Paper Presentations Session

INCREASING GOVERNMENT/PUBLIC SECTOR EFFICIENCY 

THROUGH PUBLIC-PRIVATE PARTNERSHIP

by

Mr. Abdul Rahman Turay

Chairman

National Commission for Privatisation of Sierra Leone

Introduction

Let us consider a two-sector model of the economy, government and private. In this scenario; economic theory stipulates that output in both sectors should be optimal in equilibrium.   Economists such as Wilfred Pareto, Cournot and Nash established tests for meeting this condition. In real life economies are never in equilibrium but the aim of economic management is to adopt measures, which will make them approach this position as best as possible. The disequilibria observed in most economies have been attributed largely to inefficiencies in the public sector. It is for this reason most governments as well as the Breton Woods institutions have tended to search for economic management tools that could assist in reducing the inefficiencies in the public sector in order to achieve an overall improvement in the performance of economies.

Activities in the public sector can be classified into two groups namely: industrial and non-industrial. The chief distinction being that industrial activities produce goods and services, which are traditionally, sold in markets as distinct from the output of the non-industrial category for which there is usually no tradable market.  Public sector industrial   activities are carried out by publicly owned enterprises, which before the   privatisation policies adopted by many industrial countries in 1980s onwards, accounted for significant proportions of gross domestic product, employment and gross capital formation. The different efficiencies of the public and private sectors seems generally to be attributed solely to the category of ownership

We shall now explore the analytical and empirical foundations for the doctrine of superior private sector efficiency relative to that of the public sector and attempt to establish whether this distinction is determined solely because of type of ownership.

Private Ownership

In a regime of atomistic competition economic theory stipulates that owner managers of firms tend to seek to maximise profits (present and future financial flows) and would use factor combinations at least cost so that the returns to each of the factors of production would correspondingly equal to their marginal productivity.  Additionally   the optimal level of production for each firm’s output is reached when the marginal cost is equal to marginal revenue and equal to the price of the product.  When these conditions are met then the firms’ operations are said to be internally efficient.  But some of the assumptions in this model are not usually observed in the real world. When we relax the assumption of owner manager and take into account the fact there are many shareholders of a firm’s stock and a distinct group of executives exists who manage the firms’ operations, we immediately run into the problem of the relationship between principal and agent which arises from the asymmetric distribution of information about market conditions in which firms operate. The principals who are the shareholders tend to have less information about the operations of their firms than the management who operate the companies. Therefore it is necessary for owners to design incentive schemes that can induce   managers to internalise the objective functions of the owners, namely to seek to maximise   profits of the firm as would be the case if the owners were operating the firms. However the objective functions of management do not contain profit as the sole independent variable.  Managerial utility functions have been postulated in the literature that include such variables as sales revenue growth and the level of discretionary managerial expenditure (Robin Marris Managerial Capitalism 1964 and F M Scherer 1980 Industrial Market Structure and Economic Performance). These are powerful objectives that tend to influence the degree of power and prestige, which can accrue to management.  It follows from this that the divergence of the objectives of the owners of the firm and those of management will inevitably lead to operational inefficiencies because of the possible violation of the conditions of optimal output and internal efficiency.

However the behaviour of management can be influenced in a variety of other ways. Shareholders may set a range of dividend payments to be expected each year from their investments and can then enter into contractual agreements with management requiring them to make specified minimum annual dividend payments. When shareholders dividend payment expectations are not fulfilled significantly they may, in the case of large institutional shareholders effect changes to management through changes in supervisory boards or as in the case of other share holdings that are diffused, they may just sell their shares. The mass sale of shares leads to a fall in share price and a reduction in the market valuation of the enterprise. This then makes the firm an attractive target for a takeover.   The impact of the threat of a take over on management behaviour however depends, inter alia, on the precise characteristics of the relevant capital market, including factors such as the extent of shareholder protection afforded by the regulatory and legal environment, the constraints imposed by international law and the effects of the extant fiscal regime.  It is noteworthy that the effectiveness of a takeover threat is to a large extent determined by the strength of the link between managerial effort and the probability of a takeover. Raiders often have better information relating to the performance of incumbent management than some shareholders, especially in the case where share ownership is diffused. The threat of possible loss of management control of the firm through a takeover is very often an incentive for the promotion of internal efficiency in a firm. The conclusion to be derived from this analytical approach is that takeovers generate an incentive for management.  But theoretical analysis and empirical evidence have yet to yield unambiguous conclusions about the strength and effectiveness of this managerial incentive.  Some studies have indicated the existence of market imperfections that may mar the effectiveness of capital market disciplines on internal efficiency. Ajit Singh (1971 and 1975) study of takeovers in the United Kingdom found only small differences in profitability and other measures of financial performance between companies that became take over victims and those that did not. This result did not support the thesis that poor performance tends to lead to an increase in take over threat. Singh also found that above a certain size the likelihood of a take over threat diminished sharply. The implication is that the most effective defence against a take over threat is to increase growth by acquisitions, a strategy that satisfies management’s utility function.

