Privatisation-aer

PRIVATIZATION IN AUSTRALIA: UNDERSTANDING THE INCENTIVES IN PUBLIC AND PRIVATE FIRMS.1

Stephen King, Department of Economics Melbourne University and Rohan Pitchford, Centre for Economic Policy Research Australian National University

Abstract: Privatization has been an important tool of government policy in Australia and overseas in the last two decades. We explain recent contributions to research in privatization, and apply a simple framework to ownership policy in a wide variety of Australian cases, including prisons, airports, Telstra, water and gas distribution, and ambulance services. The framework is not limited to these applications, and is aimed at providing a starting point for policy makers in their assessment of alternative ownership regimes. Our analysis is supportive of other authors, who have cast doubt on the wisdom of prison privatization, and we extend this conclusion to ambulance services and the disposal of highly toxic waste. Application of our framework also suggests that Australian privatizations may have involved excessive separation of assets. The framework also provides a basis for arguing that a key monopoly component of Telstra – the ‘wires’ component – be kept in public ownership, and access auctioned to service providers. We consider the possible pitfalls of corporatization policy, and argue that corporatized entities may operate to improve the appearance of success at the expensive of the reality.

1

This paper is a revised version of the inaugural lecture presented by Stephen King at The

University of Melbourne, May 26 1998, and a presentation to the 1998 Industry Economics conference at ANU. We would like to thank David Johnson and participants at both presentations for their useful comments.

1

1. INTRODUCTION Privatization has become a major part of Australian government policy in recent decades. Revenues from asset sales have exceeded $61 billion since the 19891990 financial year.2 Planned future sales include public electricity utilities in New South Wales, South Australia, Tasmania and a strong possibility of the sale of ACTEW in the ACT.3 The Victorian government is in the process of selling its gas distribution and retailing companies, and has called for expressions of interest for the sale of the Austin and Repatriation Medical Centre. The Western Australian government recently announced the sale of the Dampier-to-Bunbury natural gas pipeline for $2.4 billion. At the local government level, competitive tendering and the contracting-out of services such as office cleaning, park maintenance and garbage collection, has become the norm.4 A variety of new infrastructure facilities, such as the Melbourne City-Link project, are being built, owned and operated by the private sector. In some cases, these facilities will revert to public ownership at a set future date. State governments regularly ‘outsource’ prison services, so that over a quarter of Australia’s prison population will be held in private institutions by the turn of the next century.5 The Federal government has instituted a scheme to outsource job matching and employment services formerly undertaken by the Commonwealth Employment Service. Similar moves towards out-sourcing have occurred with the Commonwealth Rehabilitation service. All major airports in Australia, with the exception of Sydney, have been privatized. The federal government also recently announced the privatization of air traffic control facilities. There has been discussion of privatization of some aspects of ambulance services in Victoria, and consideration is being given to

2

Throughout this paper we use the term ‘billion’ to refer to one thousand million. For a useful

summary of government asset sales since 1990, see “A country going, going, going … private”, by Michael Bachelard, The Australian, 27 April 1998, p4. 3

See “ACT orders review of power plans”, by Mark Skulley, The Australian Financial Review,

11 May 1998, p5. 4

See Domberger and Hall (1995) for case studies on local government contracting.

5

See “A country going, going, going … private”, by Michael Bachelard, The Australian, 27

April 1998, p4.

2

the privatization of the disposal of toxic waste, specifically ozone-destroying Halon gasses. Despite the wide diversity of business that has been privatized in Australia and overseas, and the large asset values involved, it is only in recent years have economists begun to develop rigorous theories to explain the difference between public and private ownership. Grossman and Hart (1986), Hart and Moore (1990), and more recently Hart (1995), DeMezza and Lockwood (1998) and Rajan and Zingales (1998) are some of the papers that examine how the ownership of assets affects managers’ incentives. This literature has formed the basis of recent articles that compare private with public ownership. Hart, Shliefer and Vishny (1997) develop a model focussed on cost and quality trade-offs in contracted-out government services. Bolton and Xu (1997) consider ownership of schools. King and Pitchford (1998a) develop a policy taxonomy aimed at setting out the basic trade-offs between different ownership structures for a wide range of businesses. King and Pitchford (1998b) examines the governance structures often used in practice, including corporatization, with a focus on unintended effects from using imperfect indices as measures of performance. In this paper we show how recent formal research into asset ownership can better inform public policy decisions. There are two related goals. One is to criticise the sometimes ill-informed and often rhetorical or even dogmatic approaches that have been used to justify pro and anti privatization stances. The other is to apply some of the ideas from formal research to specific instances of privatization, with an emphasis on Australia. The basic methodology we use comes from the recent privatization literature mentioned above, particularly emphasising King and Pitchford (1998a) and (1998b).

3

2. The case for and against privatization A policy in search of a rationale. Much of the debate on privatization is based on rhetoric rather than research. Those in favour of privatization often assert that the profit motive makes private managers more effective than public managers. Those against, argue that the lack of social incentives in private firms makes them act against the public interest, and therefore that public ownership is preferred to private. The debate is by-and-large polarized between these two extreme views, however, it is fairly clear that advocates of privatization are winning the debate. The problem is that much privatization is being undertaken without a clear idea of the incentives that are being created in the new structures. Privatization in Australia seems to have been done without specific research into its likely effects. To a large extent, the recent wave of privatizations is simply a copy of the policy of other countries, particularly Britain. Unfortunately, the British privatizations themselves were not founded on solid research. In fact, the lack of research in the United Kingdom led Kay and Thompson (1986) to declare privatization a ‘policy in search of a rationale’. Arguments for privatization. Many of the arguments presented for privatization appear superficially attractive. However, they frequently lack intellectual rigour. For example: Efficiency: It is often claimed that the public sector is intrinsically less efficient than the private sector. While there is some empirical evidence to support this claim, it is unclear what these studies really show.6 The efficiency measures adopted in many of the studies are relevant for private sector firms. Measures of ‘net profit before interest and tax’ were used in the UK to illustrate poor public sector performance in the late 1970s.7 The Australian steering committee that monitored government business

6

See Kay and Thompson (1986) and Vickers and Yarrow (1988) for a review of the relevant

empirical literature. Donahue (1989) provides an excellent discussion on the difficulties of comparing public and private firm efficiency. 7

See Vickers and Yarrow (1988, chapter 5).

