Go back to John Quiggin's home page
Submitted June 1995
Submission to Industry Commission Inquiry into contracting out in the public sector
Visiting Associate Professor
Department of Economics
James Cook University
Contracting out is an issue faced by both private and public sector organisations. In general, the optimal boundaries of the organisation are determined by the characteristics of the set of available contracts and, in particular, by the resulting allocation of risk. In general, activities where risk may most easily be monitored internally are the best candidates for contracting out.
Two issues of importance arise in relation to public sector contracting out. First, because of the imperfect degree to which the private capital market spreads systematic risk, the rate of return on debt is significantly below that on private sector equity. Hence for capital-intensive projects, public sector ownership will normally be cheaper.
Second, under certain circumstances, public sector labour markets may have bilateral monopoly characteristics. Contracting out may be used as a strategic device to improve the bargaining position of employers, thereby achieving wage reductions. In such circumstances, the reductions in service costs associated with contracting out will represent an overestimate of the social welfare gains, which may be positive or negative.
The analytic framework developed in the paper is applied to the question of contracting out of infrastructure projects. It is argued that the turnkey contract system, in which projects are constructed by competitive tendering, but publicly owned will in general be superior to alternatives such as the so-called BOOT (Build, Own, Operate and Transfer) system.
Submission to Industry Commission Inquiry into contracting out in the public sector
The issue of contracting out is one faced by all large organisations, whether they are private firms, government departments or non-government organisations. Indeed, much of the current debate about contracting out, touches on the issue first raised by Coase (1937) -- why do firms exist at all.
As has been observed by Williamson (1975), the firm may fruitfully be viewed as a nexus of contracts. These contracts span a spectrum from anonymous market relationships (for example, when drivers employed by a firm buy petrol at a service station) through short-term or long-term contracts with other organisations (for example, a large firm arranging a contract with an oil company under which its vehicles would always use that company's service station) to full integration (a company owning and operating its own fuel depots). It is in general, impossible to define the boundaries of the organisation precisely -- for example, a McDonalds outlet may be regarded for some purposes as part of a large multinational firm and for other purposes as an independent enterprise contracting with that firm.
A natural approach to the question of contracting out is, therefore, to begin by examining those issues that are common to all types of organisations and then to consider whether government organisations have special characteristics that imply the desirability of a greater or smaller degree of contracting out than would generally be the case. The most appropriate framework for addressing general issues is that of transactions costs. Ceteris paribus, the boundaries of the organisation should be drawn at the point where transactions costs are minimised. As an alternative way of expressing the same point, boundaries should be drawn so as to maximise economies of scope.
2. Contractual relationships
Attention in this section will focus on the choice between undertaking a given activity within an organisation and contracting with an outside enterprise. It may be worth briefly considering the third option of direct market transactions. In essence, this is most appropriate when the transaction consists of the purchase of an easily definable good or service and where both the buyer and the seller are small in relation to the relevant market. The more complex the activity, the longer the duration of the relationship and the more important the buyer and seller are to each other the greater the advantages of contractual relationships.
In choosing between integration and contracting out, it is natural to focus on the insights of principal-agent theory. The central point of principal-agent theory is that if returns are uncertain and one party has private information or some other strategic advantage, it will generally be preferable for that party (the agent) to bear the risk associated with the venture. If the risk is assumed by another party (the principal) the agent will have an incentive to shirk or to divert some of the assets to private uses, then to claim that the bad outcomes of the venture were simply the result of bad luck.
The basic implication of the principal-agent literature is that, where possible, the party that has most control over risk should be the owner, that is, the recipient of the residual income. The basic conditions leading to an ideal case for contracting out are
(i) The net return from the activity is risky
(ii) The risk is most easily monitored by those undertaking the activity
(iii) The output of the activity is easily observable and contractible (that is, performance can be legally enforced)
A breakdown in any of these conditions, weakens the case for contracting out. Activities which do not satisfy condition (i) are so routine that they may be undertaken on a wages basis. Hence, assuming competitive labour markets, there is no gain from contracting out. However, contracting out may be used strategically in non-competitive labour markets (see ..)
Failures of condition (ii) generally arise when the riskiness of returns from the activity in question is related primarily to decisions made within the organisation. For example, consider the manufacture of a component for a motor vehicle. If the component manufacturing process is risky, contracting out will probably be optimal. If however, the primary source of risk relates to the production decisions of the vehicle firm, integration will normally be optimal.
In the case of public sector contracting out, the analogous problem is that of regulatory risk. If it is necessary to subject the contractor to close regulation, it is likely that the resulting arrangement will be more costly than if the activity were directly undertaken by the relevant government agency.
A failure of condition (iii) undermines the very basis of contract, since there is no way of determining whether the contract has been fulfilled. Activities of this kind (assuming they are undertaken at all) represent the core operations of the enterprise concerned.
