Many public enterprises are required by governments to undertake activities which they would not perform if they were governed by strictly commercial considerations. These activities are usually directed to governments’ social or developmental objectives. In many cases, they are a legacy from the past.

Community service obligations often involve the use o f cross-subsidies and regulatory barriers to entry which impact adversely on the economic performance o f the enterprises and users. They are often ill-defined and their costs are not specifically identified. As a result, they diminish public

accountability o f governments and enterprises.

Governments need to re-assess the continuing relevance and costs of community service obligations. If, after reconsideration, governments choose to continue to maintain non-commercial objectives, public accountability for these arrangements could be improved by their separate identification and costing. The adverse economic impact of community service obligations could be reduced by greater use of alternative delivery mechanisms which do not require regulatory barriers to entry. Options

which could be considered by governments include direct cash payments, explicit funding of the costs o f the obligations from government budgets, a levy on users, greater use o f voucher systems and contracting private as well

as public firms to deliver the services to be provided.

II Introduction

Community service obligations (CSOs) is a term which is being used increasingly to refer to the non-commercial policy objectives which public enterprises are directed to pursue by governments. These objectives generally relate to the provision and pricing of services to particular groups of consumers (for example, the disadvantaged and

those living outside metropolitan areas).

The delivery of CSOs can involve cross-subsidies1 supported by regulatory barriers which confer monopoly power on the public enterprise providing them. Consequently, CSOs can have a significant impact on the pricing practices and

1 There is no general agreement as to the most appropriate definition of a cross-subsidy. Discussion of cross-subsidies commonly employs the notion as the excess charges paid by some consumers of a given product in order to subsidise other consumers of the same product. Explicit or implicit cross­ subsidies can arise either as a result of different prices being paid for a product by different

consumers or because a uniform price is paid for a product regardless of different costs of delivery to different consumers. In a technical sense, this is not strictly adequate as it tends to merge the notions of price discrimination and cross-subsidisation. Attachment II to this appendix discusses a number of benchmarks commonly used to identify cross-subsidies.

39

operations of public enterprises, as well as wider economic effects on users, suppliers and taxpayers. The operations of the public enterprises providing CSOs may be adversely affected and made less efficient, while the overall efficiency of the economy may be impaired in its use of production resources and in its consumption decisions.

Many of the CSOs of public enterprises have been in place for decades. Their rationale has not always been made clear; in general it seems to be based on a mixture of welfare, regional development, industry assistance and national unity objectives. The question arises as to whether the original objectives remain legitimate today. It is also necessary to review the relative costs and benefits of CSOs to determine whether the associated costs imposed on public enterprises and the economy generally can be justified in today’s economic and social environment by the benefits derived from them.

A further question is whether the CSOs of public enterprises remain the most appropriate way of pursuing these types of objectives. If CSOs could be delivered without the need for cross-subsidies, then the web of regulation imposed on public enterprises to support provision in this fashion could be disentangled. Removal of statutory monopolies could result in improvement in public enterprise efficiency.

An important issue is whether the objectives underlying CSOs and efficiency of resource use are basically incompatible. Do other measures exist which could achieve social objectives at less cost than when government charges are used as the instrument of redistribution? Or is the provision of CSOs by public enterprises the most efficient way to pursue these objectives? This appendix attempts to answer such questions.

The main features of CSOs are discussed in section 12, while section 13 identifies the CSOs of major public enterprises in Australia. Section 14 then considers the economic impact of CSOs, and the problems of measuring associated cross-subsidies. Section 15 examines alternative mechanisms which could be employed to satisfy the

social objectives underlying CSOs.

12 Main features of CSOs

What is a CSO?

The term CSO seems to be used interchangeably with terms such as public policy obligations, social objectives, social goals and social commitments. Consequently, there is ambiguity as to what is included as a CSO. This was reflected in several

submissions to the inquiry.

Whatever the terminology, it appears that CSOs can be classified into two broad categories. First, there are cases where governments consider it desirable that certain services should be supplied to final consumers or industry at a uniform or an ‘affordable’ price irrespective of the cost of provision. Examples are the provision of electricity, transport, telephone and postal services to isolated regions. Second, public enterprises implement welfare and redistribution policies by granting price concessions to consumers who are deemed to be in need of some form of social support.

Both categories of services are supplied by public enterprises at the explicit or implicit direction of government. They are services which the enterprise would not provide to the community at that price if strictly commercial criteria applied.

According to Whiteman (1988, p. 1), Telecom has defined CSOs as follows:

A Community Service Obligation (CSO) arises when a government requires a business enterprise to carry out activities which they would not elect to provide on a commercial basis, or which could only be provided commercially at higher prices.

The Commission has adopted this interpretation in its approach to the inquiry. It focuses on the price and/or quantity of the good or service provided.

An alternative approach would be to also include directions from government, explicit or implicit, on the use of inputs by public enterprises. For example, Major Mail Users of Australia regard the funding of the Commonwealth Superannuation scheme and implementing government policies such as equal opportunity and Australian purchasing preferences as CSOs. Other submissions, however, excluded such

directions from their consideration of CSOs. Australian National (AN) distinguished CSOs, such as the provision of loss-making passenger services on the mainland and the operation of Tasrail, from other public policy obligations; the latter include excess superannuation contributions, long service leave, workers compensation, and participation in the Commonwealth Government’s purchasing preference policy.

There is widespread government or ministerial intervention in the commercial operations of public enterprises, particularly as they relate to the use of inputs. Matters often subject to intervention include pricing decisions, contractual obligations, borrowing terms and conditions, personnel and administration policies. While not generally regarded as CSOs, the interventions often have the objectives of providing services to parts of the community on a non-commercial basis.

Such practices raise similar issues as CSOs. They may have similar re-distributive and efficiency effects. For example, a broad policy directive to some public enterprises (including Australia Post and Telecom) is that their borrowings are restricted and they are thereby forced to fund a significant proportion of capital requirements

internally. This requirement may have an important impact on the distribution of costs between current and future users. Current customers are required to provide the funds for new capital equipment, instead of the costs being spread over the life of the equipment, as would occur if the funds could be borrowed.

Implemention of CSOs

There is a variety of ways to finance the CSOs of public enterprises. They include cross-subsidies, explicit government reimbursement of the losses incurred by the enterprise in meeting its obligations and acceptance of lower rates of return on capital. The methods are not mutually exclusive, as enterprises may engage in cross­

subsidisation, for example, and still make losses which are then either funded from the government budget or lower rates of return on capital accepted.

41

Governments may fund the CSO directly as in the case of AN, Telecom and several State government enterprises which provide concessions to pensioners. Some enterprises are required to separate the costs and revenue of CSO operations from their commercial business. The losses incurred on the CSOs may then be funded by revenue supplements from the relevant government. In some cases, there are upper limits on the amount of these supplements.

Cross-subsidisation involves financing the losses incurred by CSOs through income earned on non-CSO activities. Some cross-subsidies involve charging different prices for identical products. Many involve uniform pricing for services with differing costs of production. For many services the cost of supplying various users varies according to location and time of day of use. Cross-subsidisation between geographical areas is most prevalent in Australia, with non-metropolitan users being the major beneficiaries.

On the other hand, the presence of either uniform pricing or different prices to different consumers of the same product is not necessarily indicative of the existence of cross-subsidies and CSOs. As in the case of private enterprise, a public firm may adopt uniform pricing as a sensible commercial decision to avoid the expense of accurately pricing every unit of the good or service supplied. Discriminatory pricing, promotional pricing and pricing to capture network externalities, all apparently involving cross-subsidies, may be adopted by public enterprises as part of an efficient pricing strategy. Where enterprises adopt such pricing policies it is difficult to ascertain whether cross-subsidies, in the strict sense of prices not covering incremental costs, are present. Discriminatory pricing may not be directed at increasing the returns of the enterprise. In this case, a lower rate of return on capital is a measure of the extent of the cross-subsidy.

In other cases, the presence of cross-subsidies may indicate a policy of industry assistance rather than a CSO, although it is not always easy to distinguish between the two. For example, industry assistance considerations may be an important factor in the provision of essential services to rural areas at less than the economic cost of supply.

The implementation of CSOs is generally supported by regulatory instruments. Regulatory barriers to entry are usually required to protect public enterprises which are directed to cross-subsidise unprofitable operations or to accept a lower rate of return on their operations. If barriers to entry did not exist private firms would have an incentive to enter the market and supply the profitable services at a lower price, leaving the unprofitable services to the public enterprise (a practice known as ‘cream skimming’). This could eliminate the surplus available to the public enterprise for funding the losses on CSO services.

The redistributive and efficiency effects of implementing CSOs, and the measurement difficulties, are discussed further in section 14.

42

Non-transparency

CSOs are generally characterised by a lack of transparency, particularly those implemented by cross-subsidies. First, some CSOs are not well defined. As noted above, CSOs are often implicit, as in cases where vague legislative requirements must be interpreted by enterprise management or where the operations of the enterprises

are subject to detailed intervention by government ministers.

Second, the costs of meeting them often remain hidden. The favoured group can be provided with benefits without those meeting the costs being aware of the extra burden they are incurring. Moreover, where CSOs are funded by cross-subsidies there are severe difficulties in quantifying the extent of the transfers involved. This is

discussed further in section 14.

13 The CSOs of major public enterprises in Australia

It is difficult to list the CSOs of all the major Commonwealth and State public enterprises. The aim of this section is to highlight some of the more typical forms which CSOs take at the Commonwealth and State level and identify how they are funded.

Commonwealth public enterprises

Telecom

Under the Australian Telecommunications Corporation Act 1989 Telecom’s statutory CSO is the supply of a standard telephone service as efficiently and economically as practicable. This service must be reasonably accessible to all people and on an equitable basis wherever they reside or carry out business within Australia. Also,

there is a requirement that service performance standards reasonably meet the social, industrial and commercial needs of the Australian community.

Telecom interpreted similar directions in the previous legislation as requiring it to provide a universal service at an affordable price. This approach was confirmed in a Government announcement in May 1988 of a new framework for telecommunications services. The Minister for Transport and Communications (1988a, p. 2) stated that

one of the objectives of government telecommunication policy is:

... to ensure universal access to standard telephone services throughout Australia on an equitable basis and at affordable prices, in recognition of the social importance of these services.

Telecom has made telephone services widely available, both geographically and across income groups. Internal cross-subsidies have been used to sustain the CSO. According to the Minister for Transport and Communications (1988b, p. 38) telephone call charges - notably trunk call or STD charges - have been set at levels in

excess of the associated costs. This has allowed the fixed charges for providing basic access (connection and rental fees) to be maintained at levels below the direct cost of providing the dedicated exchange line, household wiring and telephone, so that the customer base could be expanded. It is difficult to gauge the nature and extent of

these cross-subsidies. According to Telecom, the main area of cross-subsidy is from

43

STD calls to non-metropolitan access rental and local call services (Prices Surveillance Authority 1984, p. 45). For example, the Prime Minister (1986) has noted that the cost of telephone connection in rural areas in 1986 ranged up to $25 000, but country subscribers pay little more than city subscribers for connection. Rural connection charges are not based on the cost of providing the subscriber lines but are related to ‘standard’ rental charges.

Whatever the nature and extent of existing cross-subsidies, the quality of service varies between metropolitan and non-metropolitan urban and rural areas. The main difference between metropolitan and non-metropolitan areas is the relative accessibility to other subscribers at local call rates. Metropolitan subscribers have access to a significant proportion of all network subscribers at a standard local call charge. Non-metropolitan subscribers have far more limited access at the standard call local rate. The extent of accessibility to other subscribers decreases significantly for subscribers in rural areas as subscriber density per district decreases. Other differences in the quality of service relate to fault occurrences and fault correction times.

Under new legislation introduced in 1989 the costs of meeting Telecom’s CSOs will be separately identified and will be taken into account in fixing the financial target Telecom is expected to meet. Austel has the function of monitoring and reporting on Telecom’s delivery of its CSOs.

Other CSOs provided by Telecom, although not imposed on it by legislative charter, include the provision of public telephones, free calls for ambulance, police and fire brigade emergency services, and the provision of discounts to certain disadvantaged groups as part of the Government’s welfare program. Pensioners, supporting parent beneficiaries and recipients of some other benefits are eligible for a reduction in the basic annual telephone rental. This concession is funded directly from the Commonwealth Budget.

Telecom is able to cross-subsidise without fear of loss of markets from competition because it is protected by government-imposed monopoly provisions relating to the provision of basic network services. The Telecommunications Act 1989 states that Telecom ‘has the exclusive right to install, maintain and operate telecommunications services between places within Australia’. One important reason for reserving the service solely to Telecom is to enable it to fulfil its CSOs, as the Minister for

Telecommunications and Aviation Support (Australia, House of Representatives 1989, p. 1623) stated in the second reading speech on the new legislation:

The other reason is the important need for universal service, which is a particular community service obligation placed on Telecom Australia. The Government expects Telecom to provide equitable access to standard telephone services as efficiently and economically as practicable. It is therefore necessary to allow for Telecom to average its prices and practise some cross-subsidy.

Australia Post

Australia Post is required under the new Australian Postal Corporation Act 1989 to provide a letter service available at a single uniform rate of postage for carriage within Australia, by ordinary post, of letters that are standard postal articles. Also, there is a requirement that service performance standards must reasonably meet the

social, industrial and commercial needs of the Australian community. This statutory obligation was not explicitly identified in the previous legislation governing the operations of Australia Post but had been publicly affirmed by successive governments.

This CSO is funded by cross-subsidies. The large disparity in costs of delivering letters originating and terminating within the same capital city and those sent between small and distant country areas is not reflected in different prices. This uniform pricing policy thus subsidises users who send letters to or from sparsely settled areas.

As in the case of Telecom, however, the quality of this service may vary between urban and sparsely settled regions, most notably in the frequency of mail services.

Regulatory barriers to entry supporting the pricing arrangements are very explicit; Section 29 of the Australian Postal Corporation Act 1989 provides Australia Post with a monopoly in the carriage of letters in Australia with a weight of less than 500g. The carriage of letters for which Australia Post has a monopoly accounted for about 60 per cent of revenue from mail services in 1980-81 (Albon 1985, p. 11).

Pensioners, supporting parent beneficiaries and recipients of some other benefits are eligible for mail redirection concessions. This concession is funded directly from the Budget.

Australian National

In its submission to this inquiry, AN indicated that it:

... operates two main community service obligations for the Commonwealth Government namely passenger business on the mainland and Tasrail, the Tasmanian railway system.

In contrast to Telecom and Australia Post, AN’s CSOs are not primarily funded via cross-subsidies but rather by revenue supplements from the Commonwealth Government. AN noted that in 1987-88, passenger services earned $29.2 million in revenue and received $27 million in government revenue supplements. Tasrail

earned $26.1 million from its freight operations and received $16.2 million in revenue supplements. The Government sets cost recovery targets for the services it directs AN to provide. The revenue supplements paid to AN have declined over the past two years.

Commonwealth Bank

Commonwealth legislation requires the Commonwealth Bank board to direct its policy ‘... to the greatest advantage of the people of Australia and have due regard to the stability and balanced development of the Australian economy’ (Campbell Inquiry 1981, p. 462).

45

The Commonwealth Bank is involved in a number of activities which are not purely profit motivated. The major social objective of the Commonwealth Bank would appear to be the provision of banking facilities to Australians, regardless of location. A report of the H.V. Evatt Research Centre (1988, p. 136) states:

In contrast to its private sector rivals, the CBC operates in the less remunerative rural areas. If a bank is to be found in the remote outback, it is likely to be a branch of the Commonwealth Bank.

The Commowealth Bank undertakes a number of other CSOs. The Commonwealth Development Bank (CDB) and Commonwealth Savings Bank (CSB) components of the Commonwealth Bank are required by statute to fulfil certain social objectives. Small businesses unable to obtain funds on reasonable terms from other banks may apply to the CDB for a loan. The CDB places greater emphasis than other banks on the likelihood of success of the project rather than on the ability to provide collateral. The CSB has a statutory obligation to encourage savings and promote the interest of its depositors. As such, it directs much of its lending to areas of social need such as

housing, local councils and charities.

Unlike many other government-owned enterprises, the Commonwealth Bank is not protected by monopoly legislation. It has to compete in the market with other banks. The question arises as to how the bank finances its obligations in the competitive environment.

One possible explanation is that the Commonwealth Bank accepts a lower rate of return than its competitors. As the bank provides a number of CSOs in an environment where private banks do not face similar constraints on competitiveness, it could be expected that the private banks would concentrate on the more remunerative banking services, leaving the less remunerative services involving social

considerations to the Commonwealth Bank.

The available evidence provides conflicting indications of the relative performance of the major banks. One study (Coughlin 1987) concludes that the Commonwealth Bank has lagged well behind its competitors for a considerable period with respect to return

on assets, return on equity and earnings per employee. More recent data suggest that the rate of return on shareholders funds and on assets for the Commonwealth Bank in 1985-86 and 1986-87 was comparable with that of the other major banks, although return per employee is lower than the others (Australian 3 September 1988).

A further method by which the CSOs of the Commonwealth Bank could be funded, at least until 1984, was income tax revenue foregone. Until 1984, the CDB was not required to pay income taxes or dividends. The CSB paid no income tax, but paid dividends to two States and the Commonwealth. However, these exemptions were lifted in 1984, although the CSB can still claim dividends paid to the States under the State Savings Bank Amalgamation Agreements as a tax deduction.

46

State public enterprises

Electricity authorities

Electricity authorities have as their major CSO the objective of uniform pricing. Authorities group most users into classes, generally according to end-use, and establish a uniform price for each class irrespective of the distances between generating plants and supply points. Section 54(1) of the Tasmanian Hydro-Electric

Commission Act 1944-1988 states that the Commission’s tariffs and charges must:

... provide for the same general rates of charges for electricity sold or supplied by the Commission to consumers outside the city of Hobart as are charged in the cases to consumers within that city.

The Tasmanian Department of Premier and Cabinet submitted that this uniform pricing policy was seen as both a matter of social equity and as a means of encouraging supply in the more isolated areas from an integrated system.

The Queensland Electricity Act 1976-1988, Section 64(3), states that the Queensland Electricity Commission (QEC):

... shall at all times have regard to and proceed towards the objective of progressively equalising throughout the State the prices to be paid by the consumers to whom a particular tariff applies ...

The State Energy Commission of Western Australia (SECWA) operates a state-wide system of uniform tariffs to ensure that customers in country areas generally pay the same rates for electricity as customers in the Perth metropolitan area. Hence, the charge for country users does not reflect the very considerable regional variations in costs.

Barriers to entry are again an important legislative provision to enable the electricity enterprise to cross-subsidise. In all States, legislation provides for government electricity authorities to control generation of electricity. Private generation for sale is only permitted with the approval of the relevant State authority. For example, in

Tasmania there is a prohibition (with some exceptions) on the establishment of new electrical stations, except with the consent of the Hydro-Electric Commission. The relevant legislation states:

Except as provided by or under the authority of the Hydro-Electricity Commission Act 1929, or some other Act not being this Act, it shall not be lawful for any person, after the commencement of this Act, to establish or to add to, alter, or extend any plant, machinery, or works for the generation of electrical energy, or to erect or extend any electric line for the transmission electrical energy, except energy purchased from the Commission, without the consent of the Commission. (Section 48(1))

47

In Queensland, the QEC:

... may direct, prohibit, restrict or control, or regulate in any other manner whatsoever, the supply and consumption of electricity. (Section 36(2)).

Private electricity generation in Victoria requires the permission of the State Electricity Commission of Victoria. In Western Australia no person can supply power to the public without permission from the State Electricity Commission of Western Australia.

State electricity authorities also provide a range of concessions and undertake additional expenditure in supplying electricity to specific groups of consumers. Most States provide for rebates on electricity accounts for pensioners. For example, the South Australian Government operates an electricity concession scheme for pensioners whereby it reimburses the Electricity Trust of South Australia for the concessions to pensioners. The Trust has also committed expenditure to ensure regular supplies of energy to remote Aboriginal communities. The Power and Water Authority in the Northern Territory similarly undertakes additional expenditure to supply power to remote Aboriginal communities.

The NSW Government outlined in its submission a number of CSOs implemented through the electricity authorities. It cited the Traffic Route Lighting Subsidy (TRLSS), Remote Area Power Supply Assistance Scheme (RAPAS), Baseline and CARES assistance for low income customers, energy labelling, off-peak hot water purchase assistance (OPPAS), and Pensioner Rebates and a Research and

Development Program (SERDF). These are partly funded from a special fund to which certain electricity councils and ECNSW contribute, and partly by councils directly. Not all these measures come within the bounds of the definition of CSOs being used in the inquiry. Table II lists the allocation of these special funds in 1988­ 89.

Table I I : Allocation of New South Wales special funds, 1988-89 ($ million)

Special funds (through D O E) E stim ated paym ents

by councils

E stim ated total

Pensioner rebates 2.4 25.0 27.4

Baseline 5.0 - 5.0

SERDF 2.5 - 2.5

TRLSS 3.6 4.8 8.4

RAPAS 3.15 - 3.15

CARES .50 0.5 1.0

Labelling .25 - 0.25

OPPAS .50 0.3 0.8

Total 17.9 30.6 48.5

Source: NSW Government submission, September 1988, p. 105.

In addition, transfers to rural electricity councils from a separate fund amounted to $22.5m in 1988-89, the objective of the transfers being to reduce the extent to which the smaller councils have to charge higher prices to recover higher average costs of distribution.

Water supply authorities

A common feature of domestic water supply was that, until recently, there was little attempt to relate charges to the volume of water consumed. Generally, a fixed charge based on the value of the property secured the right to use a certain quantity of water without additional payment. In an increasing number of cases, water consumption

above a minimum block is charged per unit of excess. Recently, many authorities have reduced the allowable quantity above which additional charges are made.

In some States, certain properties receive partial or total exemption from water service charges. For example, Sydney hospitals, schools, railway workshops and other properties are exempt or receive concessions for water and sewerage services. The Sydney-Illawarra-Blue Mountains Water Board is now conducting a full review of the

categories of exempt properties in order to quantify all implicit and explicit welfare payments currently being met by the Board. According to the submission of the NSW Government, it is some years since such a review was conducted.

Most States provide rebates to pensioners on water and sewerage rates, many of which are funded directly from government budgets.

Rural areas are also major beneficiaries of the CSOs of State water authorities. For example, the operations of the Victorian State Rivers and Water Supply Commission (SRWSC) have been structured to supply significant benefits to rural consumers. The SRWSC (1984, p. 6) lists as one of its major objectives:

... pricing to recover full costs and earn a return on assets except where direct subsidisation is required to meet other government objectives.

According to the SRWSC, in 1983-84 an overall loss of $673 000 was incurred in the supply to irrigation districts. In most districts revenue exceeds costs by a small amount, with the overall loss being mainly due to a $1.5 million loss in supplying the Goulburn Murray Irrigation District. In the country urban water systems in 1983-84,

26 of the 57 operated at a loss as rates and charges did not fully cover the annual operation, maintenance and capital servicing charges. The average cost of supply of water to these districts in 1983-84 was 24 cents per kilolitre, while the average charge was 18 cents per kilolitre.

Similar arrangements exist in other States.

49

Rail authorities

The non-commercial policy objectives which railway authorities are required to satisfy include:

. the operation of unprofitable lines;

. common carrier requirements involving less-than-carload freight (LCL);

. concessional rates and fares; and

. maintaining bases in certain rural towns, particularly maintenance and workshop facilities.

The various rail authorities differ in the extent to which they are required to meet CSOs and in the manner in which they are funded for performing such activities. A number of the railway authorities are separately funded for the costs involved. For example, the Commonwealth reimburses AN for loss of revenue resulting from maintaining passenger services and from carriage of passengers or freight at concessional rates. Some authorities have the discharge of CSOs taken into account for dividend calculation and performance monitoring purposes, while others receive no compensation from governments. Where no compensation is received, authorities may endeavour to recoup additional revenue from other users (that is, to engage in cross-subsidisation).

Table 12 outlines the non-commercial objectives which are established in the State and Commonwealth legislation under which the railway authorities operate.

Some non-commercial objectives of railway authorities, for example the operation of certain branch lines, are underpinned by regulations which restrict competition. Establishment of privately-owned railways requires government approval. Inter- modal competition is restrained by regulations in most States which prohibit the use of road transport for the carriage of prescribed commodities (for example, certain minerals and grains in certain areas).

Shipping services

In its report on Coastal Shipping (IAC 1988) the Commission noted that governments have pursued various non-commercial objectives through their operation of public ports. Among these have been regional development and the provision of facilities for sport and recreation. Many users of coastal shipping claimed in submissions to that inquiry that non-commercial functions were being paid for by commercial shipping. Information gathered from port authorities suggested that this was not generally the case and that costs of providing non-commercial facilities are usually

recovered from beneficiaries (for example, boating registrations and licence fees) or subsidised directly by government. However, the Commission stated that the financial reports of port authorities do not always report upon the provision of such facilities and that instances of subsidisation of non-commercial functions by the shipping industry appear to exist in some ports.

50

Table 12: Government rail authority community service obligations

R ail au th ority C om m u nity service ob ligation s A nnual c o st (1987-88) F u n d in g arran gem ent

$m

Westrail Certain passenger services 9.6 Internally funded.

AN Mainland passenger services 27 Separately funded by the

Tasmanian rail system 16.2 Commonwealth Government.

Pensioner discounts na

QR Passenger services 67.5 Funded internally, though receives

LCL freight na specific compensation from the

Uneconomic branch lines 17.3 State Government for fare

Certain livestock haulage ) concessions for school children

Certain agricultural freight ) and a few other categories (but

(eg fruit and vegetables) ) na not pensioners) and freight rate

Certain other freight ) concessions in some minor

(eg machinery and motor vehicles in single consignments)

) categories.

SRA Passenger services 101.5 Passenger and freight concession

Freight concessions 128.9 programs are separately funded

(mainly coal) from State consolidated revenue.

Pensioner discounts na

V/Line Passenger services (rail) 120 Not separately funded, except for

LCL (small freight) some payments for bus services in

Bus service (regional centres) regional centres, and from some

other government departments, ie. education.

Pensioner discounts na

Source: Submissions and Annual Reports, na not available.

The Tasmanian and Western Australian Governments have submitted to this inquiry that users of recreational port facilities usually pay fees and charges to cover the costs of services provided. The Tasmanian Government stated that there are relatively few community service functions performed by ports in Tasmania which impose cost burdens on commercial users through cross-subsidisation. The Western Australian Department of Transport stated that, generally, port authorities are not required to

undertake proposals which are not of a commercial nature. If a port is directed to specifically provide non-commercial facilities, then compensation is provided by the Government. In addition, non-commercial activities of ports undertaken at government direction may be excluded from the calculation of financial targets.

In their submissions, the National Bulk Commodities Group and the Electrolytic Zinc Co of Australia drew the Commission’s attention to the existence of cross-subsidies in the pricing of navigation aid services and port charges. The Commission has not been able to establish that these cross-subsidies exist to implement CSOs; nor has it attempted to analyse the effects of the pricing of navigation aids. In general, however, the Commission recognises that not all cross-subsidies reflect the existence of CSOs. What are apparently cross-subsidies may exist for other reasons such as discriminatory pricing policies by public enterprises in charging what the market will bear. They may

also reflect industry assistance considerations or arise because of practical design features (eg administration costs) of user-pay pricing regimes.

14 Economic impact of CSOs

Income Redistribution effects

The main beneficiaries of CSOs appear to be the users of social infrastructure, particularly users in rural areas, and certain disadvantaged groups who receive a direct benefit from paying less for the service than its cost of provision to them. They may also benefit to the extent the value of the CSO is capitalised into the value of

their assets. If, for example, a telephone was connected to a rural property after the farmer had purchased the land, the value of the land would be expected to rise, to reflect the value of this amenity to potential buyers. The provision of a telephone would also be likely to increase the rental value of the property.

Other consumers and taxpayers are forced to pay the direct cost of providing CSOs. Individual consumers pay directly through prices which are higher than the cost of supplying the service; this excess revenue then covers the losses made by supplying the favoured group at below cost. Consuming industries which use the government- supplied input intensively are disadvantaged because of the increased prices they pay for these inputs. Similarly, taxpayers pay through higher taxes if the losses associated with a CSO are financed directly from the budget or are funded in other ways, such as

revenue forgone as a result of exemptions from certain taxes and charges for some enterprises. The common feature of those who fund the subsidy is that they are dispersed and the cost to each individual is not high.

The requirement that some public enterprises undertake CSOs can help the managers and workers achieve their aims (Rees 1984), particularly where governments confer monopoly powers on their business enterprises in order to implement CSOs. Where the CSOs of the public enterprise involve the provision of a universal service via uniform pricing, or more explicit rural subsidisation leads to excessive expansion into

high cost rural areas, this helps to enlarge the enterprise and to increase the managers’ status and influence. Workers are also advantaged by higher employment and better promotion opportunities in a larger organisation. Furthermore, losses incurred by the enterprise in meeting CSOs make cost padding by the workers and

managers of the enterprise less obvious because financial results reflect a mixture of commercial and non-commercial objectives.

Politicians may also benefit because political manipulation of public enterprise CSOs may enable income to be redistributed to politically significant sectors of the community. Hartley and Trengove (1983, p. 3) stated:

Political parties can gain (and hold) office by proposing and implementing inefficient policies so long as the costs are dispersed across many voters and the policy provides significant benefits to politically influential groups.

Economic efficiency effects

The direct efficiency effects of undertaking CSOs vary according to the funding mechanism adopted. As outlined in sections 12 and 13, the most common methods of funding CSOs are by cross-subsidies and by drawing on taxation revenue. Cross­ subsidies involve a combination of consumption subsidies and consumption taxes for

different users of the same product. In contrast, funding from taxation revenue involves only consumption subsidies for selected users of a product, the cost being spread over all taxpayers rather than particular users of the product. The following discussion focusses on the efficiency effects which may result from cross-subsidies.

In terms of direct consumption, those who pay the subsidy (face a price above marginal cost) will restrict their usage of the product, even though they might value additional units more than the cost of producing them. This discourages consumption and results in a welfare loss. Conversely, those who receive a subsidy (face a price below marginal cost) will be encouraged to expand their usage of the product beyond

the point where the value to them of additional units becomes less than the cost of producing them. The result is increased consumption by the favoured group.

Resource use can be influenced by the impact of the price effects of the cross­ subsidies on services which are inputs into other industries. Those industries which use the subsidised service more intensively are able to improve their competitiveness, but at the expense of others. Those which use the higher priced service more

intensively are disadvantaged in their competition for resources with other industries.

An additional influence CSOs can have on efficiency in resource use and consumption arises from the regulatory environment in which enterprises operate. As noted above, many public enterprises are protected from competition. The absence of competition permits - indeed, is necessary for them to be able to cross-subsidise and to fulfil their

CSOs. The existence of barriers to entry will give rise to a further set of efficiency effects.

The absence of competition may encourage cost-padding practices such as relatively high remuneration for employees, restrictive work practices, and over-employment. Higher costs arising from these practices will be passed on to users of the output of

the firms by higher prices. In effect, what is intended to be a subsidised price for a service provided under a CSO of a public enterprise which practices cost-padding, may in fact be no lower than the price which would be paid in a contestable market.

Restrictions on competition from potential new entrants may also inhibit adjustment to change and slow the introduction of new techniques of production and management. Potential competition provides the incentive to use the most efficient techniques in organising production. Limiting competition thus incurs higher costs and impedes economic growth.

Public enterprises may implement pricing policies apparently involving cross-subsidies but directed at ensuring efficiency.

Discriminatory pricing is one such option. In serving a variety of markets in which the demand of some consumers is more responsive than others to changes in price, public enterprises have the opportunity to charge different consumers different prices for the same product. Provided marginal costs are covered in each market, such pricing practices serve to achieve greater use of capacity while ensuring some contribution to fixed costs.

Cross-subsidies to implement CSOs could be viewed as variants of discriminatory pricing practices by public enterprises operating in conditions of natural monopoly. Indeed, insofar as CSOs require also production at levels which would not otherwise be undertaken commercially, they could be viewed as a means of achieving more socially efficient levels of output.

A particular difficulty with this interpretation of the effects of CSOs and the associated cross-subsidies is the apparently arbitrary nature of pricing and production decisions. They appear to bear little relation to efficiency considerations, more often

being derived from political and social considerations, and there is no reason to expect that they lead to more efficient outcomes than discriminatory pricing based on the likely sensitivity of users to price changes.

Moreover, the discriminatory pricing characteristic of CSOs often involves the provision of some goods or services at less than marginal cost. Indeed, this feature distinguishes cross-subsidies, which lead to inefficiencies in production and consumption, from the rational discriminatory pricing practices of public firms which serve to enhance efficiency. However, the distinction is difficult to maintain if discriminatory pricing, while not actually involving the provision of some products at less than their cost, is used to fund CSOs through a lower rate of return on funds

rather than to maximise those returns.

Other pricing practices of public enterprises may also serve to promote efficiency and yet appear to involve the use of cross-subsidies. Pricing policy may be used to spread demand more evenly between peak and off-peak periods by charging different prices for a product according to the period in which it is used. Another pricing policy which

apparently involves cross-subsidies between users is that adopted in some industries such as telecommunications and railways, where the benefit a consumer derives from a good depends on the number of other consumers in the network. Firms may reduce prices to potential new users in order to extend the network.

54

Whatever the overall impact of CSOs on efficiency in resource use, the question remains as to whether the costs of their provision can be reduced by substituting alternative mechanisms to pursue the objectives of government. Certainly, the restrictions of consumption choice in the case of welfare services provided by CSOs,

the effects of the associated regulatory barriers and the cost padding they may encourage, together with the lack of transparency and measurement difficulties, lead to a consideration of alternative mechanisms. Some potential alternatives are examined in Section 15.

Measurement of cross-subsidies

Measurement problems

Several problems arise in attempting to measure the extent of cross-subsidies which finance many of the CSOs of public enterprises. The most basic question concerns what is to be measured. There are several possible benchmarks. These are outlined in more detail in Attachment II to this appendix. Some benchmarks emphasise the

efficiency effects of cross-subsidies, whereas others focus on the distributional and equity consequences. Concepts of what constitutes a cross-subsidy from an economic efficiency viewpoint may be at variance with attitudes in the community which focus

more on perceptions of how costs should be shared. A range of possible price schedules may be considered ‘subsidy free’, despite having widely diverse equity outcomes.

Measurement of cross-subsidies and their effects requires enterprises to have accounting systems which will produce the necessary information. Such assessments have not been the general practice. A recent government review of

telecommunications policy found that existing Telecom accounting arrangements did not enable a reliable measure of the costs or levels of cross-subsidies (Minister for Transport and Communications 1988b, p. 40).

One reason for this is that public enterprises have lacked sufficient incentive to generate the information for the identification of cross-subsidisation. Further, some governments may wish to avoid the political consequences which could flow from making transfers between sectors more explicit.

Without detailed information, it is difficult to disentangle cross-subsidies from the discriminatory pricing practices of public enterprises. As Whiteman (1988, p. 37) has noted:

One problem in quantifying the extent of the cross-subsidy is the difficulty in separating the cross-subsidy of say non-metropolitan access rental and non-metropolitan local call services by trunk calls from effects of discriminatory pricing designed to maximise profits.

