Some Topical Issues Concerning Public Enterprises
  • 1 0000000404811396 Monetary Fund
  • | 2 0000000404811396 Monetary Fund
  • | 3 0000000404811396 Monetary Fund


The growing importance of public enterprises in many developed and developing countries has led to increased concern with their operations and to increased focus on numerous issues arising from their activities.1 The purpose of this paper is to review certain issues that arise from the operations of public enterprises with particular emphasis on their relevance to macroeconomic and financial policymaking. Almost any one of these issues might in itself be the topic of a more detailed study, which suggests that their treatment in this paper is likely to be somewhat cursory.

The growing importance of public enterprises in many developed and developing countries has led to increased concern with their operations and to increased focus on numerous issues arising from their activities.1 The purpose of this paper is to review certain issues that arise from the operations of public enterprises with particular emphasis on their relevance to macroeconomic and financial policymaking. Almost any one of these issues might in itself be the topic of a more detailed study, which suggests that their treatment in this paper is likely to be somewhat cursory.

Financial variables are of paramount interest when the policy objectives are aimed primarily at influencing financial imbalances. In such cases, direct concern would focus on short-run indicators of the financial performance of public enterprises and their impact on financial markets rather than, for example, the levels of employment. However, real economic variables should not be ignored. Indeed, public enterprises often incur unsustainable financial imbalances (or deficits) in the course of attempting to influence real variables through the pursuit of noncommercial objectives. Especially in these cases policymakers should be obliged to compare the benefits and the costs and to assess the impact on the real variables in question of alternative measures aimed also at curbing the financial deficits in the short run. Furthermore, with the extension of the time horizon, real economic variables become of even greater interest. Indeed, when the policymakers’ objective is to reduce widespread cost and price distortions and structural maladjustments in production and trade, it is inevitable that the impact of public enterprises’ operations on real economic variables must be taken into account. Nevertheless, for financial policymakers paramount interest will center on the financial facet of these activities of public enterprises and on the broader impact of their operations on financial markets.

The focus on macroeconomic variables begs the question as to why and when an individual public enterprise or group of public enterprises would be of macroeconomic significance. In general, public enterprises that are operated profitably, efficiently, and in conformity with commonly accepted commercial principles would not constitute a drain on financial resources or lead to monetary expansion. Indeed, their impact would not be significantly different from privately owned enterprises carrying out similar functions. Even though in some cases these public enterprises may provide sources of quick government revenues, profitability is usually limited to a normal return to capital if operations are efficient and hence do not involve monopoly abuses. Consequently, such public enterprises are less likely to be of interest to policymakers than are public enterprises that are not operated profitably, efficiently, or in accordance with generally accepted commercial principles. The latter group is likely to constitute an ongoing drain on financial markets that may lead to inflationary credit expansion. Alternatively, it may generate less profit than efficiently operated firms. Furthermore, such operations may lead to inefficient mobilization and allocation of scarce resources or an undesirable pattern of income distribution.

The focus on the macroeconomic impact of public enterprises leads to another observation. If public enterprises’ operations are of a relatively small scale and are unimportant in the overall economy, they are unlikely to have significant impact in financial markets. This suggests that there are two distinct situations in which the financial activities of public enterprises are of potential macroeconomic impact. First, even one or a few enterprises may have a macroeconomic impact if the enterprises’ financial activities are sufficiently large relative to the economy. Petroleum enterprises, other natural resource extraction companies, and agricultural marketing companies are frequent examples of such public enterprises. Second, the situation can arise when the government’s policies vis-à-vis public enterprises in general, even though each one is small, are such as to alter the aggregate financial performance of all public enterprises sufficiently to have a macroeconomic impact. For example, output prices of public enterprises may be constrained in a probably futile effort to constrain inflationary pressures arising from other causes, such as excessive monetary expansion. This situation could arise even when there are relatively few or no large public enterprises capable individually of having a macroeconomic impact.

In the first situation any of a variety of causes may lead to the macroeconomic impact and some of these causes may be outside the government’s control. For example, depressed world commodity prices often have seriously adverse implications for important public enterprises’ financial performances. In such cases the scope for remedial action by the government to minimize the macroeconomic impact may be limited to certain cases, as for example, the domestic operations of agricultural marketing boards, where government-imposed pricing policies constrain the boards’ flexibility to adjust to external price movements. Often such pricing policies are pursued with economy-wide income redistribution objectives and may be politically difficult to reverse despite their potential macroeconomic implications. Managerial inefficiency and too large a scale of operation relative to the size of domestic demand are also frequent causes of financial losses by important enterprises. Managerial inefficiencies may be corrected by various microeconomic actions, such as changing management.

In the second situation, the government’s policies themselves give rise to the macroeconomic impact without any particular exogenous influences, and reversal of those policies may be assumed to reverse the macroeconomic impact. Numerous policies may lead to such results but most often they are policies designed to influence the prices of inputs and outputs, the selection of material and labor inputs, and the composition of outputs to be produced.


In analyzing the issues arising with public enterprises and various aspects of their operations, it will occasionally be useful to view them within the conceptual framework of the financial and economic variables that may be affected. These may be grouped into three categories: (1) the direct impact of public enterprises on bank credit and the balance of payments (the “credit impact” and the “balance of payments impact”); (2) the direct impact of public enterprises on the budget through explicit financial transactions between the government and public enterprises (the “budget or fiscal impact”); and (3) the longer-run “structural impact” of public enterprises’ operations on resource allocation and income distribution often reflected in implicit subsidies or taxes. Categories (1) and (2) will have (at least theoretically) quantifiable impacts on financial markets. Category (3) is equally important but would be more difficult to quantify. All three categories can reflect the results of the use of public enterprises as instruments of discretionary policy actions or inaction.2 Many public enterprise actions may have an indirect impact in these areas as well, and, where possible, they should also be identified and quantified.

The list of issues relating to public enterprise activities is virtually without limit. The remainder of this section will only introduce the more important issues; more detailed discussions of these issues are left for subsequent sections. In addition to the analytical framework, it is useful to categorize these issues into some common areas. Admittedly, the lines of demarcation between various issues and categories of issues are not always distinct, and, in some cases, one issue may be very closely related to other issues. Nevertheless, there would seem to be about five convenient categories into which most issues may be classified.

In many countries issues of objectives of public enterprises are of considerable operational importance. For example, it may be useful to recognize that public enterprises may have either macroeconomic or microeconomic objectives or both. Similarly, there may be both financial and nonfiancial objectives, and commercial and noncommercial objectives. To a large degree the issues of objectives involve relationships between political policymakers and public enterprise managers. Many of the difficulties that are likely to arise with public enterprises may be the result of ill specified and often conflicting objectives that government or public enterprise officials assign to public enterprises. Perhaps more than other categories, issues of objectives are likely to have an impact on the broadest array of economic variables mentioned previously.

Related to issues of objectives are issues of the method and extent of control, that is, the control over public enterprises’ activities that is exercised by the government. The methods by which control is exercised can be of great importance. For example, efforts to control by giving general policy guidance aimed at all public enterprises may have significantly different effects on the efficiency of operations and hence on profitability than efforts to control the quantities or prices of productive inputs and outputs through specific directives to managers. In general, these issues are likely to be institutional and consequently are more relevant to longer-term variables, such as resource allocation and income distribution. Issues on the extent of control would include both the actual and potential control that is exercised by the government, or even the absence of control. Invariably, these would include consideration of what is to be controlled. Should the government attempt to control real variables, such as the level of employment, the sources of supply of inputs (e.g., from another public enterprise), and the level of output? Or should the government focus its control efforts on financial targets, such as the rate of return, the level of profits, the planning and financing of investment, the distribution of profits, and the enterprise’s capital structure? Clearly, these issues are likely to be relevant to almost any time frame and could have an impact on almost any economic variable.

Perhaps more than all others, issues of pricing arise with regard to public enterprises. For a variety of reasons output prices often do not reflect the full costs of inputs employed with the result that consumption of the output is subsidized. The cost of the subsidy may be financed by a government transfer to the producing public enterprise, reduced profits and reserves for the public enterprise, increased resort to credit by the public enterprise, or some combination of these. In either case potential resources for the public sector are forgone. Less objectionable to most financial policymakers (and almost certainly less common) are public enterprises that generate unusually large profits for the public sector and are in some senses a “taxing” authority. However, the pricing issues involved are far more complex than simply generating accounting profits. For example, is it possible to ascertain whether output or input prices are at appropriate levels, and, if so, how does one do so? Are there acceptable justifications for prices that result in losses or profits? What macroeconomic objectives might be accomplished by public enterprise pricing policies? Indeed, what should be the objectives of public enterprise pricing policy and what would be the implications of different objectives for financial flows? Obviously, pricing issues are more likely than most to have a direct impact on financial variables, such as bank credit or budget transfers and receipts, and are likely to do so in the short term.