This analysis explains, for example, why government agencies have generally done very badly at operating restaurants, and why chain restaurants are normally run on a franchise basis. The return from restaurants is inherently risky, and it is difficult for an outside owner to check on the quality of service or the correctness of management decisions. Hence it is hard to stop a hired manager from shirking. The only effective disciplinary device is to make the manager bear the risk associated with the restaurant's operation. Public enterprises, such as the Civil Aviation Authority, which need to provide food and drinks for sale as an incidental part of their operations should therefore, in most cases, lease their restaurant facilities to private firms. In these cases, the lessor needs only to monitor the payment of rent. The costs associated with poor service will be reflected in reduced returns to the lessee.
The issue becomes more complex where catering services are provided, but where there is no market test of the output produced (for example, in a hospital). Here it is necessary for the agency concerned to monitor the quality of service. If this is very difficult (in particular, more difficult than directly monitoring labour inputs) integration may be superior to contracting out.
In large-scale activities, the cost of capital (or, more precisely the cost of risk-bearing) is likely to be critical in determining comparative efficiency. In general, the larger and more diversified the organisation, the lower the cost of risk-bearing.
A standard distinction in the literature differentiates idiosyncratic risk from systematic risk. Idiosyncratic risk concerns factors specific to a given project. Private mechanisms such as insurance and portfolio diversification and diversification of public sector risk through the taxation systems and the size and diversity of the public sector mean that idiosyncratic risk, qua risk, is unimportant. No risk premium should be charged on the basis of pure idiosyncratic risk. However, the existence of idiosyncratic risk is the major source of the agency problems discussed in the previous section.
Systematic risk refers to the risk associated with fluctuations in aggregate output. Market mechanisms do not work well in diversifying this risk. Calculations based on standard models of life cycle optimisation suggest that the equilibrium premium associated with systematic risk should be between 0.2 and 0.5 per cent. In fact, long data series generally show that the rate of return to buying and holding the market portfolio of stocks is considerably greater than the rate of return to government bonds. For example, Mehra and Prescott (1985) present data showing that over the period 1889-1978, the average annual yield on the Standard and Poor 500 Index was seven per cent, while the average yield on short-term debt was less than one per cent. The 'equity premium puzzle' associated with the divergence between the observed rate and that suggested by theory suggest that capital markets spread risk very imperfectly. The calculations presented by Mehra and Prescott and other writers in the equity premium literature suggest that if capital markets spread risk perfectly, the premium between the expected return on equity and that on riskless debt should be very small. The puzzle is precisely the fact that the premium is so large.
Attempts have been made to explain the equity premium by invoking alternative preference structures or risk distributions. None, however, has been generally accepted as successful. A critical problem lies in the assumption of perfect risk-spreading. A perfect capital market would fully diversify the risk associated with recessions, so that, for example, firms could insure themselves against going bankrupt in a recession. In fact, because of the potential for agency problems, such insurance is unobtainable. Assuming that at least part of the equity premium reflects imperfect risk-spreading, it would appear from the empirical evidence that the public sector mechanism of spreading risk through the tax system has significant advantages.
The superiority of the public sector in handling risk implies that the capital costs of an infrastructure project will, other things equal, be much lower under public ownership than under private ownership. Prima facie, since the real rate of return on equity is about double the real government bond rate, the capital cost of an enterprise financed purely by private equity will be about double that of a publicly financed enterprise. This differential will be reduced to the extent that the private enterprise is financed by debt, and should also be adjusted to take account of the (small) pure risk premium that would be generated by a perfect financial market.
A more precise measure of the premium associated with private financing may be derived by considering the beta coefficient from the Capital Asset Pricing Model). If the mean profits of the enterprise are denoted by m, the standard deviation by s and the correlation between profits and the aggregate income of the economy by r, we define b = rs/m. Under the Capital Asset Pricing Model, the rate of return required by equity investors in an enterprise is a linear function of b. Observe that the degree of gearing now makes no difference to the total capital costs of the enterprise. The higher the level of debt, the higher the premium demanded by equity investors, but the lower the amount of outstanding equity.
4. Gains from contracting out - efficiency improvements versus wage reductions
A large number of studies have found that the budgetary cost of providing labour-intensive services is reduced by contracting out. Domberger, Meadowcroft and Thompson (1986) and Cubbin, Domberger and Meadowcroft (1987, 1988) (hereafter Domberger et al) examine two such cases of contracting out and estimate average budgetary gains of 20 per cent. This figure has been very widely quoted, and appears to have some support from other studies. Although, there are may be some grounds for questioning the magnitude of the estimates, there appears to be a general agreement that on average, cost reductions are achieved.
