Toward a Conceptual Framework for Macroeconomic Evaluation of Public Enterprise Performance in Mixed Economies
  • 1 0000000404811396 Monetary Fund
  • | 2 0000000404811396 Monetary Fund
  • | 3 0000000404811396 Monetary Fund


Considering that operating deficits of public enterprises are a major contributor to fiscal imbalance and excessive credit creation in many countries, especially but not only developing nations, it would seem that the macroeconomic impact of the public enterprise sector has hitherto received less attention than it deserves. The primary objectives of the present paper are to stimulate increased focus on this issue and to propose a conceptual framework within which analysis of individual country situations might proceed. The framework stops short of examining alternative remedies that might be proposed by local or outside analysts in response to a given situation of imbalance.

Considering that operating deficits of public enterprises are a major contributor to fiscal imbalance and excessive credit creation in many countries, especially but not only developing nations, it would seem that the macroeconomic impact of the public enterprise sector has hitherto received less attention than it deserves. The primary objectives of the present paper are to stimulate increased focus on this issue and to propose a conceptual framework within which analysis of individual country situations might proceed. The framework stops short of examining alternative remedies that might be proposed by local or outside analysts in response to a given situation of imbalance.

The quintessential policy issue surrounding the role and size of the public enterprise sector in any country is considered in the paper to be the intrinsic social efficiency of government in organizing scarce factors of production to produce goods and services for sale. A separate, and perhaps slightly secondary issue of policy, is the public enterprise sector’s efficiency as a vehicle for accomplishing noncommercial objectives of government. The paper argues that assessment of sectoral efficiency in production presupposes careful netting out of costs and revenue losses attributable to pursuit of noncommercial objectives. These are classified into five categories—economic stabilization, economic growth, income redistribution, localization/indigenization, and miscellaneous. Issues of how to assess the public enterprise sector’s efficiency in pursuing the objectives, and whether public enterprise is the only or even the most efficient vehicle for doing so, are considered in an initial appendix.

By providing a macroeconomic picture of the comparative efficiency, and consequences of any comparative inefficiency, of the public enterprise sector, information on its fiscal and monetary impact will form an important and relatively new input into policy discussions within many governments concerning sectoral reform. It will open up a wider range of reform options than is suggested in the bulk of the current literature on public enterprise, stressing as it does micro approaches to improvement of management efficiency and application of social benefit-cost calculus to appraisal of public enterprise investments. At one extreme, some governments may wish to consider compressing the sector by discontinuing or divesting themselves of a certain portion of it, especially in branches of production apart from public utilities and others traditionally regarded as “natural monopolies.”However, it is essential that external analysts and advisors preserve ideological neutrality concerning the intrinsic merits and drawbacks of state ownership of the means of production as opposed to alternative modes of industrial organization, and the present paper adopts such a stance.

Ideally one would assess the public enterprise sector’s macroeconomic impact by comparing observed national income, and fiscal and monetary parameters with those obtaining under a baseline alternative simulated within a general equilibrium framework. However, because of the computational effort involved, such an approach will rarely be feasible; moreover, specification of baseline alternatives cannot avoid value judgments, and is thus a questionable exercise for outsiders. More feasible, and at the same time more acceptable to policymakers, will be a partial equilibrium approach in which observed parameters are compared with criterion values.

The paper argues that the most useful criterion values are those taking as their point of departure zero or “normal” profitability of the public enterprise sector, that is, costs of all factors being covered, including capital, after observed values are adjusted for monopoly profits, border prices of tradables, and costs and revenue losses (net of subsidies) associated with pursuit of noncommercial objectives. This is because the principal burdens imposed by the public enterprise sector at a macroeconomic level in many countries arise from subsidies and credit it receives from government, and credit from the financial sector, in order to compensate for operating losses.

The principal components of the public enterprise sector’s national income, fiscal and monetary parameters are listed and described in Section V. Fiscal parameters include accounting magnitudes—notably taxes, interest, and capital payments from the public enterprise sector to government, unrequited transfers and capital payments from government to the sector—as well as implicit transfers to, and implicit subsidies from government, the latter comprising tax subsidies and a pervasive subsidization of capital. On the monetary side one looks initially at gross use of credit, then goes behind this to measure sectoral impact on money supply, the level of inflation, interest rate structure, and the balance of payments, always paying particular attention to credit creation in compensation for enterprise deficits.


Of such gravity are the shortcomings that have been brought to light by recent performance evaluations of public enterprises in mixed economies, especially developing countries but not limited to them, that one would have expected a significant proportion of the burgeoning literature in this Field to add it all up and assess the impact of the public enterprise sector on various countries’ fiscal and monetary situations in the short run as well as economic growth and fulfillment of other sociopolitical objectives over the medium and long term. Among the parameters that one would want to see measured and evaluated are the shares of the public enterprise sector in nominal and adjusted financial flows from and to the central government, including implicit taxes and subsidies; the proportion of central government deficits accounted for directly and indirectly by public enterprises; the impact of public enterprise operating deficits on credit creation, money supply, inflation, and interest rates; and ultimately, the number of percentage points of growth gained (or sacrificed) as compared with some baseline pattern of industrial organization.

Even if one established, within a reasonable confidence interval, that a country’s budget deficit and money supply would have been X1 and X2 billion currency units less in year t, its inflation rate y percent less in the same year, and its gross domestic product (GDP) growth rate over the past five years z percentage points higher under the baseline regime, it could nevertheless be concluded that these represent costs of achieving certain noncommercial objectives. Public enterprises are, after all, supposed to serve social as well as economic objectives, and a certain increment in inflation or the sacrifice of some GDP growth may be regarded as a necessary, even modest, price to be paid for certain social benefits.

On the other hand, since the launching of the great debate on tradeoffs between (GDP growth and social justice, the evidence generated in many slow-growing and fast-growing economies has suggested that social benefits gained by sacrificing efficiency are often largely offset by social as well as economic losses. Hence, in some countries, one finds a number of policymakers questioning the effectiveness of public enterprises in achieving social and development goals. They look to their staff or outside consultants to measure the resultant social and economic losses and investigate alternative strategies for the future.

The current academic research is concentrating on the methodology of performance evaluation of individual enterprises, the behavior of public enterprise managers, the impact of government policies on the enterprises, and the search for ways and means of making them serve their commercial and noncommercial objectives more efficiently. Analysis of macroeconomic performance would constitute a logical extension of the research, which could then go on to examine the full spectrum of reform possibilities.


The present paper aspires to initiate the process of designing a conceptual framework by means of which government decision makers, and local as well as outside economists advising them, might evaluate the macroeconomic impact or performance of a country’s public enterprise sector by comparing the actual picture with putative outcomes of baseline alternatives. Such comparisons should indicate the magnitude of the problem and provide a basis for determining the policies to be pursued as well as the volume of resources to be enlisted in resolving it. In some cases, only incremental reforms may be suggested, in others, more extreme measures may be indicated.

While strict neutrality on the relative merits of state and private ownership of the means of production should be observed, a full range of policy alternatives on how to reduce a large budget deficit arising from operations of public enterprises in mixed economies should be considered. A reform program might confine itself to changes in management procedures in selected enterprises; it might entail new sector-wide policies in such areas as government control over managerial initiative, pricing of inputs and outputs, minimum financial returns to be required on new public enterprise investments, and for servicing of outstanding government loans; and, at the extreme, it might embrace the premise that no intermediate measures would convey as much social benefit as outright divestiture of certain state holdings.