Where there is little pressure to minimise costs, for example, because of insulation from market forces, cost padding can make it difficult to obtain a reliable indication of true cross-subsidies. It is possible for a product to cover what would be its costs under efficient operation, and yet appear to be subsidised because costs are not

minimised. In those circumstances, the subsidy from other products would more correctly be said to be flowing to the firm rather than to the product in question, the

55

subsidy being dissipated among factors of production in the form of higher costs.

Estimating how cross-subsidies between products translate into transfers between user groups encounters significantly more informational problems, namely, knowledge of the demand patterns of users. These informational requirements severely limit the practical usefulness of the concept, but without such information it is difficult to determine whether cross-subsidies are justified as an instrument of social policy.

Attempts at quantification

Because of these measurement problems, most estimates of cross-subsidies are based on definitions which are conceptually different from the strict technical definition of a cross-subsidy. In practice, attempts at measurement involve notions closer to that of price discrimination than cross-subsidisation.

In Australia, the major attempts to estimate cross-subsidies underlying CSOs relate to Telecom. When estimating cross-subsidies, Telecom uses a contribution margin approach. This approach estimates the excess of earnings over direct expenses, where direct expenses include operating expenses, such as installation, maintenance and

administrative expenses, and depreciation. These contribution margins are then expressed as ratios of direct expenses. Telecom’s estimates of these contribution margins for 1984-85 are given in Table 13. Termed proportional contribution margins, these ratios represent the percentage return made on direct expenses incurred.

Table 13 : Telecom’s estimated cross-subsidisation, 1984-85 ($ millions)

Earnings

Direct expenses (including depreciation)

Contribution margin

Access rental - metro 551 517 34

- non-metro 334 574 -240

Total 885 1091 -206

Local calls - metro 678 386 292

- non-metro 295 286 9

Total 973 672 301

Trunk calls 1469 437 1032

Source: Prices Surveillance Authority (1984, p. 45).

The cross-subsidy benchmark adopted was whatever proportional contribution margins occur in the metropolitan area (which, by definition, was considered to have subsidy free prices). Telecom defined the cross-subsidy to non-metropolitan users to

be the increase in earnings required for their access rentals and load calls to return

the same proportional contribution margins as their metropolitan counterparts (0.66 for access rentals, and 0.76 for local calls). The criterion was that neither group paid a smaller return than the other on the direct outlay made to service them. On this basis, non-metropolitan local calls were calculated to receive a subsidy of $207

million, notwithstanding a positive contribution margin of $9 million. The negative contribution margin of $240 million for non-metropolitan access rentals was translated into a $278 million subsidy received. Total cross-subsidies amounted to $485 million.

Observing that total indirect expenses are equivalent to approximately 60 per cent of direct expenses, Whiteman (1988, p. 38) applied this figure to estimate an indirect expenses charge to each of the activities. On this basis, only trunk calls operated at a significant profit; all other activities, with the exception of metropolitan local calls, made losses. Whiteman concluded that a cross-subsidy may be identified from trunk

calls to other activities, but that there was no accepted way of measuring its extent.

The cross-subsidy available to rural subscribers is further noted in a recent publication (BIE 1988) which reports a district profitability study of telephone services. The study, originally published in a report by the House of Representatives Standing Committee on Expenditure, estimated the annual profit or loss per telephone service. The study found that in metropolitan districts the annual profit per

service was $229, while in country districts there was an annual loss per service of $259.

An extensive study is presently being underaken by the Bureau of Transport and Communication Economics (BTCE). The study aims to measure basic telephone service CSOs on the basis of avoidable cost, ie the cost which could be avoided if a service were not provided. A complicating factor is the need to consider incoming as

well as outgoing calls when assessing the revenue generated. Data requirements include call pattern information, tariff schedules, estimated replacement and annual maintenance costs for loop connection to the district exchange (requiring data on distance of customers from exchange and subscriber density for each district), typical

exchange costs (based on average age, economic life, replacement cost and land value), type and number of local exchanges in each district, and estimates of costs of trunk network connections.

In the case of Australia Post in 1981, when the standard article rate was 22 cents, the Bradley Inquiry (1982, p. 42) commented on charges for standard postal articles in comparison with delivery costs:

Appropriate prices for standard postal articles might have been up to 7 cents higher for deliveries to country areas of mail originating from interstate, and up to 2 cents lower for deliveries in metropolitan areas of mail originating in the same state.

The report estimated that the total value of the cross-subsidy was about $24 million.

In relation to electricity supply in Western Australia, the Report of the Committee of Inquiry into Gas and Electricity Tariffs (1985) provided estimates of cross­ subsidisation between metropolitan and non-metropolitan regions arising from the policy of making charges within tariff schedules uniform throughout the State despite

57

considerable variations in costs of supply between metropolitan and non-metropolitan users. For 1983-84 the cross-subsidy was estimated to be $61.7 million.

The Committee also drew attention to other sources of cross-subsidisation. One was the granting of rebates on electricity consumption to pensioners and other eligible consumers, financed from higher charges on non-eligible consumers. Another was that, in most tariffs, charges did not reflect the way in which costs of supply varied with the demand profiles of customers. Because peak demand determines the

capacity of the system, and less efficient generating units are brought on line in order to meet peak demand loads, the cost of supplying electricity may vary considerably between peak and off-peak times. When tariffs do not vary with the time of day, peak users are subsidised by off-peak users.

15 Alternative redistributive mechanisms

The question arises as to whether there are less costly alternatives to achieve the government’s social and policy objectives than through the CSOs of public enterprises. Progressive income taxation, direct transfers or subsidies, in-kind transfers of goods and services and voucher entitlements are common instruments of income redistribution policies. Services or goods provided under the CSOs of public enterprises may be regarded as a form of in-kind transfer or in-kind subsidy in that specific commodities are provided to the targeted groups at reduced prices.

All these instruments have effects on efficiency in resource use. For example, progressive income taxation influences efficiency in resource use by altering the labour/leisure choice of workers; a high marginal tax at higher levels of income can have a disincentive effect on labour supplies and effort. In these circumstances, a choice has to be made between imperfect mechanisms. The question is whether any one mechanism has fewer adverse effects on efficiency in resource use and

consumption, either generally or in particular circumstances.

Cash versus in-kind transfers

In discussing mechanisms to redistribute income, the literature has generally supported the use of cash transfers. The arguments in favour of cash subsidies paid from government budgets include:

. no restrictions on consumption choice;

. regulatory barriers to entry;

• greater transparency in terms of cost and capacity to compare benefits against those costs;

. enhanced performance monitoring of public enterprises by distinguishing between commercial and non-commercial objectives; and

. better targeting to the aims of the policy objective.

The underlying basis for the belief in superiority of cash transfers is that individuals are the best judges of consumption choices to be made to maximise their own well­

58

being. The particular advantage of cash transfers is that they eliminate the direct welfare loss associated with the restriction in consumption choice which is incurred by in-kind transfers.

Both cash transfers and in-kind subsidies increase the effective income of the target group. However, in-kind subsidies have price effects in that the relative prices of goods to the target recipients are lowered, encouraging them to consume more of the good than they would if their incomes were raised with relative prices of all goods

remaining the same. If their incomes were raised by cash transfers, they may prefer to spend some of their increased income on other goods. Because in-kind subsidies restrict consumption choice they are worth less to the recipients compared with the equivalent amount of cash they could spend as they wished. In other words, the tax

cost of a cash transfer program would be lower than in-kind transfers giving the same level of benefits to the recipients.

If the intention of a redistribution program is to achieve an increased level of consumption of a specified product, in-kind subsidies are the least cost method of achieving this goal. The price effect of in-kind subsidies in this case complements the income effect in expanding consumption of the specified good.

It appears, therefore, that income redistribution programs to achieve general improvements in welfare are better served by cash transfers, while achieving desired consumption levels of specified goods are better served by in-kind transfers. The question arises as to the circumstances in which redistribution programs warrant

prescribed levels of consumption of particular goods. One rationale for such an approach is based on the existence of consumption externalities.

Consumption externalities

The notion is that the well-being of donors (typically taxpayers) is improved if they can be assured that the recipients are consuming particular goods the donors consider to be necessary (Pauly 1970; Olson 1971; Daly & Giertz 1976). Donors may be reluctant to contribute if they believe the funds are being misused by the recipient. For example, the recipient may be spending on alcohol funds intended for housing.

Where there is believed to be an externality associated with the consumption of a particular good or service by intended recipients, in-kind transfers can be used to try to ensure that a given level of the good is consumed. They may therefore be preferable to cash transfers because recipients of the latter may consume less of good

than is deemed to be socially desirable.

A basic problem, however, is to identify the consumption activities which generate external benefits. For example, services provided at reduced prices under the CSOs of public enterprises include telephone, postal, electricity and railway transport. It is not clear what characteristics of these services involve externalities which do not exist

for other goods and services, such as nutritional food and better housing.

Even if such externalities are present, the possible advantage of in-kind over cash transfers depends on the government’s ability to interpret and balance the preferences of donors and recipients. Even though cash transfers may lead to under-consumption of the desired goods, they will not necessarily be inferior on efficiency grounds to in­ kind transfers that lead to enforced changes in the desired consumption mix and

59

excessive levels of consumption of some goods. While cash transfers may not be well used from the viewpoint of taxpayers, in-kind transfers may have low values from the viewpoint of recipients. It would seem that, given the difficulty of interpreting and balancing donor and recipient preferences together with the nature of the political

process, some paternalistic overruling of recipient preferences is inevitable. Such an approach can be observed in many in-kind transfer programs (Browning 1975, 1981).

Thurow (1974) finds the consumption externality rationale for in-kind transfers unconvincing. He suggests an alternative rationale based on what he terms the ‘societal preferences’ of individuals. He argues these preferences relate to the rules

and structure of society and include preferences for the ‘rights of man’ to have some minimum standard of certain goods and services such as medical care, education and housing. Some goods and services provided under the CSOs of public enterprises, in particular telephone services, may be supported on these very grounds. Because family management ability varies, Thurow considers that cash transfers will not ensure the same minimum standard of service for each family. It is therefore necessary to provide goods and services deemed by society to be available for everyone in kind rather than via cash transfers.

Thurow’s argument is based on a concept which seems little different from that which he himself criticises - the concept of the consumption externality whereby the utility of the donor is related to that of the recipient. The deficiencies of the consumption externalities rationale would seem to apply to Thurow’s concept: there is a tendency to overrule the preferences of the recipients in favour of the donors; the range of goods and services thus provided seems quite arbitrary; and there is little regard to

the impact on efficiency of providing a potentially ever-expanding range deemed to be the ‘rights of man’. Thus, there may be the same tendency towards over supply of in­ kind transfers.

Arguments for in-kind CSOs

While the consumption externality argument does not necessarily provide a strong basis for relying on in-kind transfers in an income redistribution program, there are other reasons which may lead governments to consider maintaining in-kind transfers as one component of such a program.

The conclusion that cash transfers are preferable to in-kind transfers can be challenged on the grounds that cash transfers create greater scope for people to claim transfers intended for others. Redistribution relies upon information which individuals have an incentive to conceal. This consideration has particular relevance to those CSOs providing services in rural areas. Cash grants provided in lieu of

subsidised electricity, telephone, postal and railway services in rural areas would create significant incentives for non-rural residents to claim the grants as well. For example, if cash transfers were provided to all who live in isolated areas in lieu of current subsidised telephone services in isolated regions, there may be increased demand for postal addresses in those regions as people seek to take advantage of the offer. To obtain the in-kind transfer of a cheap telephone connection, potential imposters have to incur substantially greater set up or even relocation costs than they would if the subsidy were paid in cash.

60

The case for cash transfers appears stronger for the welfare category of CSOs. Most of the current social welfare programs are implemented by cash transfers rather than in-kind benefits. However, some studies suggest that the cost of cash-based welfare programs can be reduced by offering a welfare program package that includes both

cash payments and payments-in-kind (Nicholls & Zeckhauser 1982; Dye & Antle 1986; Blackorby & Donaldson 1988). In adopting in-kind packages, some loss of welfare is sacrificed by restricting recipients’ choices in order to improve the targeting of the redistribution process and deterring potential imposters.

Income restrictions governing eligibility for welfare programs are a primary means of deterring fraud. They are sufficient when it can be assumed that an individual’s demand for a good is a function solely of price and declared income. When demand depends on other characteristics (for example, ability to earn income, medical

condition and concealed income) that also affect marginal utility of income, policy instruments beyond cash transfers may become beneficial. In-kind transfers can be used to supplement cash transfer programs to improve the efficiency of the redistributive process where there is a relationship between characteristics such as

those above and the demand for specific goods. To achieve their objective, restrictions associated with in-kind transfers must impose substantially greater costs on potential imposters than on intended recipients.

It can be said also that governments (and taxpayers) face a ‘moral hazard’ problem in pursuing their social welfare objectives. If cash transfers are used to ensure that the less well off have access to, say, adequate housing and health services, the recipients of the welfare payment may choose to spend or Svaste’ the money on more

conspicuous consumption. In these circumstances, the welfare group could come to government with a case for more help to meet basic human needs. Providing for basic human needs by in-kind transfers such as subsidised electricity, telephone, housing, etc., is said to be a means of avoiding the moral hazard. Of course, if such

considerations were taken to the extreme, government income redistribution programs would consist entirely of in-kind transfers, severely restricting the consumption choices of the recipients and resulting in a very costly delivery system.

The policy implication of these considerations is that there is a case for redistribution programs to employ a combination of cash transfers and in-kind transfer or subsidies. Just how widely the in-kind program with its associated choice restrictions should be employed is difficult to determine since, in practice, it will depend also on

administrative costs and can only be resolved on a case-by-case basis.

Administration

One of the major advantages claimed for cash transfers is that they can be targeted easily to specific groups. In contrast, in-kind transfers through cross-subsidisation may give rise to costs and benefits which are both variable and difficult to determine. For example, in the case of cross-subsidisation of telephone services there may be various

beneficiaries depending on how distance, time-of-day and length-of-call factors are taken into account in the pricing of these services. The measurement problems associated with cross-subsidies make it difficult to determine all the beneficiaries and it is possible they may be broader than the targeted group.

61

However, targeting does not appear to be a problem for the main categories of the CSOs of public enterprises. The objectives of government policy are met by the CSOs in that they ensure access to a minimum standard of particular services for those in rural areas and they utilise the administration of social security programs (for example, the Pensioner Health Benefits card) to provide pensioners with subsidised access to certain services.

Ergas (1986, p. 105) notes that a cross-subsidy may be inexpensive to collect and disburse. There is little or no opportunity for evasion or avoidance; nor does it give rise to complex negotiations or litigation. The potential decision-making process can be limited to setting a simple rule.

In contrast, the administration of a cash transfer program can be costly, although important mechanisms for such transfers to implement CSOs are already in place, for example pensioner and family allowances. There are the difficulties in determining the level of subsidies in cases where the objective is to equalise a charge between rural and urban areas. Obviously, the cost differential to be covered by the subsidy would vary considerably from area to area and from year to year as usage patterns changed. As a result, some measurement process would have to be adopted which, as noted in section 14, is likely to be extremely difficult and time-consuming. Given the likelihood of a range of estimates depending on the methodology used, the lobbying and negotiating process over the extent of the cash transfers could be contentious and lengthy. This process would occur on a recurrent basis. In addition,there are administrative costs in excluding those for whom the subsidy is not intended.

Enhancing the efficiency of in-kind transfers

The discussion in this section suggests that there may not be a compelling case for replacing those in-kind transfers currently provided as the CSOs of public enterprises with cash transfers. Nonetheless, as discussed in section 14, there are several concerns about the use of cross-subsidies by public enterprises to fund these transfers. The

question remains whether other improvements can be made in the delivery system for these CSOs. A number of options can be considered; these include a voucher system, direct funding of enterprises for CSOs, a levy on users and contracting-out. The option of adjusting target rates of return to account for CSOs is discussed in Appendix L.

Voucher systems

An alternative to redistribution via cross-subsidies or cash transfers is the use of vouchers. A voucher is a gift of income which may only be spent on specific commodities. The advantage of voucher systems is that they can be used to implement in-kind transfers while incorporating some of the desirable features of cash transfers such as retaining an element of consumption choice, transparency and non-interference in the commercial activities of enterprises.

The particular advantage of vouchers over cross-subsidisation is that they avoid the efficiency effects of cross-subsidies and the associated regulatory barriers. As a voucher is a form of tied grant, the possibility that recipients will spend the voucher on goods other than those in question is minimised, although it is possible a secondary

market for trade in the voucher or the goods concerned may occur.

62

Voucher systems are most appropriate in circumstances where there are a number of suppliers of a service which governments wish to subsidise. A voucher system will enable consumption choices to be made between available services and thereby encourage competition between suppliers. Where alternative suppliers do not exist,

as is the case with a number of services provided under the CSOs of public enterprises, there seems little point in introducing such a mechanism. However, as natural monopoly elements in the production and distribution of these services are undermined by the introduction of new technologies, the argument for a voucher

system is strengthened.

For services currently the subject of CSOs, a voucher system seems most applicable to those provided as fringe benefits in welfare programs and where competing suppliers may exist. A quasi-voucher system already exists in the form of the Pensioner Health Benefits card which entitles holders to concessional rates on electricity, gas and

telephone rental charges, etc. Although it is not general over all States, the availability of similar concessions or rebates on electricity and gas charges provides an element of choice between energy sources for some domestic purposes. It is the practice also of most States and the Commonwealth to provide some concessions for

railway and bus travel. There would seem to be a case for the use of a voucher system to provide a set amount of concession which could be used on any transport service, whether public or private.

Administration costs will be incurred in identifying the level of transfer to particular consumers. This would be particularly difficult in the case of equalisation of charges between rural and urban areas, and could involve regular and costly reviews. For these reasons, the applicability of the voucher system to major services such as

electricity, water and sewerage, and telecommunications could be limited.

Direct funding of enterprises for CSOs

Cross-subsidisation to fund CSOs requires a regulatory environment to restrict potential entrants who might skim off the profitable services for which the public enterprise charges a premium. As outlined earlier, these regulations are likely to impair efficiency in resource use. One way of avoiding some of these effects is for the

Government to fund directly the CSOs of public enterprises. This is already done for AN, for some fringe benefits provided by public enterprises as part of Commonwealth and State government welfare programs and some other services provided by enterprises at the direction of governments. This alternative has the advantage also

of improving the transparency of CSOs. Both AN and the Australian Mining Industry Council recommended to the Commission separate accounting for non-commercial activities. It was also implied in the submission of Major Mail Users of Australia.

Direct funding of non-commercial activities which public enterprises undertake at the direction of governments has a long history in Australia, especially in respect to railways (eg Wettenhall 1987).

Extension of this approach deserves serious consideration in the light of recent reforms to develop a more commercial orientation on the part of government enterprises, and possible further reforms canvassed in Chapter 7 of this report including more precise specification of non-commercial objectives. Direct funding of

63

CSOs provides a means of reducing the efficiency effects associated with cross­ subsidies and their concomitant barriers to entry. It ensures also that the costs of the non-commercial objectives of government enterprises are subject to closer annual budget scrutiny and has the potential for public discussion of the cost and benefits of

the CSOs. Further, the cost of the redistribution of income is spread over all taxpayers rather than being limited to the users of particular goods and services. There are difficulties in monitoring the costs of operation and there are incentives for cost padding, but there seems to be little reason to expect such costs to be greater under this option compared with other means of delivering CSOs.

Nevertheless, past experience of direct funding revealed some problems which will require attention in any further application of this option. One particular difficulty arose from a combination of inadequate specification of non-commercial objectives and lack of clear demarcation between the respective responsibilities of governments and enterprises and accounting shortcomings. As a result, there was some difficulty in distinguishing the costs of politically enforced activities of the enterprises from those incurred in commercial operations. Often, compensation for enterprises was inadequate to meet all political directions and interference in the operations of enterprises. Limits on the compensation were in practice restrictive. For example, enterprises were compensated for losses on particular services or in respect of concessions on particular items, but not for political refusals to grant general increases in charges. Moreover, the case for direct funding of public enterprises to implement social objectives was challenged by many because it resulted in extensive wrangling between the enterprises and Treasury departments over compensation payments. This process seemed unnecessary given that the overall losses of enterprises were funded by governments in any case (Wettenhall 1987, pp. 48-53).

Measures to promote commercialisation of public enterprises, together with publication of political directives to enterprises, could help ensure such problems do not re-emerge in the use of direct funding of CSOs.

Direct funding is utilised to a limited extent for the fringe benefits of welfare programs and it could be extended to cover all such fringe benefits. It could be complemented by an extended use of voucher-type systems for these programs.

In the case of other complex CSOs such as those involving charges and access to water and sewerage, electricity and telecommunications (especially as they attempt equalisation of benefits between rural and urban users), some of the difficulties outlined earlier in relation to cash transfers also pertain here. These relate primarily to the difficulty in identifying the amount of funding required, given the presence of cost padding and/or commercially oriented discriminatory pricing policies. In the latter case, a problem exists in distinguishing between commercial and social objectives. This problem could be resolved by publication of specific non-commercial directives issued to enterprises by governments and restricting compensation to such directives.

A further consideration in determining the extent to be made of the direct funding option concerns the economic impact of the revenue measures used to support it. For example, the Western Australian Treasury noted in its submission that, as a result of constitutional constraints on State government revenue raising powers, those governments must rely on a number of economically inefficient and distortion ary

taxation measures to meet their revenue requirements. It stated that direct funding of all CSOs by means of such arrangements may be less efficient than the current system.

As discussed in section 14, cross-subsidies have the potential to create significant resource costs as a result of higher prices paid by particular users and the entry barriers required to support cross-subsidies. It is possible that some revenue raising alternatives devised by the States could result in more adverse effects. But the States

do have access to measures which spread the burden of funding CSOs more widely across the community than occurs in the case of cross-subsidies and also do not restrict competition in the production of goods and services. So, while it is not possible to determine in the abstract the most appropriate mechanism for each

enterprise, it seems unlikely that cross-subsidies would prove to be the most efficient method available. Moreover, resort to direct budgetary funding facilitates public scrutiny of the financial resources devoted to CSOs and the costs of raising those funds.

Levy on users ,

Another alternative for funding the costs of CSOs could be a levy on subscribers, a policy which is followed for some enterprises in the United States. This option was canvassed by the Australian Telecommunications Users Group in its submission to the inquiry. Under this system all subscribers have a line item on their telephone bills

covering the cost of telephone services provided to the disadvantaged. This requires an estimate of the cost of this service so that it can be apportioned to all subscribers, but it has the advantage of making the cost explicit and appears to have lower administrative costs.

This option would avoid the need for entry barriers as all consumers of the product, whatever their source of supply, could be included in recovery of the costs. However, consumption of the product being taxed is likely to be reduced, and the costs of those industries which use the product more intensively as an input will be raised relative to

other industries. Further, while funding of part of a government income redistribution program is extended to a broader base under this option, it may be more appropriate that the base for funding welfare programs consist of all taxpayers rather than being confined to the users of a particular service.

These considerations suggest that there is little reason, apart from their overall budgetary impact, to adopt a levy on users in preference to direct funding from the budget for the CSOs of public enterprises.

Contracting out

An alternative to cross-subsidisation as a means of providing CSOs is a form of contracting out or franchising. The Australian Bus and Coach Association, for example, submitted that the CSOs undertaken by public bus operators should be equally available to both public and private bus and coach operators. Under such a

scheme the private sector would be required to provide not only the profitable goods and services but also the CSOs. This would have the advantage of limiting the distortions in resource allocation which arise when the CSOs are financed via cross­ subsidies.

A contracting out procedure requires that the CSOs are precisely specified and that monitoring and information costs associated with ensuring performance are low. But this is not the case under many existing arrangements. CSO objectives are not well defined, and in fact there may be political advantages in keeping them poorly defined. The costs of monitoring and obtaining information necessary to judge performance of

the enterprise in meeting CSO objectives are high. This makes monitoring of performance of the enterprise difficult and may also allow cost padding even in the private enterprise.

Monitoring and information costs may extend also to the provision of profitable goods and services. Market and technological uncertainties may imply incomplete contracts which allow for adjustments as new circumstances are revealed, rather than initiating

a new round of bidding and drawing up of contracts as new specifications need to be incorporated. For more complex requirements, incomplete contracts may be more efficient than complete contracts, However, incomplete contracts often require extensive monitoring and their administration is then simply a particular form of regulation.

In addition, the existence of sunk costs may give one firm a decisive advantage in the competition for the contract, causing other firms not /to bid for the contract and leaving the incumbent firm with an uncontested monopoly. Alternatively, the inability of an incumbent firm to recover costs in the event of failure to win the franchise in later periods could lead to lower investment than is required to adequately perform the task.

For these reasons, franchising and contracting out are likely to be real alternatives in circumstances where contract specification is simple, involving standardised goods and services, and not subject to market or technological uncertainty, where performance is evenly monitored and where sunk costs are low.

16 Assessment

The CSOs of public enterprises in Australia raise several obstacles to better performance. CSOs are not always explicitly defined, and the costs of meeting them are often hidden and can be substantial. To the extent that cross-subsidies are used to implement CSOs, regulations are required to restrict potential competition.

The various methods used to fund CSOs all have implications for efficiency in resource use and consumption which need to be assessed against the perceived benefits of the objectives behind CSOs. These efficiency effects are most apparent in the use of cross-subsidies and the associated regulatory barriers to competition.

If governments wish to maintain the CSOs of public enterprises it may be possible to reduce the extent of the problems by applying certain principles to the operation of CSOs. These include making the objectives of CSOs explicit to management and the public, providing estimates of the annual cost of CSOs and considering alternative measures to implement CSOs which remove the necessity of regulatory entry barriers and increase the choices of the targeted users. Examples of the application of these principles already exist and there seems good reasons for them to be more generally applied in the implementation of CSOs.

66

The alternative mechanism most suggested for implementing CSOs is direct cash transfers to the targeted recipients. It has several advantages and the large part of government welfare programs are delivered in this way. However, there is a case for welfare packages to include some in-kind transfers, although the desirable extent of

these may require a case-by-case examination. The argument for cash transfers to be provided in lieu of the subsidised in-kind transfers made available to rural areas through CSOs is weakened by the potential incentives they would provide for non- rural residents to claim them as well.

Several alternative mechanisms can be utilised to implement CSOs. They include voucher systems, direct funding and contracting out the services to be provided. None of these mechanisms provides a complete solution to the problems encountered with CSOs. They all involve trade-offs between efficiency and other objectives.

Nevertheless, all these mechanisms could be considered as potentially part of a complementary set of measures to implement government objectives more efficiently. Direct funding of CSOs provides an avenue for ensuring that the objectives of government policy are made more explicit and providing for public scrutiny of the

costs incurred in implementing them. It is also a means of overcoming some of the more serious efficiency effects of cross-subsidies. Selected use of voucher systems and contracting out services to be provided to meet government objectives could be used to improve consumption choice.

67

ATTACHMENT I I : CROSS-SUBSIDY BENCHMARKS

The first step in identifying and measuring cross-subsidies is to establish a benchmark for subsidy free prices. These benchmarks fall into two main classes. The first, cost allocation benchmarks, define subsidy free prices by allocating production and selling costs amongst products in a firm’s output. The second, market based benchmarks, seek to define subsidy free prices in terms of what would be feasible in a competitive market. Generally, a broader range of prices would be considered subsidy free under market based benchmarks than under cost allocation benchmarks.

11.1 Cost allocation benchmarks

Under cost allocation benchmarks, cross-subsidies are defined to exist whenever prices do not return the precise costs of providing goods and services. Thus, a subsidy would be received where prices are below these precise costs, and paid where prices

exceed them.

Two main uses arise for these benchmarks. First, they are often perceived to represent fair prices, and hence may be significant to consumers or to political lobby groups. Second, if all products in a firm’s range can clearly be shown to be paying their way, the firm as a whole will be viable. Cost allocation benchmarks will often provide convenient short-cut tests of the profitability of products or product lines.

However, there are a number of sound commercial reasons why a firm might not set prices exactly equal to the precise cost (however defined) of the goods and services supplied. Thus, in many instances, these benchmarks merely help to describe how, over a range of products, firms recover their costs. Nevertheless, the benchmarks discussed in this section do underlie common uses of the term cross-subsidy.

The first part of this section discusses why factors such as information and other transactions costs, and complex interactions amongst revenues and costs, may cause firms not to price ‘perfectly’. The second part broadens this discussion to consider the treatment of costs which, even in principle, could not be uniquely assigned to one particular product.

Perfect pricing

There are a number of reasons why, in a competitive market, a firm would choose not to set prices for goods and services which precisely reflect the costs of their provision. One is that calculating the cost of supplying a product to different customers can, itself, be a costly exercise. Costs of generating and processing the information necessary to price ‘perfectly’ may cause actual prices to be different from those

hypothetical benchmark prices. Where the cost of fine tuning prices would be large relative to the price differences that would emerge, some degree of cross­ subsidisation will be efficient. Excessive fine tuning may, by raising selling costs, even increase prices to some users by more than any cross-subsidy they might otherwise pay.

This form of cross-subsidisation may be demonstrated in the case of a garage. The cost of tools and equipment is usually recovered in charges for workshop labour. Labour is often charged at a standard rate, regardless of the complexity of the task or the value of the particular tools and equipment used, eg hoists. However, where

differences in such costs are more significant and readily distinguishable, eg engine tuning, a different rate may be charged to reflect the value of skills and equipment actually used.

The appropriate level of cross-subsidisation is a matter of commercial judgment, which will be tested by the ability of the pricing arrangement to survive in the face of competition. Excessive cross-subsidisation will result in the loss of lower cost jobs to firms charging prices which more accurately reflect specific job costs, and a surplus

from those jobs will not be available to subsidise more expensive jobs.

Another reason for cross-subsidies to occur against this benchmark is that departures from uniform pricing increase the complexity of transacting business. For example, cafeterias often offer fixed price menus. Whilst this may reduce the number of staff required to service a given number of customers, the simplicity of a uniform price may

also have an attraction to customers. The appropriate level of cross-subsidisation is, again, a matter of commercial judgment which will be tested by the market. In making this judgment it is not necessary to precisely measure the cross-subsidy involved, but merely to have an appreciation of its magnitude in comparison with the value consumers place on the simplicity of a uniform price. If cross-subsidies are too

great, consumers will purchase cheaper dishes from alternative outlets which price more accurately, and the surplus generated on those dishes will be insufficient to subsidise the underpricing of more expensive dishes.

Cross-subsidies can also arise from revenue interrelationships among products. These occur where sales of one product line generate additional sales for another product line. This rationale underlies the use of loss leaders, where particular product lines are heavily discounted, possibly to below cost, in order to attract customers into the store. Another practice is for stores to stock apparently unprofitable lines. By

developing the firm’s reputation as a reliable supplier of a broad range of products, such product lines have the potential to attract more customers into the store and increase sales of all lines. The point at which the profit from extra sales is offset by the cross-subsidy is, again, a matter of commercial judgment.

This form of cross-subsidisation may also occur across markets, eg a producer of components for assembly may supply original equipment parts at below cost in order to gain the right to supply the more profitable ‘genuine’ spare parts market. Also, account must be taken of cost interrelationships among products, eg apart from

directly generating revenues, by-production processes can be profitable simply because they reduce a firm’s waste disposal costs.

Fully distributed costs

Where, as in most cases, firms have more than one output, costs arise which are not uniquely attributable to any one product, eg from shared production lines and administration. It therefore becomes unclear, even in principle, what ‘perfect prices would be. The fully distributed cost benchmark seeks to distribute such costs in

proportion to the use products make of the common facility.

However, it is rare that a clearly objective cost distribution emerges. There may be many ways of measuring use of a common facility, each apparently equally valid. For example, factory overheads could be allocated on the basis of, say, floorspace or direct labour hours - it is unlikely that either would be closely related to overheads, which may include costs as diverse as window cleaning, recruitment, and time off for a fire drill. Consequently, when this benchmark is applied, perceptions can vary considerably as to the true cost of a product.

Nevertheless, because accounting systems often cost output in a manner which absorbs overheads, information may be readily available to allow this benchmark to be used as a crude test of the profitability of a product. Clearly, if all of a firm’s products cover both their direct costs and their share (however measured) of indirect costs, the firm as a whole will cover its costs. There are, however, strong commercial

reasons why a firm would choose, against this benchmark, to engage in cross­ subsidisation.

One reason for not recovering overheads in strict proportion to use is that the viability of a product may be dependent upon the opportunity to free ride on facilities used by existing products. So long as a product covers all of the additional costs it creates for

the firm - including any fixed costs solely attributable to it - no burden will be imposed on other products. Since any revenue generated in excess of those additional costs permits a contribution to common costs, users of other goods would be better off ‘subsidising’ that product than meeting the full burden of common costs themselves.

This can be demonstrated by a simplified example of a firm in a competitive market facing the following costs and revenues for its products in the short run:

Revenues Direct Costs Margin

$ $ $

Product A 600 200 400

Product B 400 200 200

Product C 300 200 100

Total Firm 1300 600 700

Indirect Costs (600)

Profit Τ ϋ Γ

Assume that the total indirect costs of the firm are fixed regardless of output, at $600, and that the cost allocation which is considered most appropriate shares indirect costs equally among product lines at the rate of $200 each. Under the fully distributed cost criterion, Product A is profitable, Product B just breaks even, whilst Product C makes a loss - apparently receiving a cross-subsidy from Product A.

However, were Product C eliminated on this basis, the firm would lose revenues of $300, whilst avoiding costs of only $200. The firm would now just break even. However, as the indirect cost allocation to the remaining products would now be $300, Product B would also, on this criterion, now be judged unprofitable. Eliminating Product B would, in turn, place the full burden of indirect costs on the

70

remaining Product A, and the firm, despite its initial profitability, would no longer be viable.

Another reason for a firm to cross-subsidise (on the basis of this benchmark) is that some customers will be more price sensitive than others. This may reflect the income of prospective customers, or arise from the existence of alternative suppliers or substitutable products. A firm will maximise its profits by charging lower prices to

customers whose demand is more price sensitive, recovering overheads primarily from customers whose demand is less price sensitive. One common example is cinemas offering discounts to low income groups such as students, pensioners and children. Another is where coach lines operate at lower margins on routes where there is more

competition from other coach lines or from alternative means of transport.

11.2 Market based benchmarks

Whilst the cost allocation benchmarks described in the previous section underlie common uses of the term cross-subsidy, they do not provide a satisfactory economic definition. Often products which (against those benchmarks) are considered to be cross-subsidised will be shown to be paying their way when proper account is taken of

factors such as information and other transaction costs, and interactions between costs and revenues. Further, they do not provide a basis for distinguishing inefficient cross­ subsidisation from that which is efficient.

Ultimately, whether or not products or users are paying their way in an economic sense is tested by the ability of pricing arrangements to survive in a competitive market. Where products do not generate sufficient revenues to cover their costs, higher prices must be charged elsewhere to cover the resulting deficit. In a

competitive market, this would create the incentive for rival firms to supply at lower prices the profitable part of the market, whilst leaving the loss making sales to the cross-subsidising firm (cream skimming). Thus cross-subsidies, except for those of the efficient kinds discussed in section 11.1, would be competed away.

The benchmarks described in this section are designed as proxies for that market test, ie to anticipate whether prices will be able to survive in the market. This is not to say that firms in all cases explicitly calculate these benchmark prices. Nevertheless, in deciding to provide goods and services at particular prices, firms are making

commercial judgments that the conditions of these tests are being satisfied and, implicitly, that any remaining cross-subsidies are of the efficient kinds discussed in section 11.1.

