Objectives

The size and structure of the public enterprise sector vary significantly within groups of otherwise comparable industrial and developing countries. For example, in the non-socialist industrial countries, the share of public enterprise output in gross domestic product (GDP) in the mid-1970s varied from 4 percent in the Netherlands and Spain to 15 percent in Austria. Similarly, among non-socialist developing countries, the shares varied from 1 percent (Nepal) to 14 percent (Taiwan, Province of China) in Asia; 7 percent (Liberia) to 38 percent (Zambia) in Africa; 1 percent (Guatemala) to 38 percent (Guyana) in Latin America; and 4 percent (Malta) to 14 percent (Portugal) in Europe (Short, 1984). This heterogeneity—which applies not only to the size but also the structure of the public enterprise sector—reflects the range of considerations that led to the decision to undertake a particular activity in the public sector.

From the standpoint of economic analysis, public ownership has most commonly been viewed as a response to the failure of private markets to secure efficient outcomes. Market failure can occur for a number of reasons, and public production in various areas can be justified by reference to particular sources of market failure. Thus, public goods like defense and police services are provided in the public sector because they are nonexcludable (giving rise to “free-rider” problems), state-mandated environmental protection is necessary because the market does not take account of externalities, and informational asymmetries are used to justify the public provision of medical care. In the case of the traditional public enterprises, in particular those involving the use of networks (power generation and distribution, water supply, telecommunications, and transportation), the possible emergence of a natural monopolist—that is, a situation where only a single producer can exploit available economies of scale—is the principal concern.

In these areas and others, however, market failure has often tended to serve as an ex-post justification for nationalization. Moreover, while market failure can provide a strong rationale for government intervention, it does not follow that intervention must take the form of public ownership. For example, the economic objectives of nationalization can be and have been achieved through the use of regulatory controls, legal sanctions, taxes, transfers, and subsidies. To explain why the preferred mode of intervention has so often been nationalization, and why the public sector now encompasses activities that lie well outside the traditional domain, especially in developing countries, it is necessary to look to a wider concept of market failure.

An important group of arguments—some clearly of an economic character, others more of a political/ideological nature—relates to economic development and planning. In many developing countries, public production was viewed as essential given the underdeveloped nature of resources and markets. In general, private returns to investment were not sufficiently attractive to private investors, and few native entrepreneurs with investable funds were either willing to bear the risk or were capable of running modern enterprises. Also, the scale of investment required often exceeded the capital-raising capacity of the indigenous private sector. Public ownership and control of the “commanding heights” of the economy was given special emphasis in this context, in both industrial and developing countries.

Substantial social benefits were also expected to derive from the creation of public enterprises. In many cases, public ownership was thought to be conducive to the attainment of a number of social policy aims. The inability of the market to achieve distributional objectives—in particular, widespread access to essential goods and services (so-called merit goods) at reasonable prices—is a source of market failure in the standard sense. But public ownership has been ascribed wider social objectives. For example, it has been used to create employment or to prevent rising unemployment. Of particular concern have been the social costs imposed when a locality or region is dominated by a firm or industry that is experiencing financial problems. The extension of public ownership to many areas of manufacturing industry—both declining industries (those with poor demand prospects) and those suffering temporary difficulties—is often related to the adverse employment consequences of continued private ownership and the possibility of bankruptcy. The social goals of public ownership take on particular significance in developing countries, where unemployment and inequality are more readily associated with political instability.

Nationalization can also serve strategic interests. A country that retains an interest in major world industries, even where it is inefficient to do so, will not depend on potentially unreliable external sources of supply.

Public ownership has also been advocated as a political strategy, although political intent has frequently been masked by reference to economic and social considerations. While there are clear examples of nationalization (and denationalization) that reflect a prevailing political ideology, the existence of genuine economic and social benefits provide for many instances of inconsistency. As a result, there is little correlation, both across countries and over time, between political stance and the size of the public enterprise sector.

While all of the above considerations help to account for the central role currently assigned to public enterprises in both industrial and developing countries, they do not fully explain why the boundary between the private and public sectors differs so much across countries. Because so many extraneous and arbitrary factors come into play, it is impossible to predict the size and structure of the public sector of a given country, even taking into account its economic, social, political, and other seemingly relevant characteristics. Indeed, in many instances, an enterprise ends up in the public rather than the private sector largely through an accident of history.