A necessary disclaimer at the outset is that the present paper does not address the question of remedies for any given shortcomings that may be identified in a country’s public enterprise scenario; rather, it confines itself to indicators and criteria for determining whether remedies need to be sought. What specific remedies to prescribe is an issue that must be faced by the authorities of the country in question, along with any advisers they may choose to consult. Optimal remedies will take into account economic and political factors peculiar to each country’s situation. At the same time, in almost every case, consideration of potential remedies short of divestiture will necessitate resolving one or more of the five categories of issues—determination of objectives, control procedures, pricing, generation of information, and financing—outlined by Robert H. Floyd in the preceding paper in this volume.

Once the conceptual framework has been agreed, consideration must be given to ways and means of measuring the indicated parameters. This involves assessing the data base country by country and determining what resources must be applied, whether purely local or including technical assistance, to expand it. As with other categories of data, country statistics differ widely in their coverage of public enterprises. At one extreme, some governments tabulate operating account and balance sheet data for all public enterprises individually and on a consolidated basis, broken down by sectors and/or categories of enterprises. At the opposite extreme, some act as though government’s stake in the sector imposed no more responsibility for centralized financial reporting than the national statistics office bears vis-à-vis the private sector. In such cases, one may have to consider oneself fortunate to locate such bits and pieces of centralized data (apart from annual reports of the individual enterprises, with the delays to which those are often subject) as central and/or commercial bank claims on public enterprises; transfers to and from government that are identified in the budget—dividends, certain types of tax payments, and debt service in one direction and operating subsidies and capital infusions in the other; and possibly accounts of major public utilities that are sometimes found in national statistical publications.

Just as a series of adjustments are made in raw government finance data in order to compute economically relevant coefficients and indicators with respect to the interaction between government and the economy, so also the public enterprise data will have to be adjusted to give meaningful macroeconomic indicators. Some of these adjustments will not be cut and dried but will require the exercise of judgment, and will be subject to wide margins of error and uncertainty. Such adjustments involve, for example, the estimation of implicit taxes and subsidies associated with alternative pricing policies, the use of border prices to adjust operating accounts, and the attribution of costs to actions in pursuit of social objectives.

In principle every government that maintains public enterprises should be collecting and analyzing such data for itself, eventually sharing it with its international creditors for their own review and interpretation. To encourage this, the creditors should press upon governments the usefulness of the parameters described below as guides to policy, urge them to allocate adequate resources to their measurement, and, where appropriate, arrange technical assistance to support the effort. However, until such initiatives have borne fruit, outside users will have to improvise with incomplete data, seeking as always to equate the marginal returns to time invested in satisfying their wide range of data needs.

The significance of the findings that can be obtained already by analysis of incomplete nominal parameters relating to public enterprise is illustrated in R.P. Short’s concluding paper in this volume. In an analysis comparing data from close to 90 countries there can be no question of adjusting financial profits and losses for implicit taxes and subsidies, or assessing the financial burden of pursuing noncommercial objectives. Nor is it possible to estimate the impact of public enterprise deficits on price levels and external payments. However, in sketching the magnitude of the sector’s deficit, comparing it with overall government deficits, assessing the budgetary burden of public enterprise, and showing how increases in bank credit to the sector have translated one-for-one into increases in total credit outstanding, Short has pioneered in establishing highly suggestive indicators of the magnitude of the macroeconomic problem. His work should serve as an inspiration to national planners to deepen the analysis of their countries’ data and identify the proximate causes of macroeconomic imbalances in public enterprise performance as a prerequisite for the design of appropriate remedies.


The present paper, along with its two companions in this volume, focuses on the macroeconomic performance of the public enterprise sector in mixed economies—that is, economies where nonstate-owned units account for a significant proportion of output by enterprises with more than, say, 5, 10, or 20 employees (whatever standard is used in a given country to define medium-scale producers). Two key characteristics of a mixed economy with reference to the subject of the volume are that (1) the performance of private and, if relevant, other nonstate-owned enterprises provides a basis of comparison with public enterprise performance, and (2) for the production of any given set of goods and services in medium-sized and larger units, it is open to the state to consider one or more forms of industrial organization other than public enterprise, without contravening the underlying national ideology.

Many of the parameters proposed for measurement and evaluation are no less relevant to analysis of public enterprise performance in centrally planned economies. However, the economic environment in such countries, featuring the dominant role of noncommercial objectives, close integration of public enterprise and state finances, and lack of alternatives for comparison with the performance of state-owned units, changes the parameters of evaluation sufficiently to require modifications in the analytical framework.

Mixed economies can be classified in turn among developing and industrial countries. Developing countries have received the preponderant share of attention in the recent literature on public enterprise, and inasmuch as it is fiscal and monetary imbalances in these countries that currently absorb the major share of the International Monetary Fund’s (IMF) resources, inevitably the contribution of the public enterprise sector to those imbalances is currently of greater interest to the IMF than is any analogous phenomenon with respect to the industrial nations. Nevertheless, the analytical framework of the paper applies to both sets of countries, as does R.P. Short’s statistical analysis. Factors that apply preponderantly to developing rather than industrial countries will be so identified in the text.

The literature generally recognizes three conditions as characterizing a public enterprise: (1) government control over the entity, de facto or potential, which need not be synonymous with majority ownership; (2) production of goods and services for sale as the entity’s primary function; and (3) existence of a policy that revenues should cover at least a substantial proportion of costs.1

For its part, the IMF’s Manual on Government Finance Statistics divides the public enterprise sector into two components; (a) public financial institutions and (b) nonfinancial public enterprises.2 With respect to category (a), financial institutions are defined as “enterprises primarily engaged in either acceptance of demand, time, or savings deposits, or in both incurring liabilities and acquiring financial assets in the market”3 if “owned and/or controlled by the government” they are further described as public financial institutions.4

Category (b) is defined as “government-owned and/or -controlled units which sell industrial or commercial goods and services to the public on a large scale.”5 Comparing (b) with the three conditions cited above, the phrase “on a large scale” conveys somewhat imprecisely the previous notion (2) that selling goods and services is the entity’s priman lumi inn. However, the Manual definition would ac -cord public enterprise status to an entity operating under a permanent mandate to furnish its goods and services at less than a “substantial” proportion of cost.

Numerous instances can be cited of enterprises that operate profitably as either public or private entities in some countries while similar enterprises cover as little as half or even less of their financial costs in other countries, yet to avoid the political costs of large-scale displacement of labor the governments in the latter cases maintain the entities in full knowledge that appreciable cost recovery at any future time is a pipe dream.6 It would be inappropriate to exclude such enterprises from the framework of the present paper, which accordingly applies only conditions (1) and (2), namely, (1) government control, and (2) trading as the primary function.

Clearly one must look separately at public financial institutions in contexts that have little to do with the basic policy issues surrounding public enterprise. On the other hand, for some purposes, it is useful to look at both categories of enterprises together. What will be referred to in this paper as the “quintessential” issue of public enterprise—that is, the efficiency of the state in achieving social objectives as entrepreneur and manager of revenue-generating activities—is very much present in government ownership and operation of commercial banks, insurance companies, and certain types of development finance institutions.