Product cross-subsidy

This concept of cross-subsidy relates to prices for goods and services which would be capable of surviving in a competitive market. Faulhaber (1975, p. 966) proposed:

if the provision of any commodity (or group of commodities) by a multiproduct enterprise subject to a profit constraint leads to prices for the other commodities no higher than they would pay by themselves, then the price structure is subsidy free.

71

This definition gives rise to two tests:

Incremental cost test

Subsidy free prices require that each product generates revenues at least as great as its incremental cost. Incremental cost is the increase in a firm’s total costs which results solely from supplying the product in question, ie those costs which could be avoided by not producing the product. Thus, the incremental costs of a product exclude costs which are in common with other products and hence would not be avoided by deleting that product. It does, however, include those capital costs (and the required return on investment) which could be avoided by deleting that product.

Stand-alone test

Subsidy free prices also require that each product generates revenues no greater than stand-alone costs. Stand-alone costs are those of providing a product in isolation from other products, ie the existing costs of the firm, less those costs which would be avoided by deleting other products. Generally, a product’s stand-alone costs will be no greater than its direct costs plus all of the existing firm’s overheads.1 If revenue generated by any product exceeds its stand-alone costs, it is clear that it is generating

revenue which may be used to cross-subsidise other products or be taken as profit.1 2

These costs may be demonstrated by the example of a railway branch line used by freight and passenger services. The incremental cost of an single run by a passenger train would include such things as fuel used, and the additional labour and maintenance required for that run. For the passenger service as a whole, incremental costs would include the cost of passenger trains. Similar costs could be defined for freight services. For the branch line as whole, incremental costs would include the

costs of the line itself. The stand-alone costs of, say, the freight service or any individual freight run would include the costs of the line, but exclude the costs of the passenger trains.

All efficient forms of cross-subsidy discussed in section II. 1 will be considered subsidy free under this benchmark, eg the revenues generated by a loss leader would be taken to include the increased sales of other products. Further, ‘perfect prices’ (as defined in section 11.1) would, where inefficient, involve cross-subsidies, eg where a firm devoted too many resources to accurate pricing, its prices would fail the stand-alone test, since the products could be supplied at lower cost by a firm which used simpler pricing.

1 In those cases where deleting other products would result in new costs, eg disposal of wastes formerly used in by-products, those costs would have to be taken into account. 2 Whilst the stand-alone test indicates when a cross-subsidy is occurring, stand-alone costs only provide an upper bound for subsidy free prices. For example, the stand-alone cost for the supply of water to a

single house in a reticulated system would be calculated on the assumption that water would continue to be piped from a regional dam. In practice, lower cost technology could be used for a single user, eg tanks filled by rainwater or from water carted by road. Where town dwellers are prohibited from having their own tanks, they may be paying more for water than they would under a non-reticulated service. However, it is not clear from this that a cross-subsidy is occurring, since this excess cost may not necessarily benefit other users who are covering their incremental costs.

72

Both the stand-alone and incremental costs tests must hold for all products individually and for any combination of them.·* The two tests are related in that where a product does not cover its incremental costs, additional revenues must be raised on other products, causing their prices to rise above their stand-alone costs.

The distinction between the two tests arises from costs that are common to several products in a firm’s range. Multiproduct firms exist because, for many products, it is cheaper to produce them together rather than independently, eg because of the opportunity to share facilities. Eventually a point may be reached where these

advantages - termed economies of scope - are fully exploited, eg due to congestion. For outputs below this level, however, incremental costs will necessarily be less than stand-alone costs. Consequently, a range of prices may emerge which could be

considered subsidy free, sharing common costs in various ways. Many of these would, of course, be considered to involve cross-subsidies under fully distributed cost benchmarks (see section 11.1).

The presence of product cross-subsidies is important because of the role prices play in directing the use of resources in the economy. By altering relative prices, cross­ subsidies have the potential to redirect resources into production upon which the market places a lower valuation. Alternatively expressed, the efficiency of the economy in converting inputs into more valuable products is reduced.

When prices overstate costs of provision, goods and services will tend to be underutilised. Consumers will be induced to purchase alternatives which give less satisfaction, whilst producers may be encouraged to use input mixes which cost more from the point of view of the community as a whole. Where prices understate the cost

of provision, the more expensive goods and services may be used instead of a satisfactory cheaper alternative, eg subsidised landline telephone transmission might, in some cases, cost considerably more to provide than alternative technologies, such as radio, without a commensurate difference in the quality of service (Whiteman

1988, p. 42).

In addition, and potentially of greater concern, there may be productive efficiency losses which can be associated with cross-subsidies. Essentially, these result from the need to protect a cross-subsidising enterprise from firms which would be able to offer subsidy free prices. The absence of competition may permit cost padding, eg relatively high remuneration for employees, restrictive work practices, and overcapacity. Higher costs may also result from the use of inappropriate technology. In a competitive market, the cost of overcapacity or obsolescence would be borne as

capital losses by the owners of the firm, whilst in regulated markets these costs may be passed onto users as higher prices.

If a firm is not minimising costs (ie there is productive inefficiency) the totality of the firm’s output would fail the stand-alone test, since the bundle of products could be 3

3 For a group of products which forms a complete component of a system, the stand alone cost of the group would equal its collective incremental cost to the system. For example, all users of a railway branch line might cover their individual incremental costs (eg fuel and labour), but unless as a group they cover the incremental cost of the branch to the overall railway system (ie their collective stand

alone cost, including the costs of the line) they are receiving a subsidy. Nevertheless, if all users each covered their stand alone costs this would breach the stand alone test since, collectively, they would be paying for jointly used facilities several times over.

73

purchased at lower cost from an alternative, efficient firm. This would imply a subsidy from customers to factors of production.4 However, inefficiencies can often be ostensibly justified as reflecting the burden of community service obligations (CSOs).

In these circumstances, the size and direction of any cross-subsidies may be very difficult to determine, since true, efficient, incremental and stand-alone costs may relate to technology which is different from that presently used. What appears to be a subsidised price for a service provided under a CSO may, in fact, be no less than the price which would be paid to an efficient firm.

From the point of view of the firm, product subsidy free prices provide the true benchmark for the profitability of a product line; products which are being cross- subsidised will not make a positive contribution to recovery of the producer’s costs. Many prices which involve cross-subsidies under the benchmarks discussed in section 11.1 would not involve product cross-subsidies, reflecting the fact that those forms of cross-subsidisation will often represent efficient pricing decisions, eg because

information is costly to produce, excessive precision in costing the provision of goods and services will constitute a misallocation of resources, benefitting neither producers nor users.

Consumer cross-subsidy

Consumer cross-subsidy is the measure of how much one group of users subsidises the cost of another’s purchases. Consumer subsidy free prices occur where no group of users could obtain a bundle of goods and services more cheaply from another supplier. This benchmark is similar to that of product subsidy free prices, but with allowance made for the possibility of offsetting cross-subsidies among products within a particular group of consumers.

Where there is effectively one user for each commodity, eg a telephone service supplied to a particular location, consumer cross-subsidies will be equivalent to product cross-subsidies. Often, however, it is not the transfers which occur within one product which are of interest, but the overall redistributive effect of a public enterprise’s pricing arrangements. Thus while users might receive a subsidy on one product, this will be offset to some extent by subsidies paid on other products, eg the subsidy to rural users on telephone access charges may be offset to some extent by the

higher toll charges from which the subsidy is financed.

The more general relationship between product and consumer subsidy free prices will depend upon the pattern of demand across users and the breadth of the range of products being considered. For example, suppose a takeaway food outlet offers free

parking, covering the cost from higher charges on food. Clearly, food is generating a subsidy whilst the car park is receiving one.

4 In the water reticulation case discussed in footnote 2, the excess cost would involve a subsidy to factors of production, regardless of the technical efficiency with which the water reticulation service was provided, eg more contracts would be available for those supplying infrastructure, and career prospects of workers would be enhanced in a larger organisation. If those who would prefer to use

tank water contribute more to the reticulated system than their incremental cost, then other users would also benefit, through reduced charges.

74

However, if most customers prefer to drive to the store, then little or no consumer cross-subsidy would result from this product cross-subsidy. Further, even though those who do walk to the store could be seen to be subsidising those who drive, a broader view may reveal other cross-subsidies from which those who walk benefit relatively more. For example, if people who walk are more inclined to eat in rather than take away, then they could be receiving a subsidy on the cost of dining facilities from those who drive.

In a competitive market, product cross-subsidies will only survive if they do not result in consumer cross-subsidies, ie product cross-subsidies would in general have to offset each other. Prices which involve consumer cross-subsidies create incentives for a rival firm to offer the same bundle of goods at prices which do not involve cross-subsidies.5

Because consumer subsidy free prices might not be product subsidy free, the underlying usage of goods and services sold at consumer subsidy free prices may still be altered by (offsetting) product cross-subsidies. Hence, consumer cross-subsidies

are not a measure of efficiency costs. Nevertheless, to the extent that the transfers measured by consumer cross-subsidies favour particular sectors, they have the potential to alter resource use. Industries which use relatively more subsidised products can be expected to expand, but at the expense of those which use relatively more products which pay a subsidy. In the process the overall efficiency of resource use in the economy may be reduced. Further, productive inefficiencies associated with cross-subsidies may reinforce this burden on efficient, non-subsidised users.

The consumer cross-subsidy benchmark also provides an in principle basis for the measurement of the net transfers to particular households as a result of CSOs. The purpose of those CSOs might not be to redistribute income as such. It may be to offset the cost of some good or service where the price might otherwise be considered

too high for some users due to differing costs of supply (eg telephone services to remote users) or differing income levels (eg public transport). Nevertheless, such arrangements are analogous to, and could in principle be replaced by, a system of direct taxes and subsidies.

In practice, however, lack of detailed knowledge of the demand patterns of users may prevent transfers being targeted with great precision. Indeed, avoidance of the need to generate such information is one possible reason why cross-subsidies have been preferred as an instrument for the implementation of CSOs such as universal access at affordable prices. However, as a result, transfers can take place which conflict with other social objectives, eg transfers from urban poor to wealthy rural households. In summary, whilst cross-subsidies may be an administratively simple means of pursuing particular social objectives, the resulting outcomes tend to be obscured.

Whilst the consumer cross-subsidy benchmark described above is based on product cross-subsidies, the same principles could be applied to cost allocation benchmarks described in section II. 1. In particular, concern is often expressed about the perceived fairness with which costs that are mutual to a number of a public

enterprise’s outputs are shared amongst users. For example, because Telecom s

5 Even where, for most of a firm’s customers, product cross-subsidies do offset each other, there may still be a significant number of customers who, because they only purchase some of the firm s outputs, are on balance paying a subsidy. This raises the possibility of another firm supplying users wit specialised demand and undermining the viability of the multiproduct firm s overall output.

75

production process is highly capitalised and involves considerable expenditure on common costs, there will be a wide range of prices which would pass the incremental and stand alone costs tests, and hence be considered product subsidy free.

REFERENCES

Albon, R. 1985, Private Correspondence, Competition or Monopoly in Australia’s Postal Services, The Centre for Independent Studies Limited, Sydney.

Australia, House of Representatives 1989, Debates, Second Reading Speech on the Telecommunications Bill, 13 April.

Australian, 3 September, 1988.

Blackorby, C. & Donaldson, D. 1988, ‘Cash versus Kind, Self-Selection, and Efficient Transfers’, American Economic Review, vol. 78, no. 4, pp. 691-700.

Bradley Inquiry. See Committee of Inquiry into the Monopoly Position of the Australian Postal Commission 1982.

Browning, E. 1975, ‘The Externality Argument for In-Kind Transfers: Some Critical Remarks’, Kyklos, vol. 28, fasc. 3, pp. 526-44.

____ 1981, ‘A Theory of Paternalistic In-Kind Transfers’, Economic Inquiry, vol. 19, no. 4, pp. 579-97.

Bureau of Industry Economics (BIE) 1988, Regulation of Private Communications Networks in Australia, Information Bulletin 11, AGPS, Canberra.

Campbell Inquiry. See Committee of Inquiry into the Australian Financial System 1981.

Committee of Inquiry into Gas and Electricity Tariffs in Western Australia 1985, Final Report, May.

Committee of Inquiry into the Australian Financial System 1981, Australian Financial System: Final Report, (J.K. Campbell, Chairman), AGPS, Canberra.

Committee of Inquiry into the Monopoly Position of the Australian Postal Commission 1982, Report, (A.E. Bradley, Chairman), AGPS, Canberra, August.

Coughlin, P. 1987, ‘The Commonwealth Banking Corporation: A Case for Privatisation’, in Privatisation: An Australian Perspective, ed. P. Abelson, Australian Profession Publications, Sydney, pp. 204-25.

Daly, G. & Giertz, F. 1976, ‘Transfers and Pareto Optimality’, Journal o f Public Economics, vol. 5, no. 2, pp. 179-82.

Dye, R.A. & Antle, R. 1986, ‘Cost Minimising Welfare Programs’, Journal o f Public Economics, vol. 30, no. 2, pp. 259-65.

77

Ergas, Η. 1986, Telecommunications and the Australian Economy: Report to the Department of Communications, AGPS, Canberra.

Faulhaber G.R. 1975, ‘Cross-subsidisation : Pricing in Public Enterprises’, American Economic Review, vol. 71, no. 5, pp. 1083-91.

H.V. Evatt Research Centre 1988, The Capital Funding of Public Enterprise in Australia, H.V. Evatt Foundation, Sydney

Hartley, P.R. & Trengove, D.D. 1983, Who Controls Public Utilities, Discussion Paper no. 41, Centre of Policy Studies, Monash University, Clayton, Victoria.

Industries Assistance Commission (LAC) 1988, Coastal Shipping, Report no. 415, AGPS, Canberra, July.

Minister for Transport and Communications 1988a, Australian Telecommunictions Services: A New Framework, Summary, AGPS, Canberra.

____ 1988b, Australian Telecommunications Services: A New Framework, AGPS, Canberra.

Nichols, A.L. & Zeckhauser, R.J. 1982, ‘Targeting Transfers through Restrictions on Recipients’, American Economic Review, vol. 72, no. 2, pp. 372-7.

Olsen, E.O. 1971, ‘Some Theorems in the Theory of Efficient Transfers’, Journal of Political Economy, vol. 79, pp. 166-76.

Pauly, M.V. 1970, ‘Efficiency in the Provision of Consumption Subsidies’, Kyklos, vol. 23, fasc. 1, pp. 33-57.

Prices Surveillance Authority 1984, Inquiry in Relation to the Supply of Telecommunications Services: Report.

Prime Minister 1986, Speech to Northern Australian Development Conference, Commonwealth Record, vol. 11, no. 41, AGPS, Canberra.

Rees, R. 1984, ‘The Public Enterprise Game’, Economic Journal, vol. 94, Supplement, pp. 109-23.

State Rivers and Water Supply Commission (SRWSC) 1984,1983-84 Annual Report, Melbourne.

Thurow, L.C. 1974, ‘Cash versus In-Kind Transfers’, American Economic Review, vol. 64, no. 1, pp. 190-5.

Wettenhall, Roger 1987, Public Enterprise and National Development, Royal Australian Institute of Public Administration (ACT Division), Canberra.

Whiteman, J.L. 1988, ‘Defining and Quantifying Community Service Obligations - The Case of Telecom’, Economic Papers, vol.7, no. 3, pp. 33-43.

78

APPENDIX J : NATURAL MONOPOLY

It is often argued that the existence o f natural monopoly provides a rationale for government intervention both to avoid ‘wasteful’ or ‘destructive ’ competition and to limit the abuse o f market power. However, the conditions under which such competition might arise are uncommon, and the benefits of intervention are likely to be outweighed by the costs. Intervention to control the abuse o f market power may nevertheless be justified where a natural monopoly faces limited actual or potential

competition.

Public ownership and operation is only one of several approaches which could be adopted in such cases. Others include regulated private supply and mechanisms which promote competition such as contracting out, common carrier provisions, and segmentation o f operations. The relative merits of these approaches vary according to industry circumstances, suggesting the need for case-by-case assessments. Approaches which provide an explicit role for competitive pressures to promote efficient production and pricing practices are generally to be preferred to those

which do not. *

J1 Introduction

An industry is a natural monopoly if the entire output of the market can be supplied by a single firm at lower cost than by any combination of two or more firms.1 In such | circumstances, an industry structure with more than one firm would be inefficient. However, single-firm supply also raises the possibility that the monopoly will exploit

users by using its market power to raise prices.

These efficiency and distributional considerations underlie traditional arguments for government involvement in markets where natural monopoly may exist (see Chapters 6 and 7, Volume 1). In Australia, the most common form of intervention has been for government to own the industry. Another response has been for government to regulate private suppliers to prevent possible abuses of market power. Recent developments point to other approaches designed to reap the efficiency gains available to a single producer whilst limiting the exploitation of market power.

This appendix examines the justifications for government intervention in markets characterised by natural monopoly and appraises the alternative forms which such intervention might take. In doing so, it disregards other possible rationales for government intervention (such as CSOs, which are examined in Appendix I), as each rationale should be assessed on its merits.

1 A firm or an industry is said to be a natural monopoly when its cost function is sub-additive (ie C(y) + C(y*) > C(y+y’) for output vectors y and y’ (where y and y* may be the same or different products) such that y + y equals market demand. For a more technical discussion of the definition of natural monopoly, see Baumol (1977), Baumol, Panzar and Willig (1982), Sharkey (1982), or Ralph (1986).

79

The structure of this appendix is as follows. Section J2 identifies factors which might give rise to a natural monopoly, and gives some examples of industries often considered to be (or to have elements of) natural monopolies. The question of how widespread natural monopolies are in practice is addressed in section J3. Section J4 presents and assesses the natural monopoly arguments for government intervention. Alternative ways of dealing with natural monopoly situations are discussed in section J5, while concluding comments are presented in section J6.

J2 What causes natural monopoly?

Whilst the existence of a natural monopoly rests by definition on whether or not market demand can be met at lowest cost by a single firm, several factors have been identified as contributing towards its existence. These include economies of scale, economies of scope, and/or economies of organisation.

Economies of scale

Economies of scale in the production of a good or service are said to exist when an increase in all inputs yields a more than proportionate increase in outputs.

Perhaps the most commonly cited example of economies of scale is where production requires substantial unavoidable fixed costs (known as indivisibilities) which may be shared across additional units of output.2 Industries relevant to this inquiry that are often claimed to exhibit significant economies of scale include parts of railways, electricity, gas, telecommunications and water supply. Production in these industries is often characterised by large, unavoidable up-front costs.

Other sources of economies of scale include specialisation (eg through the division of labour or automation) and the reduction of uncertainty through size. An example of the latter is where, as the size of an electricity generating system grows, the risk of total system failure normally falls, enabling a lower reserve capacity per unit of output.

Economies of scope

Economies of scope exist when the cost of producing two or more products by a single enterprise is less than the cost of their separate production by a number of more specialised firms. One source of this jointness in production might be the existence of indivisibilities which, in this case, can be shared between products. For example, once a railway line is in place, passenger and freight services can be provided at a much lower cost than would be the case if each were provided on a separate line.

Economies of organisation

Economies can also arise when transactions costs for organising supply within an industry are minimised with a single supplier.3 This might be the case where the

2 The economy of scale only occurs to the extent that the indivisibility is unavoidable. A smaller plant may be less costly for smaller volumes of output (see Sharkey 1982, p.75). 3 The distinction made here between economies of organisation and economies of scope is somewhat artificial but is nevertheless useful when distinguishing between the purely technical and wider

organisational aspects of the production process.

development of an overall system requires planning and coordination between industry segments. If there are considerable difficulties and costs associated with establishing effective commercial contracts between diffuse firms, the industry may be better organised as a single vertically-integrated firm.4

J3 How widespread are natural monopolies?

Whilst there is a relationship between economies of scale, scope and organisation and natural monopoly, caution should be exercised in inferring that natural monopoly necessarily follows from the presence of such factors. Recent economic theory suggests that testing for the existence of natural monopoly involves complications not

previously recognised.5

Whether an industry is a natural monopoly depends on conditions of supply (eg technology) as well as demand (eg the size of the market). Thus, least-cost supply may vary between single-firm and multi-firm production over different output ranges, according to the level of demand. An industry that is a natural monopoly in Australia

may not be one in the USA, where the size of the market may be sufficient to warrant more than one supplier. In addition, technological change will almost certainly alter the cost structure of an industry over time (eg by impacting on the degree of economies of scope or scale in the production process). Consequently, whether or not

an industry is a natural monopoly may change over time.

Indeed, there are no simple conditions for testing for the existence of natural monopoly. A rigorous test would require a systematic comparison of production costs over relevant ranges of product mixes and output levels, for different possible organisational forms (ie various combinations of numbers and sizes of firms), and

using different combinations of technologies.6

Only a limited number of studies have been undertaken to determine whether the technical conditions for natural monopoly are satisfied. However, many studies have focused on examining the presence of economies of scale or scope in various industries, which may have some relevance to assessing whether natural monopoly

4 Various other situations may also result in the organisation of productive activities within a single firm being more efficient than through a number of firms (see Sharkey 1982, pp. 73-83). 5 The mere observation of such economies does not automatically imply that a natural monopoly exists. Also, an industry may be a natural monopoly despite the absence of economies of scale and scope.

See Baumol, Panzar and Willig (1982), Sharkey (1982), or Ralph (1986) for a detailed technical discussion of sufficient conditions for natural monopoly. 6 While no operationally useful conditions which are both necessary and sufficient for cost sub­ additivity (ie natural monopoly) are known at present, two useful sets of sufficient conditions and a

number of necessary conditions are available (Baumol, Panzar and Willig 1982). Baumol (1982) notes that there has been some excellent empirical work in this area making effective use of the known sufficient conditions for natural monopoly. Although natural monopoly is a global concept, confidence in the conclusions from tests based on these conditions will be greatest where

technological change is slow, market demand varies only within fairly limited bounds, and, in the case of a multi-product natural monopoly, demand substitution effects are limited.

exists.7 Common problems with such studies include the failure :

. to separate the effects of scale economies from technological change; . to take account of the multi-product nature of many industries (thus confusing shifts in demand with economies of scale); and . to recognise that their use of simplified engineering models may not take

account of the complexities of practical situations.

Even when hard data are available, they typically relate to industries enjoying legislated monopoly positions, thereby begging the question of whether the products would be produced differently in a less regulated environment. These studies have often yielded conflicting or inconclusive results, even in industries commonly thought to be natural monopolies (eg telecommunications).

Where studies into industries commonly viewed as natural monopolies have found some evidence of economies of scale, these economies generally relate to only a part of the industry concerned. For example, information on the electricity supply, railways, and telecommunications industries suggests that the scale economy elements apply, at most, to only a part of the market (ie electricity transmission and distribution; railway lines). As noted in the Commission’s rail and electricity studies (see Volume 4) and by several participants to this inquiry (eg Freebairn & Trace), other segments of these industries (ie electricity generation and railway rolling stock ) do not appear to exhibit scale economies.

The conclusion from this discussion is that there are a number of considerations which suggest that natural monopoly conditions are not as widespread as is often claimed. Further, past experience suggests that regulation has often extended beyond any natural monopoly segments.

J4 Natural monopoly arguments for intervention

Industries characterised by natural monopoly are most efficiently served by a single firm. This observation has led some to advocate the establishment of legislative barriers to competition from other firms as a means of ensuring least-cost production. Underlying this argument is a concern that exposing natural monopolies to potential rivals might cause social losses through the unnecessary duplication of infrastructure and/or market instability, so that any competition would be merely wasteful or destructive.

7 Studies of economies of scale in the electricity industry include: Christensen and Greene (1976), Weiss (1975), Joskow and Schmalensee (1986) and Primeaux (1975). Numerous studies have been undertaken with respect to telecommunications, principally relating to North America. Reviews of these studies include Sharkey (1982), Evans and Heckman (1983), Kiss and Lefebvre (1987) and Globerman and Carter (1988). The Institution of Engineers drew the Commission’s attention to a number of relevant references pertaining to the rail industry including Braeutigam (1979), Klein (1973), Boris (1960), Griliches (1972), Keeler (1974) and Harris (1977). These and other more recent studies including Caves, Christensen and Swanson (1981), Friedlander and Spady (1981) and Braeutigam, Daughety and Turnquist (1982) are reviewed in Dodgson (1985). The Commission is not

aware of any detailed Australian studies into the degree of economies of scale or scope applying in industries often considered to be natural monopolies.

82

A second argument for government intervention on natural monopoly grounds relates to the perceived need for some form of intervention - typically public ownership - to prevent exploitation of consumers through the abuse of market power.

‘Wasteful’ or ‘destructive’ competition

If an industry is a natural monopoly, single-firm supply can normally be expected to result from market forces, so that erecting legislative barriers to competition would be unnecessary. However, as pointed out by Telecom and the ACTU Public Sector Working Party at the Draft Report hearings, there can be exceptions. Under certain

circumstances, a natural monopoly may be vulnerable to entry of new producers even if it is producing and pricing efficiently and earning only ‘normal’ profits. Such a situation is referred to as an ‘unsustainable’ natural monopoly. In discussing such a case, Telecom (1986) stated that:

This is a surprising result because it implies that in such circumstances if a community wants the benefits of least cost production the natural monopolist must be protected. This is despite the natural monopolist being, by definition, the least cost producer.

Essentially, an unsustainable natural monopoly arises where it is impossible for the incumbent firm to charge prices which cover costs, without also providing an incentive for other firms to enter the industry.

A number of situations have been identified where an unsustainable natural monopoly may arise. One is where cross-subsidies exist because governments require that goods and services be provided at less than cost to certain groups (eg the disadvantaged or those living in rural areas). However, even in the absence of such

directives (which are not peculiar to natural monopolies), it may still be impossible for a natural monopolist to satisfy market demand without charging prices that result in some users paring more than they might to an alternative supplier. These situations include.

. where production of a single good or service is characterised by economies of scale up to a point, but then by increasing average costs (eg due to congestion), and market demand is at a level beyond the minimum cost point (but less than the level at which two firms could produce at lower cost than one firm);

. where production costs are expected to decline in the future but an existing plant with optimal capacity is expected to quickly become obsolete or below optimal capacity; and

. where two or more products can be produced at least cost by a single firm, but the cost structure and demand conditions are such that a specialised firm can undercut the monopolist on one or more products, thereby rendering other products in the monopolist’s product range unviable. 8

8 A fuller description of the conditions under which these situations may arise is provided in Attachment J1 to this Appendix. See also Panzar and Willig (1977), Baumol, Panzar and Willig (1982) or Sharkey (1982).

83

In each of the above situations, an opportunity may exist for rival firms to enter the market to supply only a subset of consumers at a price which undercuts the incumbent monopolist. However, since there would then be more than one producer in the industry, total market demand would not be satisfied at least cost. As observed by Baumol (1982), an in-principle case for shielding an unsustainable natural monopoly from competition therefore arises because:

...his market will at some time be vulnerable to partial or complete takeover by an entrant who has neither superior skills nor technological superiority and whose entrance increases the quantity of resources used up in production.

This represents the argument that exposing unsustainable natural monopolies to competition may result in waste in the form of uneconomic duplication of infrastructure and the loss of economies of scope, scale, or organisation, thereby creating additional cost burdens on consumers.

An extension of this argument is that competition may be destructive in the sense that it may lead to instability of supply with producers continually entering and leaving the industry, to the long-term detriment of consumers. This scenario has been described by Telecom (1986) as follows:

A natural monopoly subject to erosion of market share loses its capacity to be the least cost producer. In many models, the monopolist may be forced out of the market altogether. However, this does not lead to a new and efficient monopolist. Entrants will be subject to exactly the

same encroachment of markets that the original monopolist faced. Even a new monopolist will lose market share and its least cost capabilities. As a result, markets are left in a state of permanent disarray with no single firm being able to engage in optimal production. With each cycle of entrants investments are repeatedly duplicated generating massive resource wastage.

The validity of the argument for preventing entry to avoid wasteful or destructive competition is discussed below.

Assessment of the ‘wasteful’ or ‘destructive’ competition argument

While legislative barriers to entry in industries suspected of being natural monopolies are alleged to avoid wasteful or destructive competition, a number of considerations suggest that, first, the likelihood and extent of any such benefits may have been overstated, and second, that the adverse side-effects of measures designed to achieve them may exceed the expected benefits.

One reason why the benefits from erecting legislative barriers to entry may have been overstated is that wasteful or destructive entry is only possible where an unsustainable natural monopoly exists. Given the low incidence of natural monopoly generally and also the special conditions required for unsustainability, this seems unlikely to be widespread in practice. Those arguing for protection on unsustainability grounds seldom attempt to establish that these special conditions actually apply in their case. Further, unsustainability due to technical factors associated with natural monopoly is

often blurred with unsustainability due to fulfilling CSOs. Arguments for restricting competition to allow the delivery of CSOs through uniform pricing arrangements^ should be clearly distinguished from arguments for protecting a natural monopolist from wasteful or destructive competition in circumstances (such as those outlined in Attachment J l) where cross-subsidisation is unavoidable if the monopolist is to cover

costs.

The potential for wasteful or destructive competition depends not only on the cost characteristics of a natural monopoly industry being such that it is potentially unsustainable, but also that this in turn induces new firms into the industry. A number of considerations suggest that inefficient entry of new firms is unlikely to

occur in practice.

To start with, the likelihood of wasteful entry is low where significant sunk costs are involved. When entry requires significant investments which cannot be fully recouped on exit (eg specialised plant and equipment, product promotion, training of staff and negotiation of contracts), a firm will only be prepared to incur them if it expects to

survive at least long enough for these sunk costs to be recovered. If, as assumed in the wasteful competition argument, new entrants possess the same technology as the incumbent, there would seem to be no reason for a new entrant to believe that it would replace the incumbent or be able to survive competition from the entry of new

firms. Moreover, an existing supplier faced with the threat of entry seems more likely to reduce prices in order to maintain sales volume than to maintain prices - as is usually assumed in the unsustainability argument - which would make it harder for a new entrant to survive (see Hartley & Trengove 1987; Brock & Evans 1983).

Thus, consideration of the existence of sunk costs as well as the likely strategic behaviour of both existing and potential suppliers, suggest that a firm would only enter the industry if it possessed a medium to long-term cost advantage relative to the incumbent (eg because it possessed a more efficient technology or because the

incumbent was inefficient), rather than envisioning a ‘hit and run’ operation. If entry occurs because the newcomer is more efficient than the incumbent, this competition is desirable because of the resulting overall gains to the economy, rather than being

wasteful or destructive.

In relation to the destructive competition aspect of the argument, the scenario of firms continuously entering and exiting and swapping clientele seems unlikely in most conceivable cases. A possible market solution would be the use of long-term supply contracts, either as a defensive measure by the incumbent, or pre-contracting by new

entrants as a means of ensuring their ability to survive. In both cases, the likelihood of the market becoming unstable would be reduced.

Taken together, these considerations suggest that, even when the cost conditions of a natural monopoly render it potentially unsustainable, in practice purely wasteful or destructive competition is unlikely to arise. When inefficient entry does occur, it is unlikely to result in a large loss to society, since it is unlikely to continue for long

periods and is unlikely to involve significant sunk costs. 9

9 As discussed in Appendix I, the delivery of CSOs does not necessitate restrictions on competition, although it may be the most cost-effective method in some cases.

A second set of considerations relevant to this argument for barriers to entry relate to the issue of whether any benefits from avoiding wasteful or destructive competition are outweighed by costs arising from erecting barriers to entry. The efficiency- promoting capacity of exposure to competition relies in part on the entry and exit of rival firms, or at least the potential for this to occur. Indeed, this is a natural part of the competitive process whereby efficient suppliers retain their position in the market by adapting to change and meeting users’ requirements, while inefficient firms lose market share and possibly go out of business. In this sense, some element of instability is present in almost every market, reflecting their dynamic nature, and the imposition of restrictions on entry will in effect limit access by users to lower cost or better product mixes. To limit or preclude such a process is to adopt an essentially static view of what is in reality a dynamic process, and to risk forgoing the benefits from actual or potential competition.

A natural monopoly only provides the potential for single-firm production to be the least-cost way of satisfying market demand. The inherent advantages of single-firm production may be dissipated where legislative barriers prevent potential competitors contesting the market, thus blunting the incentives for a monopolist, public or private, to produce and/or price efficiently. The resulting inefficiencies, of the type documented in Chapter 4, may outweigh those associated with multi-firm production which the imposition of entry barriers seeks to avoid.

Further, legislative barriers to entry might impede the introduction of more efficient technologies which could transform an industry so that it ceases to be a natural monopoly. For example, some have argued that recent technological developments in telecommunications (such as microwave links, cellular phones, and PABXs) have substantially lessened the scope of natural monopoly characteristics in that industry. At the same time, however, others (eg Telecom, ACTU Public Sector Working Party) have argued that the natural monopoly characteristics applying within the basic network have been strengthened by other technological developments (eg fibre optic cables).

Both these claims highlight the need, in the presence of market restrictions, for the natural monopoly status of industries to be periodically reviewed in the light of changing technologies. Although some governments have recognised this need (eg the Commonwealth in respect of telecommunications) such a review process entails the danger that the protected industry will inevitably lag behind technological developments. The granting of an exclusive right to serve the market may even provide the monopolist with an incentive to retain existing technologies and actively oppose new technologies which may erode the stated natural monopoly justification.10 Even in industries where changes in technology seem limited and the

natural monopoly appears unlikely to be challenged, it does not follow that legislative

10 Some have argued that the existence of monopoly encourages rather than discourages the introduction of new technologies because the financial rewards of doing so are more likely to be retained. Others see the role of competition and the threat of new entry as providing the major stimulus to invention and innovation. While the theoretical and empirical evidence is somewhat mixed, an overall assessment (see Scherer 1980, p. 438) suggests that ‘what is needed for rapid technical progress is a subtle blend of competition and monopoly, with more emphasis on the former than the latter, and with the role of monopolistic elements diminishing when rich technological opportunities exist’.

86

barriers to entry should be imposed; they could serve merely to remove disciplines on performance.

Barriers to competition may not only stifle efficiency and innovation in the delivery of goods and services whose production is characterised by natural monopoly; their effects may carry over to other parts of the industry. Once a monopoly has been sanctioned on natural monopoly grounds, legislative barriers to competition may be

extended beyond the natural monopoly core. This may arise either through the ‘empire-building’ ambitions of the monopolist or through the failure of regulatory authorities to review the ways in which the natural monopoly status of an industry (or

parts of it) have been affected by technological change (see section J3). In addition, a monopolist may be able to exploit exclusive control over designated activities enjoying statutory monopoly status to undermine competition in other parts of the industry (eg through predatory pricing). For example, the Australian Telecommunications Users

Group submitted that Telecom’s monopoly power and financial dominance may have distorting effects on the local production of cable. The efficiency gains (if any) attributable to shielding any natural monopoly component from wasteful or destructive competition are likely to be outweighed by efficiency losses from

proscribing competition in related markets. This is particularly likely to be the case where these latter markets are more important, in terms of their impact on the competitiveness of industry, than those activities considered to be natural monopolies.

The principal conclusion from this discussion is that legislative barriers to entry are generally redundant (ie natural monopoly is generally sustainable without them). In the exceptional circumstances where it can be argued that a natural monopoly may not be sustainable in the face of competition, the potential for waste due to inefficient

entry needs to be balanced against the adverse effects on production efficiency of restricting competition. There is, therefore, a strong case for removing any legislative barriers to entry so that, regardless of whether the incumbent monopolist is public or private, it faces continuing external pressures to produce and price efficiently.