The importance of informational issues is perhaps best illustrated by the acute lack of information on public enterprises’ operations and their financial impact in many countries, A major issue is the question of what information is relevant to or necessary for various types of financial and macroeconomic analysis. For example, is information on public enterprise employment practices and wage bills needed to analyze the credit impact or the balance of payments impact? Are data on particular prices useful, or is it adequate to know only the extent of credit to public enterprises? What information is relevant to assess the balance of payments, fiscal, and structural impacts of public enterprises’ operations? These issues per se are not likely to have any significant direct impact on economic variables. However, through the availability of better information and its use by policymakers, significant improvement in public enterprise policies may be obtained that could have positive effects on various economic variables.

The final set of issues to be discussed relate to the various sources of finance. The review focuses on how the sources and types of finance may vary according to the activity actually being financed and on the implications the various sources and types of credit for public enterprises may have for credit creation, monetary expansion, and (both directly and indirectly) the balance of payments and foreign exchange availability.


Certainly among the most important aspects of public enterprises’ activities are the objectives they pursue. It has been said that “the special character of public enterprise is simply that its management is normally less subject to pressures to seek profits, and more attentive to the other objectives expressed by the agencies with which it must deal…. To benefit from public enterprise, it is above all important to create channels for careful definition of the economic and social objectives it is meant to serve.”3 A major cause of problems with public enterprises’ operations is the absence of clearly defined, consistent, and precisely communicated objectives for the managers of enterprises. This void is likely to dampen severely managerial incentives with a resulting adverse impact on the efficient internal operations of the enterprise, and it may also result in other actions by managers of public enterprises which are contrary to national interests.

An important related issue is the question of who should determine objectives. Only the government has the perspective to assess national interests and their relation to the operations of public enterprises. Consequently, insofar as a public enterprise’s operations relate to national interests, the government should determine the objectives assigned to various public enterprises. Can any a priori judgment be made as to which agencies of the government might be appropriately involved in the determination of objectives? Some countries have established ministries that supervise all public enterprises. However, these ministries have often tended to become spokespeople for the public enterprises rather than instruments for government control over the enterprises. Indeed, a broad array of ministries would need to be represented in the formulation of government objectives. Since the operations of public enterprises almost always involve financial implications, it would seem prudent to include the finance ministry. Also the central bank should be represented as public enterprises often have substantial credit requirements and often absorb or generate large amounts of foreign exchange. Since scarce economic resources are devoted to public enterprises, the planning ministry should be included in the formulation of objectives, especially those related to investment plans. Moreover, each public enterprise is usually under the auspices of at least one technical ministry, which should also be included in the determination of at least technical objectives. The presence of representatives from other agencies may also be desirable for determining objectives of particular public enterprises, but in general it is probably best to minimize the involvement of agencies not directly concerned with or affected by a public enterprise’s operations, Whether the organizational arrangements to achieve this representation should be left ad hoc or be institutionalized cannot be determined a priori. Indeed, the best arrangements may vary from country to country, but it seems reasonable that in most cases they would center in the finance ministry.

In contrast to officials of the government, managers of public enterprises should not be entrusted with assessing the national interests with respect to their firms. Apart from lack of relevant knowledge, they might also be influenced by personal or other motives narrower than the national interests. However, managers of public enterprises should be consulted about their role in carrying out assigned objectives related to national interests in order to ensure that the objective is realistic and that a particular firm is an appropriate instrument for achieving it. Furthermore, the government should avoid defining objectives relating to internal managerial or operational aspects of public enterprises that have no strategic national implications. In these matters managers should be left free to exercise their special talents to best advantage.4 It should be noted that allowing a public enterprise to operate on a fully commercial basis may be the government’s primary and perhaps sole objective. Furthermore, in some cases this would not necessarily be a bad second-best policy in view of the difficulty in specifying alternative objectives and the possible staff requirements needed to administer them.

Three overriding questions concerning objectives would seem to be of importance to the analysis of public enterprises’ macroeconomic impact. First, in the case of multiple objectives, are they mutually consistent? Second, does the public enterprise have techniques available to achieve the various objectives? Third, and most important, what is the estimated financial impact? With regard to the first two questions, since many enterprises are likely to be given multiple objectives, care should be taken to ensure that the objectives are mutually consistent or at least that when there are to be trade-offs some indication of priority among them is given. Furthermore, it will be important to ensure that sufficient and appropriate instruments are available to public enterprises to accomplish their objectives. For example, long-run growth is not likely to be consistent with short-term price stability if the latter is achieved by holding public enterprise prices fixed.

In general, the pursuit of most objectives, whether they be commercial or noncommercial, financial or nonfinancial, is likely to have an impact on a public enterprise’s financial performance that may or may not be quantifiable. Where possible, the quantification of the financial implications of nonfinancial objectives and the taking into account of their costs is crucial to the establishment of realistic overall ex ante, financial objectives or targets for public enterprises. Establishing financial objectives has several advantages. In addition to providing a guideline for ex post evaluation, financial objectives provide public enterprise managers with both guidance and incentives to improve internal efficiency. Furthermore, since a public enterprise’s potential profitability, its internally generated investable funds, and its credit requirements are all, in part, a function of the financial implications of noncommercial objectives, their quantification will also have the advantage of providing an initial indication of their financial impact. In some cases the effects may be sufficiently large to be taken into account in financial and macroeconomic policies.5

The manner in which financial objectives are defined can have significant economic implications. On the whole, economists have tended to favor specification of pricing policies rather than setting financial targets, such as rates of return or profit levels.6 First, it is argued that financial targets may obstruct allocative efficiency since allocatively efficient pricing and investment policies imply a particular surplus or deficit for a public enterprise that may differ from that set by the government. Second, investment decisions may be distorted if target and market rates of return differ. Third, financial targets may be used as a justification to charge prices in excess of allocatively efficient marginal cost prices. Finally, internal efficiency may not be enhanced if the financial targets are not enforced.

All these objections are oriented essentially to microeconomics. They are all founded in a concern for determining prices in accordance with long-run marginal costs in order to achieve efficient resource allocation and input mix. However, there are counterarguments based on both micro- and macroeconomic considerations. First, the preference for pricing rules over financial targets ignores the government’s revenue needs, and financial targets can serve as a form of indirect taxation. Second, the use of lump-sum financial targets rather than rates of return could reduce the chance that investment decisions would be distorted.7 Third, financial targets may well increase X-efficiency by more than any induced loss of allocative efficiency.8 Finally, it is certainly true that, in practice, marginal cost pricing is seldom practiced by public enterprises and such prices are difficult to identify in practice.9

In summary, the primary issues of objectives would appear to include the need to recognize, define, communicate, and evaluate any noncommercial objectives of public enterprises that are of legitimate national interest. These objectives should be feasible and, if not consistent, at least the relevant trade-offs among multiple objectives recognized and their financial impact should be at least crudely quantified. This information can then be used as an input to the quantification of financial objectives for the operations of public enterprises. While in theory there may be some preference for expressing financial objectives as pricing policies, in practice, financial targets may be just as desirable.


Should the government exercise controls over public enterprises? Who should exercise any controls, what types or methods of control should be employed, and what aspects of public enterprises’ operations should be controlled? All of these questions need to be addressed when considering control of public enterprises. Although there are few clear answers to any of these questions, many issues of control are closely related to some of the preceding discussion of objectives.

Perhaps the most obvious control issue arises from the serious problems of managerial incentives in public enterprises. It is commonly perceived that poor management resulting from inadequate incentives is largely responsible for much of the inefficiency in public enterprise operations. In turn, the absence of managerial incentives is often attributed to overly rigid controls that constrain the ability of managers to exercise their creative talents and to adapt the enterprise’s operations to changes in market conditions. This suggests that government controls over the internal management of public enterprises should be minimized in order to avoid unduly interfering with managerial decision making. It also suggests that controls should not normally be imposed on such production decision variables as the appropriate combination of factor inputs, the prices paid for those inputs, and the timing of production.

However, this does not imply that no controls should be imposed over public enterprises. Indeed, so long as public resources are committed to these enterprises there is every reason to expect that public enterprises should be at least monitored, and perhaps given some operational guidelines, to ensure the public interest is served. Furthermore, it must be acknowledged that even internal objectives and controls can have significant effects on pricing, dividend practices, and profitability of public enterprises, and, that in some cases the government might need to ensure that these are not inconsistent with macroeconomic policy considerations. Thus, there will be some inevitable conflict between the need for accountability for the public interest and the need for managerial flexibility and innovation within the public enterprise.10

At least some degree of government control may be needed to serve a variety of purposes and objectives other than coordination of macroeconomic and financial policies. Controls may be needed over investment project plans to ensure they are consistent with national priorities. Controls may be used to coordinate the activities of more than one public enterprise when there are important linkages between these enterprises. Often controls are imposed that may be intended to serve a regulatory function, such as curbing monopoly powers of public enterprises, although public enterprise control should not be regarded as a substitute for antitrust regulation of private enterprise. Finally, in order to be effective, controls may require a feedback of information that may be used as inputs to the determination and evaluation of other government policies, in particular, monetary and fiscal policies.