4

performance used financial measures such as ‘return on assets’ and ‘sales margin’ in its 1993 report.8 However, efficiency should be measured relative to the objectives that the firm is trying to achieve. If public sector managers are provided with different incentives and objectives to their private sector counterparts, then measuring public sector performance by private sector benchmarks will provide little useful information. In fact, where both private and public sector managers have similar incentives and objectives, performance differences are less obvious.9 The use of private sector benchmarks to assess public firms obscures the real issue: Which form of ownership maximises the sum of private and social surplus?

Government interference: Privatization, it is said, may improve performance by limiting government interference in company operations.10 In an ‘old style’ Australian public sector firm, business was conducted under the watchful eye of the minister. Sometimes routine decisions had to be ratified by the minister, and resulted in significant delays. New-style public sector firms are operated at arms length from the relevant minister. However, ad hoc ministerial intervention still occurs and the scrutiny of public sector firms, particularly with regards to public accountability, has a strong indirect influence on managerial behaviour. The criticism of excessive government interference begs an important question. Government interference is not limited to public sector firms. There are currently (at least) five separate bodies involved in the regulation of telecommunications, including Australian Communications Authority, the Australian Competition and Consumer Commission, the Telecommunications Industry Ombudsman, the Australian Communications Industry Forum, and the Australian

8

See Steering Committee on National Performance Monitoring of Government Trading

Enterprises (1993). 9

For example, see Forsyth and Hocking (1980) and Caves and Christensen (1980).

10

For a formalisation of this argument, see Boyco, Shleifer and Vishny (1996). Lopez-de-

Silanes, Shleifer and Vishny (1997) provide evidence for the political nature of the privatization process in the U.S.

5

Communications Access Forum.11 But most of the firms in the industry are fully or partially privatized. In the United States, the degree of government interference, if anything, exceeds that in Australia.12 The basic question often left unanswered by the critics of public ownership is whether a public firm subject to public sector style government interference, is preferred to a private firm with private sector style (regulatory) interference. Again, to answer this question we need to consider which regime leads to a higher sum of private and social surplus.13

Lack of competition: Unlike private firms, it is argued, public firms are not subject to the rigours of a competitive market and this lack of competition leads to inefficiency. This argument misses the point. Many publicly owned firms are precisely in areas of business that would not have a natural competitive market structure even if they were private. The privatization of Melbourne, Perth, Brisbane and other Australian airports has not suddenly created increased competition. Rather, these are now privately owned, regulated monopolies instead of publicly owned monopolies.14 Even when publicly owned firms operate in industries that potentially could be more competitive, this does not form the basis of an argument for transferring ownership to the private sector. Instead, it suggests that barriers to competition should be removed. Before

11

See “We’ve got the watchdogs, but who’s feeding them”, by Paul Best, The Age, 28 April

1998, page D9, for a brief description of each of these bodies. 12

For example, the forced break-up of the private company AT&T, ordered by the courts in

1984, could not occur under current Australian competition laws. For a discussion of the break-up of AT&T, see Brennan (1987) and Noll and Owen (1989). 13

A more advanced version of the ‘interference’ argument in favour of private firms is

presented by Dixit (1997), who argues that public managers must answer to a variety of conflicting ‘masters’ including politicians, the media and the electorate. Interestingly, Laffont and Tirole (1991) present the exact opposite argument. They argue that private managers face conflicting objectives, as they must answer to both shareholders and government regulators. 14

That Australia’s privatization process is creating a group of regulated private firms rather than

increasing competition, is starting to be understood in the Australian debate. See, for example, “Benign monopolies in private hands?” by Alan Kohler, The Australian Financial Review, 12 May 1998, page 21.

6

their privatization, a number of public enterprises in Australia competed with private firms. These include Australian Airlines, the Commonwealth Bank and the NSW Government Insurance Office. It is not obvious that these firms were less efficient than were at least some of their privately owned competitors.

Protection of public managers: An argument that is less-often heard today after the significant restructuring of federal and state public services over the last decade, relates to public manager incentives. Critics point out that because of job security, public managers are not forced to operate as efficiently as private sector managers. Public managers are also immune from ‘take-overs’ that can result in them being sacked. While this may have been true at one stage, it is far from obvious today. Changes to the public sector and experience with private sector schemes, such as ‘golden parachutes’, that protect incumbent managers, mean that public managers may have less security of tenure than their private sector counterparts. Even if the original argument remained true, it might constitute an argument for restructuring public sector incentives, depending on how well different incentives motivate the achievement of desired goals. It is not an argument for changing ownership per-se.

Market monitoring: Private firms, unlike government enterprises, are continually monitored by capital markets. Managerial decisions are reflected in the share price. Corporate debt holders analyse company solvency. Private owners receive dividends, appoint directors and hire or fire managers. This market monitoring, it is argued, guarantees that private managers will respond to owners’ wishes. The managers of government business enterprises face no similar constraint. While capital markets do monitor private firms, it is unclear exactly how this is reflected in firm performance. Unlisted private firms are not subject to share market scrutiny. Debt holders have only a limited say in corporate decisions. The link between share market performance and managerial rewards is often tenuous and may create undesirable incentives.15

15

See Grant, King and Polak (1996) for a survey of these issues.