Market power

A separate, but related, argument is that some form of government intervention is desirable in industries characterised by natural monopoly to prevent or lessen abuses of market power. Market power could be exercised in a number of ways such as by increasing prices, restricting output, reducing quality of service, efficiency may also

decline (eg through cost-padding) relative to what would occur if the monopolist faced the threat of competition. The potential for such abuses stems from the market power which may occur either as the inevitable result of market forces (where the monopoly is sustainable but not readily contestable) or through the deliberate barring

of entry to prevent wasteful or destructive competition where the natural monopoly is considered to be unsustainable. In such cases, governments may seek to contain such abuses through price or rate of return controls on a private monopolist, or by resorting to public provision of the good or service concerned. In Australia, public

ownership has often been seen as the most effective form of intervention.

Assessment of the market power argument

In assessing the need for government intervention to control possible abuses of market power, consideration needs to be given to the question of whether the likely

87

benefits will outweigh the costs. Intervention is not costless, and some degree of market power abuse may need to be tolerated because the cost of devoting resources in an attempt to control it would outweigh the benefits.

What is required, therefore, is a case-by-case assessment of how serious the potential monopoly abuses are likely to be. Two factors that have an important bearing on such an assessment are the existence of substitutes and the existence of actual or potential competitors in the supply of the good or service in question.

The existence of substitutes in many cases serves to constrain the extent to which a natural monopolist can exercise market power. For example, a railway monopoly is likely to find that for many freight tasks, raising prices (in the absence of regulations precluding the use of alternative modes of transport) causes some customers to turn

to alternatives (eg road) to service their needs.

The exploitation of market power by a natural monopolist may also be limited by the existence of (or potential for) direct competition. As noted previously, the threat of competition imposes a discipline on an incumbent monopolist to price and produce efficiently.

In examining the issue of market power, therefore, the important question is not so much whether a natural monopoly exists, but whether the market is contestable (ie entry and exit by rival firms, including those providing substitute products, is feasible), since this may be sufficient to limit the extent to which market power is abused.

Under certain circumstances, however, removal of legislative barriers to entry will not be sufficient to ensure that a market is contestable. In particular, the existence of unavoidable sunk, as opposed to fixed, costs 11 provides a natural monopolist with some leeway before departures from efficient pricing and production practices will

attract rival firms into the industry. Where there are reasons to believe that this leeway may be significant (eg where sunk costs provide a barrier to entry) the question becomes one of how the industry should be organised to ensure, as far as possible, that market demand is satisfied at least cost (ie that the scope for production

inefficiencies and/or monopoly pricing is minimised). As discussed further in section J5, direct public provision is only one of several possible responses.

Overall assessment of natural monopoly arguments for intervention

The preceding discussion indicates that the mere existence of natural monopoly characteristics in an industry does not necessarily justify government intervention. 1 1

11 A fixed cost is one that must be incurred in order for a firm to produce any output at all and does not vary with the level of production. A sunk cost is one that, once incurred, cannot be recouped. A cost may be fixed but not sunk (eg the purchase of a capital item for which an active secondhand market exists). While in practice there may be a substantial overlap between fixed and sunk costs (particularly in natural monopoly industries where items of capital may have no alternative uses or potential buyers), the distinction is an important one. Only if a cost is sunk, rather than fixed, does it represent a barrier to entry.

88

Rather, the appropriate policy response depends on two issues:

. Is the natural monopoly likely to be sustainable (ie can an efficient incumbent set prices which cover costs without attracting entry) and therefore how likely is wasteful or destructive competition?

. Is the natural monopoly readily contestable (ie is entry by rival firms, including those providing substitute products, feasible), or does the existence of sunk costs or other barriers to entry (including legislative ones) provide scope for the abuse of market power?

Figure J 1 draws together the main conclusions from the preceding discussion as they apply to each of four possible natural monopoly situations. The four situations represent polar extremes of what might occur, but nevertheless provide a useful reference for examining the policy implications of natural monopoly.

Figure J1: Policy implications of alternative natural monopoly situations

Sustainable natural monopoly

Potentially unsustainable natural monopoly

Market readily contestable

No case for intervention. Potential competition sufficient to ensure emergence of a single

supplier constrained to produce and price efficiently.

Wasteful or destructive competition possible, but costs of restricting entry likely to outweigh any benefits. Controls on market power unnecessary unless legislative barriers to

entry are erected.

Market not readily contestable

No case for legislative barriers to entry to prevent wasteful or destructive competition, but there may

be a case for intervention to control market power.

Wasteful or destructive competition possible but highly unlikely since new entry not normally feasible.

Thus legislative barriers generally redundant, but there may be a case for intervention to control market power.

Examination of Figure J1 suggests that the only situation where erecting barriers to entry may be warranted on natural monopoly grounds is where the natural monopoly is both potentially unsustainable and the market readily contestable. However, the foregoing discussion suggests that such a situation is unlikely and that the potential

costs associated with the removal of market disciplines are likely to outweigh any benefits from avoiding wasteful or destructive competition by preventing new entry. In general, therefore, there are good grounds for removing legislative restrictions to competition. As noted by Baumol, Panzar and Willig (1982, p. 222), who first

89

formalised the theory of unsustainable natural monopoly, there are:

...important reasons for denying emphatically that our analysis is intended to support restrictions on freedom of entry - a conclusion that is virtually the opposite of the truth.

If governments nevertheless consider that unfettered competition may be wasteful or destructive because an unsustainable natural monopoly may exist, they may judge it appropriate to retain legislative barriers to entry. Alternatively, competition may not be feasible because there are natural barriers to entry such as high sunk costs (see lower half of Figure J 1). In both these situations, there may be a case for intervention to control the abuse of market power. A number of alternative means of intervention (of which exclusive public provision is only one) aimed at avoiding wasteful or destructive competition and/or limiting the abuse of market power in natural monopoly industries are discussed in the following section.

J5 Alternative arrangements for handling natural monopoly

Whilst the mere existence of natural monopoly characteristics in an industry does not necessarily justify government intervention, the discussion in the previous sections suggests that there may be some limiting circumstances in which governments may consider such intervention to be warranted. The issue then becomes one of organising the industry so as to control the abuse of market power while retaining the

benefits of single-firm supply.

An examination of the policies applying to industries commonly claimed to be natural monopolies indicates a variety of responses. In Australia, many industries commonly considered to have elements of natural monopoly are public monopolies. In contrast, the USA has tended to rely on regulation of private monopolists through price or rate of return controls. Recent developments in the UK, USA New Zealand and Australia point to the application of approaches which promote competition in areas where it may naturally occur (eg contracting out) or where it would otherwise be

distorted in the face of monopoly in related areas (eg common carrier provisions and segmentation).

How can the relative merits of these alternative approaches be evaluated? A relevant benchmark is provided in Figure Jl. This figure shows that when a natural monopoly is both contestable and sustainable, the dual problem of realising the benefits of single-firm production (ie productive efficiency) and controlling abuse of market power (ie pricing efficiency) are simultaneously solved.12 Since this means that goods or services are likely to be provided to users at least cost, alternative forms of

intervention can be assessed according to the extent to which they reproduce this result.

As part of this evaluation, account needs to be taken of any direct or indirect costs of regulation aimed at promoting productive and pricing efficiency. Further, in evaluating alternatives, trade-offs may have to be made. For example, some approaches may provide strong incentives for productive efficiency but weak incentives for pricing efficiency, or vice-versa. In the final analysis, the preferred

12 These concepts of efficiency are explored further in Chapter 4.

90

approach will be that which incorporates incentives which, on balance, are considered most likely to enhance overall economic efficiency.

Whilst such an assessment is likely to vary according to the circumstances of particular industries, the following discussion presents some general considerations associated with three broad (but not necessarily mutually exclusive) policy approaches, namely:

. public ownership and operation; . promoting competition; and . regulated private supply.13

Public ownership and operation

Supporters of this approach argue that where markets are not contestable, public ownership and operation of natural monopolies is the most effective way of securing the benefits of single-firm production, without suffering the adverse effects of monopolistic practices. Essentially, this reflects a belief that it is easier to design and

apply incentives for a public monopoly to induce efficient pricing and production practices than to achieve them through regulation of a private monopoly.

The view that it is easier to prevent abuse of market power by a public enterprise than a private one has been supported on two main grounds. First, public enterprises are seen as more likely to act in the public interest by restraining prices to reflect the costs of supply, thereby sharing the cost savings arising from single-firm production with users. In contrast, a private monopolist, motivated by the pursuit of profits, may

be more inclined to charge what the market will bear. Second, it is sometimes suggested that it is easier for regulators to obtain the information necessary to regulate prices and production from public than private enterprises.

Both contentions are debatable. On the first point, experience overseas suggests that the concept of public enterprise managers as the custodians of the public interest is less appealing in practice than in theory. As noted by Scherer (1980, p.488):

Merely telling public enterprise managers to act in the public interest is distinctly nonoperational, for such a rule is so vague it could have a dozen plausible interpretations in any given situation. To ensure that the ‘correct’ interpretation is chosen and to prevent line managers from

advancing their own selfish interests under the discretion a broad criterion allows, countless decisions would have to be passed up the ladder for review and resolution. But this can overload management, causing decision-making breakdowns.

The second point raises an intractable regulatory issue. Private monopolies have a profit incentive to conceal information needed for regulation (eg accurate cost and demand data). Public monopolies may also have an incentive to conceal cost information in order to distribute monopoly profits within the firm in the form of

overmanning, higher salaries and wages, and overinvestment in new technology, or in order to maintain inefficient work and management practices. Indeed, as noted by

13 Further discussion of these broad approaches may be found in Waterson (1988) and Jarden Morgan NZ Limited (1988).

91

Pera (1989), regulatory bodies often have little power so long as information, technical resources, and political influence are concentrated in the public monopoly.

Even if it is presumed that altruism and government demands for information do result in allocative gains through the avoidance of monopoly pricing, the case for public ownership depends on there being appropriate incentives for public enterprises to produce efficiently. Otherwise, any potential gains from regulating prices and/or

returns may be dissipated in cost padding, over-manning etc (see Chapter 4). The very factor that diminishes the likelihood of public enterprises deliberately exploiting users by raising prices (ie the suppression of the profit motive) might also undermine the incentives for a public enterprise to minimise production costs. Moreover, legislative protection of the monopoly means that it is not possible to test whether it is a natural one, and if so whether it is achieving least cost methods of production.

The above considerations suggest that although public ownership may avoid the worst excesses of unconstrained private monopoly pricing, it may be more difficult to encourage production efficiency than is the case with a private firm. Indeed, where the costs of such productive inefficiencies are passed on to users, final prices may not differ substantially from those which a private monopolist might have charged.

Although the current incentive structures for public firms in natural monopoly situations have a number of drawbacks, there would seem to be considerable scope for improving the incentives for these enterprises to adopt efficient pricing and production practices. While it is difficult to design effective incentives to reproduce market disciplines, the discussion in Chapters 6 and 7 suggests that significant improvements could be realised by the better specification of objectives, diminished reliance on direct operational controls, improved performance monitoring, and modified intergovernmental relationships. Specific initiatives could include more efficient means of delivering CSOs (see Appendix I); setting effective performance targets (see Appendix L); and the adoption of better asset valuation systems (see Appendix M). With regard to controlling possible monopoly power abuses, alternative means include reliance on general pro-competitive measures (eg by removing existing TPC exemptions) or the establishment of specialist agencies (eg AUSTEL) to oversee prices (eg via CPI-X controls). Indeed, the Commonwealth and some State governments have introduced or foreshadowed a number of changes along these lines.

The principal difficulty with this approach is that it relies on administratively based mechanisms as a substitute for market forces in encouraging the public monopolist to service market demand at least cost. There are, however, alternative approaches which retain the benefits of single-firm production, yet harness the beneficial effects

of competition. Some of these approaches are discussed below.

Mechanisms to promote competition in natural monopoly industries

Under this approach the emphasis shifts from administratively based incentives as the means of attempting to facilitate the efficient delivery of goods and services to the promotion of contestability by removing or modifying legislative or natural barriers to entry. Promoting competition does not necessarily imply complete deregulation and unfettered competition. Rather, some regulation aimed at supporting the emergence of effective competition in at least part of the industry may still be required. Three

mechanisms which have been used in Australia and overseas to promote competition are contracting out, common carrier provisions and segmentation.

Contracting out

This approach involves the use of public and/or private suppliers to deliver goods or services to, or on behalf of governments, the contract winner normally being selected on the basis of competitive bidding. After expiry of a specified period, the contract again comes up for review.

Under this approach (discussed more generally in Appendix K), many of the benefits of competition are retained. Competition within a market is effectively replaced by competition for the right to service the market or a part of it, whilst the benefits of single-firm supply are also retained. In principle, competitive bidding for contracts

would ensure that opportunities for providing goods and services went to the most efficient operator and that operators were encouraged to maintain and improve their efficiency, or else risk losing their contract.14

The extent to which these benefits are passed on to users would be largely influenced by the particular form of contract and bidding process used. For example, if an exclusive franchise were awarded to the bidder prepared to pay the highest lump sum to government in order to gain the right of supply, any gains in productive efficiency

may be dissipated by losses in pricing efficiency as the successful bidder could be expected to raise prices in order to make a return on its investment. One way of addressing such pricing inefficiencies would be to award the contract to the supplier

who offers the lowest acceptable supply price. This approach, however, involves monitoring and information gathering costs which would reduce the available gains.

Regardless of the particular type of contract or bidding system adopted, a number of other factors may influence the degree to which the potential efficiencies from contracting out are realised in practice. For example, in some cases it may be difficult to write effective contracts because the desired level and/or quality of output cannot be precisely defined, or because market and technological uncertainties require

allowance for adjustments to contract conditions.

Another potential difficulty is that the letting of a contract for only a specified period may reduce the incentives for the successful bidder to invest and maintain capital in a manner consistent with efficiency over the longer term. Ownership of infrastructure by tenderers may also present problems in specifying tenders that do not unfairly

favour the incumbent relative to outsiders, and may also lead to difficulties associated with the management or transfer of assets.

These latter considerations suggest that in relation to existing public enterprises characterised by natural monopoly, the main scope for contracting out may apply to ‘operating contracts’. Under such arrangements, public ownership need not imply public provision of the good or service if a private contractor is more efficient in

14 As noted in Appendix K, empirical evidence in Australia and overseas indicates significant savings have occurred when certain activities previously provided publicly have been put out for tender. Moreover, savings have been realised even where the successful tenderer has been the previous public provider, suggesting that exposure to competition is an important factor in realising efficiency

gains.

93

operating the infrastructure. While separating responsiblity for investment and operating decisions may make coordination more difficult, there may be scope for more widespread adoption of this approach.

Common carrier provisions/joint ownership

This approach involves the sharing of infrastructure characterised by natural monopoly among independent firms under appropriate access arrangements. For example, under a common carrier approach sole ownership of the infrastructure

considered to be a natural monopoly (eg the basic telephone network or rail track network) remains, with other operators being afforded access on the same basis as the owner.15 Generally, common carrier provisions impose a legal obligation on the owner of the network to provide services to allcomers at a ‘reasonable’ price.

Depending on the technological characteristics of an industry, a common carrier arrangement could extend in some cases to other operators utilising the network (eg operating their own rolling stock on a railway line). In cases where the natural monopoly arises due to purely technical factors (ie economies of scale or scope),

rather than economies of organisation, it is conceivable that the associated commercial arrangements could extend to joint ownership, or separate ownership of different portions of the network by several independent firms, with appropriate coordinating mechanisms.

The rationale behind a common carrier approach is that single-firm ownership and/or operation of the means of production may not be necessary for an industry to realise economies from a technology characterised by natural monopoly.16 By imposing a common carrier requirement, uneconomic duplication of the basic infrastructure is avoided while at the same time promoting competition - not in the provision of the infrastructure itself, but in its operation and/or in other contestable areas of the industry. For example, as observed in the Commission’s electricity study (see Volume 4), if enterprises involved in electricity transmission were required by law to operate

as common carriers, generating firms could compete for the right to supply various load centres and large industrial users. Such an arrangement has the advantages of limiting the abuse of market power in components of the industry that may be natural monopolies (eg transmission) and facilitating entry into industry segments which appear to be potentially contestable (eg generation).

The extent to which these benefits can be realised, however, depends to a large degree on effective regulation being applied to the natural monopoly component of the industry, so that the owner of the infrastructure does not enjoy an unfair advantage over potential competitors in the provision of services using this infrastructure. For example, the Commonwealth Government has recognised in

setting up AUSTEL that Telecom and OTC could misuse their market power (see

15 A common carrier requirement could apply to a private as well as to a public operator. For example, the Public Utility Regulation Practices Act (PURPA) in the USA requires that power utilities (which are predominantly privately owned) make available their transmission grids to owners of co-generation facilities and to independent power producers (see electricity study, Volume 4). Similarly, private firms in the USA natural gas industry are encouraged to operate as common carriers in return for less regulation. 16 For example, the Australian Telecommunications Users Group submitted that experience in the

USA, Japan, and the UK demonstrated that there is no technological reason why the owners of telecommunications networks need to have a monopoly of messages carried on the network.

94

Minister for Transport and Communications 1988). They could do this, for example, by using profits realised on sales in the regulated market to engage in predatory pricing in the unregulated part of the market, or by favouring attachments or services which they market in the unregulated areas of telecommunications by controlling

complex technical characteristics of the public network infrastructure or through access to privileged information such as customer requirements and traffic patterns.

This suggests that an essential feature of a common carrier approach is the separation of service provision and regulatory functions currently performed by some public utilities. The aim would be to ensure a ‘level playing field’ in the areas which, through common carrier provisions, are being opened up to potential competition. For

example, the recent transfer of regulatory functions previously performed by Telecom to AUSTEL represents an important plank of the new telecommunications industry framework which seeks to increase competition in areas outside the designated natural monopoly component (eg valued-added services and customer premises

equipment). However, some participants in this inquiry (eg Link

Telecommunications) felt that AUSTEL’s present charter placed insufficient emphasis on promoting competition.

One question that arises is whether regulation designed to ensure that all users have equal access to the basic infrastructure is best achieved by reliance on general pro- competitive measures (eg TPC or PSA oversight) or by specialist industry agencies such as AUSTEL. Where the technical characteristics of an industry are complex or the industry is seen as ‘strategically’ important, there may be a case to use industry-

specific regulatory bodies. However, in most cases it would seem appropriate to rely on more general measures, unless these subsequently prove to be inadequate. Regulatory bodies with economy-wide responsibilities have the advantage that they are less likely to be ‘captured’ by the enterprises they regulate.

Another aspect of regulation that may be required under a common carrier approach is the specification and monitoring of technical and safety standards to ensure that access to a common facility by multiple users does not lead to undue safety or reliability problems. For example, one of AUSTEL’s functions is to ‘set standards where different carriers and service providers interconnect, to avoid harm to the

network and to protect the safety of users and those engaged in maintenance and operation of equipment’ (Minister for Transport and Communications 1988, p. 127) However, as noted in the Commission’s electricity study (Volume 4), reliability may not be assigned the same high priority in a more competitive environment as it is by

existing public suppliers. Such factors could be incorporated into contracts between suppliers and users.

The additional costs of regulatory bodies required to oversee potentially complex access arrangements need to be taken into account when assessing the overall benefits of a common carrier approach. In its submission to the Draft Report hearings, Telecom noted that the reported cost of administering and regulating such

arrangements in the telecommunications industry in the USA amounted to some US $1.5 billion a year. However, such costs should be viewed in context: they may well be small relative to the gains in productive and pricing efficiency engendered by the common carrier requirements. Moreover, other approaches (including public

ownership and operation) also entail administrative costs, although they are often less transparent.

95

Provided the problems of regulating the behaviour of the owners of the natural monopoly infrastructure can be overcome, the common carrier approach appears to offer considerable scope for encouraging efficiency in industries which may have some elements of natural monopoly by allowing competition in other areas. It has been applied quite widely overseas and there would seem to be considerable scope for applying it more extensively in Australia. Potential candidates nominated by various inquiry participants include electricity (suggested by PPMFA, CRA, ACM), rail transport (CRA, Freebairn & Trace), and natural gas (Esso) industries.

Segmentation

The aim of this approach is to promote competition in the industry as a whole, even where competition cannot be introduced into the segment seen as the natural monopoly core. It involves reducing or removing the opportunity for a natural monopolist to prevent or distort competition in related markets, by excluding it from

operating in the potentially competitive segments of its industry. In practice, implementation of this approach could involve legislative prohibitions or the division of an existing vertically-integrated enterprise into distinct units operating at arms- length.

The main advantage of this approach is the potential for gains in pricing and productive efficiency in industry segments outside the designated natural monopoly, induced by promoting competition in these areas. In respect of the protected natural monopoly segment, however, the same difficulties in promoting efficiency by either a

public or a private monopoly would apply.

While the need to regulate the behaviour of the monopolist in respect of the natural monopoly component of the industry would remain, a major advantage of this approach is that the need to monitor its conduct in the potentially competitive areas is removed. This substantially simplifies the regulatory task and may avoid many of the monitoring, detection and enforcement problems associated with a ‘conduct regulation’ approach such as that embodied in common carrier provisions.

To be weighed against these benefits, however, are any economies of scope, scale or organisation forgone by not allowing one firm to operate in both natural monopoly and other areas of the industry. The loss of such economies must be set against any gains from enhanced competition in contestable areas from which the monopolist is excluded. For example, while segmentation of the electricity industry may lead to increased competition in segments which may be potentially contestable (eg generation), it may also engender problems associated with the complexities and costs of devising and enforcing contracts between firms to coordinate production decisions to meet the demands of users. Regulation of technical and safety standards may also be required, but any additional costs arising in these circumstances may not be greater than those already incurred for these purposes.

One example of the implementation of the segmentation approach was the 1982 anti­ trust case in the USA, whereby AT&T was required to divest itself of its local operating companies. Implicit in this decision was a judgment that the loss of any economies of scope from operating local telephone networks in conjunction with long-distance connections was likely to be outweighed by the gains from the stimulus

96

to competition in local network services (beyond that which could be achieved through conduct regulation). Alternative judgments on this issue, however, are evident from the telecommunications policies adopted in the UK and Australia, which have relied on conduct regulation. The ACTU Public Sector Working Party went so

far as to suggest that:

Structural separation which formed the dominant regulatory tool in the United States in the last decade is now discredited and has been acknowledged as a source of poor American performance in services such as universally available packet switched services and distributed

data base services such as Videotex.

Other participants to this inquiry felt that segmenting operations of existing public enterprises would bring considerable benefits. For example, the Local Government Electricity Association of Victoria recommended th a t:

For the benefit of the industry and in the interests of the public the costs associated with the generation and transmission should be disclosed. This can best be achieved by separating the operations of distribution from the generation and transmission functions. Each of

these separate functions should be subject to public scrutiny to allow the customer to consider alternatives which may prove to be more effective and of lesser cost.

While the segmentation approach has been adopted in a number of industries overseas (eg the electricity industry in New Zealand), its application in Australia to date appears limited, although some examples do exist (eg the separation of responsibility for generation and transmission from retailing of electricity in NSW and

Queensland - see electricity study, Volume 4). A number of participants to this inquiry suggested that gains from increased competition would be realised if segmentation were extended in this industry (eg CRA suggested separating the transmission and generating functions) or applied in other industries commonly

assosciated with natural monopoly.

As with other approaches, this way of dealing with natural monopoly involves trade­ offs and requires a case-by-case assessment to evaluate the potential benefits and costs.

Regulated private supply

Another alternative to public ownership and operation of natural monopoly is to allow a private monopolist to supply the market. When there are concerns that this approach may result in the exploitation of consumers through higher prices, restricted output, or lower quality of service, it may be appropriate to impose some form of regulation aimed at modifying the monopolist’s behaviour so that it approximates that which would apply if the market was contestable. This policy approach to dealing with natural monopoly is less common in Australia than in some other countries (eg

USA). However, it is pertinent to note that in some States, the gas transmission and reticulation networks are privately owned and operated, while being subject to differing degrees of regulation, in preference to public ownership adopted in others.

97

Supporters of this approach see the major gains as deriving from the greater incentives for efficiency in production applying to private enterprises - the profit motive, the disciplines of sharemarkets and capital markets and the threats of takeover and insolvency - which do not apply in the case of public enterprises (see Appendix H). In addition, the scope for political interference in managerial decision­ making in private firms is limited compared with public enterprises. Since, on balance, the incentives for productive efficiency for private firms are more effective than the administratively-based incentives applying to public enterprises, private supply offers more chance of the cost economies of single-firm production in natural monopoly industries actually being realised.

As noted previously, however, a private monopolist also has an incentive to restrict output and to increase profits by raising prices, which may limit the extent to which the benefits of single-firm supply are passed on to consumers. These concerns could be addressed by invoking general behaviour monitoring measures such as TPC or PSA oversight or industry-specific price or rate of return controls aimed at ensuring that these benefits are passed on to consumers.

Unfortunately, regulations aimed at controlling market power tend to have direct and indirect costs of their own, particularly where they are poorly conceived or implemented. Empirical evidence (see Appendix H), suggests that such regulations seriously blunt the incentives for productive efficiency associated with private ownership. For example, where private firms are constrained to a maximum allowable rate of return on capital, decisions relating to the use of inputs are distorted, thus undermining productive efficiency (see Appendix L). In recent years, some electricity authorities subject to such regulations in the USA have experienced difficulty in raising necessary funds to finance new generation capacity (see electricity study in Volume 4).

Further, such constraints on profitability may dull the incentives associated with the sharemarket and capital markets. In some cases governments have acted to remove these disciplines entirely. For example, in some States (eg Qld, NSW, and until recently, SA) the takeover of privately owned and operated gas reticulation utilities has been prohibited.

Other forms of controlling market power by regulating prices directly may avoid some of these adverse side-effects, but also have significant deficiencies. For example, limiting price increases to the rate of inflation may provide little discipline on efficiency and engender a cost-plus environment. This has generated interest in alternative approaches, such as PI-X controls developed in the UK, whereby prices are allowed to increase each year by the movement in a relevant price index less an amount X which represents productivity gains. By allowing the regulated firm to take as profits any productivity increases in excess of X, incentives for productive efficiency are retained whilst ensuring that at least part of these benefits is passed on to users. However, this method of controlling market power also has a number of drawbacks (some of which are discussed in Appendix L, section LA). In common with other forms of regulation, these include the difficulty in obtaining the required information which private firms may have a strong incentive to conceal.

The benefits from this approach will depend heavily on the ability to design regulations which are effective in controlling market power yet do not seriously

98

impair sharemarket and capital market disciplines on private firms for productive efficiency. Again, this suggests that trade-offs are involved and that a case-by-case assessment is required.

J6 Assessment

In general, provided it is efficient, a natural monopoly can be expected to be viable in a market which is open to entry by rivals. It is possible, in some special circumstances, that exposing unsustainable natural monopolies to competition may lead to a degree of instability and inefficiency. However, this possibility does not justify the

widespread protection of industries thought to be natural monopolies. Unless Wasteful’ or ‘destructive’ competition can be clearly demonstrated to be a potentially significant problem, it is unlikely that entry which is inefficient could be prevented in advance without simultaneously eliminating the discipline which potential entry

provides and impeding beneficial and innovative competition.

Those proposing the creation or maintenance of barriers to entry would therefore seem to need to justify them, particularly in industries where the costs of such barriers are potentially high (eg those subject to rapid technological change). In the Commission’s view, it is preferable to permit the market to test for the existence of

natural monopoly than to pre-judge that it exists and preclude competition through legislation or regulation.

Where issues such as technical standards, product safety standards and reliability arise in the context of specific goods and services, the design of general regulations applicable to all firms to secure these standards should be examined on their merits. These non-price attributes of goods and services do not of themselves necessitate

public provision of the good or service.

If efficiency or market power arguments are nevertheless accepted as rationales for barriers to entry and/or public ownership, this should be clearly stated and subject to periodic review. When sanctioning the continuance or establishment of public monopolies, governments in Australia appear rarely to have critically examined

whether a natural monopoly actually exists or its possible scope, let alone whether it is likely to be sustainable or the market contestable. Moreover, with some exceptions, there seems to have been little preparedness to re-examine the natural monopoly status of industries over time. Vaguely-specified natural monopoly arguments have at

times clouded assessments of the efficacy of government interventions designed to further other goals (eg CSOs).

Even if there is sufficient evidence to suggest that an unsustainable natural monopoly exists, and that it is likely to be subject to Wasteful’ or ‘destructive’ competition, or the market is not readily contestable, there are a number of alternative policy options. Since a protected monopoly can be expected to exhibit inefficiencies in its pricing and

organisation, the impact of alternative approaches and their effects merit close scrutiny. The choice is not simply one between direct public and regulated private provision.

There are various ways in which competition can be promoted without destroying the benefits of single-firm supply in industries considered to be natural monopolies. These mechanisms include contracting out, common carrier legislation, and

99

segmentation of operations. The discussion of these mechanisms suggests that the identification of potential benefits and costs from applying them requires a detailed analysis of the industry in question and the manner in which these benefits and costs are affected under different institutional arrangements. The discussion also suggests that there are likely to be net gains from modifying the arrangements typically applying to relevant industries in Australia to allow a greater role for competition.

100

ATTACHMENT Jl: SITUATIONS WHERE UNSUSTAINABLE NATURAL MONOPOLIES MAY ARISE

A natural monopoly which faces the threat of ‘wasteful’ or ‘destructive’ competition is said to be unsustainable. This Attachment outlines the theoretical cases where such a situation may arise. For a more technical discussion of the conditions required for unsustainability see Panzar and Willig (1977), Baumol, Panzar and Willig (1982) or

Sharkey (1982).

The increasing costs problem

One case where an unsustainable natural monopoly may arise is where unit costs initially decrease with output, but then increase. An example relating to water provision is presented in Ralph (1986), based on Faulhaber (1975).

Three towns each face a stand alone cost 1 of $30 million for water provision, which would require them to spend $90 million in total. By sharing facilities, two towns could reduce the cost to $40 million ($20 million each), reducing the total spending to $70 million. Supplying all three towns together would reduce the cost to $66 million

($22 million each), making such provision the optimal solution for the community as a whole.

However, no prices exist which would allow a single system to cover costs and not involve cross-subsidies. At least one town will always be paying more than $20 million, and hence able to benefit by forming a coalition with one of the other towns, offering its partner a slightly better deal. The excluded town, faced with the $30

million cost of building its own water supply system, will then have the same incentive to induce one of the other towns to form a coalition with it, excluding the other, and so on. There will always be a town with an interest in forming a new coalition which excludes one of the towns.

Another example relates to a natural monopoly industry where single firm production will result in the lowest costs, yet demand exceeds the output level at which costs would be minimised for the most efficient plant. If the firm produced at the output level which minimises costs (leaving some demand unsatisfied) it could sell at prices

which no rival firm could better. However, as no basis emerges for customers to agree who misses out, the market would be unstable. Competing firms could offer to supply a different group of customers (which included some who were presently excluded) at the same price, disrupting the existing arrangement but being in turn

subject to the same action by yet another rival firm.

Charging a higher price to clear the market would breach the stand alone test since the same output could be sold at a lower price by an alternative firm. A similar situation would arise were the firm to attempt to supply the entire market, since to

1 Stand alone costs are those of providing a product in isolation from other products. For a detailed discussion of this and other concepts relating to cross-subsidies, see Attachment II to Appendix I.

101

cover costs it must charge a price which is above minimum cost. In either case, because there will always be some group of users paying more than its stand alone costs, a rival firm could always enter and supply only at the output level which minimised costs. A natural monopoly which is required to service all users might, therefore, not be sustainable.

Essentially, these unsustainability problems reflect decreasing costs, followed by increasing costs. Increasing costs may, however, arise from subtle sources. In particular, production costs may appear to continue to decline, yet actually be increasing when all costs of servicing users are considered, such as where additional customers would be more remote from warehouses and hence involve higher transport costs (Sharkey 1982, p. 134).

The presence of decreasing production costs means there will always be an opportunity for some customers to be offered a better deal by a rival firm, and hence production by more than one firm will not lead to a stable outcome. However, because there is always an incentive for a rival firm to enter the market, nor will production by a single firm lead to a stable outcome.

The decreasing future costs problem

Another form of unsustainability arises where production costs are expected to decline in the future. This expectation may arise from growth in demand which would allow the use of larger, more economical plant, or from anticipated improvements in technology.

Because a plant which is optimal at the time of construction will quickly become obsolete or below optimal capacity, a new firm would be able to enter with lower costs and capture the market. The concern is that, because no firm could be sure of being able to recover its investment costs, no investment will take place at all. Alternatively, there may be wasteful investment as firms prematurely scrap capital and reinvest as a defence against potential entry (Sharkey 1982, p. 157).

The specialisation problem

The unsustainability problems discussed above may occur in both single and multiproduct natural monopolies. In the case of multiproduct natural monopoly, however, a fundamentally different form of unsustainability is also possible. It is

conceivable that two or more products, each covering their incremental costs, may be produced at lowest overall cost in a single firm - implying the industry is a natural monopoly - but that a rival firm specialising in some of the products in the natural

monopolist’s range may be able to service, at lower cost, the entire market for those particular products. Users of those products would therefore under a monopoly structure be paying more than their stand alone costs.

The conditions likely to give rise to this scenario are that the firm is a natural monopoly, and that products concerned have:

relatively strong product-specific returns to scale, which would aid specialisation;

102

■ relatively weak cost complementarities, which would aid joint production; and

■ relatively strong demand substitution effects, so that the lower price for the specialist firm’s product also attracts customers away from the use of other products in the natural monopolist’s range.

Under these circumstances, the possibility arises that supply of the other products in the natural monopolist’s range will become unviable. Consequently, if such a market is open to competition it may fail to provide some products which would be valued by their potential users at more than their incremental cost (Baumol, Panzar & Willig

1982, p. 223).

103

REFERENCES

Baumol, W.J. 1977, O n the Proper Cost Tests For Natural Monopoly’, Bell Journal of Economics, vol. 8, no. 1, pp. 1-22.

____ 1982, ‘Contestable Markets : An Uprising in the Theory of Industry Structure’, American Economic Review, March, pp. 1-15.

Baumol, W.J., Panzar, J.C. & Willig, R.D. 1982, Contestable Markets and The Theory of Industry Structure, Harcourt Brace Jovanovich Inc.

Borts, G.H. 1960, ‘The Estimation of Rail Cost Functions’, Econometrica, vol. 28, January, pp. 108-31.

Braeutigam, R.R. 1979, Optimal Pricing with Intermodal Competition’, American Economic Review, vol. 69, March, p. 38-49.

Braeutigam, R.R., Daughety, A.F. & Turnquist, M.A. 1982, ‘The Estimation of a hybrid cost function for a railroad firm’, Review o f Economics and Statistics, vol. 64, pp. 394-403.

Brock, W.A. & Evans, D.S. 1983, ‘Creamskimming’, in Breaking up B ell: Essays on Industrial Organization and Regulation, ed. D.S. Evans, North Holland.

Caves, D.W., Christensen, L.R. & Swanson, J.A., 1981, ‘Productivity growth, scale economies, and capacity utilisation is U.S. railroads, 1955-1974’, American Economic Review, vol. 71, pp. 994-1002.