The actual agency of the government that should be empowered to control public enterprises would depend on a variety of factors and would be related to the agencies involved in determining objectives. For example, many technical aspects of enterprises’ operations may not have financial implications. However, in most cases they are likely to have substantial financial implications, particularly in the case of decisions about investment projects, For example, if the port authority were a public enterprise, any technical decision to expand port capacity would clearly have major financial implications. Consequently, in many such cases, both a technical ministry, such as the transport ministry, and the finance ministry would be involved in the enterprise’s investment decision. Since such ministries have tended to represent their own interests rather than those of the public enterprise and since they often view each public enterprise or problem in isolation, separate public enterprise ministries have been established in some countries to represent the concerns of public enterprises. However, as noted previously, these ministries have often tended to ignore the broader interests of the government.

The broader public interest therefore would seem to be best served by an alternative administrative organization to act as a funnel for control. There is no uniquely preferred arrangement, and the best solution will almost certainly vary with the individual country’s institutional arrangements. One interesting arrangement has recently been tried in Papua New Guinea. For any enterprise over which the Government wishes to establish some measure of control, an ad hoc planning committee is appointed, comprised of representatives of the Department of Finance, any concerned technical ministry, the National Planning Office, and the enterprise itself. The ad hoc committee is charged with defining clearly the Government’s objectives for the enterprise (with the eventual aim of quantifying their financial implications and budgetary impact) and serving as an ongoing conduit by and through which associated controls can be implemented.

Among the most sophisticated institutional arrangements and control techniques are those that have been employed in France.11 Although no effort is made here to describe all these arrangements, two general changes that were intended to be implemented after 1967 and to facilitate placing responsibility for efficient management solely on the enterprise are of particular interest.12 First, the Government was to identify clearly and to reimburse fully the costs of any noncommercial activities that it imposed on the enterprise. As a result there has been considerable progress in defining and costing the burden of social obligations and shifting them onto the Ministry of Economy and Finance.13 The efforts to make more explicit the areas in which normal commercial considerations are supplemented by additional considerations of social policy and their attendant costs or revenue losses are sometimes referred to as the concept of “transparency.”14

The second major change was the attempt to increase the autonomy and efficiency of management and to ensure concurrently compliance with government objectives through the negotiation of contrats de programme.15 These planning contracts were essentially agreements between the Government and individual enterprises. They were designed to provide management at least some flexibility with regard to pricing of outputs and inputs, input and output mixes, the distribution of operating profits, and some assurance of financing for approved longer-term investment programs. For the Government, the agreements were intended to ensure that major government objectives would be pursued and that there would be some means of assessing management’s performance following the granting of greater autonomy.16

Another method of control that has sometimes been advocated is to let public enterprises compete with other public enterprises. In practice, there have been few instances of competition with only other public enterprises. More generally it might be argued that public enterprises may be controlled by letting them compete with any other enterprises, either public or private. Although this might be expected to be effective, a few studies of this situation indicate that there is still X-inefficiency in public enterprises operating in competition with other enterprises.17 However, there are also at least some notable examples of publicly owned enterprises that have proven themselves to be efficient competitors in industries, for example, the automobile industry, dominated by private enterprises.

Leaving aside other questions of the techniques and organization of controls, focus now on the aspects of public enterprises’ organizations and operations which should be subject to government controls. The controls available to governments may take a variety of forms. Formal controls may include statutory controls that are specified in legislation, but these are often too broad in scope to be of day-to-day use. For example, formal controls may include the power to appoint the chief executive and board members of enterprises. However, in practice this control may be limited by the fact that the tenures of public enterprise managers are often longer than those of government officials. Consequently, managers may become somewhat indispensable.18 Legislation often bestows on governments, and in some cases particular ministers, the power to issue policy directives to public enterprises, especially those that are legally organized as statutory authorities. Less common is legislation authorizing governments to issue directives to publicly owned companies organized under general company legislation. In these cases the government usually operates through its representatives on the Board of Directors rather than issuing directives directly to the enterprise. Informal controls through the exercise of various pressures and moral suasion applied by government officials to public enterprise managers may, in fact, be the most effective means to influence the operations and policies of enterprises. However, informal controls have the inherent disadvantage that they may be ad hoc in nature, uncoordinated, and perhaps contradictory.

Budgetary controls are sometimes exercised over the revenues and expenditures of public enterprises. These may take many forms, including approval of budgets, or specific expenditures, or investment projects, and authority to regulate the prices charged by the enterprise. A major problem with many budgetary controls is their adverse impact on managerial incentives. For example, requirements for prior approval of specific expenditures could restrict management’s ability to react to changing market conditions. Consequently, governments have applied more general and less restrictive controls aimed at various economic variables.

More general controls could aim, for example, to achieve a certain level of profits or rate of return, or to chart the course of public enterprises’ investment by requiring that outlays for investment projects of public enterprises be subject to the same planning process as those of the government, or to restrict the access of enterprises to short-term bank credit. Such controls may be closely related to the objectives that take the form of targets and ceilings against which the actual performance of the enterprise is to he subsequently assessed. For example, an overall profitability objective could be used in conjunction with investment plans of public enterprises to determine other financial subtargets, such as their overall borrowing requirements. If projections of other sources of net credit and investment expenditures were available, an estimate of public enterprise borrowing from the banking system (i.e., credit impact) or the government (i.e., at least part of the fiscal impact) could be prepared. These same subtargets can be used as control variables. For example, by such a procedure one could establish cash limits on public enterprises’ financing requirements from all sources other than internally generated funds such as loans, equity injections, and grants. Such a system of cash limits was introduced in the United Kingdom in 1978 in order to strengthen expenditure controls.19 The use of controls on external financing requirements could, when aggregated for all relevant public enterprises, be a useful input into financial stabilization programs for both the short and long terms.

While there may be a case for establishing profit objectives and credit objectives and controls for public enterprises, there would seem to be no similar reason to establish either objectives or controls for the balance of payments impact or for the use or generation of foreign exchange by an individual public enterprise. If limits were placed on foreign exchange usage, they might inhibit a public enterprise from competing with other users of foreign exchange in the marketplace. Indeed, such an objective might be taken as an allocation rather than a ceiling. Finally, even if foreign exchange were abundant, such an objective might result in imports of a similar amount even though market forces might have normally resulted in a different composition of foreign and domestic material inputs.

These observations suggest some tentative conclusions concerning government controls over public enterprise activities. They should be minimized in order to reduce their adverse impact on managerial incentives. They should relate to clearly identifiable objectives and should be as general in nature as possible. They should aim at achieving certain measurable financial objectives that reflect the estimated financial impact of pursuing noncommercial objectives. Finally, ceilings are best suited for variables whose expansion is undesirable, and targets for those whose expansion is desirable. These considerations may be applied to one variable—prices—whose control is often advocated.


The prices that public enterprises charge for their outputs and pay for their inputs have given rise to numerous issues, but this discussion will focus on only four that seem particularly germane to the economic variables discussed previously. First, how should prices be determined and should they be set to generate a profit? Second, how should public enterprise prices be determined when monopoly elements are present, and under what circumstances, if any, should public enterprise prices be regulated? Third, are adjustments to public enterprises’ prices an appropriate or an effective technique to tax or to subsidize public enterprise outputs and inputs? Finally, what external constraints limit the flexibility of public enterprises’ pricing actions?

Public enterprise pricing policies and the techniques employed to determine their prices have been the subject of considerable discussion over the years. Much of the discussion has been within a microeconomic context and has focused on the measurement of marginal opportunity costs.20 The attention to marginal cost reflects the ever present objective of ensuring efficient use of the economic resources whén determining the prices of the goods and services produced by public enterprises. The achievement of allocative efficiency requires that the prices paid by consumers for the output of goods and services by public enterprises should cover, at the margin, the opportunity costs to society of resources used as inputs to production, basically the land, tabor, and capital employed either directly by the public enterprise or indirectly in the production of materials, fuels, and services used by the public enterprise in the production process.