Given the existence of these cost reductions, it is important to determine their source. If they arise from increases in the efficiency with which tasks are performed, there is presumably a net social benefit from contracting out. On the other hand if cost reductions arise from reductions in wages or from other transfers, the issues are much more complex. Because of its influential role in the literature and because of the generally high quality of the data and of the analysis, attention will be confined to the work of Domberger et al, who examined the contracting out of garbage collection and hospital cleaning services in Britain. In-house teams as well as outside firms were permitted to compete for contracts. Both when in-house teams were successful and when the contract went to outside firms total cost reductions were estimated to be of the order of 20 per cent.
This was lower than initially estimated gains. In the first round of contracting out of hospital cleaning services, average savings of 40 to 50 per cent were achieved, but many contractors either failed to provide adequate service or lost money doing so. To counter the potential for failure, contractors were in some cases required to post performance bonds. Because of this and the losses on early contracts, savings on later contracts were smaller. If this adjustment process was not completed at the time the Domberger et al studies were undertaken, the total gain may have been overestimated. However, this issue will not be pursued here.
Milne and McGee (1992) re-examine the question of contracting out of hospital cleaning and catering services, using data for 1985-86. They obtain higher estimates of cost savings, primarily because they treat changes in the prevailing level of wages as a source of benefits. Milne and McGee emphasise the role of wage reductions (and also losses incurred by over-optimistic contractors) as a source of cost savings for local authorities. They observe that in-house NHS staff partially or wholly forfeited bonuses (which could amount to one-third of total pay) and that similar reductions relative to pre-existing wage levels were imposed on contract staff.
Domberger et al attempted to address the question of whether cost savings represented transfers using Farrell's (1957) notion of technical efficiency. Using linear programming to estimate a frontier cost function, the gains were partitioned between those attributable to differences in technical efficiency and those attributable to other sources including reductions in wages and fringe benefits.
Domberger et al observe that the gains not accounted for by differences in technical efficiency may arise from a number of sources other than wage reductions - most notably increases in the efficiency of the input mix. This issue deserves a little further attention. Measures of technical efficiency involving the fitting of a frontier typically have the property that extreme observations will lie on the frontier unless they are strictly dominated. As in example, suppose that there are two inputs, labour and capital, a given firm is the most capital-intensive in the entire sample. As long as there is no other firm which achieves higher labour productivity with lower capital input (implying a negative marginal product of capital) this firm will lie on the frontier, even if the implied marginal product of capital is very low. By contrast, if there are a large number of firms with fairly similar capital-labour ratios, the great majority must lie inside the frontier. It follows that frontier techniques will tend to identify efficiency gains whenever firms adopt input mixes that differ from those of most existing firms.
A more critical problem is that the measured technical efficiency gains could include a substantial transfer component. The first possible source of transfer is increased intensity of effort. Indeed, the main source of efficiency gains explicitly noted by Domberger et al is the replacement of fixed 'task and finish' payments (sometimes for tasks which have become less onerous since the time rates were set) with piecework rates. Productivity gains from such changes in payment schedules will arise primarily from increased effort. Furthermore, the observation that tasks have become less onerous since rates were initially set implies that the new piecework rates will embody an effective reduction in wages in the absence of increased effort. Ganley and Grahl (1988) cite a number of cases of increases in working hours or reductions in working conditions associated with contracting out of garbage collection.
There is a tendency, criticised in Quiggin (1992), to regard output gains arising from increased effort as a free good. Such a view has no basis in neoclassical economic theory, although it may be supported by X-efficiency arguments (Liebenstein 1966). Such arguments require the existence of unexploited gains from trade, arising from distortions in the bargaining process (in addition to any element of union monopoly power). The implied claim is that the wage bargain embodies excess on-the-job leisure, in the sense that both workers and employers would prefer a bargain with higher wages and higher work intensity. There is no clear reason why bargains should be systematically biased in this way.
The argument that apparent increases in technical efficiency are more likely to reflect increases in work intensity gains strength from consideration of the tasks contracted out. Cleaning is a job with very limited capacity for either technical or organisational innovation. Cleaners typically work alone or in pairs, using equipment which has not changed substantially for decades. Furthermore, cleaners are normally paid for completion of a specified task, with no direct monitoring of time spent on the task, so that there is little scope for unions to impose restrictive work practices. The claim that there exist unexploited changes in cleaning methods permitting a twenty per cent increase in output with no increase in effort or reduction in service quality seems inherently implausible.