Christensen, L.R. & Greene, W.M. 1976, ‘Economies of Scale in U.S. Electric Power Generation’, Journal of Political Economy, vol. 84, August, pp. 655-76.

Dodgson, J.A. 1985, ‘A Survey of recent developments in the measurement of rail total factor productivity’, in International Railway Economics, eds. K.J. Button & D. Pitfield, Aldershot (UK), Gower, pp. 13-39.

Evans, D.S. & Heckman J.J., 1983, ‘Natural Monopoly’ in Breaking Up Bell: Essays on Industrial Organisation and Regulation, ed. D.S. Evans, North Holland.

Faulhaber, G.R. 1975, ‘Cross-Subsidization: Pricing in Public Enterprises’, American Economic Review, vol. 65, no. 5, pp. 966-77.

Friedlander, A.F. & Spady, R.H. 1981, Freight Transport Regulation, MIT Press, Cambridge, Mass.

Globerman, S. & Carter, D. 1988, Telecommunications in Canada : An Analysis of Outlook and Trends, The Fraser Institute.

Griliches, Z. 1972, ‘Cost Allocation in Railroad Regulation’, Bell Journal of Economics, vol. 3, Spring, pp. 26-41

Harris, R.G. 1977, ‘Economics of Traffic Density in the Rail Freight Industry’, Bell Journal of Economics, vol. 8, no. 2, p. 556-64.

Hartley, P.R. & Trengove, C.D. 1987, ‘Is the Legislated Telecom Monopoly "Efficient"?’, in Competition And Competitiveness in Telecommunications, ed. H. Ramsden, Centre of Policy Studies, Monash University, Clayton, Vic.

Jarden Morgan NZ Limited 1988, Ownership and Regulation’, paper prepared for the Electricity Industry Task Force.

Joskow, P.L. & Schmalensee, R. 1986, Markets for Power: An Analysis of Electric Utility Deregulation, MIT Press, Cambridge, Massachusets.

Kay, J. & Vickers, J. 1988, ‘Regulatory Reform in Britain’, Economic Policy, October, pp. 286-347.

Keeler, T.E. 1974, ‘Railroad Costs, Returns of Scale and Excess Capacity’, Review of Economics and Statistics, vol. 56, May.

Kiss, F. & Lefebvre, B. 1987, ‘Econometric Models of Telecommunications Firms’, Revue Economique, March.

Klein, L.R. 1973, v4 Textbook on Econometrics, Englewood Cliffs.

Minister for Transport and Communications 1988, Australian Telecommunications Services: A New Framework, AGPS, Canberra, 15 May.

Panzar, J.C. & Willig, R.D. 1977, ‘Free Entry and the Sustainability of Natural Monopoly’, Bell Journal of Economics, vol. 8, no. 1, pp. 1-22.

Pera, A. 1989, ‘Deregulation and Privatisation in an Economy-wide Context’, in OECD Ecomomic Studies, no. 12, Spring, pp. 159-204.

Primeaux, W. 1975, ‘A Re-examination of the Monopoly Market Structure for Electricity Utilities’, in Promoting Competition in Regulated Markets, ed. A. Philips, The Brookings Institution, Washington D.C.

Ralph, E. 1986, Multiproduct Natural Monopoly and Cross Subsidy, Telecom Economic Papers no.l, Melbourne, Vic.

Scherer, F. M. 1980, Industrial Market Structure and Economic Performance, 2nd Edition, Houghton Mifflin Co., Boston.

Sharkey, W.W. 1982, The Theory of Natural Monopoly, Cambridge University Press, UK

Telecom 1986, Teleconomy, Issue no. 4-86, August.

105

Waterson, Μ. 1988, Regulation of the Firm and Natural Monopoly, Basil Blackwell Ltd, Oxford, U.K.

Weiss, L.W. 1975, ‘Anti-Trust in the Electric Power Industry’, in Promoting Competition in Regulated Markets', ed. A. Philips, The Brookings Institution, Washington D.C.

106

APPENDIX K : CONTRACTING OUT OF PUBLIC GOODS AND SERVICES

As an alternative to public production or regulation of a private monopoly, contracting out offers the prospect of engendering competition in the provision of goods and services, thereby reducing costs. Considerable cost savings could be expected from a significant increase in its use by each

level o f government in Australia.

K1 Introduction

When governments decide to supply a particular good or service, they face the choice of undertaking provision themselves or buying the good or service from other suppliers. The latter approach, or contracting out, is one way of introducing an element of competition into the provision of a good or service which might otherwise be subject to public or private monopoly provision.

This appendix considers the merits of contracting out and reviews available evidence of potential savings. Various forms of contracting out are outlined in section K2, as are some advantages and possible limitations for its use. Section K3 surveys overseas and domestic experience with contracting out. The scope for its increased use in Australia is considered in section K4. A summary is provided in section K5.

K2 Characteristics of contracting out

Contracting out involves the use of public and/or private suppliers to deliver goods or services to, or on behalf of, governments. It includes the process whereby contracts are put out for tender. Contracts may require full outside production or the use of government-owned assets in production.

The tender process often associated with contracting out is designed to create competition for the supply of goods or services by attracting bids from potential suppliers. In situations where there are a number of suppliers operating in the market, a contract is usually awarded to the tenderer with the lowest acceptable

supply price. Alternatively, in markets which have natural monopoly characteristics, competition can arise for the right to supply the entire market by bidding for the temporary monopoly rights of production and/or distribution (often referred to as franchising). A franchise may be awarded either to the tenderer who offers to pay the

government the highest sum, or to the tenderer who offers the lowest acceptable consumer supply price. An example of contract franchising could be the awarding of an exclusive right to provide rail services utilising publicly owned track.

Contracting out can be applied to either all or part of an operation currently undertaken by governments. For example, a government could call tenders for the operation of its complete rail network, the operation of specific rail routes, the maintenance of rolling stock or the catering function on passenger services.

107

Potential advantages of contracting out

The main advantage of contracting out is its ability to engender competition in the supply of goods or services. Where goods or services which are consumed by governments are put out to tender, competition amongst specialist suppliers (including potential ‘in house’ suppliers) replaces protected ‘in house’ production.

Unlike public provision, contracting out provides an incentive for efficient performance, since any cost savings accrue to the service supplier during the period of the contract. There is also pressure on the contractor to maintain (and where possible improve) service quality since the contractor must be prepared periodically to defend the contract against competitors. At the same time, periodic bidding can ensure that most of the efficiency gains accrue ultimately to the public rather than to the contractor.

Contracting out thus avoids some of the inefficiencies which may be associated with public production of goods and services. In addition, because it usually results in the planning function being divorced from operational aspects of production - with the former in public and the latter in private hands - any temptation for planners to overinvest in capacity is reduced. Moreover, by requiring planners to specify more tightly the good or service to be supplied, contracting out may enhance the public policy planning and evaluation process.

Constraints on the gains from competitive tendering

The extent of inefficiency existing before the implementation of contracting out will determine the limits of any potential gains. For example, the contracting out of public activities currently undertaken in a competitive environment may offer little in the way of efficiency improvements. The potential for the highest ongoing efficiency gains is likely to be achieved in industries which are characterised by rapid

technological development yielding continuous productivity improvements.

However, the existence of transaction costs may be an important constraint on achieving net gains. The specification of tenders, evaluation of bids and the drawing up of contracts can be difficult and costly exercises. For example, difficulties may arise in specifying desired outputs and in allowing for future contingencies. In addition, the monitoring of contract performance can be difficult and the costs and uncertainties associated with these processes need to be weighed against the potential benefits.

Another problem with contracting out may be the absence of effective competition at the bidding stage. Competition could be precluded, for example, because the contract entails making a long-term commitment involving investment in a highly specialised asset (eg a railway line). Even where competitive bidding is possible for the initial contract, once a specialised investment has been made by the successful tenderer, the government and the tenderer may become locked into a situation of mutual

dependency because the transaction costs involved in the transfer of asset ownership can be high. To combat declining incentives to operate efficiently with longer term contracts (as there is no short-term threat of contract loss), there may be a need to employ more stringent performance requirements.

108

Given the above limitations, franchising involving the complete private provision of assets may be precluded in some instances. However, scope may still exist for the contracting out of various services or functions.

K3 Overseas and domestic experience

Contracting out of government goods and services is undertaken extensively in the United Kingdom. Examples of administrative functions which are tendered for include building maintenance, data processing, asset valuations and agricultural services. Various specialist management and operational functions are also put out to

tender, including legal, secretarial and design services. The United Kingdom’s independent television services have been subject to franchise arrangements since their introduction in 1954. The British Government has actively encouraged the use of contracting out through legislation requiring its use for various government

functions, such as cleaning, catering and refuse collection (Ascher 1987).

In the United States, contracting out is widely employed by municipal and State governments. An extensive range of functions, such as urban transport, highway construction, the collection of tax from defaulters, fire fighting, prisons, welfare services, child care, home maintenance and asset maintenance are put out to tender.

The US Defence Department makes extensive use of tendering to operate dockyards and to provide weapons systems, and a wide range of support services including crew training and inventory control.

In Australia, contracting out is used extensively in the private sector while governments have long used the tendering process to develop infrastructure such as roads, dams and power stations. More recently, contracting out has become more prevalent in the operational areas of public enterprises. For example, a number of

electricity authorities contract out maintenance and rehabilitation work in power stations (see Electricity Study in Volume 4 of this report). Pipeline construction and maintenance is also contracted out by the Pipeline Authority and, to varying degrees, by the publicly owned gas utilities.

For a number of public authorities the concept of contracting out would not be new as they themselves bid for work against other public and private sector organisations both here and overseas. For example, the Snowy Mountains Engineering Corporation has been involved in many overseas infrastructure developments and, in 1988, Qantas

won a tender to manage the Fijian airline, Air Pacific.

There is evidence that governments are beginning to extend the contracting out approach to administrative and similar activities of government. For example, the Commonwealth Audit Office has recently commenced contracting out some of its financial audits, while in 1988-89, the New South Wales Government announced its

intention to call tenders for the cleaning of public schools, hospitals and trains and the operation of government printing services. Both the New South Wales and Victorian Governments have also indicated that they will put out to tender the administration of

various public sector cash reserves, balances and funds. The Queensland Government has recently contracted out the management and operation of a new minimum security prison.

109

At the local government level, a recent survey of 170 South Australian and Tasmanian municipalities indicated that, in 1988, the use of contracting out was extensive, with just under 75 per cent of municipalities reporting that they contracted out at least one function. Table K1 shows the percentage of municipalities which use contracting out

for particular activities.

Table K1: Municipal use of tendering, South Australia and Tasmania

P ercentage o f m u n icip alities u sin g tenderin g

Road construction/maintenance 51.2

Household garbage collection 44.1

Operation of public halls, civic centres 27.6

Other garbage (ie commercial) 22.0

Swimming pools and beaches 17.3

Drainage 13.4

Library 9.4

Sewerage 8.7

Source: Rimmer 1988.

Empirical studies

A number of studies into the effects of contracting out have been undertaken, both in Australia and overseas. These studies have looked at actual and potential cost savings associated with contracting out, and the reasons for them. While these studies attempt to compensate for differing service standards before and after the use of contracted services, sufficient uncertainty about comparability remains to warrant caution when interpreting results.

Studies conducted in the United Kingdom suggest that, while cost savings from tendering average 20 per cent, savings can often be higher. For example, one study has found that, over a five-year period, Wandsworth Council derived savings exceeding 40 per cent for refuse services, municipal workshops, the cleaning of social service facilities, libraries and public halls and for street lighting (Ascher 1987, p. 239). Table K2 provides a summary of recently published United Kingdom data which suggest that, for a wide variety of government functions, actual savings range from 9 per cent to 45 per cent.

110

Table K2 : Percentage cost savings resulting from tendering of services in the United Kingdom

Example Source Function Cost saving

1 National Health Domberger Hospital domestic 20

services

2 Local Authorities Domberger Refuse collection 20

3 National Health CBI Report Domestic services 23

Service Catering 9

Laundry 14

4 Central CBI Report Cleaning )

Government Laundry )

Maintenance ) Catering )

Security )

23

5 Local CBI Report Refuse 45

Authorities Cleaning 31

(Wandsworth- Recreational facilities 22

Borough) Catering 13

Printing 11

Transport 26

6 Transport Hensher London buses 20

Source: Business Council of Australia 1989

Data on cost savings obtained in the United Kingdom are consistent with recent Australian data, reported in Table K3, which suggest that actual savings from the use of tendering in 1987-88 by Australian local governments range from 9 to 46 per cent. Other Australian studies have indicated possible gains from contracting out, though

these are not quantified. For example, Hensher (1989) suggests that contracting out certain metropolitan bus routes could lead to the provision of much more efficient services, based on the relative costs of existing public and private bus operators.

Studies attribute the bulk of savings derived from contracting out to productivity gains through improved use of infrastructure, capital and labour. For example, contractors frequently introduce new work techniques and technology. Such innovation is often accompanied by improved service flexibility and quality, with improved management

and organisational structures being implemented (Domberger 1988). Cost savings need not be linked to a change in the producer of a good or service. It is frequently the case that the incumbent public producer secures the tender by reducing costs in response to the introduction of competition (Yarrow 1986).

I l l

Table K3: Percentage cost savings resulting from tendering of services in Australia

Example S ou rce F unction C o st savin g

Melbourne local Rimmer Garbage 17

governments

Tasmanian and SA Rimmer Household garbage 9

local governments Other garbage 24

(170 Municipalities)

Roads/bridges maintenance 17

Recreational facilities 46

Brisbane City Rimmer Maintenance works 35

Council Capital works 24

Source: Business Council of Australia 1989

Reasons underlying the considerable variation in cost reductions evident from the experience both in Australia and overseas can be summarised as follows:

the level of inefficiency existing prior to use of contracting out;

the emergence of genuine competition in service delivery;

the quality of contracts and the effectiveness of contract monitoring and evaluation;

. the scope for the introduction of innovative approaches to management, service delivery and new technology; and

. the nature and extent of costs associated with specifying, evaluating and monitoring contracts.

K4 Scope for extending contracting out

There appears to be considerable scope for the increased use of contracting out across the three levels of government in Australia. For example, Domberger (1989) believes that a large number of government functions could be put out to tender,

generating savings in excess of $2 billion each year. In particular:

the Commonwealth Government could directly save over $234 million on recurrent expenditure, with a further $315 million from public trading enterprises and $97 million on new fixed assets per year;

State governments could save $500 million on current expenditure, $451 million by State trading enterprises and $175 million on expenditure on new fixed assets; and

. local government could save upwards of $208 million each year on recurrent expenditure, $97 million by municipal trading enterprises and $40 million from expenditure on new fixed assets.

According to Hogan (1988), possibilities for contracting out in New South Wales include coal mine operations for electricity authorities, funds management, various hospital services (such as X-ray departments and kitchen services), revenue collections for vehicle registrations and the operation of government bus routes. It is

likely that these possibilities for contracting out are not restricted to New South Wales.

There is also scope for contracting out in a number of markets which may have elements of natural monopoly. For example, tug services in certain Australian ports are operated as monopolies because of the distances between major ports coupled

with low cargo volumes handled by individual ports (LAC 1988). Where such services are presently publicly supplied, franchising could increase incentives for efficient tug operations. Under such arrangements, the government - on behalf of the community - could appropriate the ‘monopoly rents’ which might otherwise accrue to the tug

operator, or alternatively allocate the franchise to the bidder tendering the lowest supply price.

K5 Summary

Contracting out has been used extensively in the private sector for many years. It has also been widely used in the public sector in respect of infrastructure development and its extension into many operational areas of government is now being increasingly advocated.

Available Australian and overseas research suggests that significant cost savings can be achieved through the use of contracting out. The extent of potential savings depends on a number of factors, including the particular nature of the good or service and how efficiently it is currently being provided. Such savings could contribute

substantially to the improved competitiveness of Australian industry.

REFERENCES

Ascher, K. 1987, The Politics o f Privatisation: Contracting Out Public Services, Macmillan Education, London.

Business Council of Australia 1989, ‘Competitive Tendering’, Business Council Bulletin, February, pp. 10-14.

Domberger, S. 1988, ‘Competitive Tendering and Contracting Out: Some Lessons from the U.K. Experience’, Competitive Tendering and Contracting Out: Towards a Policy Agenda, Graduate School of Management and Public Policy, University of Sydney.

_____ 1989, ‘Competitive Tendering: How Much can be Saved?’, Business Council Bulletin, June, pp. 4-11.

Industries Assistance Commission (LAC) 1988, Coastal Shipping, Report no. 415, AGPS, July, Appendix M.

Hensher, D.A. 1989, ‘Competitive Tendering in the Transportation Sector’, Economic Papers, March, pp. 1-11.

Hogan, B. 1988, ‘The Potential for Privatisation in New South Wales’, Economic Papers, September, pp. 1-10.

Rimmer, S. 1988, ‘Contracting Out in Australian Local Government’, Competitive Tendering and Contracting Out: Towards a Policy Agenda, Graduate School of Management and Public Policy, University of Sydney.

Yarrow, G. 1986, ‘Privatization in theory and practice’, Economic Policy - A European Forum (Edited by G. De Minil and R. Portes), Cambridge Univsersity Press, April, pp. 323-378.

114

APPENDIX L : SETTING PERFORMANCE TARGETS FOR PUBLIC ENTERPRISES

Wide variations between performance targets currently applying to Australian public enterprises suggest that such targets have not been applied systematically to promote efficiency objectives. Recently there has been growing interest in the use o f financial targets (particularly rate of

return targets as opposed to traditional cost recovery or self-financing targets) and also of non-financial targets as a means of applying more effective disciplines on performance. Where public enterprises produce marketable goods and services, requirements to earn rates o f return comparable to those for similar activities in the private sector may provide

a strong incentive to minimise production costs and to improve productivity.

The lack of market disciplines on performance and other limitations associated with control and monitoring mechanisms, have reduced the ability o f performance targets to promote efficiency in public enterprises. Although performance targets may be less effective accountability and

monitoring mechanisms than market-based disciplines, wider use of appropriately designed targets could lead to significant improvements in the efficiency o f public enterprises.

LI Introduction

Throughout the 1980s there has been a growing interest in Australia in the use of performance targets as monitoring mechanisms for public enterprises. Initially this interest focussed on developing appropriate financial targets, but the need to supplement these targets with non-financial targets has also been recognised.

Use of real rate of return requirements received prominence in the Rae Report (1983). Following reports from the Public Bodies Review Committee in the early- 1980s, the Victorian Government adopted rate of return targeting for some of its

major public enterprises in 1983-84.1 More recently, the Commonwealth Government has indicated that ‘appropriate financial targets for enterprises constitute the linch-pin of the Government’s reform package’ (Minister for Finance 1987, p. 5). A recent report has recommended that State and Territory governments should set economic rates of return to be achieved on the market value of port

authority assets (Inter-State Commission 1989, pp. 285-8). Several other States (eg SA, NSW, WA) are currently considering the use of financial targets. Such targets have been set for nationalised industries in the UK for many years and several white

1 The affected enterprises included the State Electricity Commission of Victoria, the Gas and Fuel Corporation of Victoria, the Melbourne and Metropolitan Board of Works, the Grain Elevators Board and the Port of Melbourne Authority.

115

papers on the subject were published in the UK (United Kingdom 1961, 1967 and 1978).

One explanation for the attention devoted to financial targets is that they are seen as a means of introducing to public firms some of the disciplines applying to private firms to encourage better resource use. Proponents of this approach point to the generally low rate of return on capital achieved by public firms relative to private firms (eg the Rae Report 1983). Another explanation relates to the growing and legitimate concern (perhaps heightened by prevailing economic conditions) that public enterprises need to be provided with more guidance as to what constitutes ‘acceptable’ performance. Supporters of this view refer to the operating deficits of many public enterprises and their dependence on government revenue supplements. Further, many accusations of inefficiency with regard to pricing and investment decisions have been levelled at key public enterprises and financial targets are seen by some as a means of improving decision-making in these areas.

As the issues associated with financial targeting have received more prominence, there has been increasing recognition of the need to supplement such targets with non-financial targets in order to better assess two distinct aspects of public enterprise

performance. First, in the absence of market disciplines, some aspects of commercial performance (eg quality of service) are not amenable to measurement by financial targets. Second, when public enterprises have non-commercial objectives (eg community service obligations), the success in achieving them can often only be measured with reference to non-financial measures (eg percentage of households with telephones).

The purpose of this appendix is to examine the contribution which performance targets can make towards improving the efficiency of public enterprises and the factors affecting their usefulness in this role. The discussion draws upon both economic theory and practical experience (principally in the UK) with various types of performance target. Two types of financial target (traditional self-financing or cost recovery targets, and rate of return targets) are discussed in section L3, while non- financial targets are addressed in section L4. Some tentative findings on the ability of

financial targets of both types to improve performance are presented in section L5. First, however, some general issues relevant to both forms of performance targets are examined in section L2.

L2 Performance targets - some general issues

Regardless of the type of target applied, there are a number of factors which affect the extent to which the setting and monitoring of targets might promote better performance:

• the existence of multiple objectives; . community service obligations (CSOs); • other competitive advantages/ disadvantages; . the valuation of assets/costs; . market imperfections/externalities; • the market environment; • the system of rewards and penalties; and • the payment of dividends.

These factors are discussed below.

The existence of multiple objectives

Private firms tend to be governed by the need to make profits. Whilst managers of private firms might seek to maximise sales or pursue other goals not directly related to profit, particularly in the short run, available empirical evidence (Scherer 1980, p. 41) suggests that:

...profit maximisation provides a good first approximation in describing business behaviour. Deviations, both intended and inadvertent, undoubtedly exist in abundance, but they are kept within more or less

narrow bounds by competitive forces, the self-interest of stock-owning managements and the threat of managerial displacement by important outside stockholders and takeover raiders. To the extent that deviations do occur, they are likely to be larger when competition is weak and perhaps also when there is sharp separation of ownership

from control.

In contrast, the stated objectives of public enterprises usually encompass a wide range of goals. Hence, the performance of such enterprises can only be adequately assessed with a range of indicators. Financial targets need to be supplemented with other relevant information to enable an evaluation of not only the ‘efficiency’ but also the

‘effectiveness’ of the enterprise in achieving its goals. Non-financial targets are discussed in section L4.

Community service obligations

As discussed in Appendix I, public enterprises are frequently required to pursue non­ commercial objectives such as CSOs. Pursuit of these objectives will impact adversely on an enterprise’s ability to meet a particular financial target. There is a danger, therefore, that the fulfilment of CSOs will cloud the assessment of a public

enterprise’s performance in that part of its operations concerned with providing goods and services on a purely commercial basis. This suggests that while financial targets should be based on criteria used to judge private sector performance, explicit allowance needs to be made for the costs of providing CSOs in an efficient manner. This principle appears to have been accepted by the Commonwealth Government

(Minister for Finance 1987, p. 5) when it foreshadowed that:

Targets will be agreed with the responsible portfolio minister having regard to the impact on profitability of any ‘community service obligations’ or non-commercial objectives required by the Government.

Allowing for the costs of fulfilling CSOs may also reduce the incentive to adopt pricing structures which lead to cross-subsidisation between consumer groups. This may occur where the combination of uncompensated CSOs and a financial target forces an enterprise to mark up its prices on its remaining services. Appendix I

discusses these and other issues associated with CSOs.

117

Other competitive advantages/disadvantages

Effective performance monitoring necessitates distinguishing between decisions taken at the discretion of management and those which have been influenced by governments. In practice, this is likely to be a difficult task, as governments frequently exert significant influence over their enterprises through various controls and directives. These controls typically include loan guarantees and borrowing controls, pricing directives and/or requirements for consultation in setting prices, purchasing policies, and industrial relations/employment guidelines (see Chapter 6). Consequently, as noted by the Victorian Department of Management and Budget (1986, p. 14):

Interpreting rate of return performance data may be quite difficult especially when economic circumstances or Government policy affecting Authorities change in an unpredicted fashion.

These problems might be largely avoided if governments reduce direct controls on public enterprises to ‘let the managers manage’. Further, where access to resources is enhanced through loan guarantees and measures such as tax exemptions, it is desirable that they be removed. Such changes would promote a more neutral decision-making environment. Nevertheless, to the extent that these competitive advantages/disadvantages remain, it would seem desirable that they, and their effects, are made as explicit as possible and taken into account when setting and interpreting performance targets.

The valuation of assets/costs

Financial targets may be misleading indicators of the efficiency with which resources have been used when the cost base reflects an inappropriate valuation of assets used in producing outputs. Appendix M suggests that the relevant cost of assets is the opportunity cost or the value of assets in alternative uses. In contrast, asset valuations typically used in Australia are based on historical costs. In times of inflation, such valuations are likely to result in grossly overstated rates of return. This arises because: the true value of the asset base (the denominator) is understated; and the return (numerator) is overstated because the cost of assets consumed each year is understated.

For some enterprises, these effects may be partially offset by the historical cost valuation of some assets (eg those associated with ‘bad’ investment decisions) which grossly overstate their current market value. Nevertheless, these overstatements are

also misleading and would compensate only by chance for understatements associated with ignoring inflation.

This suggests that the introduction of financial targeting should be accompanied by the adoption of current cost accounting systems which more closely reflect economic considerations in order to provide an improved basis for pricing and investment

decisions as well as performance assessment (refer Appendix M). Moreover, in order to compare the financial performance of different enterprises, consistent conventions would be needed across enterprises. At present, however, reporting procedures vary

118

widely even between public enterprises in the same industry (see, for example, the rail and electricity studies in Volume 4 of this report).

Market imperfections

In some cases, the existence of market imperfections may also mean that a financial target becomes an inadequate measure of performance. For example, the financial target may not correspond to the economic return to society as a whole from an activity because of associated environmental costs and the intangible but real value of

an asset in place encouraging further economic development. Possible conflicts between commercial and economic efficiency objectives can also arise for public enterprises operating in so-called ‘decreasing cost’ industries characterised by significant economies of scale, where if marginal cost pricing were to be adopted,

financial losses may be incurred (refer Appendix G).

The market environment

The market environment in which many public enterprises operate may influence the degree to which a financial target serves to stimulate managerial/cost efficiencies. Lack of capital market disciplines (eg the threat of takeover or bankruptcy) means that there may be little incentive to cut costs where an enterprise is already achieving

the target level. Moreover, even where the target is set at a level above that currently being achieved, a public enterprise with significant market power (arising from restrictions on competition) may be able to meet the target simply by raising its prices or by lowering the quality (and hence costs) of the goods or services it provides.

Similarly, targets are unlikely to be effective if, as occurred in the UK, a common response to failure to meet the targets is an application for price increases (Rees 1976). This suggests that financial targets are more likely to be effective when pricing and quality of service variables are also monitored. These supplementary

performance targets are discussed in section L4.

The system of rewards and penalties

The imposition of performance targets, as with any accountability/monitoring mechanism, is likely to promote efficiency only if associated with a well-defined system of managerial rewards (eg salary bonuses) and penalties (eg dismissal) which managers know will be enforced. The political commitment to respond to the success

or failure of an enterprise in meeting targets is crucial to the effectiveness of the whole incentives environment that such targets are designed to create (refer Chapter 7, section 7.2).

In administering the system of rewards/ penalties, however, it is essential to determine whether succeeding (or failing) to reach the target was due to management policies or to external factors. Equally, however, it is important not to ‘over-adjust’ for the impact of external factors on actual performance achieved, or to categorise as

‘external’ factors which are substantially within managers’ control. Both tendencies would detract from the effectiveness of performance targets.

119

The payment of dividends

The extent to which financial targets are used as a basis for determining public authority dividend payments may also impact upon the degree to which such targets promote efficiency. Provided they are measured on an appropriate basis (ie using current cost rather than historical cost valuations), the actual return on capital achieved would indicate the maximum dividend the enterprise could currently pay without eroding its capital base. However, whilst the requirement to pay a dividend

to the government (as the return on equity to the owners) may strengthen the incentives to meet the target, there are arguments against the setting of a fixed dividend target each year or a common dividend target for diverse public enterprises. As noted by several commentators (Department of Finance 1988a; Ergas 1986; Xavier 1989a), consideration of dividends on a case-by-case basis allows public enterprises the same flexibility as private firms in determining dividend policy.

L3 Financial targets

Financial targets fall into two groups - traditional self-financing or cost recovery targets, and rate of return targets. The main forms of targets applying to public enterprises in Australia have been the self-financing and the cost recovery target. Enterprises subject to these targets are required to recover either all or a proportion of their financial costs. In recent years, greater use has been made of rate of return targets which require public enterprises to earn a return on the capital employed by them in producing goods and services. Issues relating to the design and operation of these targets and their effects on decision making by public enterprise managers are discussed below.

Traditional self-financing or cost recoveiy targets

Financial targets of various sorts have been used in the past in Australia. However, the wide variations between targets applying to different authorities (see Appendix C, Table C16), suggests that they have not been applied systematically to promote efficiency objectives. Thus in 1982, the Senate Standing Committee on Finance and Government Operations (1982, p. 101) concluded that:

...Clearly, there has been no systematic approach to the establishment of financial objectives for authorities.

As evident from Table C16, those financial targets which have been applied to public enterprises have tended to concentrate on the achievement of a narrowly specified self-financing or cost-recovery target, rather than any particular rate of return. Such targets have tended to focus on aggregate financial performance, such as a balanced- budget requirement (ie total revenues should equal total direct operating costs), or on achieving some other prescribed percentage of cost recovery. In some cases, targets have comprised a composite measure covering both cost recovery (to a specific level or for a defined set of costs such as direct operating costs) and a requirement to finance a certain proportion of capital expenditure from internally-generated funds.

120

Financial targets of this nature seem to reflect a view that if a public enterprise is covering its financial or accounting costs, then it must be performing satisfactorily. There are a number of factors which may account for this view.

. a direct analogy with the concept of profitability used in the private sector as a measure of aggregate performance;

. a concern with the macroeconomic consequences of covering the financial deficits incurred by public enterprises; and

. a concern that differing levels of cost recovery in different areas of government activity may have undesirable equity or efficiency effects (as instanced by the continuing debate over cost recovery levels in the rail and road transport industries).

Setting cost recovery targets for public enterprises imposes financial disciplines which reduce the likelihood of inappropriate supply decisions while encouraging managers to develop better control systems to reduce waste. However, the setting of such targets will only encourage improved efficiency to the extent that it necessitates the

enterprise making supply, pricing and investment decisions on the basis of real economic costs. Therefore, from the point of view of this report, a major weakness with traditional financial targets is their focus on financial rather than economic costs. In particular, as discussed in more detail shortly, targets that stipulate the recovery of

recurrent costs only (ie do not include the real cost of capital in the cost base) are unlikely to encourage efficiency.

Even if the accounting systems on which these measures are based were placed on a consistent basis and adopted more appropriate accounting conventions (eg current cost accounting), measures of aggregate financial performance are unlikely to provide a true indication of the efficiency of a public enterprise. Aggregate cost recovery

targets promote efficiency of resource use only to the extent that they encourage recovering the full economic costs of each service individually. Without sufficiently detailed breakdowns of the accounts into distinct areas of activity, it is not possible to assess whether this has occurred.

Put another way, a balanced budget does not necessarily lead to efficient resource allocation and may in fact run counter to this goal. This is because a particular level of cost recovery can be achieved even though an efficient pricing structure is not being followed. Similarly, it is possible for an efficient pricing structure to result in an

overall financial deficit (eg in decreasing cost industries). An important implication is that the pursuit of full cost recovery per se (or similar levels of recovery between closely-related industries) may not be justified on either efficiency or equity grounds.

Further problems may arise when a self-financing target relates to one particular cost element (eg covering a certain percentage of capital expenditure). Such a rule may make investment decisions subject to annual cash flow limitations and gives the capital expenditure financing ratio a prescriptive significance which it does not

deserve. As discussed in Chapter 7, concerns with possible excessive levels of investment by public enterprises are better dealt with directly by subjecting investment proposals to rigorous appraisals.

121

Recognition of the problems associated with these targets has led to attention | focussing on alternative financial target specifications which might more directly promote efficient production and pricing decisions. Particular interest has been shown in rate of return targeting.

Rate of return targets

The rationale for rate of return targets

A target rate of return is a measure which relates the income earned from an activity to an appropriate measure of the resources devoted to producing that income. Commonly-used bases are the value of equity, total assets, or turnover. Given the dubious validity of the distinction between debt and equity with respect to the public sector, proposals for rates of return generally relate to total assets or funds employed. For less capital-intensive enterprises (eg Australia Post), however, it has been suggested that turnover be used as a better measure of resources employed. Rate of return targets are expressed as a rate or ratio of values so as to provide a basis for comparing enterprises or activities of differing scale.

A rate of return which is below the target rate does not necessarily imply that those factors of production represented by the base on which the target is set (eg capital) are being used inefficiently. For example, a firm which pays too much for any of its inputs or uses them inefficiently will record a rate of return on capital below that of more efficient operators.

Since many public enterprises are predominantly engaged in producing marketable goods and services, it is appropriate for their performance to be assessed against performance criteria similar to those of comparable activities in the private sector. The adoption of a rate of return target requiring public enterprises to earn a commercially based return on their activities is seen as desirable in this context.

Indeed, if public enterprises are to make pricing and investment decisions that are efficient in the sense of obtaining maximum value from the resources used (compared with alternative uses) then the return on capital employed should match that obtainable from alternative uses available to the community. To the extent that such a return is not earned, the goods and services produced by public enterprises will be underpriced and overused.

Several participants argued that public enterprises should not earn a market rate of return on their activities, drawing a distinction between the private and public sectors based on the view that the same pricing and investment principles should not apply to both. However, the resources used by public enterprises come from the same source (ie savings) as those available to the private sector, except that the government has diverted them to the public sector. Given acceptance that the economy needs to get the best use from its resources, the government has an obligation to use these resources productively. By using these resources in activities which do not generate a commercial rate of return, the government is tying up resources which could be employed for a better return elsewhere.

Some participants also maintained that if charges do not include a rate of return, goods and services could be provided more cheaply, thus improving the international competitiveness of using industries. However, under-pricing the goods or services

122

provided by public enterprises is a very indirect means of achieving this goal. Users are diverse and include households as well as businesses. Moreover, since different industries use different amounts of government-provided goods and services, this is not a cost-effective way of achieving such a goal. It can also distort business decisions relating to the use of various inputs and increase the use of government-provided services when, if the real resource cost were taken into account, alternative inputs

might be used. The wider ramifications of such an approach for the taxation and borrowing situations of governments would also need to be considered.

Another view presented to the inquiry was that since some of the activities undertaken by public enterprises generate external benefits, these activities should not be required to generate a rate of return commensurate with that achieved in the private sector. In the Commission’s view, only a minor proportion of the services

provided by public enterprises is likely to generate such benefits. In such cases, a subsidy may be justified to increase usage to its desired level. However, the nature of the external benefit should be clearly identified and the relevant activity distinguished from the commercial activities of the enterprise before providing such subsidies.

In general, the return earned by a business is available to maintain existing productive capacity and to finance future investment. Without this return, public enterprises would have to rely entirely on borrowing or funding from consolidated revenue to finance such investment. If funds are borrowed, a rate of return at least equal to the

interest rate applying to the loan would have to be generated. Otherwise, loan repayments would need to be met from other revenue. Financing new investments or meeting the cost of outstanding loans from government revenue generally implies that additional tax revenue would need to be raised. Therefore, a policy of not allowing

public enterprises to generate a rate of return is not costless. Services are provided at a lower cost to users but at a higher cost to taxpayers. Further, raising additional taxes could reduce the international competitiveness of businesses.