However, in practice there are a number of problems in the application of the marginal cost principles. The primary problem concerns the identification and measurement of marginal costs. Marginal costs are difficult to identify because the marginal product and the inputs used in its production are difficult to identify, especially within the current technological framework and with the possibility of a number of by-products. Another problem may arise with the valuation of inputs. In competitive markets, the market prices paid by the public enterprise for labor, materials, etc., represent their opportunity costs. However, due to the exercise of monopoly power by sellers of inputs or of monopsony power by the enterprise in the purchases of inputs, in some cases (e.g., water in hydroelectric projects) opportunity costs of inputs may be remote from market valuation. Also, marginal costs are dependent on the level of technology, the peak utilization of the capacity, and the time and life of production equipment, and are therefore variable over time.

As a result of the difficulties in quantifying marginal opportunity costs, more practical pricing procedures have been sought. One approach is to adopt a trial and error pricing process aimed at setting prices in the short run, to equal the greater of (1) the running or operating costs of the plant capacity being partly used, or (2) the price required to restrict demand to that plant’s capacity.21 It would also be prudent in most cases to avoid excessive price adjustments in the short run as these may reflect only highly temporary fluctuations in demand.

Perhaps the most common method of determining prices in practice is to use some form of average accounting costs plus some markup for a profit margin. However, accounting costs are not necessarily related to social opportunity costs of resources. Accounting costs usually depict historical costs and reflect at least to some extent technological and financial history rather than current and potential usage of resources. In particular, depreciation charges based on historical costs are not likely to reflect opportunity costs during inflationary periods. Depreciation practices may also be arbitrary, may vary among industries, and may not accurately reflect either the length of an asset’s useful life or the pattern over time of its use. To address the problem of historical costs versus current costs, it has been suggested that replacement cost depreciation should be adopted in the place of historical costs. However, marginal cost pricing is based on the value in current use of existing assets and not on their hypothetical replacement in the future, and consequently replacement and marginal costs may differ. Nevertheless, periodic asset revaluations and adjustments in depreciation allowances will ordinarily bring accounting costs closer to opportunity costs and thus assist in establishing prices based on the latter.

Another difficulty is that average costs cannot always be used as a proxy for marginal costs. In a world with fixed technology, marginal and average costs in the long and short runs are equal when the scale of plant is optimum, that is, when long-run and short-run costs are simultaneously minimized. However, under conditions of rapidly changing technology, marginal and average costs are not likely to coincide in either the long or the short run. Consequently, average cost pricing may not lead to optimum resource allocation.22

Given the difficulty of measuring marginal opportunity costs and the fact that most of the alternatives to marginal cost pricing lead to inefficient resource allocation, an acceptable approach might be to set prices that would result over the medium or long term in a rate of return to capital commensurate with the returns obtainable in alternative uses as adjusted in accordance with the concept of transparency. Although not perfect, such a practice would, at least with respect to capital inputs, capture the essence of the marginal opportunity cost pricing principle.23 Since the government’s capital invested in a public enterprise is itself a resource input, if marginal cost pricing were followed it can be expected, in general, that the capital employed in the enterprise would earn a return equal to its opportunity cost just as any other input to the production process.24

This approach requires that neither monopoly nor monopsony power enables a Firm to charge output prices (or pay input prices) that do not reflect opportunity costs, with the result that profitable operations are not allocatively efficient. Pricing in the case of monopoly is particularly difficult because a profit can usually be generated even though the public enterprise is not being efficiently operated. Excessive costs or noncapital factor payments can be passed on to consumers in higher output prices without any reduction in profits. Indeed, the presence of monopolistic elements provides one of the clearest justifications for government regulation of public enterprise prices, and perhaps one of the most common examples would be publicly owned utilities.

The regulation of a public enterprise’s prices may also be justified even in the absence of the exercise of monopolistic power by the public enterprise. Often, social opportunity costs of production may differ from private costs as reflected by the market prices paid for resource inputs. For example, social costs of production may exceed private costs if a producing enterprise’s waste material pollutes the environment. Efficient resource allocation requires that all social opportunity costs, including the cost of the externality, be reflected in the output’s price. If they are not, sales and profits may exceed allocatively efficient levels.

Other exogenous noncommercial government objectives, such as social objectives relating to the distribution of income, may lead to regulation (both formal and informal) of public enterprises’ pricing decisions. It is not uncommon for the prices charged for goods and services purchased or produced by public enterprises to differ from opportunity costs of resources and materials because of efforts to subsidize the use of an input or consumption of the enterprise’s output. For example, the government may prefer to reduce the price of basic goods or services (e.g., rice or health services) rather than to provide tied grants for its consumption to the poorer sections of the community. The transparency concept requires, in order to avoid generating subsidies of unknown (and perhaps undesired) magnitudes when prices are deliberately altered for distributional or allocative considerations, that the extent to which revenues or costs are affected should be at least crudely quantified and evaluated. When financial losses are incurred, an explicit subsidy to the public enterprise may be required in the government’s budget.

Governments have also pursued stabilization objectives through puhlic enterprises’ pricing policies. Very little research has been done in studying the results of these attempts. However, the direct impact of public enterprises’ pricing decisions on overall price levels is likely to be limited and, in any case, is dependent on the size and strategic role of the enterprise. Furthermore, efforts to maintain prices at specified levels could result in reduced government revenues or increased government expenditures, leading in turn to higher government borrowing with attendant inflationary risks. It is therefore possible that the regulation of public enterprise prices as a tool of macroeconomic policy would be strongly counterproductive.25

While price adjustments may be needed to adjust for monopoly elements or for differences between social and private costs or may be desired for distributional purposes, it is not necessary that direct price regulation be employed. Even though the inclusion of a tax element in public enterprises’ prices has been practiced for generations, it has been shown that under most circumstances there may be substantial differences between a public enterprise price increase and an equal tax increase, and that for many purposes an explicit tax may be preferred. In particular, from the viewpoint of raising revenues for the government’s budget, a tax is likely to be more reliable and more predictable than a price increase and is likely to raise equal revenues with fewer distortions.26 Also, a tax would probably be the preferred method of levying a charge aimed at reflecting social costs not included in the financial costs of production, because the use of a price increase might have the undesired result that some of the proceeds would accrue to non-capital factors and material inputs, some of which may already be paid more than the value of their marginal product. Finally, tax increases are more likely than price increases to be uniform across firms and consequently might introduce fewer distortions into the pricing system. Similarly, when they can be administered, direct budget subsidy payments to the desired recipient groups are more desirable than price reductions as a method of subsidizing a particular segment of the population.

However, for some purposes other than monopoly restraint pricing action may still be preferred, primarily when public sector revenues, rather than only government revenues, are of concern. For example, if the public enterprise were operating at a loss, a price increase would probably increase the receipts of the enterprise whereas a tax might reduce its receipts and profits further. If the government had been financing the enterprise’s deficit by budgetary subsidies, a price increase might result in reduced government expenditures and smaller budgetary and public sector deficits. Furthermore, only a price increase could reduce the level of implicit subsidies that do not require the transfer of financial resources from the government to the enterprise.

Turning to external limitations on public enterprise pricing flexibility, it is obvious that domestic or international market competition would effectively constrain public enterprises to keep prices in rough alignment with competitors. In these cases the scope for achieving social and stabilization objectives solely through public enterprises’ pricing practices would be limited, but efficient resource allocation will probably be enhanced. In such cases, tariff and domestic tax policies would be the preferable instrument to alter the prices of goods in order to effect social and stabilization objectives.

The resource or material input and output prices that would reflect marginal cost pricing, other government policies, and the constraints resulting from various market conditions are shown in Table 1.27 The matrix depicts the feasible relationships between the marginal opportunity costs (MOC) of resource or material inputs and outputs and the output prices charged (OP) and the input prices paid (IP) by public enterprises. When enterprises compete in the market for inputs, the marginal opportunity costs of inputs, including capital in any form, may be represented by its free market price (MPI), or its value in alternative uses. Adjustments to opportunity cost prices may be viewed as separate and quantifiable elements of input and output prices. These include allocative adjustments for the differences between social and private costs (SC) and benefits (SB) and implicit or explicit distributional subsidies to, or taxes on, inputs and outputs (DSI, DTI, DSO, DTO).

Table 1.