7

Even if capital markets allow firm owners to perfectly control private sector managers, this does not mean that private sector firms are socially desirable. Again, the argument misses the key issue. A profit maximising private firm may not result in a higher sum of private and social surplus than an equivalent government owned firm. The case against privatization. While the common claims for privatization often lack depth, the same criticism can be levelled at many of the arguments for public ownership. Eva Cox, in the 1995 Boyer lectures, argued that public ownership benefits society by creating ‘social capital’ and a pride from collective ownership, making no mention of costs or other kinds of benefits.16 The logical end to such an argument is nationalisation. The experience in the former Soviet Union provides ample evidence of the failure of such a system to achieve more than very basic social or economic goals. Many of the opponents of privatization rely on claims that public managers will be more prepared to act in the public interest than will private managers.17 Despite the convenient juxtaposition of the word ‘public’, this defence lacks rigour. Why should a public manager behave in this way? If public managers act more in the public interest, why doesn’t the government pass a law that gives them input into the hiring of managers for private firms? To provide a rigorous argument, it needs to be explained why a given individual is differently motivated under public than private regimes. Other arguments against privatization focus on the difference between claimed and actual cost savings. Quiggin (1994) notes that the cost savings claimed from ‘contracting out’ often include a transfer from workers. While the purchaser of the service gains, the workers lose. From an economic perspective, such a transfer is not a social gain. Quiggin (1995 and 1996) also argues that the sales revenue from privatization may be less than the present value of the flow of future profits that the

16

See Cox (1995).

17

See, for example, the reaction by the AMA federal president, Keith Woollard, to the

privatization of the Austin and Repatriation Medical Centre, in “Kennett opens hospital sell-off”, by Michael Bachelard, The Australian, 16 May 1998.

8

government could earn under public ownership. His “equity premium” argument against privatization is that if the higher return accorded to private equity compared with government debt reflects an economic cost, then moving from debt to equity financing with privatization will result in a net loss. While more sophisticated, these two arguments say nothing concerning the crucial question of manager’s incentives. This is clear in the case of contracting out, where the real issue is what motivates managers in the contracted entity, and subsequently how well objectives are met. The equity premium argument, along with the formalised version in Grant and Quiggin (1998) is certainly thought provoking. However, it does not explain why the government should have control of the public asset. In the context of their argument, the government could own the majority of shares of private companies as non-voting equity. The resulting firms would essentially remain private in nature, and behave much as they do at present. Rather than defending public ownership, the arguments either support alternative modes of funding, reflect potential inequities associated with privatization, or in the case of earlier arguments, are unsupported claims.

3. Understanding Privatization The arguments for and against privatization all have at least one of two major deficiencies. The first is they neglect to specify clearly the objectives that ownership policy is meant to achieve. This is the problem with studies that compare the so-called efficiency of private and public firms. The other deficiency is that they fail to consider the basic differences between the incentives that face private and public managers. If we do not know how behaviour will differ under different regimes, it will be impossible to assess whether these regimes are likely to achieve desired objectives. We take the objective of ownership policy to be the same as for other kinds of policy such as taxation, industry, and macroeconomic policy, namely the maximisation of some measure of society’s surplus. It is convenient to divide surplus into four components. The first refers to the change in the value of society’s capital assets. Public and private managers undertake a variety of activities that alter the

9

capital stock.18 New investment in productive assets raises the capital stock while ineffective maintenance or inappropriate asset use may diminish societies asset base. We use the symbol a to reflect the change in the gross value of capital. Managerial actions affect the value of a. Society gains welfare from production but also bears production costs. These costs, represented by c are also affected by managerial behaviour. Given the level of output, welfare falls as c increases. Firm profits from managerial activity are given by a – c. Thirdly, managerial activities may spillover to the welfare of other people in society. Consider the manager of a coal-burning electricity generator. If the manager changes from a low to a high sulphur-content coal, this will increase pollution and lower general social welfare. But if the manager improves maintenance procedures pollution may fall, raising social welfare. We denote the external benefits (or harms) created by a firm by b. Managerial activity will affect the level of b. For simplicity of notation, b > 0 will refer to a social benefit; b < 0 will refer to a social harm. Finally, managerial activity will involve a personal human resource cost or benefit to the manager. More efficient firm operation does not simply occur by managerial fiat. Rather, it reflects managerial diligence and personal effort. The human resource cost of managerial activities is given by h. It will clearly have a strong direct effect on managerial behaviour. An activity that lowers managerial welfare and requires compensation is denoted by h > 0. Conversely, if an activity improves the manager’s human capital or is pleasurable then h < 0. As well as directly effecting managerial welfare, h is also part of social welfare. A change in h reflects the private and social opportunity cost of managerial activity.19

18

We use the terms ‘private owner’ and ‘private manager’ interchangeably. In other words, we

assume that a private manager will act in the interests of the owner. This is obvious for ownermanaged or tightly held firms. It is less obvious for large firms with dispersed ownership, but as noted above the profit motive that occurs under private ownership is often used as an argument for privatization. In this sense, our model starts from a basis that is favourable to privatization but shows that it need not be optimal. 19

For ease of notation, we set the ‘base’ values of a, b, c and h at zero. In other words, in the

absence of managerial activity all these values would equal zero.

10

We take the objective of policy as the maximisation of the sum of asset value less costs, and external benefits less managerial costs. Thus welfare W is given by W = a – c + b – h. This objective is deliberately chosen to include both what could be considered ‘private’ objectives – concern with asset value and both production and managerial costs – and public objectives – concern with external effects. It is considerably more difficult to establish the difference in incentives between private and public firms. Research has not yet reached the stage where these differences are fully understood. Surprisingly, it turns out that the biggest puzzle is to explain why there should be any difference between public and private ownership. This is a puzzle, because casual and more formal empiricism indicates that incentives are indeed different!