Accordingly, with regard to those fiscal indicators not encompassing credit operations of public financial institutions, this paper deals with a public enterprise sector that encompasses both nonfinancial units and a subset of public financial institutions including commercial and cooperative banks, commercial bank finance companies, insurance companies, stockbrokers, and other credit institutions that service a diversified clientele. Virtually the only public financial institutions totally excluded from coverage are the central bank and development finance companies whose sole function is to act as conduits for public credit, foreign or domestic, to a relatively small number of large-scale borrowers, public or private.7

For a discussion of additional issues that arise in applying the definition of public enterprise in concrete situations, and specifically in determining what proportion of shareholding by government qualifies an enterprise as public, the reader is referred to Section 1 of the following paper by R.P. Short.


It is useful to define five successive stages in the process of arriving at a comprehensive assessment of the macroeconomic performance of the public enterprise sector: (1) measurement of nominal parameters characterizing the sector; (2) decomposition of the nominal parameters by attribution to commercial and noncommercial objectives; (3) adjustment of the nominal parameters for implicit taxes and subsidies; (4) comparison of adjusted parameters with criterion values (partial equilibrium approach); and (5) comparison of adjusted parameters with putative or simulated outcomes of baseline alternatives (general equilibrium approach).

Measurement of Nominal Parameters Characterizing the Public Enterprise Sector

The First step in analyzing, for example, the fiscal impact of the public enterprise sector is to measure and record the various nominal financial flows in either direction between the sector and the government. Simple addition yields totals for gross flows in each direction and subtotals for related components, while simple subtraction then gives an aggregate net flow as well as net figures for the various components. The various measures can be divided by other parameters to obtain indicators such as the ratio of gross fiscal transfer to government to value added or capital invested. Findings that Country A’s aggregate net flow in year t amounted to X million currency units from government to the public enterprise sector, and that transfers from the sector to government comprised a gross return of y percent on the replacement value of public investment,8 represent initial crude measures of the sector’s fiscal impact.

Similarly, in analyzing monetary impact, one begins by computing the sector’s net use of credit (loans outstanding to public enterprises less their deposits with financial institutions). Or in studying its impact on social objectives, one looks at whatever indices can be compiled with respect to sectoral employment; purchase of goods and services within a region, including the payroll; local share ownership; etc.

Decomposition of Nominal Parameters by Attribution to Commercial and Noncommercial Objectives

Execution of this second stage in the analysis is indispensable for drawing eventual policy conclusions. The essential task is one of estimating the portion, if any, of each component of public enterprise operating accounts which is attributable to the enterprises’ following government directives in pursuit of noncommercial goals. The residual portion of each parameter is then regarded as corresponding to the enterprises’ pursuit of commercial objectives, and it is these magnitudes that enter into indicators of the sector’s intrinsic social efficiency.

The policy implications of this differentiation may be illustrated as follows: suppose it is established that the public enterprise sector in Country A over the past year contributed X (amount) to the government budget deficit and Y to credit expansion (X being included in, and normally less than Y); the portions of X and Y attributable to noncommercial and commercial objectives are estimated at X1 and Y1, and X2 and Y2, respectively. Suppose it is further estimated that the public enterprise sector must contribute amounts B and C to reduction of the budget deficit and credit expansion, respectively, in a given future period, for which the sector’s contributions to the budget deficit and credit expansion are projected at X* and Y* in the absence of corrective measures, X1* and Y1* being accounted for by pursuit of noncommercial goals and X2* and Y2* by pursuit of commercial goals.

The policy issue then becomes one of deciding what proportions of B and C to achieve via one or more of three different routes: (1) improving the management of public enterprises; (2) reorganizing the sector to enhance the role of commercial incentives (divestiture being only one of several possible approaches in this respect); and/or (3) modifying directives to public enterprises with respect to pursuit of noncommercial objectives.

It goes without saying that a comprehensive assessment of sectoral impact must encompass indicators of public enterprise performance in achieving the noncommercial objectives set for it as well as in achieving a satisfactory financial return on invested capital after making due allowance for costs and losses incurred in meeting those objectives. Appendix I discusses the noncommercial objectives and describes briefly some indicators that one might seek to measure in this connection.

The following list summarizes and categorizes government’s noncommercial objectives in establishing and maintaining public enterprises, as discussed in Appendix I.

  1. Economic stabilization

    1. Control of inflation;

    2. Food security; and

    3. Dampening economic downturns, with special reference to surges in unemployment.

  2. Economic growth

    1. Expanding absolute levels of investment, output, exports, income, and employment; and

    2. Accelerating industrialization.

  3. Income redistribution

    1. Promotion of small-scale producers through credit;

    2. Other approaches to vertical redistribution of income; and

    3. Geographical redistribution (promoting regional equity).

  4. Localization/indigenization

    1. Supply sources;

    2. Asset ownership and control;

    3. Jobs; and

    4. National policymaking.

  5. Miscellaneous objectives

Keeping in mind the distinction between commercial and noncommercial objectives helps one to deal with an ambiguity that arises in referring to the “macroeconomic impact of the public enterprise sector.” To many a casual listener this phrase would suggest the incremental impact on fiscal and monetary parameters of organizing a certain set of economic activities around public enterprises as opposed to placing them under private ownership and control. But insofar as the evaluator is merely recording nominal parameters that characterize the entities making up the public enterprise sector, one is including effects, the proximate causes of which are government policies vis-à-vis noncommercial objectives that could in many cases be applied nearly as effectively, or perhaps even more effectively, to private firms. In such cases, then, it is more precise to distinguish two different macroeconomic impacts: (1) the impact of government policy in pursuit of noncommercial objectives X, Y, and Z as implemented through the public enterprise sector; and (2) the impact of public enterprise commercial operations as a reflection of the sector’s intrinsic efficiency in generating value added through manipulation of productive factors.9

Adjustment of Nominal Parameters for Implicit Taxes and Subsidies

In this stage the analysis takes account of the fact that market or controlled prices frequently overstate or understate social opportunity costs, the latter being inflated by implicit taxes or deflated by implicit subsidies before coming to light as nominal or accounting magnitudes. An analysis aspiring to measure the impact of the public enterprise sector in real terms must thus adjust the observed values for these implicit transfers.

For example, the discussion below in “The Fiscal Impact of the Public Enterprise Sector” examines in some detail the argument that the public enterprise sector in mixed economies rarely covers the full social opportunity cost of the public capital invested in it. In an opportunity cost sense, real operating outlays are thus higher than nominal costs, and the discrepancy between accounting and real operating profit is covered by an implicit government subsidy.

Likewise, a public utility or other natural monopoly that is allowed to charge prices above levels that would cover the full social opportunity costs of production of an entity operating at an efficiency level considered reasonable by the appropriate authorities is collecting an implicit tax from consumers. If the entity’s actual costs correspond to the aforementioned “reasonably” efficient level, then the government’s pricing policy is allowing it a return in excess of capital opportunity cost; and if the excess return is being transferred to government (via dividends, excess profits tax, or some other mechanism), then the implicit tax proceeds are accruing to government. Conversely, if the entity is absorbing the proceeds, either as retained profit or by operating inefficiently and incurring unreasonable costs, then the implicit tax is accompanied by an implicit subsidy to the enterprise.

Adjustment procedures are explored in greater detail below in “The Fiscal Impact of the Public Enterprise Sector.”