Another rationale claimed for rate of return targets is that they provide an incentive for public enterprises to improve their internal resource allocation decisions and to reduce production costs (eg by eliminating inefficient work or management practices). Moreover, a rate of return target may impose a discipline on enterprises to undertake

only those investments consistent with the target. Given the capital intensity of many public enterprises and the large-scale nature of investment in public infrastructure, this ex-ante application of rate of return targets as a test discount rate for investment evaluation has potentially important ramifications for the efficiency of resource use.2

The usefulness of rate of return targets, in common with cost recovery and self­ financing (and in indeed, any other) financial target, will be affected by the factors discussed in section L2. In addition, however, some specific limitations/issues applying to rate of return targets affect the ability of such targets to promote efficient resource use. One issue relates to the possible reductions in production efficiency

that might stem from the tendency of rate of return targets to distort the efficient use of inputs. A second set of issues relates to operational aspects of using rate of return

2 The concepts of rate of return and test discount rate are sometimes used synonymously. In principle, however, the discount rate is intended to relate to returns at the margin on new investments while the rate of return refers to returns on average to past investments. The two will coincide only if all the firm’s investments are evaluated using the same discount rate, those investments offer the same

marginal return, and the actual returns turn out as expected (Johnston, Parkinson & McCray 1984).

123

targets, including determination of the appropriate rate of return and whether the required rate of return should vary between enterprises and over time.

Distortion of input use

Averch and Johnson (1962) were the first to show that an enterprise which faces rate of return regulation may employ an input mix which is not cost-minimising for the level of output that it produces. Whilst Averch and Johnson’s analysis related to controlling the exploitation of market power by private firms by setting a maximum allowed rate of return to capital, the considerable literature that has since developed on the potential effects of different forms of rate of return regulation under differing assumptions (eg differing managerial objectives) concludes that rate of return targets frequently induce inefficient input combinations. Prominent articles include Bailey and Malone (1970); Baumol and Klevorick (1970); Gravelle (1976); Gravelle and

Katz (1976); and Heald (1980). Empirical studies (eg Courville 1974; Spann 1974; Petersen 1975; Boyes 1976; and Jones 1983) generally support this conclusion.

Of particular relevance to this inquiry is whether specifying minimum rates of return expected to be earned by public enterprises is likely to result in inefficient input mixes and thus productive inefficiencies (refer Chapter 4). The framework established by Gravelle suggests that a firm required to earn a minimum return to capital will tend to use less capital relative to other inputs than it would if it did not face such a requirement. There will be an incentive for managers to choose less capital-intensive investments, to bias investment decisions in favour of those with lower initial capital costs and shorter lead times. Indeed, a report by the Price Commission (1978) into the South of Scotland Electricity Board suggested that the Board was using assets with short lives and using accounting policies which reduced its reported capital base.

The conclusion from both the theoretical and empirical work therefore is that rate of return regulation frequently results in inefficient input combinations. More debatable is the size and significance of this distortion and under what circumstances the resulting efficiency losses are more than outweighed by general cost efficiencies stimulated by a rate of return requirement. A study by Albon (1985), on the effects of

financial targets on the behaviour of monopolistic public enterprises, concluded that the overall effect is uncertain if managers behave in a manner which does not minimise costs and price output on the basis of inefficient costs. The extent and likelihood of overall gains in economic efficiency are likely to be greater the more effective are disciplines on cost-minimisation (through competition in product

markets and/or regulation of prices).

What rate of return is appropriate?

Another issue fundamental to the setting of rate of return targets is the question of what the appropriate target should be. The alternative approaches to evaluating the opportunity cost of capital devoted to public enterprises include:

. the social opportunity cost of capital (SOC); . the social rate of time preference (SRTP); . a weighted average of the SOC and the SRTP; and • the cost of funds (COF) approach.

The SOC approach takes the view that the real cost of capital invested in public enterprises is the opportunity cost of the capital resources in their next best use - assumed to be a marginal project elsewhere in the economy. On the assumption that the public sector investment is displacing private sector investment, the SOC

approach suggests a rate of return target equal to the real pre-tax rate of return from private sector capital projects.

The SRTP approach, in contrast, is based on the contention that the alternative to public sector investment is current consumption. Thus, it is argued, the required rate of return should be equal to that required by the community for deferring present consumption in favour of future consumption. Both the long term bond rate and the long term growth rate for real incomes have been suggested as surrogates for the

SRTP.

The third approach involves a weighted average of the rates derived from the first two approaches outlined above, with the weights depending upon the extent to which the resources to finance public investment are drawn from private investment or are at the expense of private consumption.

The COF approach determines a rate based on the cost of accumulating the funds which are invested in the enterprise. That is, estimates are made of the separate long term costs of debt and equity to the public sector, which are then weighted according to the debt: equity ratio used by public enterprises. This approach is the one adopted

by the Victorian Government.

In an idealised world, the approaches described above would yield an identical rate. The Department of Finance (1987, p. 3) noted:

In a simplified world the value placed by society on marginal decreases in present consumption in exchange for higher future consumption can be related to the rate of return on investment in the private sector. This is because the market interest rate will tend to equate with the rate of

return available at the margin on new private sector investment (the social opportunity cost (SOC) rate of return) and the social rate of time preference (SRTP), ensuring the optimal balance of present consumption and (investment augmented) future consumption. The

rate of interest serves the dual role of the reward for postponing consumption and the cost of funds for investment, and in this case the market rate of interest would be the appropriate discount rate.

However, amongst other things, the existence of taxation, imperfections in the capital market or product markets, and government intervention in interest rate determination can all lead to significant divergences in the rates suggested by each method.

Estimates of average real pre-tax rates of return earned by private sector firms in Australia have generally been around 10 per cent per annum. Estimates in this range have been supported by Clare (1982), Johnston, Parkinson and McCray (1984), the Treasury (1983), and the Department of Finance (1987). This last paper, however,

suggested that the real pre-tax rate of return actually achieved has averaged around 6 per cent per annum since the mid 1970s whilst the required rate of return on new

125

investment has averaged around 12 percent over the same period. However, on the basis that market forces will, in the long run, ensure that average ex-post and ex-ante rates of return should be much closer, it suggests a 10 per cent real rate of return as an appropriate medium-term objective, although it considers 6 per cent ‘reasonable in current circumstances’. Ball and Davis (1984) have pointed to Australian Graduate School of Management estimates of a negative average real rate of return during the

1970s. It should be noted, however, that the 1970s were characterised by unanticipated inflation rates, access by capital markets to funds arising from OPEC surpluses and negative real rates of return in major capital markets. As noted by the Department of Finance (1987), estimates can vary even within the SOC approach due to the application of different methodologies (eg whether national accounts or stock market data are used) and whether ex-ante or ex-post measures are considered. Most estimates under the SOC approach, however, are in the range of 6 to 10 per cent.

Estimates under the SRTP approach (using both the long term bond rate and the long term growth of real incomes) have been in the vicinity of 2 to 3 per cent per annum. The cost of funds approach used by the Victorian Government has yielded a required rate of 4 per cent per annum. The 4 per cent rate was derived from estimates of the long-term costs of debt (3 per cent) and equity (5 per cent) to the public sector (using time series analysis spanning more than one hundred years), weighted by a debt equity ratio of 50 : 50 reflecting the extent to which these forms of finance are utilised by public enterprises.

Depending upon the approach taken, the annual rate of return requirement could be between 2 and 10 per cent, and possibly more. Clearly, different rate of return requirements will have significant consequences for pricing and investment policies undertaken by public enterprises and for resource allocation within the economy generally.

The conceptual approach adopted is therefore critical, and the issue remains contentious. Moreover, it is not clear that any one approach is appropriate in all circumstances. Nevertheless, the Commission considers that, in the case of public enterprises which provide clearly marketable goods and services (the charges for which are the main focus of this inquiry), the SOC approach offers the most suitable benchmark for performance. The case for requiring public enterprises to earn a rate of return comparable to relevant private sector counterparts is particularly strong when they face actual or potential competition from private sector suppliers.

Uniformity o f the rate o f return target

Another important question is whether a target rate of return should be applied uniformly across all public enterprises or varied according to individual circumstances.

One school of thought is that a common rate of return requirement is generally appropriate. For example, the Rae Report (1983, p. 98) argued:

On resource allocation grounds it is hard to see any compelling reason why in the longer term the rate, however it is determined, should vary among authorities although there are clearly a number of social and equity considerations that must be addressed and there would obviously

126

be some problems in implementing a general rate of return requirement.

The present rate of return targets applying to major public enterprises in Victoria are set at a uniform rate. However, the government has conceded that non-uniform rates might be applied in the future (Victorian Department of Management and Budget 1986, p.14):

The approach adopted in Victoria in the implementation stage is to use a common target rate of return which authorities should aim to achieve over different time periods. Refinements can be considered and introduced if appropriate at some later time. For example, different target rates could be specified for each authority, in recognition of the particular characteristics - average interest rates, appropriate risk

margins and debt: equity ratios - of the individual authorities.

The approach proposed by the Commonwealth Government also involves some degree of fine-tuning of rate of return targets (Minister for Finance 1987, p. 21):

In considering financial targets the Minister will have regard for the trading conditions in which the enterprise operates, for its relative commercial and market strengths and for the extent to which, on the basis of Government policy, it is required to meet any ‘community

service obligations’ and to observe residual central controls designed to achieve other non-commercial objectives of the Government which adversely affect its profitability.

The view that rate of return targets should be varied between individual authorities is perhaps the majority one - but considerable differences of opinion exist as to which factors justify an adjustment to the ‘standard’ rate. Issues raised by the Department of Finance (1988a, 1988b) include:

. risk;

. community service obligations; and . specific market and general economic conditions.

. Risk

In the private sector, the rates of return required by investors will increase with the perceived risk of the returns offered by alternative investments. It is risk-adjusted rates of return (ie nominal rates of return less an appropriate risk premium) which will be equalised across investments in a competitive market. Clearly, if differences in

risk between public and private sector investments are ignored, an inefficient allocation of resources between the two sectors will result. This implies that there is a need to specifically adjust required rates of return in public sector projects according to the risk of the investment in order to place them on the same basis as those in the private sector (where risk is already implicitly incorporated by the market).

Some commentators have argued that this adjustment factor should be zero, since governments undertake a large number of projects and the risks should offset each other. This argument, first raised by Samuelson (1964), was supported by Arrow and

127

Lind (1970) who stated that public sector projects should be evaluated at the risk-free rate provided that the distribution of returns being evaluated was statistically independent of returns to the economy as a whole. Others, such as Officer (1983, p.48) have argued that the idea that government can diversify away all risk is an absurd notion since governments are unable to control all relevant variables and in consequence there is some risk that is non-diversifiable.

More recently, the Department of Finance (1987) has reconciled these seemingly opposing views by making use of the Capital Asset Pricing Model (CAPM) to show that returns to public sector projects independent of the economy as a whole should be discounted at the risk-free rate. Thus, the argument is not whether rates of return should be adjusted for risk, but what the level of risk in public sector projects is.

If however, it is accepted that the government cannot diversify away all risk, and that rates of return for individual enterprises should be individually adjusted for risk, some estimate of this risk premium needs to be made. As noted by the Department of

Finance (1988a), one approach would be to use the average risk premium implicit in the rate of return in the corporate sector as a whole (using the CAPM model). However, while uncertainties about economy-wide variables such as prices, wages growth and demand may affect public and private enterprises equally, other risks of a

more industry-specific nature may cause the risk profile of returns on a public sector investment to vary significantly from the average.

To account for such enterprise-specific risks, the Department suggests that rates of return earned by private enterprises in the same field of activity be used as the benchmark. However, a major problem with this approach is that many public enterprises do not face direct private sector competition. Under these circumstances, reference could be made to returns earned by similar enterprises operating in other States or countries. There are difficulties in making appropriate comparisons. In particular, different accounting conventions (eg valuation of assets), operating constraints (eg community service obligations), taxation provisions, or general market environments may all undermine the validity of the comparison, and be difficult to isolate and to measure. When domestic or overseas comparisons are unavailable or considered unreliable, the Department suggests that a fallback position would be to take the average pre-tax return in the private sector and adjust it for any demonstrable differences in risk or tax status (actual or imputed) for the activity relative to the corporate sector norm.

. Community service obligations

As noted in section L2, there is a case for adjusting financial targets to allow for the cost of delivering CSOs when setting financial targets. While the Commonwealth Government has apparently accepted this view (Minister for Finance 1987), the mechanism for achieving it has not been specified. A review of rate of return targeting in Victoria (Victoria, Parliament 1988) recommended that the Government consider setting different rate of return targets for each enterprise which take CSOs

into account. The Department of Finance advocates, as a general rule, an adjustment not to the rate of return target itself, but to the accounting base on which it is measured. For example, in the case of CSOs, it suggests adding the net loss of revenue attributable to the CSO back into the figure forming the numerator of the rate of return calculation, and increasing the valuation of assets forming the

denominator by the difference between replacement value and recoverable cost as shown in supplementary statements to the accounts. Such an approach, involving the explicit valuation of CSOs, would have important transparency advantages over making ad hoc adjustments to the target rate of return.

. Specific market and general economic conditions

Some commentators such as the Department of Finance appear to support adjustment of the rate of return target itself in some circumstances (eg to reflect the business climate that is expected to apply over the period of the target and to reflect the particular risk and tax status of the enterprise). However, the ‘risk-related’ factors would already be incorporated into the benchmark rate derived under the SOC

approach.

Nevertheless, as noted by Officer (1986, pp. 8-9), in interpreting ex-post rates of return:

It is not critical if the cost of capital (required rate of return) set by management turns out to be significantly different from the rate of return which was ultimately achieved. What is critical is how the returns from a company’s investments compare with the returns

achieved by alternative investments... The implication is that it is misleading to set a required rate of return as a control mechanism and then conclude that the system has got out of control if that required rate of return is not met. The firm or investments have only ‘gone out of

control’ if the return that they have realised has fallen relative to returns realised for investments of the same risk class. It is only in these circumstances that investment and perhaps management decisions should be changed to ensure a more efficient allocation of resources. In

short, the appraisal of the firm by the market place is not done in a vacuum, it is done in the context of the return achieved or expected by alternative investments.

Moreover, in interpreting an individual enterprise’s rate of return as a measure of performance, Davies and Mclnnes (1982, p. 32) have warned that:

Some variations in income will arise even if there is no change in the efficiency with which a concern is being run. Such variations can stem from changes in market conditions completely outside the control of management. The managers of nationalised concerns have less scope

for manoeuvrability than managers in the private sector. There is no escape (for concerns whose activities are narrowly prescribed by statute) from stagnating or declining markets into the growing and profitable sectors of the economy.

. Assessment of the uniformity issue

Whilst adjustments to the rate of return may be justified in a conceptual sense, in practice it is important that such adjustments be carefully assessed and fully documented. There is a danger that, once exceptions to the base rate are allowed, excessive allowances for the cost of CSOs, or excessive discounting of the rate of

return target, could provide a cloak behind which inefficiencies might occur. Moreover, the scope for political interference in the management of the enterprise may be increased.

Even if these problems are avoided, a more fundamental question is whether the benefits from such fine-tuning outweigh the costs. It is quite possible that the costs associated with the process of examining the individual circumstances of every enterprise (both the direct administrative costs and the possible efficiency costs alluded to in the preceding paragraph) may outweigh any allocative inefficiencies which would flow from conceptually inappropriate rate of return targets.

Reviewing rate of return targets

A further question that arises in rate of return targeting is how often the target should be reviewed. Officer (1986) has suggested that the lag between variations to the rate as they are set in a discontinuous process by bureaucracy is a major disadvantage of rate of return regulation (indeed, of any regulation). In the private sector, required

rates of return are continuously adjusted on the basis of realised rates. The failure of a regulatory authority to adjust to changing market rates may result in rate of return requirements having varying degrees of effectiveness on the enterprise, causing undesirable effects on the enterprise’s planning and budgetary policy. In similar terms, the Department of Finance (1988a, pp. 14-15), notes that:

The return earned by public enterprises is likely to vary with changes in business conditions in the industries concerned and as a result of variations in overall business activity. It would be undesirable, and could well be counter-productive in competitive markets, if enterprises were forced to vary mark-ups in order to counteract the effects of general business fluctuations merely to achieve a fixed rate of return target that did not reflect these fluctuations.

The Department suggests that these problems can be minimised by specifying the rate of return in terms of an average achieved over a period of years, rather than year-by­ year targets. However, recognising that the target can be rendered progressively less relevant by changes in the structure of the economy if the period is too long, it

suggests a three year horizon as a reasonable compromise. A similar approach has been adopted by the Victorian Department of Management and Budget (1986, p. 14) which stresses that:

Above all, it should be appreciated that the policy of setting prices with reference to a desired rate of return on assets is long term in nature. It is not anticipated that prices should be varied in a way which would ensure that the target rate of return will be met in every year... Furthermore, periodic reviews of the target rate of return will be necessary to ensure that the target remains both realistic and appropriate.

A parliamentary review of the Victorian Government’s rate of return guidelines (Victoria, Parliament 1988) supported this approach, suggesting that ‘some further consideration of the 4 per cent real rate of return target may be appropriate’ and that ‘some consideration could, perhaps, be given to setting differential rates for each

130

authority’. A similar process of periodic review is evident from a recent announcement by the UK Treasury that, in view of the higher rates of return now being earned in the private sector, the required rate of return on new investments in nationalised industries is to increase from 5 to 8 percent (Economist, 1989).

L4 Non-financial targets

Whilst appropriately-based financial targets can play an important part in measuring and encouraging public enterprise efficiency, it is apparent that they have limitations which suggest there are dangers in relying exclusively on them to monitor performance of public enterprises. To a large degree, however, these limitations arise

because of special characteristics applying to public as distinct from private enterprises. In particular, the existence of non-commercial objectives and lack of direct competition in the markets in which many public enterprises operate suggest the need for financial targets to be supplemented by appropriate non-financial targets to achieve three principal objectives:

. enable evaluation of performance with respect to non-commercial objectives (eg CSOs) by reference to relevant non-financial measures (eg percentage of a target group for a welfare or social service receiving the benefit);

. ensure that financial targets cannot be met simply by raising prices or reducing the quality of service, by monitoring these aspects of commercial performance; and

. enable the introduction of indirect competitive pressures by way of ‘yardstick’ comparisons with other similar enterprises, either in Australia or overseas.

Non-commercial objectives

In relation to evaluating performance in achieving non-commercial objectives, it is desirable to set targets closely related to these objectives. One prerequisite, therefore, is that these objectives be clearly specified to facilitate the development of output measures to indicate whether targets are being achieved. When combined

with an explicit financial costing (as discussed previously), this would seem to provide a sound basis for assessing the efficacy of existing approaches to delivering CSOs compared with feasible alternatives.

Non-financial aspects of commercial performance

As noted in sections LI and L2, the effectiveness of financial targets in promoting better performance is likely to be increased if quality of service and pricing variables are also monitored. The monitoring of these variables may also be necessary to lessen the scope for abuses of market power where governments have given public

enterprises exclusive trading rights through the imposition of restrictions on competition.

Quality of service variables

A limited number of public enterprises in Australia already publish information relating to these non-financial aspects of their performance (eg reliability of service

131

indicators), and conduct surveys of their consumers to isolate areas of inadequate performance or to identify opportunities for providing a wider range of services or improving existing services (eg greater emphasis on timeliness). To date, there appears to have been little external involvement by governments in the selection of appropriate indicators or their use in the formal accountability/monitoring process. Consequently, little information on these variables is available and there is a risk that enterprise managers opting to report performance against such measures will report only those indicators which show their performance in the best light.

In the absence of direct competition, one means of trying to replicate market disciplines on non-price aspects of performance could be to establish minimum reporting guidelines in this area. Such information could also be used to assess whether the range and quality of services offered by an enterprise was diminished as a

means of achieving a financial target. It may be worth considering the possibility of involving consumer or user representatives in the development of non-financial targets and in ongoing communication of their interests to enterprise managers through formal consultation processes. A number of port authorities, for example,

have sought to develop such processes.

Pricing targets

The concern that public enterprises with market power may be able to meet financial targets simply by raising prices suggests the need for prices to be monitored. In Victoria, for example, the authorities currently subject to rate of return targets are required to achieve them while also containing increases in their charges below CPI

movements. A recent review (Victoria, Parliament 1988, p. 124) suggested that the effectiveness of rate of return targets would be greatly strengthened if prices set by authorities producing marketable goods and services for a direct charge were subject to surveillance by an independent prices authority operating within the framework of the rate of return targets set by the Government. Increasing interest is being shown in more sophisticated price capping arrangements, notably the so-called Price Index minus X (PI-X) mechanism.

In the UK this mechanism has been used to regulate British Telecom, British Gas and the Airports Authority. It has been recently adopted in the US by the Federal Communications Commission for its international carrier AT&T. In July 1989 a similar price-capping arrangement was applied to a basket of ‘specified reserved services’ provided by Telecom and OTC. The price cap, which will operate in conjunction with a new financial target requirement, will restrict Telecom and OTC from increasing the prices of the basket of services by more than the CPI minus 4 per cent.

The PI-X mechanism was initially intended as a substitute for a rate of return requirement. PI-X regulation limits or caps average price increases and allows the enterprise to take as profits any productivity increases above X. However, in the longer term the profitability of the enterprise is reviewed and the X factor altered to

ensure that any future excess profits are passed on to consumers. This potential to earn higher profits through productivity increases above X is a crucial incentive for the success of the mechanism.

The optimal duration of the price cap formula is subject to a trade-off between conflicting incentives. With a short period price cap or where there is provision for an early review, it is likely that higher profits will be passed on to consumers through an increase in the size of X. Hence the shorter the duration of the formula the less the

incentive for the enterprise to minimize costs. If a longer review period is utilised or if there is no provision for an early review, uncertainty is greater and the price cap may not be workable because of sudden changes in the market. Hence the duration of the formula and the provision for an early review require some balance between the likely effects of the options. The government has decided to adopt a three-year period for the Telecom and OTC price caps. Because OTC is vulnerable to exchange

rate movements and to other international factors, Austel may at OTC’s request advise the government on the need to adjust the value of X prior to expiry of the three-year period.

If the PI-X mechanism is applied to a profit maximising private operator, no other regulatory measure is required (Littlechild 1988). However, when it is applied to a public enterprise such as Telecom, the incentive structure may not be sufficient to ensure productivity gains. In this instance, as in the case of Telecom and OTC, PI-X

may be used in conjunction with a minimum rate of return target. The Bureau of Transport and Communications Economics (1989) has also suggested that for public enterprises achievement of all available efficiency gains under the combination of a PI-X mechanism and rate of return requirement may also require some form of

managerial incentive such as an income bonus based on achieving rates of return in excess of those anticipated.

An advantage of the PI-X mechanism over the simple CPI price capping arrangement is that it can be better targeted to the relevant industry/enterprise. This may be achieved by relating the price and productivity components of the price cap to a price index for the industry’s product mix and assessed potential for productivity

improvement respectively. However, the difficulties in establishing an appropriate allowance for these components should not be underestimated. Because an index of industry prices may be within the control of the enterprise, it may be preferable to use the CPI rather than an industry price index. Indeed, the CPI is used as the benchmark in the UK and for Telecom and OTC. In addition to overcoming the strategic behaviour of the regulated enterprise, the CPI has a number of other advantages. It is

readily available, administratively simple to apply and publicly acceptable.

The choice of X in the price cap relies on the profitability or target rate of return for the enterprise and the potential for efficiency gains. If X is set too high then the target rate of return cannot be achieved, alternatively if the enterprise can easily meet its price target, it will be under little if any pressure to increase efficiency.

Productivity improvements with the introduction of PI-X rely on the existence of some opportunity for cost reduction before its implementation. If the enterprise is operating efficiently prior to the implementation of the price cap, reductions in average costs may be limited. Technological constraints in some industries may also

limit the potential for on going improvements in productivity. However, as discussed further below, it may be possible to promote more efficient pricing and production outcomes by way of a ‘competitive yardstick’ in an industry otherwise isolated from competition by market restrictions.

1 7 7

In a similar fashion to the way in which financial performance by other enterprises in the same industry overseas or interstate could be used as a benchmark for setting financial targets, non-financial measures of performance achieved elsewhere could be utilised to promote efficiency through ‘yardstick’ competition (Kay & Vickers 1988).

For example, in setting PI-X targets, an X factor set with reference to ‘best practice’ overseas or interstate could be used in an attempt to increase the discipline such a target imposes on performance. As part of this process, enterprises would need to establish reliable data bases to aid the development of agreed targets and the subsequent monitoring of performance against these targets. Variables covering the technical performance of county councils in the NSW electricity industry have been

included in recent performance agreements with the State government.

One difficulty with this approach, however, is in measuring productivity gains actually achieved. Typically, enterprises provide information on labour productivity, rather than total factor productivity, with the latter being fundamental to the design and monitoring of improved price capping arrangements as well as assessments of operational efficiencies. Telecom has recently sought to measure and assess its total factor productivity performance as part of its CPI-X price capping arrangement. Another problem is the ability to compare like with like, and the danger that making adjustments for particular circumstances of enterprises also provides scope for

manipulating the target to make it less effective.

Implementation of non-financial targets

In the past, non-financial targets have played little role in performance monitoring processes applied to public enterprises in Australia. In more recent times, however, there has been a growing realisation of the need for non-financial measures of performance, particularly as complements to financial targets. Shand (1989 p. 182) for example, has stated that:

...there have been marked improvements in GBE annual reporting in all Australian Governments in recent times. In particular there has been a much greater emphasis on reporting on performance both of a financial and efficiency/production type. Both Victoria and NSW have emphasised the importance of non-financial preformance indicators.

Performance information or indicators of a high standard are currently provided in many GBE annual reports including those of Telecom and State Electricity Commission of Victoria, which have both in the past won good awards in the Australian Institute of Management Annual

Report Awards Scheme.

Given that the trend towards non-financial performance targets is relatively new, the question arises as to how this process can be developed so as to ensure that such targets are effective. One important requirement would seem to be ongoing involvement in the process by appropriate representatives of governments (see Xavier

Yardstick’ competition

134

For example, the UK White Paper (1978) on nationalised industries called for the production and publication of key performance indicators, including; Valid international comparisons, together with performance indicator of service and labour

productivity, where relevant’. However, several authors (Likierman 1979; Perks & Glendinning 1981) have noted that the initial response of the UK nationalised industries to this call was very patchy. Mayston (1985) noted that after six years in

which to produce relevant performance indicators, the response remained both incomplete and uneven, particularly with regard to international comparisons and quality of service indicators. Moreover, as noted by Barrow and Wagstaff (1989), government departments have acknowledged that the indicators developed to date,

focusing on input usage and throughput rates, shed relatively little light on the question of efficiency.

This suggests that in implementing non-financial targets, their effectiveness is likely to be increased if:

. an ad hoc approach is avoided by selecting indicators that are directly relevant to the enterprises’ objectives;

. there is external involvement in their selection and subsequent monitoring;

. adequate and reliable data is available to meaure non-financial targets; and

. targets are set in the context of a well-designed analytical framework.

Considerable research has been undertaken, particularly in the U K to try to improve on existing performance indicators by using statistical and other quantitative techniques (eg Barrow & Wagstaff 1989). There would seem to be scope for drawing on this research and UK experience to date, to improve the effectiveness of non- financial performance targets. Even if more effective targets were to be

implemented, however, their use should be conditioned by their inherent limitations such as measurement difficulties, scope for manipulation and difficulties in developing useful comparative data for assessing performance or designing meaningful yardsticks to promote competition.

L5 Summary

A hallmark of public enterprises is that they often operate in environments characterised by limited discipline on performance exercised through product or capital markets. A need therefore arises for administratively based mechanisms to perform these functions.

Financial targets, because of their widespread use in the private sector as tools for decision making and performance evaluation, have been viewed by some as potential substitutes for the disciplines associated with markets (eg shareholder dissatisfaction, takeovers and bankruptcy). However, setting financial targets does not expose public

enterprises to the same pressures as those faced by private firms. The absence of competitive pressures means that administratively based mechanisms are subject to various limitations. Moreover, public enterprises may be required to discharge various non-commercial objectives as part of their charter.

135

In using financial targets these limitations need to be recognised and adequate account taken of factors that could cause such targets to be inadequate indicators of performance and/or to lessen their effectiveness in improving the incentives for managers to price and produce efficiently. Important considerations in this regard include: the need for appropriate asset valuations; the need to take account of CSOs; and the need to remove any remaining factors which advantage or disadvantage public enterprises relative to private sector counterparts. Non-financial performance

measures have a role in supporting financial measures through the monitoring of price/quality of service variables for public enterprises with monopoly power. In addition they are needed to monitor the effectiveness with which non-commercial objectives (eg CSOs) are discharged and to promote greater production efficiency through yardstick competition (eg international or interstate comparisons of technical efficiency).

Where such factors are successfully addressed, setting rate of return targets for public enterprises could enhance economic efficiency. Nevertheless, governments face a number of conceptual and practical difficulties in setting targets and interpreting performance in relation to them.

The difficulties in designing effective administratively based mechanisms to provide incentives for public enterprises to perform better are severe and may call for a re­ assessment of the wider institutional environment in which these enterprises operate. Indeed, given the difficulties of devising effective accountability/monitoring mechanisms (demonstrated by the UK experience over the last twenty years or so), the questions of modifying or removing barriers to contestability of markets and ownership need to be seriously considered.

If greater exposure to market forces is rejected, the question is how administratively based mechanisms can best be applied to promote efficient production and pricing decisions. Of the two types of financial targets examined, rate of return targets seem preferable. However, there are dangers in relying on any one measure - such as a rate of return on capital - as the sole indicator of performance. Public firms generally have objectives encompassing commercial as well as non-commercial functions, and the extent to which these objectives have been achieved cannot be assessed by reference to a single measure. Moreover, without market-based disciplines, managers may have greater scope to manipulate resources to achieve specified targets. Nevertheless, rate of return targets may well prove to be a more effective approach to the problem of public enterprise monitoring and accountability, particularly if applied in conjunction with appropriate non-financial targets and other complementary reforms (eg use of current cost accounting, adoption of a system of rewards and sanctions, and improved investment appraisal techniques).

REFERENCES

Albon, R. 1985, ‘The Effects of Financial Targets on the Behaviour of Monopoly Public Enterprises, Australian Economic Papers, vol. 24, no. 44, June, pp. 54-65.

Arrow, K.J. & Lind R.C. 1970, ‘Uncertainty and the Evaluation of Public Investment Decisions’, American Economic Review vol. 60, no. 3, June, pp. 364-78.

Averch, M. & Johnson L.L. 1962, ‘Behaviour Of The Firm Under Regulatory Constraint’ American Economic Review, vol. 52, no. 5, December, pp. 1052­ 9.

Bailey, E.E. & Malone, J.C. 1970, ‘Resource Allocation and the Regulated Firm’, The Bell Journal of Economics and Management Science, vol.l, no. 1, Spring, pp. 129-42.

Ball, R. & Davis, J. 1984, Appropriate Financing and Investment Policies For Telecom Australia, Report prepared by the Centre of Management Research and Development for Telecom Australia and included in a submission to the Prices Surveillance Authority, March.

Barrow, M. & Wagstaff, A. 1989, ‘Efficiency Measurement in the Public Sector : An Appraisal,’ Fiscal Studies, vol. 10, no. 1, February, pp. 72-97.

Bates, R. & Fraser, N. 1974, Investment Decisions In the Nationalised Fuel Industries, Cambridge University Press, London.

Baumol, W.J. & Klevorick, A.K. 1970, ‘Input Choices and Rate-Of-Return Regulation: An Overview of the Discussion’, The Bell Journal of Economics and Management Science, vol. 1, pp. 162-90.

Boyes, W.J. 1976, ‘An Empirical Examination of the Averch and Johnson Effect’, Economic Inquiry, vol. 14, no.l, pp. 25-35.

Bureau of Transport and Communications Economics 1989, CPI-X Price Control in Australian Telecommunications, Paper no. 137, May.

Clare, R.W. 1982, ‘Discounting in Australian Public Sector Project Analysis’, Paper presented to the 11th Conference of Economists, Adelaide, August.

Courville, L. 1974, ‘Regulation and Efficiency in the Electricity Industry’, The Bell Journal of Economics and Management Science, vol. 5, no. 1, Spring, pp. 53-74.

Davies, J.M. & Mclnnes, W.M. 1982, ‘The Efficiency and the Accountability of UK Nationalised Industries’, Accounting and Business Research, vol. 13, no. 49, Winter, pp. 29-41.

Department of Finance 1987, The Choice Of Discount Rate For Evaluating Public Sector Investment Projects, A Discussion Paper, November.

137

1988a, Rate o f Return Reporting for Government Business Enterprises (GBEs): Implementation Issues, A Discussion Paper, May.

1988b, Rate of Return Reporting for Government Business Enterprises (GBEs): A Conceptual Approach, Paper prepared for BTCE Round Table Meeting: Government Business Enteprises, 14 July.

Economist, 1989, ‘State Industries: The Rate for the Job’, 8 April, p. 72.

Ergas, H. 1986, Telecommunications and the Australian Economy: Report to the Department of Communications, AG PS, Canberra

Gravelle, H. 1976, ‘Public Enterprises Under Rate-Of-Return Financial Targets’, Manchester School o f Economic and Social Studies, vol. 44, no. 1, March, pp. 1-16.

Gravelle, H. & Katz, E. 1976, ‘Financial Targets and X-Efficiency in Public Enterprises’, Public Finance, vol. 31, no. 2, pp. 218-34.

Heald, D. 1980, ‘The Economic and Financial Control of UK Nationalised Industries’, The Economic Journal, vol. 90, no. 358, June, pp. 243-65.

Inter-State Commission 1989, Waterfront Investigation: Conclusions and Recommendations, 2 vols, AGPS, Canberra, March.

Johnston, M.N., Parkinson, M.L. & McCray, L.T. 1984, ‘A Rate of Return Requirement For Public Authorities’, paper presented to the 11th Conference of Economists, Adelaide, August.

Jones, F.W. 1983, Input Biases Under Rate o f Return Regulation, Garland Publishing Inc, New York.

Kay, J. & Vickers, J. 1988, ‘Regulatory Reform in Britain’, Economic Policy, October, pp. 286-351.

Kiss, F. 1988, ‘Productivity and Price Cap Regulation’, paper presented to International Telecommunications Society Seventh International Conference, Massachussets Institute of Technology, Cambridge, Mass, June 29-July 2.

Likierman, A. 1979, ‘Performance Measures for the Nationalised Industries’, Accountancy, May, p. 117 and October, pp. 91-3.

Littlechild, S. 1988, ‘Economic Regulation of Privatised Water Authorities and Some Further Reflections’, Oxford Review o f Economic Policy, vol. 4, no. 2, pp. 40-68.

Mayston, D.J. 1985, ‘Non-Profit Performance Indicators in the Public Sector’, Financial Accountability & Management, Summer, pp. 51-73.

Minister for Finance 1987, Policy Guidelines for Commonwealth Authorities and Government Business Enterprises, White Paper, AGPS, Canberra.

Officer, R.R. 1983, ‘Accountability and Monitoring the Performance of the State Enterprise’, in State Enterprise and Deregulation, Centre of Policy Studies, Monash University, Special Study no. 5, pp. 43-59.