Incentives and ownership. An understanding of the nature of this puzzle is gained by considering the conditions under which there is no difference between private and public regimes. Suppose the government has a measure of asset value ar, and a measure of current costs cr, where the superscript r indicates a reported value. Assume that these measures are verifiable by a court so that they can be used either in incentive contracts for public managers, or in regulatory rules that affect private managers. A basic result in the recent privatization literature is that if complete contracts can be written over relevant production variables, then there will be no difference between public and private ownership.20 In the context of our example, suppose that measurement of the relevant variables is perfect, so that ar = a, and cr = c. Then we can replicate the outcomes under private ownership in a public firm, and outcomes under public ownership in a private firm. To see why, note that the government can set an incentive contract for a public manager based on a and c, T(a,c). It can also set regulatory rules for a private firm denoted by R(a,c). In the absence of an incentive contract, a public manager will simply receive a fixed wage w enough to cover human resource cost h, and gains

20

See Schmidt (1990) and Hart Shliefer and Vishny (1997), for example.

11

utility u0 = w – h. A public manager has no claim over the firm’s assets, which are owned by the government, and is not personally liable for the firm’s costs. Thus a public manager does not directly receive a or pay c. In contrast, in the absence of regulatory rules, a private owner receives the true value of their assets and is liable to pay the true costs, and so enjoys the utility u1 = a – c – h. But the government can get the public manager to face exactly the same payoffs as a private owner. It simply sets T(a,c) = a – c – w. The government rewards the public manager according to increases in asset value and ‘punishes’ the manager for increases in costs. The manager receives utility T(a,c) + u0 = a – c – h, the same as an unregulated private owner. The government can also make a private owner behave like a public manager by taking away the incentives to raise asset value and limit costs. The regulatory contract R(a,c) = c – a + w achieves precisely this objective. In other words, the government can indemnify the private owner from bearing any increased costs but ‘fine’ the private owner whenever asset value is increased. The private owner receives utility R(a,c) + u1 = w – h. Their payoff is now identical to a traditional public manager. In fact, this argument generalises. With perfect and verifiable measures of asset value and cost, any outcome that can be achieved under private ownership and regulation can be replicated with public ownership and an appropriate incentive scheme, and vice versa. If there is any basic difference between public and private firms, it must be that governments do not have perfect measures of asset value and cost, so that in general ar ≠ a and cr ≠ c. Imperfect performance measures, by themselves, do not explain the difference between public and private firms. In addition, the residual rights that public and private managers enjoy must be different, where ‘residual rights’ mean the rewards that managers receive in the absence of regulation and incentives. This was, of course, built into the values of u0 and u1 given above. Our basic thesis is that the difference in the degree of public accountability measures determines the difference in residual

12

rights of private and public managers.21 Stringent public accountability measures make it very difficult for public managers to capture a share of asset value, either through direct diversion, or pet projects. Activities that even hint of value diversion are often curtailed, including the buying and selling of physical and financial assets, and the payment of ad-hoc bonuses.22 In contrast, private managers have much more freedom to undertake potential value diverting activities. A key difference between ownership regimes is the beneficiary of increases in the net value of the assets that underlie the business. A public sector manager, at the end of their tenure, has no claim on these net assets, as they belong to the government. In contrast, a private owner retains the assets and has the right to sell them and receive the value of the assets through this sale. Where ownership and management are separate, side payments can be made to reward managers in a variety of ways. In summary, compared to private owners of an asset, public managers have less incentive to undertake personal activities and investments that will increase net asset value. In the simple example presented above, private manager’s incentives are in line with the surplus u1= a – c – h, while public manager’s incentives depend only on the personal cost – h.23 Determinants of ownership policy: a basic case. As we have argued, ownership affects incentives, and a reasonable objective to judge the performance of different ownership regimes is to compare the values of social welfare W = a – c + b – h that are generated. The reader will notice that social surplus differs from private surplus by the amount b, the external net benefit that the business generates. Therefore, a basic difference between public and private

21

See King and Pitchford (1998a) for details.

22

As a simple example, before the management of the government-owned Port of Brisbane

Corporation can allow an employee to undertake work-related overseas travel, they must apply for approval from multiple government departments. The approval process takes approximately 40 days. 23

Note that – h can be positive below a certain level of activity and negative afterwards,

indicating that the manager gets some benefit from work up to a point. The benefit of work can come in a variety of ways, including benefit from an addition to human capital that accrues as a higher salary in the future, or simply enjoyment of work.

13

ownership is the effect these regimes have on external benefits. We use the term external benefits generically, however, and note that they may be positive or negative externalities. We are now in a position to assess the performance of private and public regimes, at least at a basic level. We will do this through a series of examples. To start, consider a simple natural monopoly firm. Such a firm faces no direct competition from close substitutes in either a public or a private regime. A private owner has incentives to undertake activities that lower production costs and boost private profit. A private owner also has an incentive to undertake activities that extract consumer surplus from the firm’s customers by reducing the level of output. This reduction in output leads to an allocative inefficiency so there is an external harm (b < 0) and this harm increases as output falls. The private manager enjoys a commercial gain from lowering production costs and from cutting back output. In contrast, a public manager gets no commercial gain from decreasing output or reducing production costs. Therefore, the private manager will generally deliver lower costs, lower output and higher prices than will the public manager. The private firm therefore generates higher profits and lower external benefits than does the public firm. To assess whether or not to privatize, we must weigh up the gain in profits against the loss in external benefits. The advantage of an approach that clearly spells out the objective and the private and public manager’s incentives is that it enables us to explicitly consider the important trade-offs that are inherent in the privatization decision. The generic case of a monopoly is a useful warm-up example. We can apply the model as it currently stands to highlight the implicit trade-offs in recent Australian privatization decisions. Consider Prisons. Suppose that net asset value a – c is increasing in some “cost reducing” innovation i. However, the quality of prison services is measured by q which is decreasing in i. Following Hart et al (1997) quality includes rehabilitation possibilities, humane treatment, or training of guards, among other things. A decrease in prison quality may lead to social harm so that b declines as q declines. With q (and hence b) decreasing in i, we capture the effect that cost cutting can lead to neglect of quality. A private firm has greater incentive than a public firm to undertake cost