Another factor that may influence private enterprise management behaviour is the threat of bankruptcy by the firm’s creditors if they believe that the market value of a firm’s assets is less than its liabilities. There are two observations to be made about this hypothesis.  Firstly, if managers believe the probability of a firm’s survival is small regardless of any remedial actions they may take, they are likely to decide to enjoy more managerial discretion in the short run than take action to promote internal efficiency.  Secondly, because managers determine the level of debt firms carry, they would choose debt equity ratios that are sub-optimal in terms of shareholders’ interest.  This low-level debt strategy eases the bankruptcy constraint thus weakening the incentive for promoting internal efficiency and thereby increasing management’s own utility function.  But when there is intense product market competition and a depressed demand for the firm’s products the threat of bankruptcy can influence managerial behaviour significantly.  

Public Ownership

Managers of public sector enterprises are agents, initially for the supervisory minister responsible for them, the associated civil servants and ultimately the general public who elected the government.  As in the private sector the relationship between agent and principal is dependent on the institutions and structures of the relevant economy. However there are three noteworthy institutional arrangements as outlined below, which distinguish public from private enterprises.   These distinct features are also the reason for the observed relational difference between the management of public enterprises and their agent, the government compared to that of shareholders and management in the private sector.  

v   Government, the agent do not seek to maximise profits

v  There are usually no marketable ordinary shares of public enterprises as there is only one owner, and therefore there is no market for corporate control

v  There is no direct equivalence to the bankruptcy effect on financial performance. Very often lame ducks are allowed to continue with heavy subsidies that are a drain on public sector finances.

Economic analysis suggests that governments demand that management of public enterprises seek to maximise the public interest which, borrowing from welfare economics, is defined as the sum of   producers ‘and consumers’ surpluses. Optimal output conditions would require that the enterprises’ operations are such that no change in output can occur without making someone worse off even with feasible lump sum transfers and that such output is priced at marginal cost.  Governments may wish to attach different weights to the components of producers’ and consumers’ surpluses because of income distribution considerations. The low-income elasticities of demand for the goods and services of some utilities result in low-income households tending to account for a substantial percentage of the sales of these utilities. Government may therefore wish to attach extra weight to consumer interests for income distributional reasons. Similarly a government concerned only with domestic welfare in evaluating the welfare generated by the activities of a privately owned firm may discount profits that may be accrued to overseas shareholders.

There are no potential threats to bankruptcy of public enterprises due to the subsidies provided by government but the bailouts, financed either by taxation or government borrowing, are costly to the economy.  Both policy options may have distortionary effects on the economy and could violate the conditions or Pareto Optimality that should be satisfied to meet the governments overall objective of social welfare maximisation.  Consequently the authorities would have to therefore design a monitoring system for public sector enterprises to ensure that they are, at least, financially viable. The supervisory ministries gather information on the activities of the public enterprise by demanding periodic returns on various quantifiable performance data and may use it to set appropriate incentive structures, including profit related bonuses. Alternatively they may use the data to fire management with poor performance records. On the other hand supervisory ministries may abandon the arms length relationship between the principal and the agent and get directly involved in the decision making process of the management of the public enterprises. 

