Privatisation: Theory and Evidence

Agreat wave of privatisation has swept the world in the past two decades, embracing the industrial economies, the transition economies of East Europe and large parts of the less developed world, and it continues to roll on. It is interesting, however, that the basis in theory for such a vast movement was by no means unambiguous to start with. Moreover, a sizable enough body of empirical evidence, on which hypotheses about the impact of privatisation could be tested, became available only several years down the road. So much of the initial impetus to privatisation entailed a leap in faith, and, as happens all too often in the development of knowledge, attempts to explain its impact have followed on the heels of widespread practice. 

Contrary to popular supposition, neither the theory nor the empirical evidence on privatisation provides unqualified support for the belief that privatisation leads to outcomes superior to those under public ownership. The theoretical literature, while pointing to the potential benefits of private ownership, also underlines the many conditions required for such benefits to materialise.

As for the empirical evidence, while a large body of literature has emerged that purports to establish that private ownership and privatisation make for superior performance, the evidence is actually quite mixed. Further, the methodology used in many of these studies is widely perceived to suffer from a lack of rigour.

This paper is intended to provide a bird’s eye-view of the large body of theoretical and empirical literature that has emerged on privatisation. The paper is organised as follows. In Section II, we review the theoretical considerations underlying privatisation. Section III surveys the empirical evidence, citing some important studies. Section IV concludes by spelling out possible implications for privatisation policy.

II
Theory

– to promote increased efficiency,
– to raise revenues for the state (and thereby to bridge fiscal deficits),
– to reduce government interference in the economy and promote greater private initiative, and
– to promote wider share ownership and the development of the capital market.
Of these, the first objective, the promotion of efficiency in running commercial organisations, has been the dominant motivation. There is a sense that public ownership somehow leads to lower levels of efficiency than are possible under private ownership. Inefficient enterprises, in turn, are seen as creating other problems such as pre-emption of government revenues (badly needed for investment in social sectors in the less developed countries) through subsidies or recapitalisation of uncompetitive firms in the economy.

All this is now virtually taken as axiomatic and is part of the conventional wisdom, but it is noteworthy that neo-classical theory does not have much to say about firm ownership. It dwells instead on the importance of market structure in generating efficient outcomes. If anything, the theory suggests that, in instances of market failure that cannot be entirely rectified through Pigouvian taxes or subsidies, there is a case for public, rather than private, ownership to meet overriding social objectives.

Subsequent literature, drawing on property rights and public choice theory has, however, come up with a number of reasons why private ownership might be superior. The property rights school of thought focuses on the agency problems under the two forms. Public choice theory emphasises that outcomes are sub-optimal where organisations are saddled with multiple and conflicting objectives.

The property rights school contends that agency problems are more acute under public ownership. Managers will perform only if they are monitored and incentivised [Jensen and Meckling 1976]. From the standpoint of the property rights school, managers in the public sector lack monitoring and incentives. These, in turn, stem from the fact that ownership is diffuse and hence property rights in the shape of rights to profit ill-defined. Very often, the firms are not publicly traded and hence not vulnerable to the threat of takeover and this reinforces the poor level of monitoring. Thus, from the perspective of the property rights school, inefficiency in PSUs arises from the failure to clearly assign property rights.

Or, to put it differently, politicians and bureaucrats, who are vested with the job of monitoring on behalf of the larger public are not as good at monitoring or designing incentive systems as shareholders in a private company (very often, institutional shareholders who perform the monitoring role on behalf of small investors).

Another reason managers in the public sector lack incentives to perform is that they do not fear bankruptcy; thanks to the ‘soft budget’ constraint, managers in the public sector can expect to be bailed out by public funds [Kornai 1980].

Public choice theory complements what the property rights approach has to say about relatively inefficiency in the public sector by focusing on the behaviour of politicians and bureaucrats. Unlike their counterparts in the private sector, managers in the public sector might lack focus because they are expected to pursue a variety of objectives, not all of which are calculated to maximise profit [Shleifer and Vishny 1996].

Multiplicity of objectives arises from the fact that public sector managers are answerable to different constituents, such as legislators, civil servants and ministers, each with its own objective. In particular, politicians, who are answerable to constituents such as labour, would tend to push public sector managers to pursue objectives, such as an increase in employment, that militate against profit maximisation.