14

reductions. A private prison will deliver higher profit and lower quality than will a public prison. A similar trade-off is readily apparent in the privatization of other services. Consider Ambulance Services or Toxic Waste disposal. If q refers to the quality of ambulance service, or the care taken to dispose of toxic waste, precisely the same trade-off is generated. The important point in each of these three cases is that if the social harm generated by privatization is sufficiently high, then privatization is not a good idea. One could conceive of toxic wastes that are costly to dispose of and where a private firm cuts corners. However, a public firm has no concern with spending a lot (in our simple example), and so will be less likely to cut corners. Privatization is a bad idea if the waste is of extreme toxicity and the fall in care sufficiently large under private ownership. The experience in New Jersey in 1987 and Los Angeles in 1996 provides evidence for this concern. Private firms were employed to dispose of medical waste. This usually involves high temperature incineration. But to minimise costs, the private firms found that it was cheaper to simply dump the waste at sea. When the waste began to wash up on local beaches, questions were raised about the desirability of cost minimisation through dumping waste near the shore.24 Similarly, if private ambulance services suffer a significant degradation in quality, then public ownership might be preferred. While we are not trying to suggest that these ownership policies are optimal, the model brings to light the kinds of factors that should be considered in practice in a careful examination of the costs and benefits of privatization.25 Ownership policy with multiple assets. Our basic framework consists of three assumptions. Private managers can capture private value; public managers are motivated by personal rewards, and society’s objective is to choose an ownership structure to maximise the sum of profits

24

See “Medical-waste mess leads to fees for health professionals”, by Shelby Grad, The Los

Angeles Times, 29 December 1996, Metro p1. 25

To fully consider, say the privatization of the ambulance services would involve far more

analysis than presented here. The key point is that this analysis needs to be undertaken. The costs as well as the benefits of different ownership structures need to be fully understood.

15

and external benefits less all costs. We can easily apply this framework to the case of ownership policy for multiple assets. In this situation, the set of possible regimes is greatly expanded. Suppose there are just two assets. As above, productive use of an asset can generate benefits or harms that affect outside parties. In addition, there may be spillovers between firms. That is, the activities of one firm may raise or lower the profits of a second firm, a2 – c2 (and vice versa). The regulatory and ownership issues facing a government are clearly more complex with multiple assets. Should certain assets stay together, that is, be managed in one firm, or should they be separated and managed in separate firms? If the assets are kept together should the firm be public or private? If assets are separated, which should be placed in private hands and which are best in public ownership? While considerably more complex to analyse, one result that we can state clearly is that when private and social gains are aligned, private ownership is optimal. Suppose the government is considering privatization of a group of assets, whose use in production generates either positive or only small negative external effects, and that there are positive spillovers between the activities associated with these assets. Such assets should then be privatized under the umbrella of one company. Positive spillovers are internalised by the firm, and the commercial incentives of private ownership ensure that external benefits are high (or external costs are not too high) compared with public ownership.26 It will clearly not conveniently happen that all practical cases fit into this category. However, the case of unified private ownership of multiple assets provides a useful benchmark against which to weigh the costs of other regimes, such as public ownership of one or more of the assets involved, separate private ownership, or mixed ownership with unified management.

26

Commercial research and development joint ventures provide a simple example.

There are often positive spillovers for firms engaged in commercial research and positive externalities for society in general. The desirability of encouraging research and development joint ventures in such circumstances has been reflected in US legislation. See Jorde and Teece (1990).

16

Telecommunications Consider the case of telecommunications in Australia. Should Telstra be retained as a single company or should it be divided into a number of separate companies? Suppose a privatized Telstra were vertically separated, into a ‘wires’ company and a retail telephone company. There may be a loss of economies of scope between the upstream wires operation and the retail operations (with reference to our benchmark, this is a loss due to the fact that positive spillovers will not be internalised). That is, certain kinds of technical coordination would be lost. What are the potential external benefits from vertical separation of Telstra? With separation, the wires company and the retail phone company would find it harder to coordinate on anti-competitive activities that harm downstream competitors. A number of telecommunications companies in Australia have complained that Telstra, as an integrated company, can distort competition by charging different external and internal prices or by restricting local network access. In our framework, this anticompetitive behaviour may result in negative external effects. Separating Telstra may reduce anti-competitive behaviour, but this must be weighted against a reduction in profits due to a reduced ability to account for positive spillovers between assets. Should Telstra be wholly privatized, or would some other form of ownership be desirable? An alternative to the full privatization of Telstra is ownership of the wires portion of the company by government while privatizing the retail telephone company. As discussed, this could cause problems of technical coordination, and consequently reduce positive spillovers. What would be the benefits? It is widely recognised that it is the wires portion of a telephone company that gives monopoly power. We have argued that a public monopoly is less likely to pursue profit. Softer behaviour with regard to profits might enable the following kind of arrangement. The government could tender for construction of new cable networks and for the operation and maintenance services of the wires company. Spectrum in the cable could be auctioned much as mobile phone spectrum is auctioned in Australia. Parties, including Telstra retail that bid for cable spectrum could well benefit from lower prices. However, Telstra retail would no longer own the engine of its monopoly power.