138

____ 1986, Financial Targeting for Public Enterprise: The Criterion of a Rate of Return to Capital, University of Melbourne, Graduate School Of Management, Working Paper no. 70, May.

Perks, R. & Glendinning, R. 1981, ‘Performance Indicators Applied to the Nationalised Industries’, Management Accounting (UK), October, pp. 24-26 and December, pp. 28-30.

Petersen, H. 1975, ‘An Empirical Test of Regulatory Effects, The Bell Journal of Economics, Spring, pp. 111-26.

Price Commission 1978, South of Scotland Electricity Board: Price Increases in the Supply of Electricity, HC 535, HMSO, London.

Rae Report. See Senate Select Committee on Statutory Authority Financing 1983.

Rees, R. 1976, Public Enterprise Economics, Weidenfeld and Nicholson, London.

Samuelson, P.A. 1964, ‘Discussion’, American Economic Review, ‘Papers and Proceedings, vol. 54, no. 3, May, pp. 93-6.

Scherer, F.M. 1980, Industrial Market Structure and Economic Performance, 2nd Edition, Houghton Mifflin Co., Boston.

Senate Select Committee on Statutory Authority Financing 1983, Statutory Authorities o f the Commonwealth: Financing, vol. 1, Report, (P. Rae, Chairman), AGPS, Canberra, September.

Senate Standing Committee on Finance and Government Operations 1982, Statutory Authorities o f the Commonwealth, Fifth Report, (P. Rae, Chairman), AGPS, Canberra, September.

Shand, D. 1989, ‘Corporatisation or Privatisation Will Alter the Legal Requirements for External Reporting and Will Have Implications for External Accountability’, Paper presented to the Australian Society of Accountants National Accountants in Government Convention, Melbourne, March 2-4.

Spann, R. 1974, ‘Rate-Of-Return Regulation and Efficiency in Production: An Empirical Test of the Averch and Johnson Thesis, Bell Journal of Economics and Management Science, Spring, pp. 38-52.

The Treasury 1983, Public Monopolies: Telecom and Australia Post, Submission to the Committees of Inquiry into Telecommunications Services in Australia and into the Monopoly Position of the Australian Postal Commission, Treasury Economic Paper no. 10, AGPS, Canberra.

United Kingdom 1961, The Financial and Economic Obligations of the Nationalised Industries, White Paper, HMSO, Cmnd 1337, London, April.

____ 1967, Nationalised Industries : A Review of Economic and Financial Objectives, White Paper, HMSO, Cmnd 3437, London, November.

____ 1978, The Nationalised Industries, White paper presented to Parliament by Chancellor of the Exchequer, HMSO, Cmnd 7131, London.

Victoria, Department of Management and Budget 1986, Public Authority Policy and Rate o f Return Reporting, Information Paper no. 1, October.

139

Victoria, Parliament 1988, Economic and Budget Review Committee: Controls over Commercial Authority Debt Levels, Report no. 24, May.

Xavier, P. 1989a, ‘Economic and Commercial Principles for Determining Government Trading Enterprise Charges: How Compatible Are They?’, Paper presented to a seminar on Electricity Tariffs convened by the NSW Inquiry into electricity tariffs, January.

____ 1989b, ‘Minding Everybody’s Business : Performance Indicators for Australia Post’, paper presented to the 1989 Australian Conference of Economists, University of Adelaide, 10-13 July.

Xavier, P. & Graham, B. 1987, ‘Financial Targets for Commonwealth Business Enterprises: Are They Appropriate and How Should They be Determined and Measured?’, Paper presented to the 16th Conference of Economists, Surfers Paradise, 23-27 August.

APPENDIX Μ : VALUING THE ASSETS OF PUBLIC ENTERPRISES

Public enterprises in Australia typically employ accounting systems based on historical cost valuations o f assets. This appendix describes the main problems which reliance on historical cost accounting create both for rational economic decision-making by managers and performance

evaluation by governments. These deficiencies could be reduced by supplementing existing systems with current cost accounting (CCA) systems which more closely reflect economic costs. To date, however, progress towards implementing CCA in public enterprises in Australia has been generally slow and ad hoc. Experience with nationalised industries in the

UK suggests that introduction o f CCA in Australia could be facilitated by the consideration o f simplified CCA systems and a program of public education. There is a strong case for adopting a coordinated approach through the development and application o f agreed common guidelines.

Ml Introduction

There has been considerable debate about the need for improved accounting and reporting procedures by public enterprises in Australia and elsewhere, in the interests of obtaining better information for decision-making by managers and performance evaluation by governments. In recent years the issues have been discussed in papers

dealing with both conceptual and practical aspects (eg Barton 1983; Graham & Xavier 1987). Inquiries into the relevance of traditional accounting practices for the purpose of economic evaluation have been held in a number of countries, including Australia, the United Kingdom, the United States, Canada and New Zealand.

Whilst various deficiencies in public enterprise accounting have been identified in these reports, particular attention has focused on the issue of asset valuation. Specifically, there has been much interest in the potential for improving performance by adopting accounting procedures based on current rather than historical cost.

Support for valuing the assets of public enterprises according to current cost principles reflects a recognition that the efficient use of the community’s resources requires decisions based on the economic costs of using resources in alternative activities. If financial information on public enterprises bears little relationship to

economic reality, decisions on pricing, supply and investment are likely to lead to a misallocation of resources. In this regard, current cost asset valuation systems provide more relevant measures of costs for the purposes of pricing and investment decisions than do historical cost systems normally used by public enterprises throughout Australia. 1

1 In Australia, the issues have been reviewed in the Mathews Report (1975); and the Rae Report (1983). The most notable recent overseas review is the Byatt Report (1986) in the UK. Earlier reviews include the Richardson Report (1976) in New Zealand and the Sandilands Report (1975) in the UK.

141

Providing information useful to managers in making economic decisions is, however, only one objective of an accounting system. Others include the provision of financial information to various users, some of which (eg taxation authorities) are entitled to

have information prepared to satisfy their specific requirements. Reporting to users is governed by a number of conventions, regulations and legal requirements. Simultaneous satisfaction of these objectives may, in fact, require the maintenance of accounting information utilising several different bases for valuing assets, reflecting

the varying requirements of users. Consequently, this appendix does not necessarily advocate the abandonment of existing historical cost systems. Rather, its purpose is to draw attention to the important need to have available and to publish additional information based on valuations of the assets of public enterprises in a way which is conducive to better economic performance. In essence, the aim is to emphasize the value of maintaining information on changes in the economic value of an enterprise’s

asset base.

Whilst CCA is not practised widely in the private sector in Australia, having accounting systems which provide information on economic costs is especially important in the case of public enterprises. First, the typically long lives and capital­

intensive nature of many of the assets of these enterprises mean that accounting systems based on historical costs, by understating the true cost of capital services used in a given period, give a misleading picture of performance. More rapid rates of inflation since the early 1970s have exacerbated this problem. Because of their size,

inappropriate pricing/investment decisions on the part of public infrastructure providers are likely to have serious consequences for performance of the economy as a whole. Second, there is no share price reflecting the market’s day-to-day assessment of the enterprise’s performance in managing resources, and no avenue by which

market disciplines (eg through sale of shares, takeovers etc) can operate to ensure resources are used efficiently. Thus, the accounting records of public enterprises are critical to internal management decision-making as well as for external evaluations of performance. They become even more critical when the enterprise faces little or no competition in its product market, which might otherwise provide a check on inefficient pricing or investment decisions.

In the last few years, a growing concern with efficient resource utilisation and with growing operating deficits has led to proposals that public enterprises be required to earn a rate of return on capital which approximates that earned by private sector counterparts (see Appendix L). Reasonably up-to-date valuations of assets are required for achieving accountability in terms of such rate of return targets.

The main problems associated with historical cost accounting systems, both for internal decision-making within public enterprises and for external monitoring of performance, are discussed in section M2. Section M3 outlines the main alternative asset valuation systems. A brief overview of current developments in Australia and issues associated with the implementation of current cost accounting are discussed in section M4, drawing on experience in the UK and Australia. Finally, section M5 offers the Commission’s assessment of the asset valuation issue and suggests some possible approaches to reforming the accounting information systems used by public enterprises in Australia.

142

M2 Problems with historical cost accounting valuations

Historical cost accounting principles value assets according to the written-down original purchase price or construction cost. Problems with this approach stem from its inability to provide relevant information for current or future economic decision­ making. The general deficiencies of historical cost accounting in this regard have

been described (Graham & Xavier 1987, p. 19) as follows:

It has long been recognised that, particularly where inflation is significant, accounts drawn up on historical cost conventions are misleading especially where assets are long-lived. Balance sheet figures of original cost do not represent the value of assets to the business, profit and financial trends are misrepresented. If accounts are to show

resource use and economic performance they must allow for general inflation, for fluctuations in specific prices and costs, and for technical progress resulting in changes in the value of capital equipment.

The failure of historical cost systems to take changes in prices into account leads to problems both for internal decision-making within enterprises and for external review.

The internal public enterprise environment

Historical cost accounting poses a number of inter-related problems for managers of public enterprises. The absence of an appropriate information base acts as a major constraint on the ability of managers to make properly informed decisions on pricing, investment and dividend payments to government.

Misleading balance sheet valuations

At the heart of the problem is the fact that, under historical cost accounting, the balance sheet will fail to reflect the level of the inflation-adjusted economic cost of resources employed in the business. For example, suppose an asset is bought in year 1 for $100 000. Given an annual inflation rate of 10 per cent, and assuming the

replacement price of an equivalent asset increases at the same rate, the value of a similar new asset at the start of year 2 (in year 2 dollars) would be $110 000. The actual value of the original asset will be less than this, however, because it has experienced one year of wear and tear. If the annual depreciation associated with

such wear and tear was 5 per cent, and was fully reflected in the market price of the asset, its market value would be $104 500. Thus, even after just one year, the historical cost valuation (net of straight-line historical cost depreciation of $5000) of $95 000 would represent a significant undervaluation of the economic value of the

asset. Clearly, if this type of figuring were extended to cover the full life of public sector assets, which often include items with lives exceeding 20 years, the cumulative effect of failing to take price changes into account is likely to yield misleading information. The divergence between historical cost valuations and true economic

costs may be considerable (with the divergence being greater the higher the rate of inflation and the longer it has been since the asset has been revalued - if at all).

Some examples of the large divergences which can arise between the written-down book value of assets under historical cost valuations in actual use and those under

143

various types of current cost systems are set out in Table Ml. In the case of Telecom, this understatement of the average written-down cost of assets in service amounts to some $4.8 billion. Similar divergences were documented in the Rae Report in 1983. For example, the report noted that an asset revaluation performed by OTC in 1982 resulted in the book value of fixed assets being increased by some 72 per cent from $109 million to $189 million (Rae Report 1983, pp. 265-70).

These examples demonstrate that use of historical cost valuations will, in general, result in a substantial understatement of the economic cost of assets used in a business.2

Erosion of the capital base

Setting prices based on historical cost measures of depreciation will generally result in an erosion of the capital needed to maintain the operation of a business. Suppose that the asset quoted in the previous example has an expected life of 20 years. Under a straight-line depreciation schedule, the annual accounting depreciation charge on the historical cost book value of $100 000 would be $5000. In contrast, the true cost of using the asset for the first year would be $5500, representing the fall in the market value of the asset due to physical deterioration between year 1 and year 2.3 Failure to recover the true cost of utilising the asset means that at any point in time the enterprise will not have recovered sufficient funds to enable it to replace the asset at inflation-affected market prices. Thus, the market price of the one year old asset ($104 500) plus the $5000 recovered during the year would be insufficient to purchase an equivalent new asset which now costs $110 000. This gap would get progressively wider in later years although the problem may only become apparent when the time comes to replace the asset.

The widespread use of historical cost valuations makes this a common problem throughout the public sector. For example, the Commission’s studies (see Volume 4), identify it as a primary problem with traditional accounting systems used by many electricity and rail authorities in Australia.

It should be noted, however, that introduction of more appropriate accounting systems will not, by itself, be sufficient to ensure that the capital base is not eroded. This can be seen by extending the hypothetical example by another year. In year 2,

2 However, in some cases rapid technological change or grossly inappropriate investment decisions may result in historical cost valuations exceeding the opportunity or economic cost of the asset concerned (or at least partially offsetting the general price movement effect). For example, the market value of an obsolete asset may be close to zero. 3 Under the assumptions outlined in the previous example, the depreciation charge in the first year

under a current cost system would indeed be $5500 (ie $110 000 divided by 20). In practice, it is unlikely that a straight-line depreciation profile would accurately reflect the rate at which an asset was used up; that this inherent loss in value was reflected exactly in the market price; or that expectations of movements in the asset-specific price index (including those reflecting technological change) were precisely fulfilled. True economic cost could only be ascertained by reference to actual market value. Nevertheless, where it is necessary to implement some form of extrapolation between discrete market valuations, systems which attempt to incorporate movements in asset prices are likely to yield better

estimates that those which do not. A further corollary is that where there is more likelihood of error in forecasting asset-specific price movements (eg where the asset is subject to rapid technological change), more frequent market valuations need to be undertaken.

Table Ml: Financial aggregates under alternative accounting systems2

H isto rica l c o st acco u n tin g

O p eratin g cap ab ility m ain ten an ce SA P l b

Prop rietary ap p roach es to cap ital m aintenance

F in an cial aggregate

RR R rep ortingc CC A l d Real CCAe

1. Average Written-down Value of Assets in Service ($m) Telecom 11433 16 209 16 209 16 209 16 209

OTC 726.8 972.3 972.3 972.3 972.3

Australia Post 528.4 942.3 942.3 942.3 942.3

SECV 8 335 11455 11455 11455 11455

2. Profit Before Finance Charges and Tax ($m) Telecom 1 392.1 972.8 944.8 972.8 980.1

OTC 147.4 129.4 126.4 129.4 128.5

Australia Post 37.1 25.0 14.9 25.0 29.4

SECV 947.6 649.1 630.2 649.1 659.5

3. Profit to Equip Holders ($m)‘ Telecom 464.6 45.3 656.8 386.7 224.2

OTC 80.1 61.8 85.6 39.5 (66.1)

Australia Post 30.9 18.8 22.3 23.5 61.4

SECV 121.8 (176.7) 177.2 588.8 848.0

4. Rate of Return on Assets (%)8 Telecom 12.2 6.0 5.8 6.0 6.1

OTC 20.3 13.3 13.0 13.3 13.2

Australia Post 7.0 2.7 1.6 2.7 3.1

SECV 11.4 5.7 5.5 5.7 5.8

5. Opening Equity ($m) Telecom 2 354 6 958 6 958 6 958 6 958

OTC 265.2 557.6 557.6 557.6 557.6

Australia Post 122.4 500.1 500.1 500.1 500.1

SECV 548 4 210 4 210 4 210 4 210

6. Rate of Return on Equity (%)“ Telecom 19.7 0.7 9.4 5.6 3.2

OTC 30.2 11.1 15.4 7.1 (11.9)

Australia Post 25.2 3.8 . 4.5 4.7 12.3

SECV 22.2 (4.2)1 4.2 14.0 20.1

ended 31 March 1987, and those for the SECV to the financial year 1987-88. b Reflecting accounts prepared on the basis of Statements o f Accounting Practice - Current Cost Accounting (SAP1), issued by the Australian Society of Accountants and the Institute of Chartered Accountants (1983).

(cont’d)

145

c Reflecting accounts prepared on the basis of Accounting Policy Statement No. 1 issued by the Victorian Department of Management and Budget (1987). d Reflecting accounts prepared on the basis of the New Zealand accounting standard CCA1. e Reflecting accounts prepared according to the Real CCA system advocated in the Byatt Report

(1986) and Barton (1983) and described in Graham and Xavier (1987), and Xavier (1989). f After adjustment for finance charges, gearing, real holding gains,taxes, and extraordinary items, g (2) Profit before finance charges and tax divided by (1) Average written down value of assets in service. h (3) Profit to Equity holders divided by (5) Opening Equity, i IAC estimate derived from data presented in Xavier (1989). Sources: Xavier & Graham (1987) and Xavier (1989).

annual inflation of 10 percent will have increased the price of a similar new asset to $121 000, although the original asset will only be worth 90 per cent of this, or $108 900 (since by year 2, 10 per cent of its productive capacity would have been used up). Under CCA, the depreciation charge for use of the asset during year 2 will be $6 050 (ie 5 per cent of 121 000). If this amount is recovered from users during the year, and added to the $5500 recovered in year 1, the enterprise will have $11 550 in funds, plus an asset now worth $108 900, or total assets of $120 450. This falls short by $550 of the $121 000 required to replace the asset. If, however, the enterprise has usefully employed the $5500 funds raised in year 1 to return 10 per cent annually, it will have available this $550 as its return on this investment. Thus, in addition to implementing current cost depreciation, maintenance of the capital base also requires that internally-generated funds are reinvested (in productive capacity or other investments) to earn a rate of return at least equal to the asset specific inflation rate.

Erosion of the capital base through underpricing of capital services used up in production may be exacerbated when the use of historical cost accounting encourages the payment of excessive dividends. For example, the fact that the recorded depreciation charge underestimates the true economic cost of the assets used up in providing goods and services means that the reported gross operating surplus of the enterprise is overstated. Dividends paid by public enterprises to governments out of such illusory surpluses may erode the capital base itself. This would reduce the ability of the enterprise to maintain its operations into the future, unless technological change offsets the effects of inflation so that assets can indeed be replaced at or below their historical cost.

Pressures to overinvest

Another consequence of failing to set prices to recover the true costs of using assets is that inappropriate information may be generated on which to base investment decisions. To the extent that prices are set ‘too low’ because of the understatement of depreciation, demand for the goods or services produced will be stimulated. In attempting to satisfy current and future levels of demand, there may be pressures on managers to overinvest, with consequent biases towards capacity expansions rather than the better use of existing capacity.

The external environment

The use of historical cost accounting principles by public enterprises also causes problems for external parties trying to monitor their performance. In particular, it

146

presents difficulties for governments interested in assessing the efficiency with which these enterprises operate, including the appropriateness of the prices they charge (particularly for those which possess some degree of market power). Another difficulty created for governments is in determining the appropriate dividend which

could be collected without impairing the operations of the enterprise. Consumers are also affected through the distorted price structures which historical cost valuations are likely to induce.

Performance monitoring

Under historical cost accounting, performance measures such as overall surplus/deficit or rate of return on assets or on equity may be misleading. Operating surpluses will tend to be overstated because of the underestimation of the cost of utilising assets, and rate of return measures may be poor indicators of the efficiency

with which the enterprise has used the community’s resources if the base on which the return is calculated grossly understates the true economic cost of the resources involved (see Appendix L).

An indication of the extent to which performance measures based on historical cost accounting may differ from those which more closely reflect actual performance is revealed in Table Ml. For example, in 1985-86, the historical cost accounting profit (before finance charges and taxes) for Telecom was $1,392 billion, compared to a

figure under alternative current cost systems of between $944.8 and $980.1 million. Similarly, Telecom’s recorded rate of return on assets varies from 5.8 to 12.2 per cent in the same year depending on the cost system used; its return on equity varies from 0.7 to 19.7 per cent. Similar variations occur in respect of recorded financial

aggregates for OTC, Australia Post and the SECV. Clearly, reliance on historical cost measures poses severe problems for performance monitoring and for assessing the capacity of enterprises to pay dividends to the government.

Price regulation

Many government agencies provide goods and services in markets where they face little actual (or potential) competition. Such a market environment may afford the enterprises some discretion in setting prices. To prevent the possibility that users may be over-charged, governments may impose some form of price regulation. Inevitably,

this involves determining what is the appropriate price for a good or service. If prices are based on historical rather than economic costs, the regulated price set for the public enterprise will lead to inefficient consumption and investment decisions, and may undermine the ability of the enterprise to continue in business.

Effects of distorted prices on users

Prices based on historical cost will give inappropriate signals to users about the true costs of production, and may encourage overconsumption. Where prices of services provided by public enterprises are ‘too low’, using industries are, in effect, being assisted at the expense of taxpayers or other industries which are not given the

opportunity to employ the excess resources used by the public enterprise. Thus, inappropriate pricing of government-provided goods and services will hinder the efficient allocation of resources throughout the economy.

147

In addition to the adverse effect prices based on historical cost may have on the efficient allocation of resources, the practice may also give rise to inequities between past and current users of publicly provided services. As noted above, historical cost pricing will generally fail to generate sufficient funds to permit replacement of infrastructure at prevailing (ie inflation-affected) market prices - although this often only becomes apparent towards the end of an asset’s life when replacement becomes

necessary. The question arises as to whether it is equitable to require current consumers to pay for the full cost of replacing the infrastructure simply because of failure to fully recover costs from previous generations of users. In this regard, the NSW Irrigators’ Council submitted that it was quite inequitable for the present generation of landowners to shoulder the whole burden of replacing assets owned by the Department of Water Resources (eg channel and pipe networks), and suggested that government should meet this cost because ‘it is Governments which have failed to establish reserves or means of covering this need in the past’ (NSW Irrigators’ Council submission, September 1988, p. 3). The fact that mistakes have been made in the past, however, is no rationale for perpetuating them so that, provided such infrastructure is commercially justified, the full cost should be recovered from users. Where this would lead to large divergences between new and existing charges, however, there may be a case for phasing in such price increases.

M3 Alternative valuation systems

In view of the problems caused for both internal and external parties from use of historical cost systems by public enterprises, there has been widespread agreement, in principle, that a move towards accounting systems which more closely reflect true economic costs is desirable. This has engendered the development of a range of alternative valuation systems.

The alternative accounting systems that have been proposed can be distinguished according to:

. the basis of asset valuation;

. the capital maintenance concept employed (eg maintenance of operating capability or the maintenance of financial capital); and

. the unit of measurement used (eg nominal dollars or a unit of constant purchasing power).

The discussion in this appendix is restricted to accounting systems employing asset valuation methodologies which attempt to update book valuations of assets to better reflect their economic values. A fuller discussion of the relative merits of a range of alternative systems may be found in Barton (1983) and Graham and Xavier (1987).

The economic value of an asset is that which would emerge if the asset were traded in a fully competitive market. This will be the value that, taking into account the expected stream of future cash flows and disposal value of the asset, will generate a rate of return to the buyer equal to that earned if the resources were used elsewhere in the economy.

Of the two asset valuation bases, only current cost accounting (CCA) systems attempt to adjust the original prices of assets to reflect factors such as general and specific price movements and technological developments which result in changes to the market value or economic cost of the asset. In this regard, the Byatt Report (1986, vol. 1, p. 6) concluded:

...the principles of CCA asset valuation are correct as a basis for measuring economic costs. The overall principle is that assets should be included in the balance sheet at their value to the business... The value of assets to a business means what potential competitors would find it worth paying for them, even if the competition is hypothetical. This will be the net replacement costs of a Modern Equivalent Asset if

the asset would be worth replacing or the recoverable amount if it would not.

While current cost asset valuation is increasingly being accepted in principle, there has been considerable disagreement about its implementation. One set of issues - discussed in section M4 - relates to practical implementation difficulties (eg lack of active secondhand markets for many assets). A second set of issues relates to the

appropriate capital maintenance concept to apply. The latter issue is about determining the division of end-of-period capital between the initial capital base and the income for the period. The debate essentially revolves around whether the

appropriate capital maintenance concept to use is operating capability or financial capital.

Maintenance of operating capability

Under this approach income is the amount which can be distributed while maintaining an enterprise’s physical capital or operating capability intact. Generally, operating capability is measured in terms of being able to produce a given volume of output.

The hypothetical example presented earlier can be used to illustrate this concept. Suppose that the asset bought in year 1 for $100 000 is of a type subject to technical progress such that the market value of an equivalent new asset falls during the year to $80 000. Assuming that the annual depreciation rate of 5 per cent is fully reflected in

the market, the one year old asset would now be worth $76 000. If the asset generates a net cash flow of $15 000 for the year, the recorded profit for the year would be $11 000 (after allowing $4000 for depreciation to maintain operating capability).

The emphasis under this approach is upon price movements in specific assets rather than general price movements. Thus, a profit of $11 000 is recorded despite the fact that the enterprise has, by virtue of holding an asset of falling value, suffered a loss in its general purchasing power. However, the operating capability approach does provide useful information to managers on the extent to which revenues have been

sufficient to enable an enterprise to continue to provide goods and services on an ongoing basis. Thus, as long as the enterprise generates a net cash flow of at least $4000, it will have raised enough funds to permit replacement of its capital assets in the future (thus avoiding the erosion of its capital base as described in section M2).

While the above example illustrates the basic concept underlying this version of CCA, several variants of this approach have been proposed. In particular, there has been considerable debate on the appropriate treatment of monetary assets and liabilities, and on whether the manner in which the enterprise is financed should be taken into account in determining income. The basic question is whether owners benefit from an enterprise holding liabilities, fixed in money terms, during times of inflation, and if so, how this gain should be measured.

One view, reflected in the Rate of Return Reporting (RRR) version of CCA adopted by the Victorian Government, is that the holding gain should be measured with reference to movements in the general price level index, and be fully credited to the profit and loss account (Victorian Department of Management and Budget 1987). Another view (implicit in the New Zealand accounting standard CCA1 and the now withdrawn UK standard SSAP16) is that the enterprise only benefits if the borrowed funds are invested in assets which appreciate, and so holding gains on monetary liabilities should be measured with respect to specific price movements of the assets being financed (Xavier 1989; Richardson Report 1976; and Sandilands Report 1975). In contrast, the SAP1 approach supported by the Australian accounting bodies (Australian Society of Accountants & The Institute of Chartered Accountants 1983) excludes holding gains on loan capital from the measure of profit by transferring such gains to a reserve account.

Whilst debate continues as to which of these (and other) approaches is the most appropriate, what is clear is that the particular approach adopted can have a great impact on reported financial results. This is particularly so for public enterprises which typically have very high gearing ratios. For example, as demonstrated by

Xavier (1989), and shown in Table Ml, the return to equityholders for the SECV in 1987-88 was $177.2 million under the RRR approach compared to a loss of $176.7 million under the SAP1 approach, with the entire difference of $353.9 million attributable to the inclusion of holding gains on loan capital. Furthermore, the surplus recorded under RRR exceeds even the return recorded under historical cost principles of $121.8 million, supporting Xavier’s assertion (1989, pp. 74-5) that for most private enterprises the additional depreciation charges associated with current cost accounting:

...would usually result in a deterioration in the business result, compared with one calculated on a historical cost basis. However, long term liabilities which are financed by loans comprise such a major component of the debt structure of a public enterprise, that notional gains calculated on these liabilities and taken through the profit and loss account could far exceed the additional depreciation calculated under the replacement cost method. Consequently, the public

enterprise could show a vastly improved business result without an accompanying increase in its cash resources and in its capability to finance part of the inflated cost of financing new and replacement fixed assets from internal sources.

150

In view of these concerns, a recent review of RRR (Victoria, Parliament 1988) suggested that further refinements may be warranted, giving specific consideration to:

possible inconsistencies in the capital maintenance concept and the method of determining holding gains on monetary liabilities by a general price index rather than by asset-specific price indices. Consideration should also be given to the implications of any possible

overstatement of returns on equity for dividend payments and for cash flow in the authorities.

Maintenance of financial capital

In contrast to the operating capability approach, this approach seeks to maintain intact the purchasing power of the shareholders’ investment. The basic difference between the two approaches is the treatment of real holding gains (or losses) on operating assets. Whilst these gains (or losses) are incorporated into the restatement

of initial capital under the former approach, they form part of income when the financial capital maintenance concept is used.

Employing the previous example again, income under the financial capital maintenance approach would equal the difference between the firm’s actual command over goods and services at the end of year 1 (ie $91 000) and the amount that would be needed to maintain the same purchasing power as it had at the start of

year 1 (ie $100 000 in year 1 dollars or $110 000 in year 2 dollars (given a general inflation rate of 10 per cent per annum)). The enterprise would therefore record a loss of $19 000 under the financial capital maintenance approach.

Again, several variants of this approach have been proposed (Barton 1983; Graham & Xavier 1987).

Assessment of the alternative approaches

The example used to illustrate the two main versions of CCA systems highlights the fact that reported financial results may vary greatly depending upon which capital maintenance concept is chosen, particularly if technological change is such that price movements for specific assets differ considerably from the general movement in prices (see Table M l for some examples). Given this divergence of results, which is

the most appropriate version to apply to public enterprises?

The answer to this question depends on the purpose to which the information is to be put. In this regard the Byatt Report (1986, vol. 1, p. 21) observes:

...both operating capability and financial capital maintenance have their place. Management of a continuing business will emphasise the need to avoid distributing funds required to maintain operating capability.

Investors will want to calculate the real rate of return after the maintenance of the real value of their capital for comparison with returns available elsewhere. In general, information about both profit after operating capability maintenance and profit after real financial

capital maintenance is relevant for the different purposes of the various

151

users of accounts. Whether it is appropriate to select one basis or the other as dominant in presenting accounts depends on a range of factors determining the applicability of the concepts, the availability of alternative sources of information, the quantitative difference between the two concepts and the specific role envisaged for the accounts. Comparability between businesses in a similar position will also be a consideration.

The same report suggests the following factors be considered in selecting the appropriate capital maintenance concept:

. whether the nature of the business is a continuing one;

. whether or not there is an adequate capital market for shares in the business which would provide continuous information (other than accounts) to investors on their investment;

. the role of the accounts in relation to pricing policy; and

. the rate of technical progress.

Most businesses (particularly those within the public sector) aim to be of a continuing nature. Indeed, some public enterprises may even be specifically charged with maintaining a given level of operating capability. The first factor thus suggests prominence be given to the maintenance of operating capability.

The second factor, however, suggests that for public enterprises, lack of capital market disciplines underlines the need for reliable measures of economic performance. According to the South Australian Treasury (1988):

The public sector does not possess the equivalent of a capital market which provides investors (ie Parliament) with information on the performance of agencies and, therefore, financial statements which

report an income figure which is arrived at after maintaining physical capital may be inadequate in assisting Parliament to gauge the real return on funds invested in an enterprise.

In relation to the third factor, it has been observed that where a public enterprise does not operate in a competitive market and is a price-maker rather than a price- taker, there may be a need to set pricing targets based on cost of supply (see Appendix L). The continuing cost of supply includes a normal profit on investment after financial capital maintenance. It can therefore be argued that where public enterprises face price regulation, it is particularly important for accounts to be available showing performance on the basis of financial capital maintenance.

Finally, the choice of capital maintenance concept will be influenced by the rate of technical progress. As illustrated in the context of the hypothetical example presented earlier, reductions in capital costs due to technical progress will tend, in highly competitive markets, to overstate profits (after allowance for the maintenance of operating capability).

On the basis of these considerations, many commentators favour the adoption of a system of CCA that uses a real financial capital maintenance concept (generally known as the Real Current Cost Accounting or RCCA system). This concept has been supported by the Byatt Report in the UK and several Australian commentators

on the issue (Barton 1983; Graham & Xavier 1987; Xavier 1989). Some State Governments (eg SA, Victoria) also appear to favour this approach. However, it has also been recognised that the maintenance of operating capability concept also provides important information to managers. This has led to the widespread

suggestion that whichever approach is preferred, information on the other should also be supplied (eg in supplementary notes to the accounts).

M4 Implementation issues

Even if a particular CCA valuation system is accepted as the preferred approach, there may be considerable practical problems associated with its application. As noted by the South Australian Treasury in its submission, CCA has not been widely adopted in Australia and little information is available on problems associated with its use.

Victoria has made an attempt to implement an accounting system more attuned to economic costs, with the development of RRR. This approach was adopted by five major commercial authorities - the SECV, the Melbourne and Metropolitan Board of Works, the Gas and Fuel Corporation of Victoria, the Port of Melbourne Authority

and the Grain Elevators Board in 1985-86, whilst the Rural Water Commission also prepared supplementary RRR financial statements in 1986-87. As noted previously, however, the RRR approach has been subject to some criticism (Hogbin 1985; Graham & Xavier 1987; Xavier 1989).

Whilst some individual enterprises use CCA systems, only Victoria has published a comprehensive policy on this issue. However, the Commission understands that some other States (eg NSW, SA, WA) are beginning to addressing the CCA issue in conjunction with the possible application of financial targeting. In a similar vein, a recent report (Inter-State Commission 1989, pp. 285-8) recommended the

development of an appropriate asset valuation base for port authority assets to facilitate rate of return targeting. The WA Treasury submitted that port authorities in WA now use CCA systems and that CCA is likely to be extended to other authorities if rate of return targets currently under consideration are adopted. In contrast, the

Tasmanian Government stated that current cost accounting is not used by government agencies in Tasmania. This appears to reflect a concern that such a system of accounting would result in excessive charges to users and higher than warranted dividend returns on capital to governments. Similar reservations were expressed by

some other participants to this inquiry.

At the Commonwealth level, implementation issues are arising as the Government begins to apply the rate of return targets foreshadowed in the 1988 May Economic Statement. It has been reported that only OTC has undertaken a comprehensive asset revaluation, and other public enterprises are resisting moves towards CCA

(Shand 1989). Some Commonwealth public enterprises (eg AN) did, however, indicate to this inquiry that they had commenced to revalue their fixed assets over a period of five years in accordance with government guidelines.

153

Potential problems in obtaining current market values

The most obvious potential problems which may arise in implementing CCA are in obtaining reliable current cost data. Such difficulties may arise when trying to determine market values/replacement costs for assets which are unique and for which no ready market exists and which may only have a value in their existing use (eg dams).

Whilst such valuation problems may seem at first sight to be intractable, this does not appear to be borne out by practical experience. For example, a follow-up report by the UK Treasury (HM Treasury 1988) on the experiences of implementing current cost valuations in public enterprises as recommended in the Byatt Report suggested

that, whilst many problems remained, significant progress had been made towards estimating current valuations, even in industries such as water supply, electricity supply and coal mining. Similarly, as noted by the South Australian Treasury in its submission, United States experience in the wake of replacement cost disclosure

requirements of the Securities and Exchange Commission also indicates that replacement cost data can be calculated. At the Commonwealth level, an exercise involving asset revaluation estimates for public enterprises including Australia Post, AN, Qantas, ANL, the Pipeline Authority, the Snowy Mountains Hydro-Electric Authority, Telecom and OTC conducted in the context of the Rae Report (1983, pp. 265-270) suggests that the problems are not insurmountable. Furthermore, the introduction of CCA in parts of the Victorian public sector appears to have been achieved without encountering major valuation problems by making use of three broad approaches to valuing assets in current cost terms (Victorian Department of

Management and Budget 1986, p.21):

. current market selling price of an asset - this approach estimates the current cost determined by reference to an accessible net market value of the asset. It applies where there is an active market for the product;

. current reproduction cost : this approach estimates the current cost by reference to the cost per unit of productive capacity of reproducing or replicating the asset. It applies when the asset being valued would be replaced at balance date by a similar asset; and

. current replacement cost : this approach estimates the current cost by reference to the cost per unit of productive capacity of the most appropriate modern replacement facility. It applies when the asset being valued would be replaced at balance date

by a different asset (in terms of scale and/or technology) having a similar productive capacity.

Simplified CCA systems

A further consideration is whether it is more cost-effective to implement the favoured CCA system or a less costly alternative which retains most of the benefits from adopting current cost valuations. Regularly revaluing all the assets of an enterprise is

154

likely to be time-consuming and costly, and raises questions such as whether valuations should be targeted to particular assets and undertaken at a frequency influenced by the expected rate of change in their value, rather than annually.