The hierarchical process of monitoring the activities of the management of public sector   industries has its own built in dysfunctionalisms. It is difficult to reconcile the assumption of maximisation of public interest when the civil servants and politician groups of supervisors display different utility functions that are distinct from that set for management and may include arguments such as monetary rewards, effort levels and power.  Politicians’ objectives also reflect their expectations of direct electoral benefit which may be frustrated by the asymmetrical distribution of information between the supervisory politician themselves and the relevant social groups such as voters, consumers and workers in the affected public enterprises who may be affected by policy changes introduced by the politicians. This asymmetrical information distribution between the politician and the civil servants who actually monitor the performance of public enterprises also leads to ministers internalising the utility functions of the civil servants and become captured by the ethos of the civil service bureaucracy. It is often the case that politicians may rather support a policy of price reductions on sensitive goods and services below marginal cost, if such action can be directly linked to them by the  electorate than pursue fiscal remedies through taxation, which may have similar income distribution effects but may  not be so appreciated by the electorate. Civil servants behaviour in respect of controlling the activities of managers of public enterprises is influenced by the size of the relevant ministerial department and the rents they can earn from their stewardship.  The size of a ministerial department is positively related to the magnitude of its budget, which could serve as an effective tool of control by a minister. But this is a variable that also enhances the stature of ministers among colleagues thus having positive effects on ministerial welfare, hence the capture phenomenon described earlier.  Given the rapid turnover in ministerial departmental responsibilities and the absence of benchmarks against which public sector management performance may be contrasted, there are no compelling reasons for a minister to search for a monitoring system that could provide an incentive for the management of public enterprises to improve their efficiency.

Apart from methodological difficulties relating to the measurement of public interest there seem to be three reasons why sub optimal results may occur given the framework of control employed for monitoring the activities of the management of public enterprises as follows:

v   The substitution of political for social objectives

v  Tendency for direct political involvement in managerial decisions instead of setting up appropriate managerial incentive structures which should form the basis of evaluating management performance

v   Internal inefficiencies in the bureaucracies, which is accompanied by inefficient levels of activity. 

From the above discussions we can safely conclude that, in general, public sector incentive structures are flawed and tend to have imperfections at each level of the supervisory hierarchy. They are therefore ineffective in terms of influencing the behaviour of public sector management.   Public interest theories of political decisions cannot provide an adequate analytical framework for the behaviour of the principals and agents in the public sector.

The   analysis suggests that the relative performance of publicly and privately owned firms in respective of allocative and internal efficiency depends on a range of factors that includes the effectiveness of the monitoring systems, the degree of competition in the market, the extant regulatory policy and the technological progression of the industry. Evaluation of the welfare implications of privatisation will necessarily therefore depend on an empirical assessment of each of these various factors. 

Private and Public sector enterprises comparative efficiency 

Methodological inadequacies make direct comparisons of the relative efficiencies of enterprises in the public and private sector domains difficult. It would be useful and didactic therefore to extend our investigation of these comparisons by including other explanatory variables besides type of ownership. We shall do so by   examining   empirical studies that have used financial analytical tools to make direct comparisons of the relative efficiencies of both types of firms.  Even when data specification has not been a problem, statistical tests have not been sophisticated enough to account of the interacting non-separable effects of ownership, competition, and regulation on incentive structures and consequently on the performance of firms. 

The empirical examination of the relative performances of public and private firms has been extensively carried out in the US where the two types of ownership frequently co-exist in similar market conditions and cover firms in the generation and distribution of electricity, water and refuse collection. In the case of electricity, studies by Meyer 1975; Pescatrice and Trapani 1980; Fare et al 1985 concluded that after adjustments for differences in output mixes and input prices public sector utilities typically have lower unit costs than those privately owned.  However a study by the Edison Electric Institute 1985, stated that in Europe electric utilities have easy access to capital and if this is adjusted for, ownership seemed to have little effect on internal efficiency. Studies by Peltzman1971 and De Alessi 1977 on the US pricing behaviour of public and private electric utilities found that time of day pricing (multi-part pricing) which was more common in private utilities can be expected to lead to higher allocative efficiency   It was established that the jurisdictions in which the quality of regulation is deemed high tend to have multi-part tariffs.  Public utilities in France and Britain were early pioneers of peak load pricing systems.  Studies on US water utilities tend to suggest similar results. Crain and Zardhoohi 1978, found that regulated private water utilities were over-capitalised and had higher labour productivity and that on balance, had lower unit costs compared to those of public utilities.  A later paper by Bruggink 1982, found in favour of public enterprise on unit cost criteria.  The general conclusion to be reached from these results is that where firms face extensive regulation and little product market competition it is difficult to support one type of ownership over another.