17

Victorian gas reform Gas reform in Victoria provides another example where we can apply our model. The Victorian government has vertically separated gas transmission from distribution and retailing. It has also horizontally separated distribution into regional monopolies. The state government is in the process of privatizing these gas businesses, and is creating a ‘spot-market’ to facilitate gas trading. In terms of our benchmark unified private firm, it is as if policy makers have decided that the benefit from internalising spillovers is outweighed by some external cost. It is far from obvious that these changes are the best choice. At present, almost all the natural gas sold in Victoria comes from a single source – the ESSO/BHP joint venture that controls the Bass Strait gas fields. With only a single upstream producer, the addition of a spot market downstream may offer only cosmetic improvement. A spot market with one seller is still a monopoly! Breaking assets up into transmission, distribution, and by regions means there will be many buyers. However, being regional monopolies in gas, it is unclear how these buyers can compete with oneanother. As we have commented, it is quite possible that downstream reforms may actually lead to higher costs – again due to spillovers not being internalised – and therefore less efficient transmission and distribution. If separation leads each part of the vertical chain of production to add its own margin, then final gas prices may rise unless there is rigorous regulation.27 The application of our framework raises questions as to whether these complex reforms are necessary. To complete the analysis, we need to ask what external costs are avoided by the reforms. Is there some reduction in monopoly power? In Victoria, most gas is for residential rather than industrial consumption. Therefore, there are ready substitutes for many residential uses of gas. Electricity can be used for cooking, hot water and heating, as is the case in other Australian states. Solar power, coal, oil and wood are also substitute household fuels. If the electricity market is competitive, it is hard to see the source of monopoly power, and hence

27

That vertical disaggregation may lead to higher prices is a well-known result in economics,

called ‘double marginalisation’.

18

external harm, that is avoided by the complex Victorian gas reforms. Therefore public ownership of the gas distribution system appears unnecessary. The Victorian gas industry may be a case where there is little conflict between private and social objectives. A private gas utility may be unable to exercise market power without losing market share to electricity. It is far from obvious that there are any spillovers from private gas consumption and production to wider public welfare. If private incentives and public welfare are aligned then unified unregulated private ownership is optimal. The Victorian government may be wasting taxpayer’s money by restructuring and regulating the gas industry. Rather, it could have simply privatized the former Gas and Fuel Corporation and allowed interfuel competition to provide the correct social incentives.28 Airports Australian airport privatization provides another example of the separation of assets. When the United Kingdom privatized its airports, it sold them as one integrated operation. In Australia, the federal government decided to sell the airports separately. There are important operational spillovers between airports. If a plane is delayed leaving Melbourne airport then this can lead to significant rescheduling at the destination airport – say Brisbane. Further, it is hard to regulate to avoid delays that benefit the originating airport but are otherwise unjustified. No regulator would force a plane to take-off if the airport claimed that it was currently unsafe to do so due to weather or congestion.

28

The ‘gas crisis’ in Victoria in September 1998, brought about by a fire in the ESSO plant at

Longford, may affect future Victorian gas reform. This crisis was moderated by the ability to transport gas from New South Wales over a recently completed interconnector between Albury and WaggaWagga. The construction of this interconnector followed competition reforms that removed historic limitations on gas trade. This suggests that interstate gas trade may provide an ‘insurance’ benefit to gas consumers. To the degree that this represents a positive social benefit from privately profitable activity, it improves the case for privatization together with access reforms to aid interstate gas sales. For a slightly longer discussion of gas reform, and a general discussion of regulatory issues in vertical production chains, see King and Maddock (1996).

19

If a privatized Melbourne airport acts to lower its costs by not properly scheduling air movements, or by downgrading maintenance which leads to delayed departures, then this will raise the costs to the privatized Brisbane airport. It will also place an extra burden on passengers. To avoid these spillovers it may be desirable to have joint ownership of the airports. In this way, when the owner of Melbourne airport imposes costs on Brisbane airport, the same owner who creates the costs bears the costs. What are the external benefits of separate ownership? Some have argued that separation is justified as a way of reducing monopoly power. However, competition between airports might be quite limited. Airports are characterised by large fixed costs, and at Australian population levels, there has tended to be only one major airport in each city. Thus, competition between airports is by-and-large restricted to competition between destinations. While it is not possible to undertake a definitive analysis in this paper, we argue that to justify separate ownership, the claim that it will lead to increased competition needs to be examined closely. If competition is sufficiently high, though this seems doubtful, separate private ownership would be justified. If there were very little competition, joint ownership would be preferred. Water The water industry also illustrates the dangers of separating assets. In Melbourne, for example, the body that controls the headwaters and water transmission, Melbourne Water, has been vertically separated from water distribution. The water distributors have also been horizontally separated into local monopolies. This separation may offer scope for ‘game-playing’ between different parts of the production chain. Suppose consumers notice an increase in water turbidity. To whom do they complain? The water distributor is likely to blame the transmission company, claiming that they receive ‘dirty’ water. The transmission company will blame the distributor, arguing that the turbidity is due to a broken pipe or high leakage levels. It will be difficult for the consumer or a regulator to correctly allocate liability for a reduction in water quality. Because of the spillover between the vertical stages of production in the water industry, separation is likely to result in lower levels of maintenance and a reduction 20

in water quality over time. Each party only bears part of the cost of a reduction in quality due to lower maintenance and only receives part of the credit when they increase quality by improving maintenance. As a result, separated companies will neglect maintenance compared to an integrated company. This will hold under either private or corporatised public ownership. Whichever ownership policy the government adopts, it is likely to be better to keep the assets together. Summary We have considered four cases involving the break-up of assets: the Telstra privatization, Victorian Gas, Australian Airports, and Victorian Water. Our framework admits consideration of seemingly more exotic forms of ownership. Suppose there is a government-owned national park located beside a private resort. The private resort benefits from the national park’s proximity and is interested in the maintenance of the park. In other words, there are positive spillovers from the park to the resort. However, handing ownership of the park to the resort may lead to undesirable spillovers in the reverse direction. It may be difficult to prevent the resort owners from exploiting market power in pricing the entrance fee to the park, for example by excluding non-resort users from the park. It may also be difficult to prevent the eventual encroachment of the resort into the park. If the park is retained in public ownership, resort owners will find it harder to pursue these undesirable goals. However, a public park manager will not internalise the spillovers from the park to the resort. This suggests that public ownership of the park, private ownership of the resort, but management of both by the resort owner might be the desired solution. Ownership policy is not as simple as our analysis portrays. However, we feel our framework is a useful starting point for practical policy. Even within the simple framework, there are potentially complex and diverse regimes that can be considered. And the applicability of any single regime will vary according to the external benefits, spillovers between assets, profitability of business, and basic motivations of managers. It is hard to escape the conclusion, that ownership policy is best done on a case-by-case basis, and done carefully.