Comparison of Adjusted Parameters with Criterion Values (Partial Equilibrium Approach)

In this next-to-last stage, one takes the leap into normative analysis, comparing adjusted public enterprise sector parameters with criterion values as a basis for judging whether the former are “adequate,” or conversely as a basis for identifying measures that should be taken to “improve” them. The comparison makes sense only after the parameters have been decomposed and adjusted as per the two preceding subsections. With respect to the attribution between commercial and noncommercial objectives, analysts adhering to the present paper’s emphasis on the intrinsic social efficiency of public enterprise as a producer of goods and services will focus on comparisons featuring the parameter values attributable to public enterprise commercial operations. The criterion values may be historical values for the public enterprise sector itself; they may be measures of analogous parameters for a set of private firms; they may be based on international comparative data; or they may represent merely theoretical norms.

For example, one may ask whether the public enterprise sector’s commercial operations, reflected in parameter values from which the impact of supplemental costs and/or revenue losses attributable to pursuit of noncommercial objectives has been netted out, are generating more or less before-tax profit now, per unit of value added or capital stock valued at replacement cost, than they did five years ago; or how these ratios compare with analogous ones for private firms.

The single most prevalent criterion value for assessing the commercial performance of public enterprise in a partial equilibrium framework is zero profitability, with adjusted revenues covering the social opportunity costs of all inputs and productive factors including capital. (Such a position is sometimes described as one of “normal” profitability, with “supernormal” profits absent.) Capital comprises, of course, not only the stock of fixed capital valued at replacement cost but also inventories and other working capital, whose importance normally exceeds that of fixed capital in the ease of enterprises engaged primarily in marketing.10

Consistency requires that if capital is to be valued at its social opportunity cost, then so also must labor. For those subscribing to the doctrine of a near-zero social opportunity cost of unskilled labor in economies or localities afflicted by severe unemployment or underemployment this opens up the possibility of discounting a sizable portion of public enterprise wage bills. On the other hand, analysts concerned about fiscal and monetary stability will justifiably incline to approaches that stress the disutility of effort and/or the benefits obtainable via potential alternative uses of resources devoted to enhancing the consumption of public enterprise employees. According to these approaches, endorsed by the present paper, downward adjustment of public enterprise wage and salary bills to account for divergence of social opportunity cost from market rates should not exceed a small fraction reflecting generation of tax payments via the workers’ consumption expenditure.

Use of the zero profitability criterion value reflects an implicit comparison with a private-sector, free-market standard, on the presumption of neoclassical theory that private firms operating under conditions of pure competition exactly cover their opportunity costs, with subnormal profits leading to exit of firms and supernormal profits attracting new entrants. Similarly, according to theory, firms of apparent high profitability that persist in that state are in fact merely) compensating for risks, covering the opportunity cost of extraordinary entrepreneurial skills, or earning scarcity rents attributable to factors such as favorable location.

The preoccupation with public enterprise operating deficits in many analyses dealing with problems of the sector reflects a tacit, if not explicit, acceptance of the zero profitability norm. (On the other hand, given pressures of time and the absence of information, most studies rarely distinguish the components of public enterprise parameters attributable to pursuit of commercial and noncommercial objectives.)

Apart from looking at commercial performance, one may also compare the efficacy of the public enterprise sector with standard indices, however determined, as a vehicle for government pursuit of the noncommercial objectives listed above under “Decomposition of Nominal Parameters by Attribution to Commercial and Noncommercial Objectives.”

Comparison of Adjusted Parameters with Putative/Simulated Outcomes of Baseline Alternatives (General Equilibrium Approach)

Normative evaluation of the public enterprise sector’s macroeconomic impact culminates in this stage, where the attempt is made to estimate the net impact of sector operations on key economic and social parameters as compared with the operations of an assumed baseline industrial organization. Here is where such ultimate questions are posed as: how much more (or less) is the government budget deficit, given the public enterprise sector as it operates today, than it would have been had the baseline situation prevailed? How much have public enterprise operations contributed to money supply, inflation, and the cost of borrowing, as compared with levels that would have prevailed in the baseline situation? How much more (or less) employment is there; how much greater (or smaller) is the Gini coefficient of income inequality than it would have been: and how much more (or less) disposable income is generated in disadvantaged regions?

To the extent these questions can be answered, the analyst will have succeeded in measuring both types of macroeconomic impact identified on page 46. That is, he or she will know how effective the public enterprise sector is, compared with the assumed baseline industrial organization, as society’s agent in implementing noncommercial objectives; and an index of the comparative efficiency of its commercial operations will have been obtained.

The key problem with conducting this kind of analysis, of course, is the heroic assumptions one must make about the baseline situation. What kind of government policy does one assume? In lieu of establishing public enterprises outside the standard core of power, water supply, posts and telecommunications, the railroad, national airline, etc., should one assume that government would have adopted an entirely passive approach to industrial development? To what extent (or under what circumstances that must be further assumed) would the market have stepped in and created economic activity to fill the putative vacuum left by governmental abstention? What would have been the fiscal and monetary consequences of the alternative private sector expansion (or nonexpansion)?

There can be no unambiguous answers to these questions; rather, the analyst must model one or more baseline alternatives, making the assumptions explicit, and compare the actual outcome, given the public enterprise sector as it functioned in the real world, with the outcome predicted on the basis of those assumptions. It is highly desirable that economists in governments making substantial use of public enterprise should conduct such exercises. Given the complexity and, in most countries, novelty of the approach, this is an area in which staff economists in international development institutions could provide strategic help. Yet so value-laden are the assumptions underlying the various alternative models that one imagines only with difficulty an IMF or World Bank staff report undertaking to lay out a comprehensive baseline alternative with which the performance of Country A’s public enterprise sector should be compared and the sector’s net impact on fiscal, monetary, and social aggregates assessed.

* * * * * * * * * * *

Following the methodological discussion in this section, the remainder of this paper focuses on specific indicators of the public enterprise sector’s macroeconomic impact. The organization of the remaining segments reflects in part the logical progression of the five stages above and in part a focus emphasizing fiscal and monetary stabilization over economic growth and the efficiency of public investment.


The three subsections of this section will examine as many different categories of economic aggregates:

  1. Summary accounts of the public enterprise sector;

  2. Indicators of fiscal impact; and

  3. Indicators of monetary impact.

Summary Accounts of the Public Enterprise Sector

This subsection considers the estimation of the public enterprise sector’s shares in national product, investment, savings, and capital stock, along with various measures of its resource balance vis-à-vis the rest of the economy.

Aggregation of standard accounting magnitudes across the sector as a whole, differentiated by industrial groupings, or, if useful for policymaking, a few large enterprises, constitutes the First step in constructing these macroeconomic indicators. Details of the format will vary among countries with different accounting systems, but the basic elements entering into the computation of the national income accounts should be present. The most severe obstacles to preparation of accounting aggregates arise because (1) in many countries public enterprises have varying financial years, and (2) the uniform standards and formats (if any) prescribed for public enterprise accounting have rarely been refined to a sufficient level of detail to remove variations in procedure.

Table 1 illustrates a consolidated public enterprise account for the West African nation of Mali, a country falling well within the United Nations “least developed” category whose government is nonetheless, as regards production of this type of data, a long jump ahead of many featuring per capita incomes several times as high. 11

Table 1.