The studies by Kitchen 1997, Savas1977 and Steven 1978 found that privately owned firms tendering for refuse collection contracts in the US tended to have greater internal efficiency but that this was significantly attributed to competition.  Savas found that the gap between the unit costs levels of private and public firms was closed by competition induced by the process involved in tendering for contracts.   A study by Caves and Christensen 1980, on the relative performance of the two Canadian railway companies, one private and the other publicly owned, concluded that public ownership is not inherently less efficient than private ownership but that the inefficiency associated with the public sector results from isolation from effective competition.  A study by Boardman and Vining concludes that privately owned firms out- performed publicly owned firms when both operate in an effective competitive environment. Observations of the relative high frequency with which privately owned firms win competitive tenders is consistent with the general presumption, which favours privately owned firms largely because of the efficiency derived from operating in markets which are relatively competitive.

Most empirical work in the UK has concentrated on competitive market structures where the two types of firm ownership co-exist. Pryke 1982 looked at firms operating airlines, ferries, hovercraft and the sale of gas and electrical appliances. Using profitability and output data as proxies for internal efficiency he concluded that in each case private firms tended to be superior to those in the public sector.  Similar results were reached in studies by Forsyth et al 1986 on airlines, Bruce 1986 on ferries. Yarrow and Rowley 1981 found a slight deterioration in the productivity of the British Steel industry following nationalisation coupled with significant declines in market shares and in the rate of diffusion of new steel making processes. A UK Audit Commission study in 1984 on refuse collection costs reported wide variations in public sector performance costs with the better public enterprises showing lower costs than those of private firms .The UK reports seem to support the view that when competition is effective, the available evidence indicates that private enterprise is generally preferred on both grounds of internal efficiency and, subject to the absence of other substantial market failures, on the social welfare criterion.  In competitive markets it is possible to find public sector enterprises that are efficient but those that survive tend to be fewer.

Finally let us now examine the results of two surveys of empirical work on the relevant performance of firms in the public and private sectors.  Borcherding et al examined over fifty studies from five countries and came to the conclusion that most of the studies agreed with the notion that public firms have higher unit cost structures.  But Millward’s 1982, survey of North American literature find no evidence for private enterprise superiority.  Borcherding et al report that out of more than fifty studies surveyed in only nine did private enterprise fail to out perform public enterprise.  Millward studied a further four studies which supported this result.  The studies agree that the presence or absence of competition may be an important determinant of a firm’s performance. This strongly suggests that opening up a market to competition is crucial in promoting improved economic performance.  This is in line with what Kay and Thompson 1986 asserted that in at least some cases liberalisation without ownership change will generate substantial improvements in productive efficiency.  Borcherding et al seem to concur with this view when they conclude that it is not so much the transferability of ownership but the lack of competition that leads to the less efficient production in public firms.

On balance one can sum up the findings of the empirical studies by stating that ownership is a necessary but not sufficient criterion for the determination of the relative      efficiency of firms. Changes in the structure of property rights are likely to have significant repercussions on the behaviour of firms.  Managerial incentive structures result from a complex interaction of factors such as ownwership, the degree of product competition and the effectiveness of the prevailing regulatory regime in each economy.  Serious flaws in the public sector incentive systems lead to weak, conflicting and ineffective control of public enterprises.  This makes for relatively incompetent and inefficient management and poor performance of the institutions public sector enterprise management operate compared to those in the private sector.   

Conclusions

Where product markets are competitive it is more likely that the benefits of private   sector monitoring reflected in improved internal efficiency will exceed any accompanying deterioration in allocative efficiency. This view has been generally borne out by empirical studies of the comparative performance of public and private firms.  In the absence of rigorous product market competition the balance of advantage is less clear-cut and much will then depend on the effectiveness of the regulatory framework.  It can therefore be postulated that joint private-public sector partnerships may enhance the efficiency of the public sector enterprises if the product markets are competitive and there is a relevant regulatory framework that is designed to promote effective competition. Therefore in addition to encouraging joint public-private enterprise partnerships governments should liberalise their economies and maintain regulatory institutions that would ensure barriers to competition are removed. Performance contracts should be adopted as benchmarks for judging the efficiency of the management of the public sector. Where public sector goods and services can be better produced by the private sector, their production should be hived off and offered for tender to both the private and public sector enterprises. This policy should also apply to the output of the non-industrial sector such as medical, education and of other departments of the civil service if some activities can be hived off and put out to tender by the private sector without risk. Franchising can also be adopted as another useful tool for promoting competition