In the case of garbage collection, there is more scope for technical innovation (for example through the use of large 'wheelie' bins). However, such changes would probably have been captured explicitly in the detailed data available to Domberger et al. The simplest source of increased labour productivity in tasks of this kind is speed-up -- in this case, requiring collectors to run faster. Up to a certain point, speed-up yields efficiency gains as well as transfers, since the truck and driver are used more efficiently. However, speed-up can easily be pushed to the point where the risk of acute injury or repetitive strain injury increases at a rate that outweighs any benefits of more effective capital utilisation. Contracting out may well create incentives for employers to disregard these costs, either because of moral hazard associated with workers compensation insurance or because of the availability of workers who are unaware of the risks of injury.
A second source of gains which is likely to appear in the measured improvement in technical efficiency is that arising from tax evasion. The opportunities for evasion and avoidance are increased by contracting out. Public sector wage employees have less opportunities for evasion than any other group of income-earners. By contrast, contractors and their employees are in a very good position to evade taxes, especially if, like cleaners and garbage collectors, they work non-standard hours. The evidence reported in Tanzi (1982) indicates that evasion is insignificant among government employees and highest in the small business sector. This is, of course, what would be expected given the incentives and opportunities faced by different groups.
The existence of employees who are concealed from the income tax system is unlikely to be reported to local authorities, especially if workers compensation insurance as well as income tax is being evaded. The output of these employees will be measured as a productivity gain. In this context, it is noteworthy that the in-house teams studied by Domberger et al showed much lower gains in technical efficiency and much larger gains arising from wage reductions. In-house teams would, at least initially, consist of workers already known to the authorities. Hence, it is likely that they would find it difficult to achieve cost reductions through tax evasion and would, instead, find it necessary to accept explicit cuts in wages and conditions.
In the Australian context, the relatively stringent controls on tax evasion introduced since the mid-eighties suggest that the gains from tax evasion are likely to be smaller than in countries where the informal sector is less tightly controlled. This suggests that gains from contracting out will be smaller than would be estimated on the basis of UK evidence or of Australian evidence from the period prior to the introduction of the Tax File Number and Prescribed Payment Systems.
In summary, it seems likely that at least part of the budgetary gains associated with contracting out services such as cleaning and garbage collection are derived, in the first instance, from reductions in wages or, equivalently, from increased intensity in the work required to earn a given wage. (Some of these losses may be passed on to central governments through tax evasion). It is then, of interest to consider, whether such transfers are socially desirable. In some cases, such as cleaning services, where the scope for increased efficiency is limited, the majority of the budgetary gains may be derived from transfers.
Tax evasion is often regarded as a transfer. However, there are deadweight losses associated with tax evasion, including the resources dissipated in public efforts at tax enforcement and private efforts at concealment of untaxed income. Thus, to the extent that the measured benefits from privatisation are due to tax evasion, it is likely that the true social benefits are negative. Hence it is reasonable to confine attention to the question of whether the wage reductions associated with contracting out are socially desirable.
5. Public Sector Wage Policy
A number of advocates of contracting out have argued that if contracting out generates wage reductions, such reductions must be socially desirable. Domberger, Meadowcroft and Thompson (1988, p 89) state
If it turns out that the best case that can be made for the retention of a public sector monopoly in refuse collection is that it provides a basis for the expropriation of monopoly rents, then this seems to us to be a compelling argument in favour of the introduction of competition rather than the reverse.
From an ethical viewpoint, this statement relies implicitly on a form of marginal productivity ethics. The 'expropriators' in this context are garbage collectors and cleaners. Unless one is committed to marginal productivity ethics, it seems difficult to develop much moral outrage over the possibility that such workers are earning 25 per cent in excess of the market wage for their labour.
From an efficiency viewpoint, it would normally be assumed that the setting of wages in excess of the market rate would produce distortions and welfare losses, although these would be second-order in magnitude. A standard welfare triangle analysis, assuming a unit elasticity of demand for labour would suggest that the removal of a 25 per cent distortion in wages would yield a welfare increase of around 3 per cent. However, the validity of this analysis is far from clear in the context of public sector wage and employment policy.
The public sector is a very large employer, accounting, directly and indirectly, for as much as a third of the workforce in many developed countries. For many categories of workers, such as teachers and health workers, the public sector is the principal employer. As a result, the public sector enjoys considerable monopsony power. Obviously a policy of exploiting that monopsony power to maximise the consumer/taxpayer surplus associated with the provision of public services would be unlikely to promote social welfare. Such a policy would imply the adoption of excessively capital-intensive techniques and the provision of suboptimal service levels, with a resulting suboptimal level of employment.