Regulation and incentives

21

We argued above that without perfect measures of asset value and cost, and by the nature of residual rights of public and private firms, there is a role for ownership policy. However, we have so far left out an important ingredient. Suppose that the reported measures of asset value ar and costs cr are imperfect, but at least yield some information about the underlying true values a and c. Given performance measures, albeit imperfect ones, several natural questions are raised. How well will regulated private firms behave? Are incentive schemes for public managers a good idea? Corporatization – where the government requires a public firm to operate like a private firm – is one example of a public sector incentive scheme. The manager of a corporatized public firm is likely to be replaced or demoted if certain cost, output and dividend targets are not achieved. We define a corporatized firm, as one where the government retains assets after a manager’s tenure is over, but in which the manager is rewarded according to the proxy ar – cr, and so receives utility u0 = ar – cr – h. Rather than facing the actual net present value of profits, the corporatized manager is rewarded according to imperfect measures of net asset value. Given the incentives outlined above, a natural question is whether a corporatized public firm really behaves like an unregulated private firm. Notice that the manager of a corporatized public firm is rewarded for reported increases in asset value and decreases in cost. The manager will therefore have strong incentives to manipulate these reports rather than change actual costs and asset value.29 In fact, a corporatized public firm may behave quite differently to unregulated private sector firm. To see this divide profits a – c into two parts, ac – cc that is current profits and af – cf that is future expected value. Managerial activity today affects both current profits and future firm value. Suppose that the government has a good measure of current profits arc – crc but the measure of future value arf – crf is poor or absent.30 A

29

A simple example of report manipulation involves claims about the rigging of waiting lists in

Victorian public hospitals. See The Age, 15 July, 1998. 30

One of the functions of a stock market is to provide a measure of future value. Corporatized

public firms are, of course, not sold on the stock market, and so the government does not have the benefit of such a measure.

22

corporatized public sector manager will be rewarded for increasing short-term profits by cutting current costs, but will face little sanction if this action undermines long term value. For example, the manager may act aggressively against current competitors – more aggressively than would a private firm – in order to increase current profit. The manager may have little interest in firm reputation and long-term customer relations. The manager may also have an incentive to neglect long-term maintenance. This will reduce current costs, albeit at the expense of long-term value. But the manager is rewarded for short-term cost reductions and, unlike a private owner, does not bear the burden of lower long-term value. Compared to a private owner, a corporatized public manager will be biased towards short-term performance. A corporatized public firm can appear to be highly successful. In fact, by the measures used by the government, it will often appear more successful than will an equivalent private firm. But this reflects inadequate performance measures, not actual performance. If the corporatized firm engages in behaviour that improves measured (but not actual) performance but that reduces social welfare, then it may be better to privatize the firm completely. This managerial behaviour to improve corporate appearance is simply the public sector analogue of the distortion that occurs under private sector regulation. It has long been recognised that regulated private firms will engage in a variety of inefficient conduct.31 Regulation alters the incentives that face private owners just as corporatization alters the incentives that face a public manager. Both ownership and regulation effect performance. A public sector manager with an incentive scheme as part of their employment contract will behave differently to a public sector manager who receives a flat wage. The owner of a private firm subject to stringent regulatory controls on profit will engage in different activities compared to a private owner of an unregulated firm. When considering the optimal ownership policy for a firm, the government must simultaneously consider any additional incentives or constraints that will be placed on managers. For example, consider the ownership of Australia Post. This is currently a corporatized publicly owned firm. If the government believes that it is most desirable

31

See for example Carlton and Perloff (1994) for a discussion.

23

for Australia Post to operate to maximise profits, then unregulated private ownership will be the optimal policy. A private owner of Australia Post will have strong incentives to maximise actual rather than reported profit. At present, however, Australia Post faces a number of community service obligations. Australia Post is required to provide a minimum level of service to remote and rural communities. It is also required to deliver a standard letter anywhere in Australia for a fixed price. A private firm may find it unprofitable to service some rural and remote communities. If the government wants to continue with these obligations then it would have to place tight regulations on a private operator of Australia Post. The private owners would have incentives to undermine this regulation and a private regulated Australia Post would behave very differently to the current government firm. Is it desirable to privatize Australia Post? To answer this question the government must trade-off between a public firm with imperfect corporate incentives and a private firm with imperfect regulation. The optimal ownership regime will depend on the effectiveness of the incentives created by corporatization and regulation. It may be better to move Australia Post to private hands, but this requires a careful analysis of both the current ability to provide incentives for the public managers to operate efficiently and the potential regulations that the government can use to moderate the behaviour of a private firm.

4. Concluding comments In this paper, we have demonstrated how formal models of privately and publicly owned firms can be applied to Australian examples. Our models are based on two key principles; (1) The objective of ownership and regulation policy should be clearly spelled out (2) The different incentives under private, public, corporatized and regulated firms must be elucidated. These two principals seem obvious, yet we have found that the standard debate on ownership policy often ignores them. Importantly, public performance cannot simply be compared to private performance without considering objectives. While public ownership gives relatively flat commercial incentives, this can be beneficial if the firm in question is undertaking socially harmful activities. 24