Public Enterprises: Output and Input Prices

article image

When public enterprises compete with other purchasers in markets for resource and material inputs and must therefore pay at least the going market prices, the only way in which the government could influence a public enterprise’s decision with regard to the hiring and mix of these inputs would be to impose a tax or charge on its payments for the input or to subsidize its use by the enterprise. However, since the public enterprise is a price taker, the incidence of the tax or subsidy will be on the profits of the public enterprise. When the markets for inputs are competitive, the only other feasible variations from the market price are a subsidy applying either to all enterprises or to only publicly owned enterprises and a tax applying to all resource inputs. In the absence of such taxes or subsidies, the price paid by the enterprise will equal the input’s marginal opportunity cost (i.e., IP = MPI).28

Similar constraints affect the determination of the prices of outputs by public enterprises. A public enterprise whose output markets were export oriented would not be able to impose a social or distributional charge in addition to the competitive world price or the world cartel price since buyers would not be willing to pay these charges and would switch to other sources of supply. Furthermore, there is no obvious reason why an exporting country would want its enterprises to attempt to subsidize foreign buyers either because of distributional or social externality considerations. Therefore, neither subsidy nor tax elements are likely to enter into output pricing considerations when the markets are predominantly foreign, and the output price would equal the world competitive or cartel price.

If a public enterprise’s output is oriented mainly toward domestic markets, and if the public enterprise faces either foreign or domestic competition, in the absence of protective tariffs many of the same considerations apply. Although a public enterprise could presumably implement a subsidy by charging a lower price than its competitors, this would not be a particularly effective subsidization technique since the subsidy would accrue only to the customers of the public enterprise and would result in a lower return to the enterprise. Furthermore, in the absence of a rationing mechanism, such subsidies would be open-ended and could lead to high levels of subsidization without any guarantee that the public enterprises’ customers are the members of society that the government wishes to subsidize. Finally, in the absence of a budgetary grant such a subsidy would result in a reduction in profits and would represent a departure from marginal cost pricing for at least one input, namely, equity capital.

If outputs are sold primarily in domestic markets, and if the enterprise is a monopoly, social and distributional charges or subsidies can be implemented by varying output prices from opportunity cost levels. However, if the financial effects of these variations are not explicitly budgeted, they would directly affect the profitability of the enterprise in a perverse manner. For example, it would make little sense for a social charge levied on an undesirable or luxury product of an enterprise to result in higher profits to the enterprise. Consequently, even when tax and subsidy policies can be effected through the price mechanism atone, there is still a strong case for taking them explicitly into the budgetary and tax process.

In summary, the discussion indicates that in the absence of specific reasons to the contrary, public enterprises should expect to operate profitably and should price their outputs and purchases accordingly. Direct government regulation of prices would seem appropriate in only limited circumstances, primarily when monopoly elements are present or when existing prices are inappropriate. In general, explicit taxes and subsidies are the preferred instruments for achieving allocative, distributional, and stabilization objectives. Finally, external constraints will in many circumstances limit the effectiveness of using public enterprise prices to achieve noncommercial objectives. However, these practices are rarely observed which suggests that either public enterprises themselves have not achieved efficient internal operations, or that governments may not be pursuing rational tax, subsidy, or pricing policies with regard to public enterprises, or both.


One of the most striking aspects of public enterprises is the general lack of information about them. Governments often do not have balance sheets, operating statements, or other current data on their operations. Even when some data are available, they are often not comparable among various public enterprises, Little information is available outside individual countries. For example, in the Government Finance Statistics Yearbook, 1981 there are no data for any country on transfers from government to public enterprises, and useful data on transfers from public enterprises are limited to about 30 countries.29 International Financial Statistics publishes data on credit to public enterprises for only about ten countries. The general lack of information greatly restricts the ability of the government to exercise effective control over individual enterprises and especially to analyze their collective impact on macroeconomic aggregates.

There is no clear explanation for the lack of information. However, two possibilities arise. First, the growth of the public enterprise sector and the proliferation in the number of public enterprises has often occurred in an ad hoc manner such that the desirability of information on their operations may not have been initially evident. Consequently, few efforts would have been made to develop information sources. Second, in many cases private enterprises were transformed into public enterprises when governments bought controlling interests from shareholders of existing joint stock companies. In cases of public enterprises organized as joint stock companies, there may be legal barriers to the provision of information. Furthermore, the government’s links to the actual management of such enterprises may be somewhat ill defined. Often the only link is the government representatives on the Board of Directors of such companies. Typically these representatives themselves may not even be aware of the needs of the government for various statistical information.

There have been no surveys to date reviewing the routinely available information on public enterprises and their operations in a broad range of countries.30 There have been a few country-specific studies that have assembled some available data, but these have been entirely ad hoc and follow no consistent pattern. This almost total lack of generally available data has doubtlessly hampered efforts to undertake research in almost any aspect of public enterprises’ economic impact. This is probably particularly true when data on more than one public enterprise would be required.

The data requirements of governments and other users interested in the analysis of the economic impact of public enterprises would seem to fall into essentially two basic categories. To the extent that specific objectives are assigned to certain enterprises and to the extent that the government wishes to exercise control over some or all activities of particular enterprises, data of a microeconomic nature would be required. This suggests that the data required might be rather more detailed than merely broad aggregates. However, for purposes of macroeconomic analysis, such micro-oriented data from individual enterprises would be of interest only if it had a significant macroeconomic impact. This suggests that, in such cases, microeconomic data would be of interest only for a few large enterprises. For example, an agricultural marketing board might provide large subsidies to agricultural producers, food marketing enterprises might provide large subsidies to consumers, and oil or other natural resource companies’ operations may involve the collection of a natural resource rent. Any of these may have significant effects on the macroeconomic aggregates, and data users may have an interest in more micro-oriented data than might ordinarily be expected in order to analyze microeconomic variables with macroeconomic impacts.

More generally, for macroeconomic analysis, interest would focus on data that would be useful as inputs to formulating and evaluating macroeconomic policies. The statistics of interest would not necessarily deal with the operations of any one particular enterprise, and their preparation would entail compilation of bits of data from virtually all public enterprises. Such information would include various aggregates such as the generation and usage of foreign exchange by the public enterprise sector, the projected and actual net financial flows between the government and the public enterprise sector, bank credit to the sector, and various other aggregates relevant to macroeconomic policymaking and evaluation. If such indicators were available, they would permit some assessment of the overall impact of the various government policies vis-à-vis the public enterprise sector, such as constraints on output pricing adjustments. Clearly, the compilation of such data would be a task well beyond mere ad hoc occasional efforts, and the unavailability and importance of such statistics suggest a need to establish an information reporting system that would facilitate the collection and compilation of data on an ongoing basis in many countries.

An information system would aid the authorities in formulating and evaluating economic and financial policies by regularizing the flow of information to the analytical staffs responsible for studying economic conditions and advising policymakers. Any information reporting system would need three main elements: (1) the sources from which data may be obtained for processing; (2) an agency with responsibility for processing the data into aggregates relevant to macro-policymaking purposes; and (3) the recipient users of the data. In addition to the fairly standard data, such as the level of profits or losses and the various rates of return to various capital and equity portions, other statistics that would be useful as inputs to formulating and evaluating macroeconomic policies should be generated routinely by the system along with possible interpretations of their implications. The generation of such data should result in substantially enhanced use of available information by enabling policymakers to shift from their present focus on the microeconomic measurements relating to the internal efficiency of particular enterprises and to focus in addition on the macroeconomic impact, including efficient mobilization and utilization of resources for development.

In summary, for various reasons little information is available on the activities of public enterprises in most countries. This merely compounds the other difficulties governments encounter in their efforts to analyze and control public enterprises. It is important, therefore, that information systems be developed in countries to generate routine and timely data for use in evaluating both the micro- and macroeconomic impact of public enterprise operations.


Perhaps no other issues surrounding public enterprises are of more direct relevance to macroeconomic and financial analysis than are those concerning the Finances of public enterprises. The range of financing issues is broad, and it is useful to divide them into essentially three groups for purposes of this paper. First, there are issues that relate primarily to the initial financing of an enterprise or to the subsequent financing of major projects or investments of the ongoing operations of enterprises. Second, some issues are of interest primarily because they may have implications for credit creation, monetary expansion, and indirectly for the balance of payments. Third, some issues are of interest because they may have direct implications for the balance of payments.

The question of how an enterprise is initially financed revolves mainly around whether equity or debt financing is employed. In the broadest sense capital from all sources and in any form is essentially homogeneous and pricing policies should be established to yield a return equal to its opportunity cost. Any differences between the returns to different forms of capital should reflect identifiable causes. For example, the financial return to equity capital might include compensation to the shareholders for their greater exposure to risk or their ownership rights to resource rents that would not be reflected in the return-to-loan capital. Nevertheless, public enterprises may, and often do, prefer equity capital because they look upon it as entailing little or no cost.31 However, there are more practical differences between debt and equity capital that may influence the choice between the two. In particular, there is a legal obligation in the case of debt to pay the interest due (although this is often not rigorously enforced) which in turn requires the generation of a positive cash flow from the enterprise’s operations. This suggests that for public enterprises in construction or in gestation and for some unprofitable enterprises, in particular those in decreasing cost industries, there may be a preference for equity financing from the point of view of cash flow requirements. However, the preference should not be interpreted as justifying lower prices for the eventual output than would result from appropriate costing of all capital.