Performance

A growing body of evidence claims to show that when the public and private sectors can be compared in terms of the cost of producing similar outputs, the private sector outperforms the public sector. For example, Borcherding, Pommerehne, and Schneider (1982) summarize the results of a number of studies covering a wide range of activities (including air, bus and rail transport, electric and water utilities, and insurance) in the United States, the Federal Republic of Germany, Australia, Canada, and Switzerland—countries with allegedly similar social and political institutions—that support this view. But such results should be treated with caution. Given the objectives of nationalization, it should not be surprising that there are few examples of truly comparable public and private sector activities and enterprises. Attention therefore focuses on specific aspects of comparability; however, as the focus of attention changes, the results of such comparisons also tend to change.³

The results may also change over time, both in the long term and the short term. For example, impetus was given to creating municipal enterprises in Italy at the end of the nineteenth century by studies comparing similar private and public enterprises which clearly showed inefficiency in the private sector (Marchese (1985)). However, while once viewed as a model of successful nationalization, it is alleged that the public enterprise sector in Italy is now riddled with inefficiency and corruption (Martinelli (1980)). The relative inefficiency of the private sector in the United Kingdom was discussed by Pryke (1971) whose later work (1982), on the other hand, showed the private sector to be more efficient than the public sector. But perhaps the major limitation of such comparisons arises from the fact that public enterprises are assigned multiple objectives—including social obligations to deliver essential services, sell at below cost (which may involve cross-subsidization), and provide employment—and to the extent that these objectives must be traded off against commercial objectives, public enterprises are bound to appear less efficient in terms of the criteria by which private enterprises are judged.

Despite the inconclusive nature of the evidence, it is difficult to believe that existing public enterprises are not capable of improving efficiency significantly, be it in the public or the private sector. Moreover, increasing budgetary support for public enterprises suggests that their performance has been deteriorating, particularly in developing countries.⁴ To reverse this trend, the major sources of inefficiency need to be identified.

Problems

Many of the early proponents of government ownership argued that socialized industry could be self supporting and economically successful only if it were freed from political interference. While public enterprises should be accountable to government, day-to-day decision making should be left to enterprise managers. This has been referred to as the “arm’s length principle.” In practice, public enterprises are subject to a wide range of statutory and administrative controls, as well as to less formal modes of intervention. Government influence extends well beyond that necessary to ensure that enterprises fulfill their economic, financial, and social objectives. Indeed, a significant part of the problem is that politicians can influence the objectives of public enterprises; in particular, less compelling non-commercial objectives are substituted for economic, financial, and more immediate social objectives. Notwithstanding the claim that some countries have experienced a recent improvement in the relationship between politicians and public enterprises (see, for example, Posner (1984), in many countries it is unlikely that politicians can be persuaded to interfere much less than in the past.

It has also been suggested that in choosing to locate an activity in the public sector, market failure has given way to bureaucratic failure. For example, property rights (or agency) theory suggests that, because they do not have access to shared information, governments (the principals) face difficulties in providing appropriate incentives to public sector managers (their agents) and in monitoring their performance. Managers are therefore given less discretion than their private sector counterparts and so choose a relatively quiet life (Alchian (1965)). They will perform only to the level necessary to meet the performance standards set for them, and these may be modest compared with the potential of the firm or industry concerned. From a different perspective, public choice theory suggests that public managers can secure more pay, power, and prestige than their private sector counterparts by forming coalitions with civil servants in supervising ministries that result in increased budgets (Niskanen (1971)). Indeed, budget maximization becomes an end in itself, and other objectives—both commercial and noncommercial—have to be conceded to achieve it. While these two theories imply different behavior on the part of public sector managers—and so far the available evidence is incapable of distinguishing between them—both theories predict that public production will be relatively inefficient.

Political interference and bureaucratic failure are probably the principal sources of inefficiency associated with public ownership. There are, however, other important sources of inefficiency. For example, with government backing public enterprises cannot go bankrupt, and they do not face the risk of take-over; they are not, therefore, forced to observe the financial discipline imposed on the private sector. Specifically, public enterprises either do not have to borrow on the private capital market, or, if they do, government guarantees or the assumption of government backing results in their being favorably treated relative to private enterprises. This, of course, accommodates the inefficiency resulting from political interference and bureaucratic failure. It also allows public sector unions to exploit their power to interrupt the supply of essential goods and services to secure pay and conditions that are out of line with those in the private sector. In developing countries, it is also argued that limited human resources are spread too thinly over large public enterprise sectors. The above problems suggest that public enterprises will perform badly in terms of productive efficiency, because they are likely to have higher production costs at a given level of output than comparable enterprises in the private sector.

It is also alleged that public enterprises fail to achieve allocative efficiency, because they have little incentive to respond to consumer demands; the quantity, quality, and other characteristics of goods and services provided by public enterprises are not those most valued by consumers. While public ownership per se may lead to productive inefficiency, it can result in allocative inefficiency only when associated with considerable monopoly power—which is often granted by statute—or when some other form of protection from competitive pressures—usually the result of inappropriate financial and trade policies—is implied.⁵