The two principles give us a simple way of gaining insights into possible effects of ownership policy. When social and private objectives are aligned, private ownership is best. When social and private objectives conflict, it is necessary to weigh up the gain in profits from private ownership, to the rise in social costs. Careful consideration should be given to privatization of prisons, and potential privatization of emergency services and toxic waste disposal. Our analysis, as well as that of other authors (Hart, Shliefer and Vishny (1997)), suggest that in some cases quality of service could be sacrificed for costs savings in a private entity. Ownership policy for multiple assets is made clearer using unified private ownership as a benchmark. We argue that privatization policy in Australia, with some exceptions, may be splitting up assets that might be best left together. The benefits from separating firms should be subject to significantly more scrutiny than they are at present with Airports, Water, and Gas. Specifically, the social harm that the unified entity might cause must be made explicit. In the case of Telstra, there are grounds for retaining a wires company in public ownership, because the wires company is a major source of monopoly power. The social harm in this case, comes about because entry to the wires market is exceedingly difficult and costly. Even if entry could occur, it is likely to result in wasteful duplication of capacity. Auctioned access could well reduce the price to intermediate service providers, and free entry of such providers would result in reduced consumer prices. While it is useful at first pass to separate them, ownership policy and regulatory policy go together. In practice when considering privatization, a public firm with managerial incentives should be compared to a private firm with regulation. We have argued that corporatization can have some pitfalls. If performance is based on imperfect measures, this can give managers perverse incentives. An example is shortterm profit seeking at the expense of long term value. The effects of such incentives should be weighted against the benefits of corporatization – often argued to be the adoption of a more professional culture than the old-style public firm. Governments that make ownership decisions, either public or private, without considering keeping in mind the two key principles are likely to mistakes. Unfortunately, it is far from clear that governments take account of the relevant tradeoffs and complexities in their privatization decisions. In particular, governments at 25

both the state and federal level in Australia appear to pay little attention to the reality of privatization, preferring to follow their own rhetoric.

References Bolton, P. and Xu, C. (1997) “Ownership and competition: an application to schools”, Mimeo. Boyco, M., Shleifer, A. and Vishny, R. (1996) “A theory of privatization”, Economic Journal, 106, 309-319. Brennan, T. (1987) “Why regulated firms should be kept out of unregulated markets: understanding the divestiture in United States v. AT&T”, Antitrust Bulletin, 32, 741-93. Carlton, D. and Perloff, J. (1994) Modern Industrial Organization (2nd. ed.), Harper Collins, New York. Caves, D and Christensen, L. (1980) “The efficiency of public and private firms in a competitive environment: the case of Canadian railroads”, Journal of Political Economy, 88, 958-76. Cox, E. (1995) A truly civil society (1995 Boyer lectures), ABC Books, Sydney. De Meza, D. and Lockwood, B. (1998) “Does asset ownership always motivate managers? Outside options and the property rights theory of the firm”, Quarterly Journal of Economics, 113, 361-386. Dixit, A (1997) “Power of incentives in private versus public organizations”, American Economic Review, 87, 378-82. Domberger, S. and Hall, C. (eds.) (1995) The contracting casebook: competitive tendering in action, AGPS, Canberra. Donahue, J. (1989) The privatization decision: public ends, private means, Basic Books, Forsyth, P. and Hocking, R. (1980) “Property rights and efficiency in a regulated environment: the case of Australian Airlines”, Economic Record, 56, 182-85. Grant, S. and Quiggin, J. (1998) “Equity premium puzzle and the privatization paradox", mimeo, ANU, Canberra. 26

Grant, S., King, S. and Polak, B. (1996) “Information externalities, share-price based incentives and managerial behaviour”, Journal of Economic Surveys, 10, 1-22. Grossman, S and Hart, O. (1986) “The costs and benefits of ownership: a theory of vertical and lateral integration”, Journal of Political Economy, 94, 691-719. Hart, O. (1995) Firms, contracts and financial structure, Oxford University Press, Oxford. Hart, O. and Moore, J. (1990) “Property rights and the nature of the firm”, Journal of Political Economy, 98, 1119-58. Hart, O., Shleifer, A. and Vishny, R. (1997) “The proper scope of government: theory and application to prisons”, Quarterly Journal of Economics, 112, 1127-61. Jorde, T. and Teece, D. (1990) “Innovation and cooperation: implications for competition and antitrust”, Journal of Economic Perspectives, 4, 75-96. Kay, J. and Thompson, D. (1986) “Privatisation: a policy in search of a rationale”, Economic Journal, 96, 18-32. King, S. and Maddock, R. (1996) Unlocking the infrastructure: the reform of public utilities in Australia, Allen and Unwin, Sydney. King, S. and Pitchford, R. (1998a) “Private or public? A taxonomy of optimal ownership and management regimes”, Department of Economics Research Paper number 625, The University of Melbourne. King, S. and Pitchford, R. (1998b) “Ownership, regulation and incentives”, mimeo. Laffont, J. and Tirole, J (1991) “Privatization and incentives”, Journal of Law, Economics and Organization, 7, 84-105. Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. (1997) “Privatization in the United States”, The RAND Journal of Economics, 28, 447-471. Noll, R. and Owen, B. (1989) “The anticompetitive uses of regulation: United States v. AT&T (1982)”, in The Antitrust Revolution, (J. Kwoka and L. White eds.), Scott Foresman, Glenview Illinois. Quiggin, J. (1994) “The fiscal gains from contracting out: transfers or efficiency improvements”, Australian Economic Review, 27/3, 97-102. 27

Quiggin, J. (1995) “Does privatisation pay?”, Australian Economic Review, 28/2, 2342. Quiggin, J. (1996) Great expectations: microeconomic reform and Australia, Allen and Unwin, Sydney. Rajan, R. and Zingales, L. (1988) “Power in the theory of the firm”, Quarterly Journal of Economics, 113, 387-432. Schmidt, K. (1996) “The costs and benefits of privatization: an incomplete contracts approach”, Journal of Law, Economics and Organization, 12, 1-24. Steering Committee on National Performance Monitoring of Government Trading Enterprises (1993) Government Trading Enterprises performance indicators 1987-88 to 1991-92, Industry Commission, Canberra. Vickers, J. and Yarrow, G. (1988) Privatization: an economic analysis, MIT Press, Cambridge, MA.

28