Consolidated Financial Statements, Public Enterprise Sector, Mali. 1976

(In millions of Mali francs)

article image
article image
Source: Republic of Mali. Ministère de Tutelle des Sociétés et Entreprises d’Etat, Analyse de la Situation Financière du Secteur d’Etat, au 31 Décembre 1976, Bamako. October 1S77. Tables 5, 11, and 14.The formal of these financial statements is modeled on that of Richard Goode. “Nonfinancial Public Enterprises” Appendix 1. Table 2. “Financial Analysts of Operations of NPEs.” IMF Fiscal Affairs Department, May 15, 1980 (unpublished). Major departures from Goode’s format are: (1) introduction of turnover and indirect taxes as determinants of operating income; (2) relabeling ol item A(2)a. “use of purchased goods and seivices;” (3) placement of nonoperating income and expenditure following computation of after-tax operating profit, and (4) introduction of sources and applications of funds statement.The source of the Malian data does not provide a basis for distinguishing depredation and other provisions, nor For interpreting item 2(a)iii,. miscellaneous current applications (“contributions diverses”), which presumably contains any dividend payments or redemption of government equity. Clarification of these items would be desirable for subsequent economic interpretation of the accounts (of Table 2).

The aggregates of primary interest from the viewpoint of the national income accounts may be grouped under three headings, as is done in Table 2, defining the public enterprise sector’s shares in national product, investment, and saving, respectively. Also of interest, particularly for purposes of cross-country comparison, as per the following paper by R.P. Short, are the ratios of the respective public enterprise parameters to GDP.

Table 2.

Computation, Based on Enterprises’ Financial Statements, of the Public Enterprise Sector’s Share in National Product, Investment, and Saving

article image

United Nations, Statistical Office, A System of National Accounts (New York, 1968)

Conventionally referred to as GDP at market prices.

Conventionally referred to as GDP at factor cost.

The United Nations publication reserves the term “saving” for retained income. This study favors the concept of “gross saving” inclusive of depreciation to maintain the validity of the national income accounting identity: investment – saving = imports – exports, wherever investment is treated on a gross basis, as is customary in developing economies.

The aggregates lend themselves to various manipulations with a view to assessing the resource balance between the public enterprise sector and the rest of the economy. Beginning with the national income parameters, the ratio of public enterprise savings (gross) to investment measures the share of capital expenditure that is internally financed; subtracted from unity it gives the share of public enterprise investment financed by the rest of the economy and the outside world.

A first approximation to overall resource balance is given by the sector’s gross expenditure, current (excluding depreciation) plus capital, less sales; since the subtraction nets out intrasectoral transactions these do not have to be estimated. Looking at the resource balance from varying points of view, one might want to examine a variant where the deficit is reduced by adding nongovernment capital transfers to sales, on the ground that capital resources which the public enterprise sector attracts competitively from the private sector represent less of a “burden” on the rest of the economy. In some cases a further reduction by expanding revenues to include current government transfers could be justified on the supposition that these compensate for services the public enterprise sector provides in pursuing government’s noncommercial objectives.

Also of interest is the public enterprise sector’s current balance; to arrive at this one might begin with an operating balance, given by sales revenue less expenditure on noncapital goods and services, then add receipts of current transfers to the revenue side and interest, dividend and direct tax payments to the expenditure side. Variants including and excluding depreciation may be calculated to distinguish between the sector’s real and financial balance.12

Two major adjustments are required to convert public enterprise sector aggregates obtained from financial accounts into magnitudes that reflect social opportunity cost. First, the application of depreciation rates determined for tax purposes to historical investment cost is a poor guide to the annual charge that should be made against output to express the cost of maintaining existing productive capacity. Even a revalued capital stock reflecting current replacement cost is an imprecise measure of productive capacity under conditions of rapid technological change. Still, given even the lowest inflation rates prevalent today, it constitutes an economically more significant base for computing depreciation than does historical cost; and construction of any alternative would be subject to debate and uncertainty. Altering tax-based depreciation rates to accord with asset lives that reflect current rates of obsolescence compensates for part of the imprecision associated with any cost-based estimates of capital stock.

Upward adjustment of depreciation/consumption of fixed capital does not affect the computation of value added/GDP or gross saving, since it is offset one-for-one by reduction of another component of value added and saving, namely, be fore-tax profit/operating surplus (or by increased losses); however, it widens the gap between gross and net measures of output and investment.

The second major adjustment required is to recompute before-tax profit/operating surplus by revaluing at border prices plus indirect taxes all tradable goods and services sold and purchased by the public enterprise sector. When applied in the computation of manufacturing sector GDP in countries that are establishing high-cost import-substituting industries behind high protective barriers, this adjustment has the effect of reducing the sector’s share of GDP by several percentage points. In not a few industries, it converts a respectable value-added figure, derived from the financial accounts, into a negative sum (i.e., the industry earns or saves less foreign exchange than it consumes via current inputs and utilization of investment goods and services). In many countries, the adjustment will likewise create negative value added in some public enterprises and significantly reduce the sector’s share of total GDP.

Of no less interest than the public enterprise sector’s share in annual investment flows is the value of its capital stock. This is essential for determining, inter alia, the implicit capital subsidy as part of the public enterprise sector’s Fiscal impact. (See the discussion in the “Fiscal Impact of the Public Enterprise Sector,” below.)

Computation of capital stock is perhaps the most complex of the statistical exercises involved in tabulating aggregates of the public enterprise sector. But in light of the critical role of this parameter in public enterprise performance evaluation, it is appropriate for international agencies to try to sensitize governments to its significance, and urge that local staff resources be devoted to measuring it and keeping the measure up-to-date. A primary component of updating is, of course, annual revaluation to compensate for inflation.

One of the major problems in estimating capital stock is how to take inflation into account in computing depreciation. Appendix II outlines a procedure for this based on the perpetual inventory method, which enables one to build up the capital stock estimate from constant-price investment series (using prices of the year whose annual flows are being analyzed), additional parameters being the asset lives of major capital goods categories and the growth rates of capital stock and value added over an initial period when the two rates can be assumed to be approximately equal.

The constant-price investment series should in turn be adjusted for divergences of nominal prices of capital assets from social opportunity cost. Perhaps the most prevalent cause of downward divergence is overvaluation of the exchange rate, sometimes involving maintenance of a multitiered rate structure giving preference to capital goods imports, especially those of public agencies.

A divergence in the opposite direction arises where investment is financed by foreign capital inflow that is tied to a particular activity, such that its opportunity cost is the reverse flow of service payments rather than benefits arising from an alternative use should the activity in question not be carried out. This divergence is also discussed in Appendix II, along with a suggested approach to correcting for it.

In conclusion, once the national income and capital stock aggregates for the public enterprise sector or subsets of it have been computed in absolute terms, it is a simple matter to calculate them as proportions of GDP or (with respect to savings and investment) as proportions of the national or subsectoral totals of the variables in question. These coefficients begin to have policy significance—for example, the statement that Country A’s public enterprise sector accounted for 60 percent of gross capital formation in 1982, in the context of other evidence about the economy, may suggest either that Country A’s investment and growth would have been very much less without the leading role played by the sector or, conversely, that the sector appears to be crowding out private investment. Even greater policy significance attaches to Findings that internally generated resources (depreciation and retained earnings) of the sector in question covered, say, 20 percent of its investment expenditure, meaning that 80 percent had to be provided by government and the private sector: that the sector’s aggregate deficit vis-à-vis the rest of the economy amounted to 5 percent of GDP; or that it ran a current surplus equivalent to 2 percent of GDP.