There may also be some preference by public enterprise managers for equity capital in cases of enterprises whose operations are unprofitable as a result of deliberate government policy to subsidize outputs or inputs. By avoiding the necessity of generating cash for interest payments, the enterprise can more easily finance any subsidy. However, there are caveats to this conclusion. Unwanted and perhaps even unknown subsidies may also develop and be masked by financing practices that permit low or negative profits. Indeed, these practices may provide welcome excuses to the managers of inefficient enterprises to cover their own shortcomings. Consequently, when such practices are employed, it is especially important for the government to evaluate fully the desirability and extent of every subsidy and its costs in terms of forgone profits or interest payments so that the level of subsidy is not de facto left to the enterprises’ managers or to the whims of the market.

There are other arguments over the debt-to-equity ratio that are generally spurious. First, there is no uneven economic burden on an enterprise with a relatively high debt-to-equity ratio due to high interest expenses. This misconception results from failure to consider both interest payments and dividends as payments to capital. This also suggests that variations in the dividend rates of different public enterprises resulting solely from variations in the debt-equity ratio are not necessarily meaningful comparisons. Finally, in the case of private enterprises it is sometimes useful to manage the debt-equity ratio in order to preserve the control of certain equity shareholders. However, with public enterprises, both equity and loan capital are usually provided by the government or by other government-controlled enterprises, such as public financial enterprises. In such cases, preservation of equity shareholder control has little significance.

Closely related to debt versus equity issues is the question of the appropriate distribution of profits, if any, to the government in its role as stockholder. Assuming that public enterprises’ profits should be subject to the same taxes as those of private enterprises, and given the previous discussions with regard to pricing, and investment planning, an appropriate dividend policy, as well as some inappropriate ones, is clear.32 The amount of dividend will be the result of pricing policies that reflect opportunity costs and other government objectives, approved investment expenditures of the enterprises, and their needs for working cash balances. Thus, an appropriate and consistent dividend policy will not imply that a fixed proportion or absolute amount of profits should be transferred to the government from every profitable public enterprise. Indeed, dividends may not even be necessary in certain circumstances. Profits and depreciation allowances may be left with an enterprise if it has investment projects approved by the government on which these funds can be disbursed in the reasonably foreseeable future. However, if the profits of an enterprise (plus the cash equivalent of depreciation allowances) are more than the expenditures expected on approved projects in the reasonably near future, then the excess profits would be an initial approximation to a reasonable dividend. Nevertheless, in order to enhance incentives to efficient managerial practices, consideration should be given to allowing firms to retain some profits within the overall framework of national priorities. However, there are two strong exceptions to these remarks. To the extent that public enterprise profits represent natural resource rents or monopoly profits, it is reasonable to expect that this should be transferred fully to the government.

Turn now to the second category of financial issues—those that are of interest because they have implications for credit creation, monetary expansion, and, indirectly, the balance of payments. To the extent that public enterprises employ bank credit for normal commercial operations, their credit demands may have as much merit as those of private enterprises. The substantive questions of merit revolve around whether public enterprises have any easier access to credit than do private enterprises and whether credit facilities are made available for noncommercial purposes. It is almost axiomatic that for numerous reasons public enterprises probably do have somewhat easier access to credit than do private enterprises. Even when the government does not explicitly guarantee a public enterprise’s debt, banks almost certainly regard a public enterprise’s borrowings as having at least the implicit guarantee of the government. Much the same point could probably be made with regard to suppliers’ credits. Furthermore, there are frequent examples of bank financing being provided to cover the operating losses of public enterprises which result from the pursuit of noncommercial objectives. And, furthermore, in many countries the financial institutions themselves are public enterprises that are consequently likely to be particularly sympathetic to the credit requests of other public enterprises. In some cases, this de facto credit rationing in favor of public enterprises may distort the allocation of national savings such that public enterprises are able to obtain credit for uses with lower social returns than alternative private sector activities.

Two essential questions appear to summarize the issues with regard to credit granted to public enterprises in excess of its requirements for normal commercial operations. First, is any excess a reflection of the pursuit of deliberate, government-endorsed noncommercial objectives or is it a reflection of inefficiency in the enterprise? Second, what are the implications of the excess for credit creation and availability and how does it differ according to whether the credit is provided to the enterprise or to the government? The First questions require an effort to identify and to qualify the financial impact of the government’s noncommercial objectives, and to access residually the inefficiency of its commercial operations. The second questions will in a large measure depend on how the central bank acts to accommodate the borrowing of public enterprises and what credit instruments may be rediscounted. Such policies would vary from country to country and from time to time, but in general one might expect that debt of public enterprises is less likely to be rediscountable (and, hence, less likely to be monetized) than the direct debt instruments of the government. This suggests that, at least in the short run, credit demands of public enterprises in excess of their normal operating requirement are more likely to result in crowding out private sector borrowers and in increases in the costs to private borrowers than in expansion of total credit.

In the longer run, the result is less clear but seems more likely to be reversed. It is unlikely that financial institutions would be willing or able to carry in their asset portfolios constantly increasing and/or unrepayable debts of public enterprises. The viability of the financial institution requires that these debts be repaid. However, the debtor enterprises would be able to repay only with the assistance of the government which in turn will increase its use of bank credit through either deposit drawdowns or direct borrowing. In the latter case the conversion from credit to the enterprises to credit to the government involves potential monetary expansion if a rediscountable debt instrument is issued.

The final category of issues related to financing includes issues that are of direct relevance to the balance of payments. Essentially two distinct issues appear relevant. First, there is the potential impact on the balance of payments of public enterprises’ borrowing from either domestic or foreign sources and their access to development financing. Second there are the foreign exchange resources that might be potentially available from a public enterprise’s exploitation of some form of monopoly power or rents, usually natural resource- rents.

Public enterprises can and often do borrow substantially in international capital markets. According to partial data compiled by the International Bank for Reconstruction and Development (IBRD), gross borrowings in international markets by public enterprises totaled $25.0 billion in 1978.33 However, even these data are not comprehensive; they include only borrowings in the form of foreign bonds, Eurocurrency credits, and bank lending. Loans from international agencies, governments (including on-lending), and suppliers’ credits are not included. Nevertheless, these data are believed to represent about 70 percent of gross foreign borrowings of public enterprises.

Two types of problems may be encountered in regard to foreign borrowing. First, these borrowings have been increasing, and while by and large they are intended to finance expansions of new projects, there may be cases where the proceeds are used for current purposes, The implications of such borrowing are clear in that the latter type tends to be inflationary. Second, the controls on these would appear to be inadequate in some cases. For example, in Portugal, public enterprises do not need the approval of the Ministry of Finance and can borrow after informing the central bank. Variations of this are to be found in other countries. In Ivory Coast, although regulations require consultation it would appear that the relevant authorities (the Cause Autonome d’Amortissement) have not been very effective in preventing borrowing from external sources without its approval. It would also appear that requirements for prior approvals have in some cases given rise to perverse effects. For example, in Mexico, state enterprises placed orders with foreign suppliers (by resorting to suppliers’ credits) rather than with domestic suppliers in part because less time was required. It is therefore essential that budgetary and other mechanisms be strengthened to permit an overall coordination and at the same time prevent unintended adverse consequences.

The second set of issues relating directly to the balance of payments pertain to public enterprises that generate foreign exchange through foreign sales. However, the normal commercial results are of less interest than situations in which the exporting public enterprise generates substantial amounts of foreign exchange through exercise of some form of monopoly power or the exploitation of natural resource rents or both. The most obvious examples are, of course, the various oil exporting and other natural resource producing and exporting companies, many of which are fully or partially government owned. The issue in such cases is the appropriate distribution of foreign exchange among the enterprise, the government, and any private sector shareholders. While private sector owners certainly have a claim to a proportional share of all profits, there is no obvious reason that they should have an a priori claim to any earned foreign exchange. Of course, the exporting enterprise itself might be given some preferential allocation of foreign exchange if it were believed necessary to ensure that its operations could be continued. However, this does not imply that it or any other exporting enterprise should be permitted to retain foreign currency balances in excess of normal commercial requirements. Consequently, one could ordinarily expect that essentially all foreign exchange proceeds would be converted into local currency through official channels.

To summarize, as long as all forms of capital yield their opportunity cost, the choice between debt and equity financing is of secondary importance. Bank credit to public enterprises is of concern primarily insofar as it exceeds normal commercial requirements. In the short run such credit is more likely to crowd out other potential borrowers, but in the long run it seems more likely to be inflationary. Furthermore, public enterprises are often large borrowers in international markets, and more effective controls on these borrowings seem desirable. Finally, whenever a public enterprise generates foreign exchange earnings, balances in excess of normal commercial requirements should be converted into local currency through normal channels.