Since contracting out has been shown to reduce wages and conditions below those that typically result from bargaining between unions and public employers, the possibility of monopsony power is relevant to any assessment of the desirability of contracting out. In itself, a commitment to contracting out need not imply any exploitation of monopsony power. Provided the units of work that are contracted out are small in relation to the relevant labour market, and that each contracting decision is undertaken without consideration of the effect on other contracts, the public sector will behave in a competitive fashion. However, there is no guarantee that these conditions, and in particular the second of them, will be fulfilled. Governments have very frequently used their own wage bargains as an instrument of policy, and public sector employers making 'excessive' wage settlements have been subject to considerable pressure. In the British case considered here, the imposition of contracting out coincided with a series of measures including rate-capping, abolition of local authorities and direct intervention in local government decisions by a central government overtly hostile to public sector employees. These actions were defended as an attack on the monopoly power of public sector unions. However, in a bilateral monopoly situation it is difficult to tell where monopoly ends and monopsony begins. There is, therefore, every possibility that the effect of contracting out is to exploit monopsony power and to drive wages below the socially optimal level.
The existence of high rates of unemployment complicates the issue here. Suppose for simplicity that high unemployment results from wages in the private sector being fixed above the market-clearing level and that this is taken as unchangeable, so that public sector wage policy must be developed in a second-best context. Then the second-best optimum will involve a lower public sector wage, and a higher public sector employment level, than the first-best optimum in which the private sector wage is flexible.
A policy of contracting out aimed at minimising the cost of provision of a predetermined level of public services, or at maximising the consumer/taxpayer surplus associated with the provision of services will imply that both wages and employment are below the first-best optimal levels. In both cases, the existence of monopsony power gives government an incentive to set employment levels below the optimum.
A public sector policy of this kind will tend to move the wage share of GDP and the average real wage closer to the equilibrium level, but will nevertheless result in higher levels of unemployment, and greater welfare losses, than a policy that fixes wages and employment levels at the first-best optimum. This is a situation in which the aggregate-level reasoning characteristic of much discussion of unemployment may be seriously misleading.
The situation is even worse if unemployment is generated by restrictive macroeconomic policies. In this case, the successful exploitation of monopsony power depends on the maintenance of high levels of unemployment. A relaxation of macroeconomic policy will generate a public sector wage breakout which is likely in turn to provoke a new tightening of fiscal policy. The governments role as an employer creates a bias in favour of excessively high rates of unemployment.
6. Contracting out of infrastructure projects
We may usefully distinguish three aspects of a typical infrastructure project in which there is a choice between public and private sector approaches. These are
The relative importance of the different aspects varies greatly, although the term 'infrastructure' normally implies a large capital expenditure and hence a substantial role for construction. For example, the 'operation' of road infrastructure involves little more than routine maintenance. On the other hand, in areas such as airports, operational
The most common practice in Australia has been for construction to be undertaken by private sector firms under a system of competitive tendering. Core operations of publicly owned infrastructure have mostly been public although peripheral operations (hospital cleaning, airport gift shops) have been contracted out. There are cases of nominally privately-owned but publicly operated infrastructure projects (for example, leaseback arrangements on power stations), but these almost always arise out of attempts to evade Federally imposed borrowing limits rather than from considerations of efficiency.
In the past, it was common for public infrastructure projects to be constructed by government departments on a day-labour basis. In general, this has proved less satisfactory than the alternative of competitive tendering. In essence, this is because it is relatively easy to ensure that the private constructor bears most of the risk associated with the project, and therefore has strong incentives to overcome the moral hazard problem. By contrast, the incentives for individuals within a government department to minimise costs are relatively weak and diffuse.
The main difficulties are associated with the possibility of the private constructor going bankrupt. In this event, the public will bear both costs of delay in the project and increased expense. The risks may be mitigated if the constructor is required to post some form of performance bond, giving the party providing the bond an incentive to monitor the firm's financial status.
It is also necessary that there should be a capacity to verify that the project has been completed according to the specified standard. These monitoring difficulties also arise in the case of public sector construction but are mitigated by the absence of strong incentives to cut costs. In most infrastructure projects, this does not appear to be a major difficulty, but there are exceptions such as the construction of nuclear power stations.
It is useful, although difficult in practice, to draw a distinction between peripheral and core operations. In essence, the notion of core operations refers to those activities that must be undertaken by the owner of the project either because they are very difficult to monitor or because they involve very large risk of loss to the owner. Thus, for example, the operation of generating equipment would probably be a core activity for an electricity enterprise, whereas cleaning would be a peripheral activity. By definition, private sector operation of core activities is in appropriate unless the project is privately
There is a widespread consensus that contracting out of peripheral operations yields significant savings. The figure of 20 per cent, based on the Domberger et al studies of the UK is commonly quoted. However, two main objections may be made to this estimate. First, Paddon (1991) has argued that savings have declined as the process of contracting out has proceeded, with recent contracts yielding savings around 6 per cent. Second, many of the apparent gains may in fact represent transfers arising either from reductions in wages and working conditions (Ganley and Grahl 1988) or from the increased potential for tax evasion associated with many forms of contract employment relative to public sector employment (Quiggin 1994).