Fiscal Impact of the Public Enterprise Sector

Under this heading, the paper takes as its point of departure financial flows between government and the public enterprise sector, excluding from the analysis only payments at fair market value—or, in the absence of competition, social opportunity cost of production—for goods and services rendered. For purposes of subsectoral analysis, flows within the public enterprise sector are also taken into account, although these cancel each other out when the sector as a whole is examined in relation to government. The following categories of flows are listed, classified as accounting magnitudes or implicit transfers, and then examined individually:

Flows from the public enterprise sector to government

Taxes and royalties.

Increase of arrears in government payment for goods and services.

Payments of return on investment:

  1. Interest; and

  2. Dividends.

Redemption of capital:

  1. Amortization of government loans; and

  2. Redemption of equity.

Lending to government by nonfinancial public enterprises.

Expenses and revenue losses arising from pursuit of noncommercial objectives.

Discrepancy between fair market value—in the absence of competition, social opportunity cost of production—and actual prices paid by government.

Flows from government to the public enterprise sector

Purchase of equity.

Debt transactions;

  1. Lending by government;

  2. Increase in arrears of tax payments to government; and

  3. Repayment of government borrowings from public enterprise sector.

Interest on government borrowings.

Unrequited transfers to the public enterprise sector.

Tax subsidies:

  1. Implicit tax revenue retained by public enterprises; and

  2. Conventional tax subsidies.

Capital subsidy.

Flows within the public enterprise sector

Purchase of equity.

Debt transactions.

Payments of return on investment.

Cross-subsidies, notably provision of goods and services below fair market value or, in the absence of competition, social opportunity cost.

Fair market value of payments for goods and services may be defined as the mean price prevailing in arm’s length transactions of comparable scale during the relevant period; to the extent that government purchasers are exempt from payment of sales and excise taxes and the like, it should be measured net of these. Payments at other than fair market value involve implicit subsidies in one direction or the other that must be taken into account in drawing up the Fiscal balance. Without competition to define fair market value, the presence of subsidies would be signaled by prices diverging from the social opportunity cost of production of the goods and services in question. Estimation of either fair market value or social opportunity cost within an acceptable margin of uncertainty may, of course, involve substantial effort.

An ultimate measure of Fiscal impact is the net balance of all relevant flows. The most obvious criterion value for assessing sectoral performance in this connection is zero—that is, if there is a net flow from the public enterprise sector to government, one may say the sector’s Fiscal impact is positive; a net flow in the opposite direction signifies a negative impact; and if flows in both directions cancel out, one may say the impact is neutral.

In general, however, this naive approach has limited policy significance. For example, the flow of investment capital from government to the public enterprise sector in a given year will be heavily influenced by the public investment program being implemented at that time, the optimal composition of which has little to do with the factors that influence the reverse flows of taxes and returns on previous investment. Moreover, the flow of capital from government is sensitive to the preferences of capital sources for lending directly to public enterprises as opposed to two-stage lending through government.

Accordingly, this paper classifies the flows into various categories and manipulates them to obtain indicators that can be compared with appropriate criterion values. Flows from the public enterprise sector to government are listed in the subsection below, those from government to the public enterprise sector on page 68. The first part of each of these subsections lists flows that comprise accounting magnitudes available in financial reports, while the flows described in the second part, corresponding to the second and third stages of analysis outlined in Section IV above, consist of implicit values requiring estimation.

The subsection “Financial Flows Within the Public Enterprise Sector,” deals briefly with flows between public enterprises, particularly cross-subsidies. A final subsection, the “Fiscal Impact of the Public Enterprise Sector: Conclusion,” summarizes the components of the fiscal balance, both accounting magnitudes and implicit values, in tabular form.

An assumption underlying the presentation is that computation of any indicator bearing on the commercial efficiency of public enterprise, and in particular any indicator on which public enterprise performance is to be compared with that of the private sector, has adjusted all relevant parameters for costs and revenue losses attributable to pursuit of noncommercial objectives.

Financial Flows from the Public Enterprise Sector to Government
Accounting Magnitudes

Taxes and royalties. This heading deals with the following indicators (throughout this section, the superscript P refers to the public enterprise sector, the superscript F to the private sector, and the subscript t to a year):

TAXi,tP= total government receipts from tax/tax category i in yeart;TAXi,tPand TAXi,tF= payments by the public enterprise and privatesectors, respectively, of tax/tax categoryi in yeart (category 1 representing direct taxes);p˙= rate of price inflation from yeart 1 to year t;PRFtP and PRFtF= accounting profits of the two sectors in yeart; andVALtP and VALtF= value added in the two sectors in yeart.

The indicators are:13

TAXi,tPTAXi,t=public enterprise sector’s  proportional  share  in  total  government receipts  from  tax/tax  category  i  in  year  t.TAXi,tPTAXi,tTAXi,t1PTAXi,t1=change in that sharef rom year t1 to year t.TAXi,tPTAXi,t1PTAXi,t1P=proportional change in nominal tax payments of public enterprise sector.TAXi,tP/(1+p˙)TAXi,t1PTAXi,t1P=realproportional change in sectoral taxpayments.TAX1,tPPRFtP/TAX1,tFPRFtF=proportional comparison of ratio of direct tax paymentsin year  t to  accounting profit as betweenpublic enterprise and private sectors.TAXi,tPVALtP/TAXi,tFVALtF=proportional comparison of ratio of tax payments tovalue added between public enterprise and privatesectors.

Later in this section, it will be indicated how the comparison of TAXI PRF ratios (tax payments as a proportion of accounting profit) may lead to imputation of an implicit tax subsidy to the public enterprise sector.

Tax categories (subscript i) whose coefficients one might wish to scrutinize under this heading include (i) direct taxes (comprising company/corporation income taxes, excess profits tax, payroll taxes, land and real estate taxes, licenses); (ii) taxes on domestically consumed goods and services other than import duties, sales and excise taxes, which are excluded here—first, because the same or larger amounts could be collected from final consumers if the commodities were imported in finished form and, second, because their point of collection is arbitrary (import duties could be collected from private importers and included in prices charged to public enterprises, and excises could be collected from private retailers, in both of which cases public enterprise tax payments would be correspondingly lower for institutional reasons having nothing to do with operating efficiency); and (iii) export taxes and royalties on natural resource-based exports.

Increase in arrears of government payments for goods and services supplied by public enterprise. (Decrease =—.) This item, denoted ARRtP, could also be classified with public enterprise sector loans to government (subsection, “Lending to government by nonfinancial public enterprises,” below). One can define several ratios as indicators of the public enterprise sector’s exposure under conditions of Fiscal stringency. For example, ARRtP/DEFt denotes the proportion of the year t government deficit financed by increased arrears to public enterprises; and ARRtP/SALtP represents the proportion of sales to government not compensated by timely payment, compared with the analogous ratio for the private sector, ARRtF/SALtF.

Payments of return on investment. The key items here are:

INTtP= interest payments on government loans to public enterprises;andDIVtP= dividend payments on government equity investment in the sector.

The point of departure in analyzing such payments is to compare them with the value of the investment on which they are earned. The obvious criterion value for the nominal return is unity plus the year’s inflation rate, times the difference between the annualized social opportunity cost of the public capital employed by the public enterprise sector and its direct tax payments (tax payments are subtracted, because social opportunity cost, based on the social discount rate used in benefit-cost analysis of public investments, is computed gross of these whereas interest and dividend payments, of course, exclude them).