Until recently public enterprises have generally been viewed as individual entities that had no particular role as a stabilization policy instrument in unplanned economies. In reality, this view had little merit since public enterprises’ pricing and employment policies were frequently manipulated by policymakers, although usually with more concern for political tenure than for macroeconomic stabilization. With the growth of the public enterprise sector in many developing countries, there has been an increasing awareness that public enterprises collectively can, and often do, affect the economy as a whole either as a result of government policies vis-à-vis the sector or because a few enterprises are relatively so important in the economy. This has given rise to questions concerning the appropriate policies for governments to follow with regard to the public enterprise sector and has raised the question as to what issues are of relevance to macroeconomic policymakers?

This paper has reviewed a variety of issues concerning public enterprises and their financial and macroeconomic impacts. No effort has been made to resolve those issues which remain the subject of controversy. However, where a general consensus has been reached, at least in theory, on an issue (e.g., marginal cost pricing), the results of that consensus have been presented. No effort has been made to this point to assess the relative importance of any of the issues reviewed. However, it would seem useful to conclude the paper with at least the personal assessment of the author as to the relative importance of these issues.

Progress is probably most advanced in the area of pricing issues. At least in theory the appropriate procedures are clear. Indeed, there is even a growing expertise in the quantification of allocative and distributive adjustments. Nevertheless, probably the single most important outstanding issue of pricing, and an issue that should receive more attention in the future, is that of what techniques should be applied in practice over a broad spectrum of public enterprises in order to approximate the marginal cost pricing objectives. Closely connected with these would be techniques to quantify the financial costs of the government’s noncommercial objectives. Insofar as the latter may facilitate use of lump-sum limits on costs, the problems of the unit price subsidy (or tax) element may be reduced.

Probably the issues of highest priority for future work are those concerned with channels of control and information flows. Governments are increasingly being forced to accept the need to impose more financial discipline on their enterprises if for no other reason than to avoid undesirable results in financial markets and on macroeconomic variables. It is clear that existing systems have not coped well with these needs and it is also clear that governments generally have little understanding of either the policies they wish to pursue or the information they require to evaluate them, or the macroeconomic variables that will be affected by these policies. Until these rather basic, if somewhat institutional, matters are resolved, advances in more technical issues are unlikely to yield any significant improvement in the macroeconomic impact of public enterprise operations.


  • Cassese, Sabino, “Public Control and Corporate Efficiency,” in State-Owned Enterprise in the Western Economies, ed. by Raymond Vernon and Yair Aharoni (New York: St. Martin’s Press, 1981).

    • Search Google Scholar
    • Export Citation
  • Chu, Ke-Young and Andrew Feltenstein, “Relative Price Distortions and Inflation: The Case of Argentina, 1963–76,” Staff Papers, International Monetary Fund (Washington), Vol. 25 (September 1978), pp. 45293.

    • Search Google Scholar
    • Export Citation
  • Floyd, Robert H., “Equivalence of Product Tax Changes and Public Enterprise Price Changes,” Staff Papers, International Monetary Fund (Washington), Vol. 28 (June 1981), pp. 33874.

    • Search Google Scholar
    • Export Citation
  • Floyd, Robert H., “Some Aspects of Income Taxation of Public Enterprises,” Staff Papers, International Monetary Fund (Washington), Vol. 25 (June 1978), pp. 31042.

    • Search Google Scholar
    • Export Citation
  • Foster, C. D., Politics, Finance, and the Role of Economics: An Essay on the Control of Public Enterprise (London: Allen and Unwin, 1971).

    • Search Google Scholar
    • Export Citation
  • Gillis, Malcolm and Charles E. McLure, Jr., “Standards of Conduct for Public Enterprises,” in Richard Musgrave, Fiscal Reform in Bolivia: Final Report of the Bolivian Mission on Tax Reform (Cambridge: Harvard University, 1981), pp. 54649.

    • Search Google Scholar
    • Export Citation
  • Heald, David, “The Economic and Financial Control of U.K. Nationalised Industries,” Economic Journal (London) (June 1980), pp. 24365.

    • Search Google Scholar
    • Export Citation
  • Henderson, A. M., “Prices and Profits in State Enterprise,” The Review of Economic Studies (1948–49), Vol. XVI (1), No. 39 (Clevedon, Avon, United Kingdom), pp. 1324.

    • Search Google Scholar
    • Export Citation
  • India, Government of, Bureau of Public Enterprises, Ministry of Finance, Performance Aims and Financial Targets of Central Government Public Enterprises 1982–83 (New Delhi: April 1982).

    • Search Google Scholar
    • Export Citation
  • International Bank for Reconstruction and Development, Borrowing in International Capital Markets, May 1980 (Washington: IBRD, 1980).

  • International Monetary Fund, Government Finance Statistics Yearbook, Vol. V, 1981 (Washington: IMF, 1982).

  • Jenkins, Glenn P., “An Operational Approach to the Performance of Public Sector Enterprises,” Harvard Institute for International Development, Development Discussion Paper No. 47 (November 1978).

    • Search Google Scholar
    • Export Citation
  • Jones, Leroy P., Public Enterprise and Economic Development: The Korean Case (Seoul: Korea Development Institute, 1975).

  • National Economic Development Office, A Study of U.K. Nationalised Industries, Background Paper No. 2, Relationships of Government and Public Enterprises in France, West Germany, and Sweden (London: National Economic Development Office, 1976).

    • Search Google Scholar
    • Export Citation
  • Pryke, Richard, “The Comparative Performance of Public and Private Enterprise,” Fiscal Studies, Vol. 3 (Oxford: July 1982), pp. 6881.

    • Search Google Scholar
    • Export Citation
  • Pryke, Richard, The Nationalised Industries: Policies and Performance Since 1968 (Oxford: Martin Robinson, 1981).

  • Sheahan, John B., “Experience with Public Enterprise in France and Italy” in Public Enterprise: Economic Analysis of Theory and Practice, ed. by William G. Shepherd and Associates (Lexington: D.C. Heath and Company, 1976).

    • Search Google Scholar
    • Export Citation
  • Shepherd, William G., “Objectives, Types, and Accountability” in Public Enterprise: Economic Analysis of Theory and Practice, ed. by William G. Shepherd and Associates (Lexington: D.C. Heath and Company, 1976).

    • Search Google Scholar
    • Export Citation
  • Sweden, Government of, Department of Industry, State Business; Public Enterprise Experience in the EEC, prepared by Metra Oxford Consulting (Oxford: 1978).

    • Search Google Scholar
    • Export Citation
  • Turvey, Ralph, Economic Analysis and Public Enterprises (Totowa, New Jersey: 1971).

  • Vernon, Raymond and Yair Aharoni, State-Owned Enterprise in the Western Economies (New York: St. Martin’s Press, 1981).


Mr. Floyd is a Senior Economist in the Middle Eastern Department of the International Monetary Fund.


For the purposes of this paper, a public enterprise may be defined as any government-owned or controlled unit that produces and sells industrial, commercial, or financial goods and services to the public. Although a more rigid definition seems both elusive and undesirable in view of the variety of legal and organizational forms encountered in various countries, in order to distinguish public enterprises more clearly from governmental and quasi—governmental bodies, such as school districts and charitable organizations, it is perhaps useful to require that the revenues of public enterprises should be more or less related to their output and that at least some day-to-day-operational autonomy should be in the hands of the managers of the enterprise rather than the ministerial authorities.


Issues may also be viewed in the context of the time frame in which they may be relevant. Issues could be categorized by whether they have long-term or ongoing importance, or are relevant mainly in the medium run or the short run. Issues of continuing interest focus primarily on the institutional arrangements for determining and implementing government policy and on data indicating the direct impact of the public enterprise sector on domestic and external financial markets. Furthermore, in the longer run, not only the level of financial variables but also their structure and the appropriateness of the government’s general policies toward public enterprises with regard to allocative and distributional objectives (and in some cases even a few significant individual enterprises’ practices, e.g., energy pricing) should be taken into account. It is in this context that the various supply-side effects of public enterprises become most important. Issues of short-term relevance would include primarily those related to the impact of public enterprise operations on stabilization. These could include the direct use of public enterprises for noncommercial objectives such as price stabilization, employment, and a more favorable balance of international payments.


John B. Sheahan. “Experience with Public Enterprise in France and Italy.” in William G. Shepherd and Associates, ed., Public Enterprise: Economic Analysis of Theory and Practice (Lexington: D.C. Heath and Company, 1976), p. 176.


For a discussion of the importance of specifying objectives see Leroy P, Jones, Public Enterprise and Economic Development: The Korean Case (Seoul: Korea Development Institute, 1975), pp. 140–52.