The owner of a project may most usefully be described as the party bearing the residual risk. There are strong grounds for public sector ownership of most infrastructure projects. As has already been argued, the superiority of the public sector in bearing pure risk is illustrated by the existence of the large 'equity premium' between the government bond rate and the rate of return demanded by private equity holders. In essence, every dollar of private equity in a project doubles the eventual cost relative to the public debt finance it displaces.
In many cases, moreover, agency considerations also favor public ownership. The risk associated with many infrastructure projects depends far more on public policy decisions than on the management skill of the operator. Consider for example the position of a prospective purchaser of Kingsford-Smith airport or a prospective operator of the proposed Badgery's Creek airport. The profitability of these enterprises will depend far more on government policy decisions with respect to aircraft noise, international aviation decisions, transport links and the like than on the skill with which the airport is managed.
A similar argument applies to road projects. Construction firms have no special knowledge concerning traffic flows and no special expertise in financial management. Governments are much better placed to handle both of these sources of risk. They have large and diversified portfolios and, as a result, can borrow at low rates of interest. Governments are also much better placed to handle risks associated with traffic flows. Not only do governments have access to large quantities of information on traffic patterns, but they are continuously engaged in making decisions that will affect future traffic flows. Unless they are contractually bound, governments have no incentive, in making these decisions to take account of the effects on the profitability of private road operators. Obviously, private investors would be unwilling to give hostages to fortune by making investments under such conditions for rates of return similar to those obtainable from government bonds or even from ordinary equity investments.
It follows that the private operator must either demand a large risk premium in addition to the usual equity premium or must demand ironclad guarantees of favorable treatment. In practice both avenues have been pursued. Where there are competing firms of course, the difficulties are multiplied, since favorable treatment for one necessarily damages the others.
In practice, governments engaged in BOOT projects have typically ended up with the worst of both worlds. In order to secure private participation they have had to give implicit guarantees both of a favorable future policy stance and of a commitment to rescue the private operator if things go too badly wrong. Thus taxpayers have ended up bearing most of the risk. On the other hand, the residual riskiness of the project has meant that investors have demanded a rate of return far in excess of the governments's normal cost of borrowing.
Finally, if the service provided is essential, the government must, in effect guarantee the viability of the firm providing it. The standard processes of bankruptcy are too slow and too liable to lead to loss of service provision in cases of this kind. But the owners and managers of the firm can only be made to forgo the benefits legally available to them under bankruptcy provision if the state offers more favorable treatment (for example, a buyout at prices above the 'fire-sale' value that would be realised in bankruptcy). Hence the maintenance of essential service implies some level of public financial guarantee.
In fact, it appears that in many allegedly private infrastructure projects, the public sector bears all or a large part of, the residual risk. The difficulty of determining where the risk lies has been illustrated by the NSW Auditor-General's attempts to qualify the public accounts in respect of the Sydney Harbour Tunnel. While effective public ownership is perhaps inevitable in the light of the economic considerations discussed above, it is likely that cases of this kind involve a public assumption of the risks of ownership combined with private capture of the benefits.
D BOOT projects
A number of recent private sector infrastructure projects have operated on the basis referred to by the acronym BOOT (Build, Own, Operate, Transfer). On the basis of the analysis set out above, all such projects should be viewed with grave suspicion. Two main objections may be made.
First, in the absence of strong economies of scope between construction on the one hand and operation and ownership on the other, there is no reason to tie the construction of the project to the subsequent ownership and operation. Assuming that private sector control was desirable in all three phases, but that the project was a matter of public concern, it would be preferable for the government to put the construction out to tender, then separately solicit bids for the rights to own and operate the project. In this way, it could be clearly established either that there is no public sector subvention or that any such subvention was transparent and available to all firms involved in tendering for the project.
The second objection relates to the transfer phase. If private sector ownership and operation is indeed superior, there is no rationale for eventual transfer to the public. Under these assumptions, the project could be financed at lower cost (in terms of the present value of the charges levied on users) if
These theoretical objections have been borne out by experience with BOOT contracts. They have been almost uniformly written in a manner unfavorable to the public and involving far greater expense than the standard approach of publicly financed construction under conditions of competitive tendering. Because of the secrecy associated with most such projects, it is likely that many of them contain contingent obligations on the public sector that have not yet been revealed.
In summary, BOOT projects appear to offer 'something-for-nothing'. As with most 'something-for-nothing' deals, BOOT projects will prove very profitable for their promoters and very costly for the buyers, in this case, the Australian public.
E Choosing between private and public ownership
The analysis presented above suggests intuitions about the types of activities that are likely to favour private involvement. To determine the optimal approach in specific cases, it is necessary to apply the tools of cost-benefit analysis. In essence the critical test is whether the present value of the flow of benefits associated with the project will be greater with or without private involvement at a given stage.