The objection may be interposed here that determination of the appropriate discount rate by which to compute social opportunity cost of use of public resources is subject to a considerable margin of uncertainty, and arguments are advanced for widely differing values. On the other hand, the consensus of modern social opportunity cost theorists is that the low rate of social time preference by virtue of which society collectively sets aside a much greater proportion of available resources to benefit future generations than the average citizen would agree to independently is not an appropriate discount rate for government expenditure.

The vast majority of estimates focusing on returns available from application of public capital to expenditure opportunities available at the margin yield values ranging from 10 percent to 20 percent in real terms (net of inflation). Since few public enterprises in mixed economies yield anything approaching such returns to their governments on commercial operations alone, the thrust of public enterprise evaluation against the criterion value in question is clear and a few percentage points of uncertainty in the social discount rate do not alter the conclusion that the public enterprise sector in most countries enjoys a hefty capital subsidy.14

The ideal measure of the capital stock to which the social discount rate is applied for purposes of calculating annualized opportunity cost is the net present value of the stream of social benefits obtainable from the assets in question under conditions of “normal” efficiency. Historical cost, even when inflated as a proxy for replacement cost, is only a rough index of this ideal measure, which treats all past investment as a sunk cost. There is much to be said for it in theory, but its estimation across the public enterprise sector as a whole poses insuperable problems. A more practicable candidate for denominator in the rate of return calculation, notwithstanding the major computational effort involved, is the constant-price value of the public enterprise sector’s capital stock computed according to the methodology outlined in Appendix II. As noted in the subsection on the “Comparison of Adjusted Parameters with Criterion Values,” above, for use in the present context, capital stock is computed gross of working capital (mean level during the period in question) and net of construction in progress or investment still in a start-up phase.

Specifically with regard to government equity investment, it could be argued that the criterion value for comparison with return on investment should include an element of compensation for the risk and uncertainty associated therewith. Opposing this would be the consideration that, particularly in the current inflationary environment, an equity investor expects part of his return by way of capital appreciation; hence, not even the most demanding finance minister would call for dividend payments corresponding to the full opportunity cost of government’s equity.

This argument does not apply to loan capital. However, nominal financial charges that governments impose on their public enterprises are almost invariably below the social discount rate plus inflation, owing in large part to a widespread conception that one accelerates development by charging entities that carry out socially “desirable” activities less than they would have to pay to borrow on the open market, particularly considering the risk premium that the market would have to collect from many of the activities in question. Thus, direct loans to public enterprises from government treasuries, as well as loans from government-controlled development banks and other financial intermediaries, more often than not carry nominal interest rates that are below the inflation rate and, hence, are negative in real terms.15

In most developing countries this picture is not confined to government-run credit facilities, since government regulation of the private financial sector holds nominal interest rates in the formal money market below levels that would prevail under competition, which one would expect to compensate for inflation and risk while approximating the real social opportunity cost of capital.

Flows to developing countries from public credit institutions likewise contain subsidy elements ranging up to and even exceeding 90 percent in the case of long-term (up to 50-year), low-interest (below 1 percent or even zero) loans when service obligations are discounted by anticipated export price inflation plus a real capital opportunity cost of 10 percent or more. Such institutions generally require that portion of their assistance destined for commercial-type activities of public enterprises to be on-lent by the host government on “economic” terms, but the outcome of a bargaining process in which most recipient governments (and, of course, the participating enterprises) strive to keep the on-lending terms relatively soft is normally an effective interest rate well below the social discount rate.

Flows from private foreign lenders usually carry terms much closer to the domestic opportunity cost of capital, however, even in this case, the following caveats apply: (1) such loans, particularly the subcategory of supplier credits, sometimes carry a hidden subsidy in the form of source government guarantees that soften the terms of repayment; (2) a portion of the true financing cost is sometimes hidden in inflated prices of capital equipment acquired under other than international competitive bidding arrangements; and (3) it is in this category that one most frequently encounters nonfungible transfers, altering the cost calculus as described in Appendix II. (Asset purchase costs are deducted from the value of capital stock, and annual interest and dividend payments are added to the product of the social discount rate times the residual capital stock.)

In conclusion, there is no mixed economy whose public enterprise sector would not display a substantial shortfall against the proposed criterion value.

Perusal of budgets and investment programs over a number of years in several countries reveals that early optimism about the proportion of these that could be financed out of cash returns on government’s investment in public enterprises has faded. Today it is the rare national treasury that finances more than a trivial proportion of its investment budget from this source, other than via dividend payments more appropriately classified as royalties on natural resource-based output.

This suggests, as a second indicator with respect to cash returns on government investment in public enterprises, a comparison between actual outturn of INTtP+DIVtP and planned or budgeted values, INTtP*+DIVtP*, suitably inflated to current prices. The latter figures may be available in summary form in annual budget and/or multiyear plan documents.

A related concept that applies to loan investment only is a comparison between the outturn of INTtP and the sum of products of loan balances outstanding times the respective contractual interest rates. Any discrepancy represents default of interest owed to government by the public enterprise sector.

Regardless of how far interest and dividend payments to government fall below the social opportunity cost criterion value, it may still make sense to examine a time series of the proportion that government is recovering in cash out of the annualized cost of its investment in the public enterprise sector. A marked trend upward or downward might be considered significant with respect to the authorities’ diligence or lack thereof in imposing financial discipline on the sector.

It is worth repeating here that the present conceptual framework stops short of prescribing remedies for shortcomings in public enterprise performance, and specifically with regard to payments of return on investment, it does not advise that every government should as fast as possible reorganize its public enterprise sector to cover the full opportunity cost of past investment. Depending on factors specific to a given country, it may well be that social goals would best be served by, in effect, writing off much or all past investment in public enterprise while stiffening the criteria for approval of new investment proposals. Indeed many countries, both developing and industrial, are at one or another stage of implementing such a policy, and many are being supported in that direction by their international creditors. It is nevertheless the paper’s thesis that evaluation of sectoral performance hitherto is a valid if not exclusive criterion for assessing the sector’s likely contribution to achievement of macroeconomic policy objectives in future, and that measurement of current returns on past investment is a major component of such evaluation.

Redemption of capital. Under this heading, the following terms are defined:

AMRtP= amortization payments in year ton government loans topublic enterprises; andREDtP= transfers in redemptionof government's equity investmentin the sector.

The second of these terms represents a comparatively rare phenomenon, as the movement is overwhelmingly in the opposite direction—government increasing its equity (or providing outright grants in lieu of equity) either to assist in expansion or to cover operating deficits. Where an enterprise’s cash flow would permit it to redeem shares without prejudice to financing its own socially beneficial expansion, most governments are inclined to channel the surplus into their own debt instruments or to tax it away via royalties, excess profits tax, etc.

Two indicators suggest themselves with respect to amortization of loan capital: first, a comparison of outturn with contractual obligations—payment arrears being represented by a negative result, AMRtPAMRtP*<0—suggests problems of cash flow, lax financial discipline, and the like. Second, the presence of significant discrepancies between AMRtP and depreciation allowances on account of the facilities financed by the public loans in question indicates either financing of medium-to long-term investments with excessively short-term capital or, conversely, a preferential arrangement that gives the borrowing enterprises financial leeway to increase investment out of cash flow or incur operating deficits. Either finding suggests an irregularity with respect to conventional standards; however, either practice may also be justified under special circumstances.