The importance of determining financial objectives and assessing financial performance of individual public enterprises is discussed in Glenn P. Jenkins, “An Operational Approach to the Performance of Public Sector Enterprises,” Harvard Institute for International Development, Development Discussion Paper No, 47 (November 1978).


For a more detailed discussion see David Heald, “The Economic and Financial Control of U.K. Nationalised Industries.” Economic Journal (June 1980), pp. 245–65.


In India, lump-sum targets are used for projected losses, but rates of return are used for projected profits. See Performance Aim and Financial Targets of Central Government Public Enterprises 1982–83 and 1983–84, Government of India, Bureau of Public Enterprises, Ministry of Finance (New Delhi, April 1982).


X-efficiency refers essentially to the quality of managerial decisions and actions. The internal operations of public enterprises would be X-efficient if it were not possible to increase profits through reductions of inefficient managerial practices, such as employment of excess labor or payment of wages in excess of levels prevailing in the private sector. Such practices are thought to be common, especially in public enterprises in many developing countries.


It could be argued that managers would prefer financial targets so long as traditional managerial variables, for example, input and output prices and output levels, are left lo management’s discretion. In contrast, if they are fixed externally, managers are likely to prefer pricing targets to possibly unachievable financial targets.


For a discussion of some aspects of these problems, see William G. Shepherd. “Objectives, Types, and Accountability” in William G. Shepherd and Associates, ed., Public Enterprise: Economic Analysis of Theory and Practice (Lexington: U.C. Heath and Company, 1976), pp. 33–47; and C.D. Foster, Politics, Finance and the Rote of Economics: An Essay on the Control of Public Enterprise (London: Allen and Unwin. 1971).


For a complete description of French control arrangements see National Economic Development Office, A Study of UK Nationalised Industries, Background Paper No. 2, Relationships of Government and Public Enterprises in France, West Germany and Sweden (1976).


In 1967, a working group prepared a study (the Nora Report) of the control apparatus which has had considerable influence over French control efforts for several years. The Nora Report essentially advocated clear definitions of, and distinctions between, the responsibiiities of the public enterprise, its technical or sponsoring ministry, and the Ministry of Economy and Finance.


National Economic Development Office, op. cit. It should be noted, however, that others have questioned the extent of progress. Cassese states that, “In France, the Nora Report proposed reversing ‘the burden of proof,’ requiring the government financially to motivate and compensate the public enterprises. However, in France, as in Italy, the compensation mechanism has been purely theoretical.”See Sabino Cassese, “Public Control and Corporate Efficiency,” in State-Owned Enterprise in the Western Economies, ed., by Raymond Vernon and Yair Aharoni (New York, 1981), p. 155.


For further discussion see Sweden, Department of Industry, State Business; Public Enterprise Experience in the EEC, prepared bv Metra Oxford Consulting (Oxford, 1978), pp. 168–70.


The contrats de programme were also aimed at measuring managerial efficiency through inclusion of a comprehensive indicator of performance referred to as the productivité globâle des facteurs.


Such contracts were negotiated with the railways and the electrical industry in 1969. By 1977 all of these agreements had lapsed. However, four new agreements were signed in 1979, and in 1981 the Government announced its intention of expanding the use of contracts. Many difficulties were encountered with these contracts, but most essentially reflected difficulty by both the Government and the enterprises in defining, quantifying, and abiding by long-term commitments to provisions relating to the various objectives of each party. For example, enterprise managers would find it difficult to agree to estimates of a wage hill projection in the absence of agreement from often powerful labor unions. Apparently, the Government also had similar problems since it retained the right to alter price and investment decisions as necessary to regulate the economy. Recently, the Government of Senegal has begun to use contracts with some of its public enterprises, referring to them as contrats-plan.


See Richard Pryke, “The Comparative Performance of Public and Private Enterprise,” Fiscal Studies, Vol. 3 (July 1982), pp. 68–81; also Richard Pryke, The Nationalised Industries: Policies and Performance Since 1968 (Oxford: Martin Robinson, 1981).


See William G. Shepherd and Associates, op. cit., p. 44.


Cash limits have been criticized as encouraging the use of self-finance, possibly with the result that unjustifiable investments might be undertaken when self-financing is available and economically justifiable investments forgone when self-financing is not available. However, there is no a priori reason to assume that improper investment decisions necessarily follow appropriate financial decisions. Indeed, the two should be essentially independent with each public enterprise’s investment plans being made in comparison with macroeconomic constraints and with the competing demands for investment funds from other public enterprises.


Any use of resources by public enterprises involves a cost in terms of forgone opportunities for potential alternative private and public uses of these resources; these costs are often referred to as social opportunity costs. The opportunity cost of any resource input is defined to be its value in its next best alternative use.


Ralph Turvey, Economic Analysis and Public Enterprises (Totowa, New Jersey, 1971), p. 74.


A more exhaustive catalog of difficulties that are encountered when attempting to apply marginal cost pricing in practice can be found in A.M. Henderson, “Prices and Profits in State Enterprise,” The Review of Economic Studies (1948–49), Vol. XVI (1) No. 39, pp. 13–24.


Henderson, in proposing a similar pricing technique, described the task of choosing an alternative to marginal cost pricing as “… an ungrateful task, for there is no possibility of being right, it is merely a matter of producing an acceptable compromise which shall not be too grossly wrong.” See Henderson, op. cit., pp. 20–21.


For overall profitability the measurement that is relevant for economic purposes would not distinguish between various legal forms that investment capital might take. All capital should be expected to earn its opportunity cost, and consequently, as an initial approximation the relevant measure of income includes all forms of return to capital, in particular both profits (either distributed or retained) and interest payments for long-term borrowings. Similarly, the relevant measurement of capital includes all long-term assets, including all equity (including retained earnings) and long-term borrowings. Profits should be defined after deduction of replacement cost depreciation allowances but before income tax payments. Capital should be measured after asset and liability revaluation where necessary to compensate for inflation and after deducting cumulative depreciation allowances. A more substantive discussion of relevant measurements of payments to various forms of capital can be found in the second paper in this volume, “Toward a Conceptual Framework for Macroeconomic Evaluation of Public Enterprise Performance in Mixed Economies,” by Clive Gray.


Chu and Feltenstein have estimated that in Argentina government transfers to cover public enterprise losses were proportionately ten times as inflationary as the financing of private enterprises’ losses through commercial bank borrowings primarily because it is assumed that only in the former case are the losses translated into high-powered money through central bank financing of the government deficit. See Ke-Young Chu and Andrew Feltenstein, “Relative Price Distortions and Inflation: The Case of Argentina, 1963–76,” International Monetary Fund, Staff Papers, Vol. 25 (September 1978), pp. 452–93. Another aspect of stabilization policy is the tendency in some countries to use public enterprises as an employer of last resort of unnecessary or unqualified labor. Such policies would have an adverse impact on public enterprises’ finances and would lead to inefficient production of goods and services.


Robert H. Floyd, “Equivalence of Product Tax Changes and Public Enterprise Price Changes,” International Monetary Fund, Staff Papers, Vol. 28 (June 1981), pp. 338–74.


A matrix similar to that depicted in Table 1 was developed by Malcolm Gillis and Charles E. McLure, Jr., in “Standards of Conduct for Public Enterprises,” in Richard Musgrave, Fiscal Reform in Bolivia: Final Report of the Bolivian Mission on Tax Reform (Cambridge: Harvard University, 1981), pp. 546–49. It has been altered as necessary to conform to the purposes of this paper.


The input pricing relationships in Table 1 do not reflect situations in which the public enterprise docs not employ opportunity cost pricing. Suppose that a public enterprise chooses to pay workers a higher wage than they could earn in alternative private employment. In such a case the relationship would be IP = MPI + SI where Sl is a subsidy to workers. With mobile factors that are not fully employed it is possible that the enterprises may pay less than the market price (IP<MPI) because the latter exceeds the factor’s opportunity cost (MPI>MOC).


International Monetary Fund, Government Finance Statistics Yearbook, Vol. V, 1981 (Washington, D.C., 1982).


The difficulties encountered in assembling and using data on public enterprises have been noted in the paper in this volume by R.P. Short, “The Role of Public Enterprises: An International Statistical Comparison.”


See the observations concerning the attitudes of public enterprise managers discussed in Raymond Vernon and Yair Aharoni, State-Owned Enterprise in the Western Economies (New York: 1981), p. 16


Robert H, Floyd, “Some Aspects of Income Taxation of Public Enterprises,” International Monetary Fund, Staff Papers, Vol. 25 (June 1978), pp. 310–42.


IBRD, Borrowing in International Capital Markets, May I980 (Washington. D.C, 1980).