Consider for example a road construction project. Such a project might be undertaken in several different ways
(i) Construction by public day labour, financed by a general petrol tax
(ii) Construction by competitive tender, financed by a general petrol tax
(iii) Construction by competitive tender, financed by a toll
(iv) Construction by a private owner with the right to levy a toll
Any two of these methods could be compared using the standard tools of benefit-cost analysis. The critical issue is the choice of discount rate. On the basis of the arguments presented above, the appropriate discount rate is the real government bond rate, with a small adjustment for systematic risk (say 1 per cent).
F The need for transparency
One of the major advances in public policy over the seventies and eighties was the move to transparent decision-making. Time-honored rationales for secrecy were tested and found to be little more than self-serving rationalisations generated by bureaucrats and politicians to protect their actions from public scrutiny. Openness in government is essential both to promote improved decision-making and to avoid corruption.
Similar moves are desirable in the private sector, and more particularly in the corporate sector, where the relationship between managers and shareholders is analogous to the relationship between politicians and citizens. The case that the public interest is served through commercial secrecy is no more soundly based than the analogous arguments in the public sector. Although it is natural that firms will try to keep secrets, there is no general advantage to be gained from secrecy. While any one firm would be disadvantaged if its private information were revealed, there would be a general gain from increased transparency.
One of the most pernicious aspects of recent private infrastructure projects has been the use of claims of 'commercial confidentiality' to remove the decision-making process from public debate. This has undoubtedly led to bad decisions that might have been avoided by a more open process. Although no cases of criminal corruption have yet been proven, corruption is an inevitable consequence of secret decision-making processes involving large commitments of public money. It should be a condition of private involvement in public infrastructure projects that any claims to commercial confidentiality relating to the terms of agreements between the private and public partners.
7. The self-financing rationale
One important rationale for private infrastructure projects is that, since such projects will only be undertaken if they are profitable, private sector involvement is a guarantee against the occasional tendency of transport ministers and other politicians to wasteful monument-building. However, this rationale is only applicable if the services generated by the infrastructure project are sold in competitive markets, and are not characterised by significant externalities. If these conditions are not met, then the capacity to generate profits is neither a necessary nor a sufficient condition for a project to be socially desirable.
A Toll-financed road projects
One major class of externalities is network externalities. Roads, railways and harbours are part of a larger transport network, and must be evaluated in that light. The benefits of a new freeway are experienced not only by drivers on the freeway itself but also by those on the alternative routes from which traffic is diverted. These drivers pay nothing, but enjoy less congested and less dangerous roads.
On the other hand, there are external costs that are not included in the usual estimates of construction costs for roads and airports. The cost of land for roads is rarely counted as part of construction costs, although, on some estimates, as much as 30 per cent of urban land is taken up by roads. Also, costs such as noise and air pollution are not taken into account properly. On the whole, increased road construction tends to raise noise and pollution, but some projects, by relieving congestion, may reduce pollution. To sum up, the ability of a road project to generate sufficient toll revenue to finance its construction is neither necessary nor sufficient to show that the project is a good one.
In this case, the pricing mechanism itself may substantially reduce the benefits of the project. Some bridge tolls may be justified as a congestion tax, but tolls on new roads generally have the effect of forcing more traffic onto old ones, leading to worse congestion and more accidents. From an efficiency viewpoint, if tolls must be imposed, it would be better to impose them an old roads and leave new ones tax-free. But, in the absence of a general system of road pricing, isolated tolls are almost always inferior to a general system of user charging through fuel taxes and registration fees.
As well as being inefficient, toll funding is usually inequitable. Some new roads are subject to tolls, others are not, and there is no clear rationale. The arbitrary nature of toll funding tends to compound old injustices. For example, political pork-barrelling led to the inland New England Highway getting National Highway status at the expense of the more heavily-used Pacific Highway. Instead of rectifying this injustice, it is now suggested that the unfortunate users of the coast road should be forced to pay a toll to finance improvements to the Pacific Highway, while still contributing through their taxes to the maintenance and upgrading of the inland route.
In summary, unless their is a significant element of monopoly power, toll-financing will reduce the net benefits of road projects and will capture only part of those net benefits. This implies that only in cases where the benefits of new toll roads massively exceed the costs will toll finance provide sufficient revenue to finance their construction, raising the question of why such projects have not already been constructed. The predictions of economic theory in this respect have been very clearly borne out. Toll-financed road projects have generally involved either a significant public subsidy is provided (NSW M5 motorway, proposed airport rail link) or a provision by which the private contractor is allowed to levy tolls on existing roads constructed with public funds (Sydney Harbour Tunnel, Tullamarine freeway).