The experience in the UK offers some guidance in this regard. A major finding of the 1988 review (HM Treasury 1988) was the importance of distinguishing between price- makers and price-takers.The review suggested that with price-makers, there is a strong case for sophistication in valuing assets because of the relevance of capital

costs as a factor in setting prices. Accurate, up-to-date current cost information is needed for regulatory purposes. With price-takers, however, prices are set in competitive markets. Consequently, for them there may be a case for simplified procedures to reduce complexity (eg developing modified historical cost accounts by

indexing for general inflation and using generally representative depreciation profiles).

In a similar vein, the South Australian Treasury has suggested an interim approach to implementing CCA, given that the asset valuation process may be expensive and time­ consuming. This approach represents a simplified version of current value accounting

which, although lacking the refinements of the CCA system, captures the essential elements. In supporting this approach the South Australian Treasury (1988) has observed:

A considerable amount of work is still necessary in some agencies to value assets. As an initial step in the introduction of real accounting systems, it may be desirable to concentrate on the revaluation of major assets. This would have the advantage of minimising the time needed

to introduce a new accounting approach and although it would, in the early stages, provide results which were somewhat crude, those results would be likely to be superior to anything being provided currently. The quality of information would improve progressively as all assets were

revalued and those values incorporated in the financial accounts.

Concentration on the major assets also appears to underlie the Commonwealth Government’s approach to the asset valuation issue, as set out in draft guidelines published by the Department of Finance (1988). These guidelines emphasise that ‘the application of full current cost accounting is not advocated in these Guidelines.’

Indeed, the sentiments underlying the guidelines are perhaps best described in another discussion paper (Shand 1989, p. 180) which states that the guidelines are ‘not advocating a search for some theoretical CCA precision, but rather an attempt to get a "better feel" for financial performance’. The Guidelines require that all tangible

non-current assets (as defined in the guidelines) be revalued on a five-year cycle. Non-current assets with a cost below an agreed threshold and operational assets (eg loans to clients, shares or securities in companies) are excluded from the guidelines. Exclusion of these assets may reflect a view, as expressed by the Treasury (1988,

p. 13), that:

It is the return on durable assets which is of most relevance : financial enterprises aside, the amount of investment by Government business enterprises in monetary assets is generally relatively small and special adjustments are only needed when determining real economic income,

not nominal.

155

The Treasury also appears to favour a system that simply applies end-year adjustments to historical cost systems to update asset values. Moreover, it proposes a further simplification of adjusting the accounting depreciation on total assets, rather than making a replacement cost adjustment (to estimate market value) of every asset. This would require an assumption of the average physical rate of depreciation, and an

average specific inflation rate, for all the enterprise’s durable assets. However, an important consideration would seem to be whether accounts are used mainly for annual performance evaluation or for ongoing internal decision-making purposes.

The development of these modified accounting approaches begs the question as to the circumstances under which modified historical cost accounting systems amount to a practical application of the main principles of current cost accounting. In this regard, a UK report (HM Treasury 1988, p. 25) suggests that the modified historical cost systems need to incorporate:

. appropriate revaluations of all assets;

. regular updating of revaluations or extrapolating of valuation trends between valuations; and

. consistent treatment of revalued assets in profit and loss accounts and balance sheets.

The need for education

Another important conclusion of the UK review of the implementation of CCA in government entities was that there was a general need for simplification in order to aid understanding and acceptance (HM Treasury 1988, p. 15). Despite its shortcomings, historical cost accounting is relatively simple and its principles and practical application readily understood and accepted by most professional accounting bodies, governments and statement users. In contrast, CCA systems appear to cause some confusion. The South Australian Treasury (1988) noted that:

A study in the US carried out by Arthur Anderson revealed that directors of a large number of companies surveyed questioned the reliability, usefulness and cost of providing the data. Similarly, an Australian survey yielded results in which a significant number of respondents felt that the benefits were not worthwhile or did not give a realistic and accurate reflection of financial performance and position.

A common misconception of CCA is that it represents a move by governments to broaden their taxation base and to force existing and future users to pay for facilities again. A second objection frequently raised to adopting CCA in public enterprises is that this imposes a requirement which does not apply to their private sector counterparts. For example, in implementing the Commonwealth GBE reforms, a Department of Finance official (Shand 1989, p. 181) has noted that:

Ongoing discussions are taking place with others including Telecom and Australia Post but their general view is that they should continue to report in accordance with ‘commercial practice’ eg use historical cost

156

reporting, except for the provision of information on revaluation of land and buildings required under the Companies Act. For example both Qantas and Australian Airlines argue against disclosing the current value of their aircraft in their financial statements because their

competitors do not.

However, Shand (1989) notes that industry analysts can very readily work out the current value of aircraft and that inadequate reporting by private firms does not in itself justify their public sector counterparts doing the same.

Moreover, the argument overlooks the fundamental point raised in Section Ml that the absence of external monitoring mechanisms which apply to private firms (eg sharemarket) leaves accounting records as the only source of performance evaluation

of public enterprises. As noted by Shand (1989, pp. 180-1):

To the extent that a GBE uses historical cost in its financial statements, it is not reporting useful or valid information on financial results and position (regardless of what any competitors may do)... it is clear that,

whatever the private sector reports in its published financial statements, it does not measure its performance on the basis of historical cost rates of return ... Presumably or hopefully all these GBEs already have their own (non historical cost) financial performance measures by which they

judge their performance, even if it is not reported externally. As such any reporting task should not be too onerous.

This discussion suggests that to reap its full potential benefits, implementation of current cost accounting in Australian public enterprises would need to be accompanied by a program of education so that managers, government and the public understand the concepts involved. Understanding of CCA may also be facilitated by the publication of a reconciliation of accounts prepared on a CCA basis with those under the more familiar historical cost basis. Indeed, the RRR statements of Victorian authorities are currently prepared as supplements to historical cost accounts but are audited. Moreover, as noted by a parliamentary committee, it may be that before RRR statements are presented exclusively, there will be a period of time when

the historical cost based statements are included as supplementary (Victoria, Parliament 1988, p. 127). The committee recommended that ‘each commercial authority provide in its annual report 10 year summaries of Profit and Loss and Balance Sheet data on both historical cost and RRR basis, as far as this data is readily

available’. The Hydro-Electric Commission of Tasmania also submitted that, because of its statutory obligations, any results it reported in CCA terms would probably achieve only the status of supplemtary information.

The need for consistency

Another major issue in implementing asset valuation systems that more accurately reflect economic costs is the need for consistency. Efficient resource allocation may be hampered if some enterprises in the public sector were to move towards CCA systems whilst others retained existing practices. If left to their own devices,

individual enterprises may be tempted to adopt accounting systems which show their performance in the best light, rather than ones which are more conducive to rational economic decision-making. As noted by Heald (1980, p. 257), ‘... it is clear that any

coherent system for the control and accountability of the nationalised industries cannot allow accounting policies to be entirely left to the managers of individual industries.’ A number of participants to this inquiry (eg Australian Chamber of Manufactures, CRA) supported the establishment of a standard set of accounting practices to apply to all public authorities.

The growing realisation of the desirability of current cost accounting in public enterprises suggests that the time for developing a consistent approach is fast approaching. Indeed, the South Australian Treasury (1988) observed that:

Now would seem to be an opportune time to actively develop the accounting policies, accounting systems and the philosophies needed to introduce MFPM [Meaningful Financial Performance Measurement], The need to lift performance of public entities, and the developments being undertaken in agencies, give a sense of urgency to the need to develop a common set of accounting policies which will apply to the public sector; there is a risk of ad hoc and possibly inconsistent developments

There is no unambiguous prescription of the best approach to adopt to the current cost valuation issue. Nevertheless, experience in the UK offers some lessons that might serve as a useful guide for Australia. The 1988 UK Treasury Report (HM Treasury 1988, p. 17) concludes:

The [previous Treasury] Report .. advocated an evolutionary approach to developing current cost accounts. The debate will continue. The accounts of nationalised industries will continue to evolve in the light of changing circumstances and the need to measure their costs and their economic performance in ways which are easily comprehensible to Parliament and the public.

Australia is generally far behind in such an ‘evolutionary process’. There is an important and obvious need for the supplementation of historical cost accounting systems used by most government entities in Australia with current cost systems. Moreover, there is a strong case for a co-ordinated approach, perhaps facilitated by a joint Commonwealth/State committee to oversee implementation of such systems.

One participant (CRA), suggested a government task force supported by accounting standard bodies as a means to achieving uniform standards.

Relationship with other public enterprise reforms

Implementation of current cost accounting in public enterprises, whilst worthwhile in its own right, should be seen as complementing other reforms designed to improve performance. For example, the success of rate of return targets in improving the monitoring and accountability of public enterprises is largely dependent on having an appropriately valued asset base on which to measure the rate of return (see Appendix L).

One possible consequence of introducing current cost accounting, in the absence of other initiatives aimed at improving the efficiency of public enterprises, could be to increase significantly the prices charged for government-provided goods and services.

158

For example, the Major Mail Users of Australia (Major Mail Users of Australia Ltd submission, September 1988, p. 4) expressed a concern that the effect of revaluing the assets of Australia Post every five years:

...initially, will be probably to double the book value of the assets. If Australia Post is considered to be operating efficiently at present, the application of its current rate of return figure to the revalued asset figure would then double the amount Australia Post would be required

to generate.

These effects are likely to be used as an argument against the adoption of CCA. Viewed in a more positive light, however, pricing according to the true value of resources used up (rather than their historical cost) increases the pressure on governments and managers to negate the consequent price increases by pursuing efficiency gains in the enterprise’s operation (eg ensuring the service is produced at least cost by eradicating inefficient management/work practices). Consequently, the introduction of CCA systems may also provide an added incentive to increasing

efficiency in other aspects of an enterprise’s operations, by making the true economic costs of providing goods and services clearer. Further, under CCA the true costs of making additions to capacity are made more apparent thereby encouraging the more effective use of existing capacity. In addition, CCA facilitates the application of more

appropriate charges, thereby lessening the likelihood of inter-generational inequities.

M5 Assessment

Whilst the historical cost valuation of assets may satisfy existing legal/taxation requirements, it is not well suited to fulfilling another important objective - that of providing information relevant to rational economic decision-making and performance evaluation. Without information on economic rather than historical

financial costs, inefficient pricing and investment policies and misleading indicators of performance are likely to result. There is now widespread agreement that, on their own, historical cost systems provide limited information about economic costs and that the associated adverse effects are likely to be severe in the public sector,

particularly where assets are long-lived and public enterprises enjoy monopoly positions.

The deficiencies of historical cost systems have led to calls for the introduction of current cost asset valuations in public enterprises in Australia. There is, however, an active debate on which particular version of CCA should be adopted. This debate does not provide an excuse to delay the supplementation of existing historical cost

systems typically used in public enterprises throughout Australia. Even modified historical cost systems are likely to provide better information for decision-making than existing historical cost systems.

To date only the Victorian and Commonwealth Governments appear to have taken practical steps towards implementing CCA systems, although the Commission understands that some other States (eg SA, NSW, WA) are presently considering the issue in conjunction with the application of financial targeting. Other States do not

have an explicit policy in relation to the use of current cost accounting in their enterprises. The limited use of CCA in Australia to date suggests that a broad

159

program of public education of the benefits of CCA is likely to be a necessary prerequisite to its widespread introduction.

Experience overseas and in Australia (eg Victoria) indicates that it is practical to apply modified CCA systems. However, the nature of modifications can clearly vary. In order to avoid ad hoc choices of approaches that do not facilitate the provision of relevant and consistent information, there is a strong case for adopting a coordinated approach through the development and application of agreed common guidelines. This could be facilitated by an intergovernment review conducted in conjunction with

the accounting bodies.

REFERENCES

Australian Society of Accountants & the Institute of Chartered Accountants 1983, Statements o f Accounting Practice - Current Cost Accounting (SAP1), Sydney, November.

Barton, A.D. 1983, ‘Performance Measurement and Rates of Return for Public Sector Enterprises’, Research Lecture in Government Accounting, Australian Society of Accountants.

Byatt Report. See HM Treasury 1986.

Committee of Inquiry into Inflation Accounting 1976, Report, (I. Richardson, Chairman), Government Printer, Wellington, New Zealand.

Committee of Inquiry into Inflation and Taxation 1975, Report, (R.N. Mathews, Chairman) AGPS, Canberra.

Department of Finance 1988, Draft Guidelines for Asset Valuation in Government Business Enterprises, July.

Graham, B. & Xavier, P. 1987, Inflation Accounting for Australian Public Enterprises - Economic Rationale and Financial Implications, Swinburne Institute of Technology, Faculty of Business staff Paper no. 37, Melbourne, July.

Heald, D. 1980, ‘The Economic and Financial Control of UK Nationalised Industries’, The Economic Journal, vol. 90, no. 358, June, pp. 243-265.

HM Treasury 1986, Accounting for Economic Costs and Changing Prices, A report to HM Treasury by an Advisory Group, (I.C.R. Byatt, Chairman), HMSO, London, (2 volumes).

____ 1988, Accounting for Economic Costs and Changing Prices, A Situation Report by the Treasury, January.

Hogbin, G. 1985, The Rate of Return and Public Authority Dividend Requirements, and Grain-Handling Charges in Victoria, A Study prepared for the Victorian Farmers and Graziers Association, September.

Inflation Accounting Committee 1975, Report (F. Sandilands, Chairman), HMSO, London, Cmmd. 6225, September.

Inter-State Commission 1989, Waterfront Investigation: Conclusions and Recommendations, vol. 1, AGPS, Canberra, March

Mathews Report. See Committee of Inquiry into Inflation and Taxation 1975.

Rae Report. See Senate Select Committee on Statutory Authority Financing 1983.

Richardson Report. See Committee of Inquiry into Inflation Accounting 1976.

Sandilands Report. See Inflation Accounting Committee 1975.

Senate Select Committee on Statutory Authority Financing 1983, Volume 1 - Report, (P. Rae, Chairman), AGPS, Canberra, September.

Shand, D. 1989, ‘Corporatisation or Privatisation Will Alter the Legal Requirements for External Reporting and Will Have Implications for External Accountability’, Paper delivered to Australian Society of Accountants National Accountants in Government Convention, Melbourne, 2-4 March.

South Australian Treasury 1988, Discussion Paper: Meaningful Financial Performance Measurement in the South Australian Public Sector, Adelaide, 19 August.

The Treasury 1988, Economic Income and Rate o f Return Reporting of Government Business Enterprises.

Victorian Department of Management and Budget 1986, Public Authority Policy and Rate of Return Reporting, Information Paper No. 1, October.

1987, Rate o f Return Reporting, Accounting Policy Statement No. 1 Issued by the Comptroller-General, August.

Victoria, Parliament 1988, Economic and Budget Review Committee: Twenty-fourth Report to the Parliament: Controls over Commercial Authority Debt Levels, May.

Xavier, P. 1989, ‘Economic and Commercial Principles For Determining Government Trading Enterprise Charges: How Compatible Are They?’, Paper presented to a seminar on Electricity Tariffs convened by the NSW Inquiry into Electricity Tariffs, 31 January.

Xavier, P. & Graham, B. 1987 ‘Financial Targets and Dividend Requirements for Commonwealth Government Business Enterprises’, paper presented to the 16th Conference of Economists, Surfers Paradise, Queensland, 23-27 August.

162

APPENDIX N : OVERSEAS DEVELOPMENTS IN RELATION TO PUBLIC ENTERPRISES

Concern over the performance o f public sector bodies, particularly the operation of public business enterprises has increased over the past decade. Much o f this concern has focused on the insulated and restrictive

environment in which many public enterprises operate. As a result, many governments have introduced significant changes.

This appendix examines the experience of three countries - the United Kingdom, New Zealand and Canada Initial approaches in the first two countries focused on administrative measures to achieve reforms, accompanied by some increase in competition. Subsequently, the UK

Government privatised and is continuing to privatise many of its enterprises, a path now being followed by New Zealand. The emphasis of the Canadian program is also on privatisation.

A common thrust in the current approaches o f the three countries is increased exposure to competition of their public enterprises, regardless of whether ownership is subsequently transferred to the private sector. The manner in which the difficulties which arise in this process are being

resolved can provide valuable lessons for Australia’s reform process.

The extent to which governments overseas are involved in the provision of goods and services varies from country to country. While historical, geographical, political and social differences amongst countries can limit the usefulness of international comparisons, it can nevertheless be instructive to examine overseas experience in relation to reform of public enterprises. In order to minimise some of the more

obvious problems of drawing parallels, the Commission has examined the United Kingdom, New Zealand and Canada, three countries having a number of characteristics in common with Australia.

N1 United Kingdom

Background

Following the Second World War, many UK industries were nationalised including electricity, coal, gas, transport, automotive and steel. The performance of the nationalised industries compared unfavourably with that of other industries and by the end of the 1970s, they were increasingly seen as contributing to the country’s economic ills.

At the time of the election of the Thatcher Government in 1979, nationalised industries accounted for about one-tenth of gross domestic product (GDP) and one- seventh of total investment in the economy. They employed about 1.5 million people, and dominated the power, transport, communications, water and sewerage, iron and

163

steel, and shipbuilding sectors of the economy.

The approach

The UK Government’s response in the early eighties was two-fold. First, it set out to improve the performance of the nationalised industries within the public sector. This was done by setting clearer financial targets and implementing improved performance monitoring mechanisms, while providing the industries with as much commercial

freedom as possible. This policy achieved marked increases in productivity and profitability in some industries, and was seen as a necessary first step by the Government prior to the implementation of further reform. Subsequent reform

measures included the transfer of profitable industries to the private sector, a process that has been labelled ‘privatisation’. The principal aims of the UK privatisation program are to improve efficiency, reduce the public sector borrowing requirement and government involvement in enterprise decision making, and encourage share ownership (particularly amongst employees).

The mechanics employed in each privatisation are complex and vary from case to case. The main options being used are a direct sale or public float on the Stock Exchange, either by a fixed price offer or by a tender offer with a minimum price. In

sales of its enterprises on the stock market, a Government aim has been to maximise sales proceeds and to widen share ownership. However, in such sales, most share issues were substantially oversubscribed and the share price increased dramatically after listing, suggesting that the shares were underpriced. Special arrangements were often made to encourage small investors and employees to participate in privatisation sales. In cases where shares were offered for sale, employees were typically given a preference.

The use of regulatory bodies to control the operation of industries after privatisation has been a feature of the process in the UK, particularly in relation to industries having elements of natural monopoly. In the case of British Telecom (partially privatised in 1984) and British Gas (privatised in 1986), specialist regulatory bodies,

the Office of Telecommunications (OFTEL) and the Office of Gas Supply (OFGAS), have been created to regulate the privatised industries, particularly in relation to price increases.

Another feature of privatisation in the UK has been what is commonly referred to as the ‘golden share’, whereby the Government retains specific powers over the future ownership, control or conduct of a privatised company.

The sale of public enterprises has been one of several components of the UK Government’s policy designed to reduce its involvement in the economy. Other elements have included the contracting out of activities previously performed by local authorities - eg rubbish collection, catering, cleaning, construction and maintenance work. Another element has been the subsidised sale of public housing to tenants in order to extend private property ownership.

The outcome

The Government has claimed that its privatisation program has been successful. The contribution of nationalised industries to GDP and the number of persons employed

164

by them has been almost halved. Nearly all the privatised industries are reporting higher profits and the number of people owning shares has trebled to 9 million since 1979.

During the period 1979 to 1983, proceeds from privatisation were around 500 million pounds per annum, derived mainly from firms already operating in a competitive environment, eg British Aerospace, Britoil, Cable and Wireless, and Enterprise Oil. Following the re-election of the Thatcher Government in 1983 the pace of the

privatisation program increased substantially and proceeds from sales were expected to approach 5 billion pounds annually by 1989.

Privatisation has included nearly all the UK’s oil and gas interests,

telecommunications, civil aviation and airports, buses, road freight and aerospace manufacture. Receipts from sales to the Government had reached 17 billion pounds by the end of 1988, and around 655 000 jobs had been transferred to the private sector.

There has been, however, increasing criticism of the privatisation program, particularly in relation to natural monopolies. Particular areas subject to criticism are British Telecom and British Gas, both of which have already been partially or fully privatised, and the electricity supply industry and the water industry which have been

foreshadowed as privatisation targets by the Thatcher Government. Criticisms have principally centred around the degree of government regulation that is required after privatisation, with many questioning the economic benefits of converting a government monopoly to a private monopoly.

This view is best summarised in the following quote:

...the problems of organisation and control in utility industries such as telecommunications, gas, electricity, and water are amongst the most difficult in the field of microeconomic policy. Indeed, our view is that under private ownership there are conditions in which they become so acute that public ownership is to be preferred. When there are massive economies of scale and scope, high entry barriers, or externalities, private ownership performs poorly. The incentive and opportunity to

exploit consumers threatens allocative efficiency, and the lack of competitive benchmarks leads to internal inefficiency and slack.The fact that public ownership is also far from perfect in these

circumstances reflects the inherent difficulty of economic organisation in such industries (Vickers and Yarrow 1988, p.428).

Another criticism of the privatisation program relates to the Government’s practice in some industries of maintaining a ‘golden share’ or, effectively, control over the conduct of a privatised company. As political interference is seen as a barrier to efficient operations, where governments retain substantial shareholdings in privatised

industries, the potential for such interference remains.

Main lessons

The UK experience highlights the importance of having public enterprises operating efficiently, irrespective of whether privatisation is being considered or not. This is

because of the important linkages between the public sector and the rest of the economy. Where privatisation is being considered, sale proceeds are likely to be enhanced if the organisation to be privatised is operating soundly.

The UK experience also highlights the difficulties of devising an appropriate regulatory setting, particularly for public utilities, as a prelude to privatisation. In areas where this has been addressed (eg gas and telecommunications), the

Government has opted for industry-specific regulatory bodies but the outcomes have been the subject of considerable criticism. Some critics have expressed the view that competition has not increased, services have deteriorated and the public interest has been ignored.

A further lesson relates to the manner of privatisation. The practice of making fixed price offers rather than tendering has raised criticisms that shareholders are gaining a once-off benefit at the expense of taxpayers, resulting in a substantial redistribution of

income. Hence, the specific privatisation approach needs to be closely linked to the objectives if unintended outcomes are to be avoided.

N2 New Zealand

Background

State enterprises undertake one-fifth of New Zealand’s investment and produce 12 per cent of national income. In considering the performance of existing State enterprises and State trading activities, the Government concluded in 1985 that their performance and hence their contribution to national income could be increased. The review by the Government concluded that the shortfall in performance by these

enterprises was caused by a number of factors, the key ones being:

. multiple and conflicting objectives;

• excessive bureaucratic controls on managers leading to a lack of managerial autonomy;

• legislative protection from competition; and

• lack of managerial accountability.

The approach

To deal with these fundamental problems and constraints on State enterprise performance, the Government considered that a new approach to the operation of State enterprises was necessary. The Government’s State-owned enterprise policy, announced on 12 December 1985 by the Minister of Finance, contained the following

points:

the responsibility for non-commercial functions would be separated from major trading State enterprises;

managers of State enterprises would be given the principal objective of running them as successful business enterprises;

166

• managers would be given greater responsibility and accountability for the day- to-day operations; and

. the environment in which State enterprises operate would more closely resemble that faced by private enterprise so that commercial criteria can be used to provide a fair assessment of managerial performance.

The government also announced that the policy would be applied to the trading activities of the Department of Lands and Survey and the Forest Service, as well as the Electricity Division, State Coal Mines, the Post Office and the Civil Aviation Division. In 1986 the State-Owned Enterprises Bill was passed.

The Bill aimed to strike a balance between the powers of Ministers to control the direction of policies pursued by State enterprises and the need to ensure that the boards of such enterprises were given the ability to manage their operations commercially, free from day-to-day political control. The Bill had the dual objectives

of facilitating the improved economic performance of State enterprises and enhancing their accountability. These objectives were to be achieved through a process termed ‘corporatisation’, involving a clearer specification of objectives, greater management autonomy and accountability and improved performance assessment. The legislation

specifically provided for explicit funding by government in cases where enterprises are required to undertake non-commercial activities.

A fundamental principle of the new provisions was that the new operating framework for State enterprises ensured that they would neither enjoy advantages nor suffer disadvantages in relation to private sector commercial organisations. In this regard competition was to be injected into the markets of State enterprises wherever possible. They are to make a commercial return on the funds provided by the

taxpayer, and they are to pay taxes on the same basis as private business. Also, the Commerce Act 1986 ensured that State enterprises would be subject to the same disciplines as private organisations in respect of monopoly or anti-competitive behaviour.

In the Economic Statement of 17 December 1987 (New Zealand Government 1987), the Government announced its intention to sell a number of government businesses, including certain State enterprises. The statement provided two main principles

which would underscore the business sales program. These were that:

. the benefits and desirability of transferring government businesses and commercial assets to private ownership would be assessed on a case-by-case basis; and

. statutory restrictions on competition were to be removed and sales revenue would be maximised.

The sales criteria include the provision that the sale must make a positive contribution to government revenue and that any sale would further the Government’s economic and social goals. Additionally, it was announced that all State-owned businesses and commercial assets may be transferred to the private

sector and the Government would attempt to create an environment of competitive

167

neutrality prior to the sale. Through these sales, the Government hopes to raise NZ$14 billion by 1992, which represents approximately one-third of New Zealand’s current public debt.

Maintaining the ‘public interest’ appears to be a major consideration in the Government’s privatisation program, with considerable attention being devoted to ensuring an appropriate regulatory framework before sales take place. Not yet settled is the appropriate form of regulatory body to oversee a privatised industry; options being considered are an industry specific body or an expanded role for the Commerce Commission (which currently has functions similar to Australia’s Trade Practices Commission).

The outcome

Over the last four years the principal goal of the New Zealand Government has been in ensuring that its enterprises contribute in a more positive way to economic growth. Policies have focused on the organisational reform of State enterprises by giving them

more autonomy and greater financial accountability, while at the same time endeavouring to make them operate in a competitively neutral environment.

The 1988 Budget speech (Minister of Finance 1988) announced results of the first year of operation of many of the enterprises and compared their results to their predecessors. Included were Postbank (NZ$30 million profit compared with a NZ$51 million loss the previous year), Airways Corporation (NZ$4 million profit

compared to a NZ$5 million loss), Coal Corporation (NZ$4 million profit compared to a NZ$76 million loss), New Zealand Post (NZ$72 million profit compared to a NZ$38 million loss) and the Forestry Corporation (NZ$54 million profit compared to

a NZ$71 million loss).

To date the New Zealand Government has received around NZ$2 billion from the sale of State businesses. Those sold include Petrocorp, NZ Steel and Air New Zealand. Other government-owned businesses on the sales agenda include international airports, non-core assets of New Zealand Railways, Coal Corporation,

Post Office Bank, Rural Bank and Finance Corporation, Forestry Corporation, and the Shipping Corporation.

On the basis of the available information, achievement by the Government of its objectives of turning loss making operations into profitable ones and of selling assets to allow a reduction in the national debt, appears to be well advanced. However, assessing the degree to which efficiency of the economy has been improved is not as easy. For example, the approach of selling public enterprises to existing corporations (such as the sale of Air New Zealand to a consortium which includes Qantas), may well reduce competition in the absence of pro-competitive measures such as increasing access to the market by other carriers.

Main lessons

The New Zealand experience highlights the wide breadth of government activities to which commercial operating principles can be applied with a view to improving their efficiency. It also highlights that with a strong commitment by government, major

changes in the operating environments of public organisations can be implemented in

a relatively short time.

Of interest also is the apparent resolution of the generally perceived conflict between introducing commercial operating practices and the need for public authorities to fulfil non-commercial objectives. The New Zealand approach is based on the assumption that if the objective of privatisation is to maximise revenue from asset

sales, placing targeted organisations on a sound footing is a prerequisite to the sale.

Other aspects of interest are whether the main method of assets sales to date (ie by tender) will ultimately reduce competition and whether regulatory functions necessary to maintain the public interest will be undertaken by industry specific-bodies or a single body having wider scope.

N3 Canada

Background

At the federal level there are currently 54 parent Crown corporations and 117 wholly owned affiliates, accounting for C$60 billion in assets and about 190 000 employees (about 2 per cent of total employment). At the provincial level, the public enterprise sector is larger than that at the federal level, in terms of both assets and employees,

mainly because of the substantial resources involved in province-owned electricity utilities.

The development of Crown corporations in Canada arose from a number of factors. Principal among these was the small population (currently 25 million) and the large land area (10 million sq.km). As a result, a number of vital commercial activities, such as transport and energy, were beyond the means and resources of the private

sector to supply, thereby encouraging government involvement.

A policy of privatisation of Crown corporations was announced by the Mulroney Government in its May 1985 Budget speech. The main rationale for pursuing a privatisation policy appears to be a belief that:

. in many cases, the original reasons that lead to the establishment of a

particular Crown corporation may no longer be applicable; and

. Crown corporations are not as effective as privately owned organisations, insofar as they tend to be slow in decision making and adverse to risk taking.

Perhaps more importantly, constraints since the late 1970s on new expenditure meant that measures for new programs were largely dependent on resources being released from existing programs.

The Budget speech set out a number of basic principles for privatisation. These included the need for close consultation with management, unions and the provinces, and a commitment not to sell at distressed prices. As well, the possibility of selling large corporations in stages and the question of foreign ownership were also to be

taken into account.

169

The approach

Following the May 1985 statement to undertake the privatisation program, a Cabinet task force with a secretariat at the Treasury Board was established. This arrangement centralised decisions about selection of candidates and methods for privatisation, but

left the actual execution of the sales with line departments.

Progress under this approach was slow and in December 1986 the status of the Canadian program was raised by the creation of the Office of Privatisation and Regulatory Affairs (OPRA). As a single portfolio responsible for privatisation, OPRA’s role is to control and implement the four essential elements of the privatisation process:

. selecting the enterprise to be privatised; . examining the public policy implications; . recommending to Cabinet how to sell the entity; and . actually selling it.

The process adopted is a case-by-case assessment of Crown companies to see whether government ownership any longer serves a public policy purpose. Where this is the case, an important consideration is whether the policy objective can be achieved by regulation or by some more cost effective means.

Other considerations include ones of competition policy, such as whether competition would be reduced if the Crown corporation were to be acquired by a competitor. Assessments are also made of the likely viability of the corporation under private or

mixed ownership, the regional and community impacts of the sale and consistency with other policy objectives.

When Cabinet has approved the sale of a corporation, the essential legal, financial and legislative steps are taken, culminating in the actual sale and the transfer of ownership to the private sector. An associated process is the post-sale monitoring of the organisation, as well as its customers and suppliers, in order to gauge lessons for future sales.

The outcome

The value of assets privatised by the Mulroney Government since 1985 is around C$6.0 billion. Crown corporations privatised in recent years include Canada Arsenals, a munitions manufacturer; Canadair Ltd, an aircraft maker; and Teleglobe Canada, Canada’s international telecommunications company. At present, privatisation of four enterprises is underway, including Air Canada (of which 45 per cent was sold in September 1988 for C$3.2 billion).

In a survey by OPRA in 1988 of recently privatised Crown corporations, including Canadair Ltd and Teleglobe Canada, opinions were sought on the major commercial benefits that had been achieved following privatisation. These were said to include:

. an increase in the freedom to aggresively compete in existing and new markets;

170

• easier access to financial markets and an expansion of access to foreign markets and joint ventures;

. an opportunity to provide a better and wider range of services to existing and potential customers;

. an ability to lower the cost of their goods or services; and

■ the creation of a new image for the corporation which increased confidence both inside and outside the companies.

Privatisation by provincial governments varies greatly in its extent. In general, the focus of privatisation efforts has been on selling Crown enterprises that were already competing in the private sector. The provinces which are most advanced in this regard are British Columbia, Quebec and Saskatchewan. Additionally, British

Columbia has engaged in some contracting-out of services previously provided by government departments (eg maintenance of bridges and highways). To date, about C$4 billion worth of assets have been sold, compared with the total assets of provincial Crown corporations of around C$140 billion.

The Federal Government’s privatisation program also involves the contracting out to the private sector of activities previously carried out within the government sector. For example, travel arrangements have been contracted out to private firms; some functions previously performed by the Department of Energy, Mines and Resources

in the area of surveys, mapping and remote sensing have been contracted out; and Canada Post Corporation has recently begun to contract out full postal services to the private sector on a franchise basis.

Main lessons

The experience of Canada as regards privatisation is relatively recent. However, it would seem to have particular relevance to Australia as both countries have three levels of government. There appear to be two main lessons for Australia from the Canadian experience to date.

The first is that a case-by-case approach to reform appears essential to allow appropriate consideration of all relevant issues, including alternatives to any remaining public policy requirements. The Canadian case-by-case approach largely accords with the approach already being taken in Australia, although the added

dimension in Canada is that the starting point in each case appears to be the questioning of the need for continued public ownership of Crown corporations rather than improving efficiency whilst retaining government ownership.

The second lesson is that if governments are serious about improving economic and operational efficiency, either through privatisation or other means, the use of a single independent body seems an effective way to proceed. That is, vesting control of the necessary reviews and the implementation process in such a body is more likely to

overcome some of the problems that have emerged (eg. lengthy negotiations and poor communications with the public, management of the companies being sold and the employees involved), than if the process is left largely to the organisations themselves.

171

N4 Assessment

Each of the countries examined has privatisation programs underway. In the case of the UK and New Zealand, the programs were preceded by reforms aimed at improving the operating efficiency of government business organisations through administrative means and, where possible, by injecting competition into the markets of the organisations concerned. Like Australia, the measures employed in the UK and New Zealand are ones designed to replicate as far as possible a private sector operating environment.

The motivation behind the privatisation programs in each of the countries examined appears to come principally from two sources. The first has been an attempt to reduce budget deficits (or increase revenue), and the second a desire to increase

economic efficiency.

There is recognition in each of the countries concerned that the detail of particular reforms needs to be assessed on a case-by-case basis and that the main issues to be resolved are the appropriate regulatory framework and the manner of selling the assets.

The processes for considering such matters differ somewhat between the countries examined, ranging from a highly centralised approach in Canada to one that has been developed on a case-by-case basis in the UK. Differences are also apparent in the approaches taken to regulation and asset sales. The UK, for example, has industry- specific regulatory bodies (a path being followed by Australia with the recent establishment of AUSTEL) while New Zealand is still exploring whether a single body can oversee all industries. New Zealand sells its assets mainly by competitive tender to existing private corporate entities rather than offering them for sale to the public by flotation of shares in the UK manner.

It is extremely difficult to say which, if any, of the reform approaches adopted overseas would be appropriate for Australia. Nevertheless, as Australia moves along the path of public enterprise reform, it is as well to be aware that overseas experience can illustrate a range of problems that may be encountered and the various approaches that have been tried to address them. Australia could benefit by close observation of overseas experience.

REFERENCES

Γ"

New Zealand, Minister of Finance 1988, Budget, Part 1 : Speech, Government Printer, Wellington.

New Zealand Government 1987, Economic Statement, Government Printer, Wellington, December.

Vickers, J. and Yarrow, G. 1988, Privatisation: An Economic Analysis, The MIT Press, Cambridge US.

1 780b44 107802 89/21 648 Cat. No. 89 1783 8