Broadly speaking, both the approaches suggest that behaviour and hence performance of managers will differ in the public and private sectors because the objective functions are different and also the constraints are different. Neither is good performance incentivised in the public sector nor is bad performance penalised through takeover or bankruptcy.

The proposition that agency problems are necessarily so much more acute in the public sector than in the private sector as to make a difference to performance has not gone unchallenged. Stiglitz (1997) makes the point that in all big firms, whether in the public or the private sector, managers enjoy considerable discretion. Since managers can appropriate only a small fraction of any improvement in productivity, in neither sector do managers have any incentive to design good incentive structures.

However, it is also true that the effort required to design incentives structures is not so great as to impose a major impediment in either case. As Stiglitz puts it, “It seems extremely implausible to think, in either case, of managers mulling over whether to exert the little effort to design a good incentive structure that will make the organisation function better, carefully balancing the returns they will obtain with the extra effort they will have to exert.” Stiglitz cites the extraordinary performance of public enterprises in certain provinces of China to make the point that economic success is possible even under conditions in which property rights are ill-defined.

The Sappington-Stiglitz theorem (1987) shows that conditions under which privatisation could fully implement public objectives of equity and efficiency are extremely restrictive. This is reflected in some of the choices societies have made for public production. It is difficult, for instance, to design Pigouvian taxes or subsidies to attain the ‘right’ level of risk-taking or innovation or to attain objectives in public education such as ‘social integration’. The theorem also demonstrates that an ideal government could do a better job of running an enterprise itself than it could through privatisation – although, in practice, such an ideal government may be hard to come by.

Martin and Parker (1997) put forward a number of other reasons why the suggestion of superior performance implied by the property rights and public choice literally need not be taken at face value. They argue that it is simplistic to suppose that public and private sector companies are two distinct categories, each with a particular organisational design and orientation.

In fact, there is a continuum of organisational types ranging from the archetypal government bureaucracy (such as the Foreign Office) through various types of agencies (such as public corporations) and private sector firms heavily dependent on government contracts (such as defence firms) to limited liability companies and small businesses. At the same time, all types of firms are subject to some degree of state influence on account of regulations, taxes and macro-economic management. It seems more plausible, therefore, that “public and private sector firms may have similarities that define their behaviour and performance more clearly than the presumed differences”.

Further, the authors make the point that too much should not be made of the lack of incentives, especially pecuniary incentives, in the public sector and the effect of this on the performance of public sector managers. Managers could have broader motives than the usual self-seeking ones. They also quote a study by Fama (1980) showing that the labour market may capitalise performance in managerial remuneration creating incentives for performance quite irrespective of ownership. Further, they also cite studies that suggest that the behaviour of politicians and bureaucrats may not be as divorced from the public interest as is generally supposed.

Finally, the authors underline the point many have made about the imperfect nature of the market for corporate control. The failure of managers to maximise shareholder wealth and, in particular, any tendency on their part to help themselves to over-generous salaries and perquisites is, in theory, disciplined by the capital market. Shareholders will sell under-performing shares, causing prices to fall and creating conditions for a takeover by another firm. In practice, however, the market may not operate this way because of a number of factors: the transaction costs of trading shares, significant information imperfections, diffused share ownership, the failure of takeovers to deliver promised improvements in performance and profit, etc.

All this has led to the contention that what matters is not ownership so much as competition. [Vernon-Wortzel and Wortzel 1989]. However, it has also been pointed out that, while competition undoubtedly contributes to efficiency gains, the existence of a publicly-owned firm as the incumbent, might deter other firms from entering the market, no matter that competition is permitted. Moreover, since real competition means not only freedom to enter but freedom to fail, the existence of PSUs may not conduce to meaningful competition [Sheshinski and Lopez-Calva 1998].

Given the many ifs and buts about the theoretical merits of privatisation, researchers have had to turn inevitably to the evidence on the ground in order to arrive at conclusions. But the empirical evidence on ownership and efficiency, contrary to impressions that might have been created in the popular press, is by no means unambiguous, least of all where less developed countries are concerned, as the next section outlines.

Table : British Privatisation- Summary
of All Comparisons with the Results for
the NAtionalisation Period