In summary, toll financing, as well as being both inequitable and inefficient, is unlikely to provide sufficient revenue to fund road projects. Once the special assistance required to make such projects viable is included in the package, it is almost impossible to assess whether the projects are cost-justified.
B Natural Monopolies
In many cases, notably including airports and water supply, the provision of infrastructure is a natural monopoly services generated by the infrastructure project are sold in noncompetitive markets. This raises the possibility that monopoly power will be exploited. Perhaps more importantly, projects may be undertaken as a means of securing and extending monopoly power, rather than because they are socially beneficial.
In general if a project generates sufficient returns to cover its costs it will be socially beneficial even if there is a costly element of monopoly pricing. However, project selection may be biased by the desire to maximise monopoly rents or to forestall competitive entry. The recent debate over pay television policy reveals many possibilities of this kind.
8. Regulation and contracting out
In cases where the market for the service provided by an infrastructure project has significant monopoly elements, it is necessary to design some scheme of regulation. The most common response, following the UK approach, has been CPI - X regulation. That is, some index of the firm's prices is required to rise by no more than the change in the CPI less some 'efficiency dividend' X. The CPI-X approach was initially advocated as a transition measure in cases (eg telecommunications) where it was assumed that a fully competitive market would eventually emerge.
The CPI-X approach has merits in its transitional role. In the long term, however, the approach is untenable and must be replaced by some form of direct price surveillance or rate-of-return regulation. The main difficulties with this approach are
(i) There is an incentive for reduced service quality. Direct regulation of service quality implies much greater government involvement in management.
(ii) X is normally set for a fixed period. The firm has an incentive to practise cost-padding in order to negotiate a smaller X in subsequent periods
(iii) There is potential for cross-subsidisation with lower prices being offered to groups allied with management.
An alternative approach, rate-of-return regulation has been favored for monopoly utilities in the US. In rate-of-return regulation, the firm's costs are monitored and prices set in such a way as to yield a commercial rate of return on the capital employed. The rate-of-return approach has also been criticised for encouraging cost-padding and excessively capital-intensive methods of production.
Benchmarking may be used as a supplement to either rate-of-return or direct price regulation. The idea of benchmarking is to find comparable enterprises in other jurisdictions and to set prices (or allowable rates of price movement) on the basis of the performance of the benchmark enterprises.
For both public and private organisations, the critical issues in determining the optimal degree of contracting out relate to the proper allocation of risk. As a general principal, governments should bear risk related to changes in regulatory policy and risks related to capital-intensive project, while risk specific to the operation of particular activities should be contracted out as far as possible.
Much of the measured gain from contracting out has arisen from reducitons in wages and increases in work intensity. The net social gain from contracting out in such cases is much smaller than the measured gains. In particular, if governments require decentralised bargaining while effectively imposing a common stance on employers, contracting out may reduce wages and employment below the socially optimal level.
Coase, R. (1937), 'The nature of the firm,' Economica 4, 368-405.
Cubbin, J., Domberger, S. and Meadowcroft, S. (1987), 'Competitive tendering and efficiency: The case of hospital cleaning', Fiscal Studies 8(3), 49-58.
Domberger, S., Meadowcroft, S. and Thompson, D. (1986), 'Competitive tendering and efficiency: The case of refuse collection', Fiscal Studies 7(4), 69-87.
Domberger, S., Meadowcroft, S. and Thompson, D. (1988), 'Competitive tendering and efficiency in refuse collection: A reply', Fiscal Studies 9(1), 80-5.
Farrell, M. (1957), 'The measurement of productive efficiency', Journal of the Royal Statistical Society (120).
Ganley, J. and Grahl, J. (1988), 'Competitive tendering and efficiency in refuse collection: A critical comment', Fiscal Studies 9(1), 86-9.
Liebenstein, H. (1966), 'Allocative efficiency vs X-efficiency', American Economic Review 56, 392-45.
Mayer, C. and Meadowcraft, S. (1986), 'Selling public assets: Technical and financial implications', in Kay, J., Mayer, C. and Thompson, D. (ed.), Privatisation and Regulation: The UK Experience, Clarendon Press, Oxford.
Milne, R. and McGee, M. (1992), 'Compulsory competitive tendering in the NHS: a new look at some old estimates', Fiscal Studies 13(3), 96-111.
Quiggin, J. (1992), 'Free lunches in the case for privatisation', Economic Analysis and Policy 22(1), 1-18.
Quiggin, J., (1994), Does privatisation pay ?, Australia Institute Working Paper.
Tanzi, V. (ed.) (1982), The underground economy in the United States and Abroad, Blackwell, Oxford.
Williamson, O. E. (1975), Markets and Hierarchies, The Free Press, New York.
Go back to John Quiggin's home page