Lending to government by nonfinancial public enterprises. It is appropriate here to distinguish between medium- and long-term lending on the one hand, and purchases of treasury bills and other short-term instruments on the other hand as a component of normal liquidity management. To be sure, to any given program of medium-term and long-term lending mandated by government there will always be a corresponding equivalent rate of increase in purchase of short-term instruments, maintained over an adequate period of time.

Lending to government, denoted by LNDtP, suggests no special indicator other than the fact that a significant positive value means one of two things: either (1) certain components of the sector, from which we exclude financial institutions for purposes of this heading, are in a cash surplus position that enables them to help finance other government programs, or else (2) fiscal stringency has led government, after exhausting the resources of the financial sector, to compel public enterprises to buy its bonds or otherwise lend to it against their commercial interest.

Case (1) may raise a question as to whether government should not be appropriating the surplus through special levies in lieu of indebting itself with the public enterprise sector. Alternatively, it may be a matter of particularly creditworthy enterprises, such as petroleum producers, mining concerns, or electric power and telecommunications companies, borrowing abroad in excess of the residual of their investment needs less internal cash flow; here, it often makes economic sense for the enterprise to go ahead and borrow as much as it can, meanwhile investing the surplus in general government obligations.

Implicit Transfers

Two items deserve mention under implicit financial transfers from the public enterprise sector to government.

Expenses and revenue losses arising from pursuit of noncommercial objectives. EXPi,tP denotes the expense or revenue loss incurred by the public enterprise sector in year t as a result of pursuing, by government directive (explicit or implied), a given noncommercial objective, represented by the subscript i (see list on page 45 which is amplified in Appendix I).

A host of methodological issues arise in measuring the financial burden of pursuing noncommercial objectives. How to estimate prices that would have prevailed in the absence of inflation control measures? How to measure optimum food stocks from a commercial viewpoint, as opposed to stocks held on government directive to ensure an extra margin of security? What are the extra costs of a crash industrialization program initiated by government, as compared with the baseline investment program that the pre-existing public enterprise sector would have carried out in accordance with commercial objectives? What are the real differences in costs between operating in less developed regions benefiting from government preference and operating in locations that would be chosen on the basis of commercial criteria?

Moreover, there may be honest disagreement as to whether pursuit of a given objective does, in fact, impose a financial burden on a public enterprise. What, for example, is the cost of localizing a management position (assuming as the baseline alternative that a foreigner would have occupied the post in any private facility established in lieu of the public initiative)? Nominal emoluments are normally far lower. Who would be bold enough to quantify the offsetting loss of revenue incurred as a result of the local manager’s allegedly inferior experience, skills, market contacts, etc.?

A standard approach to estimating EXPp would be first to ask enterprise managers for documentary evidence concerning regulatory controls imposed on them, on the one hand, and, on the other hand, instructions received with respect to positive acts of resource allocation. The next step would be to sift through accounts to net out the impact of controls and instructions that the managers claimed had affected adversely their operating results. Such netting out involves not only subtracting costs and revenues attributable to the relevant measures, but also considering the disposition of the incremental operating surplus that would be generated by their cancellation.

For example, fiscal transfers identified above—tax payments or returns on government investment—might well increase. Alternatively, if the effect of the netting out is to reduce a pre-existing deficit rather than create or increase an operating surplus, an assumption is required, preferably buttressed by advice of management, as to what source(s) of deficit financing would most likely be released.

The exercise cannot, however, stop at documented measures of control. Enterprise managers may well attribute a portion of any losses to oral instructions, to decisions of their own in pursuit of general policy guidelines, or perhaps merely to their own supposition as to what course of action would be most in the public interest. These attributions must be verified with government officials, who will often counter that the enterprise managers are merely covering up for bad business decisions or inept management. Ultimately some exercise of the analyst’s judgment cannot be avoided.

A few countries have progressed in government-public enterprise relations to the point where, within the framework of a periodic (annual or multiannual) agreement,16 government acknowledges responsibility for covering a certain level of losses which the enterprise anticipates incurring in pursuing noncommercial objectives. Even where this approach is adopted, however, the authorities who negotiate such arrangements on government’s behalf may not be free to acknowledge the noneconomic motivations underlying certain instructions—for example, using an enterprise’s payroll as an instrument of political patronage, as distinct from the broader goal of curbing unemployment in society at large—so that the analyst may be obliged to augment the estimates of the financial burden. Still, under the pressure of negotiations leading to the agreement the parties normally gather much data that are useful in the analysis.

Introducing the EXPP term into the fiscal balance does not trigger an offsetting entry in the reverse flow from government to the public enterprise sector, although the sector may already be receiving sufficient cash or implicit tax subsidies to cover the sum of its EXPps.

Below-market supply of goods and services to government. LOStP (“LOS” stands for “loss”) symbolizes a positive difference (if any) between the fair market or opportunity cost value of goods and services that the public enterprise sector supplies to government in year t, and what government actually pays for them. Having excluded the fair market or opportunity cost value of transactions from fiscal impact, one is obliged to enter this discrepancy as a flow from the sector. The criterion value is, of course, zero; any positive value of LOSp is tantamount to imposition of an extralegal tax.

The reverse flow embodied in the accordance of preference to public enterprises in government procurement may be more common, however, cases of cereals marketing agencies supplying foodstuffs to civil servants below purchase and handling cost are numerous.

Financial Flows from Government to the Public Enterprise Sector
Accounting Magnitudes

Explicit financial flows from government to the public enterprise sector are grouped under three headings: (1) purchase of equity, (2) debt transactions including return on loans (from the public enterprise sector to government), and (3) unrequited transfers (grants and subsidies). Flows analogous to those in the subsection, “Financial Flows from the Public Enterprise Sector to Government,” carry the same notation, except that the superscript “G” denotes payments by, or changes in the financial assets of, government.

Purchase of equity (denoted by EQItG). Two indicators are of interest here: first, debt/equity ratios for the public enterprise sector as a whole, various categories of enterprises, and the sector’s current investment program or individual segments of it. Public enterprises and their individual projects are frequently undercapitalized, as governments strive to carry out ambitious investments with a minimum commitment of budget resources. Prudent debt/equity ratios differ sectorally and among categories of activities within a sector. Once a mean criterion value has been determined for the set of activities currently under implementation, divergence of EQItG from that value (almost invariably in a downward direction) increases the probability of future claims on the budget and the money market, unanticipated in the activities’ initial planning, to forestall bankruptcy.

The second indicator pertains to the allocation of government transfers to public enterprises among equity, outright grants, and loans. Indiscriminate inflation of equity (or loan capital) on the occasion of every bailout operation enlarges the capital base against which returns are compared and may eventually create a misleading picture of low social profitability if the transfers merely served to compensate for pursuit of noncommercial objectives or were required during an atypical period of inferior management.

Debt transactions. Three items may be identified here.

LNDtG denotes lending by government to the public enterprise sector. Relevant indicators are analogous to those under purchase of equity. In addition it is of interest to distinguish insofar as possible between external loans on-lent directly to public enterprises and all other government lending to the public enterprise sector. The first is a wash transaction with respect to its impact on the near-term fiscal balance, even though gross magnitudes are correspondingly enlarged and a government liability is created subject to some degree of probability that the exchequer will be called upon to make it good. (This latter point applies also to direct borrowings of public enterprises which are guaranteed by government, otherwise treated here as a monetary parameter under “The Monetary Impact of the Public Enterprise Sector” below.)

TARtG denotes increase (decrease = –) in arrears of tax payments from the public enterprise sector to government. It enters into the fiscal balance equation only insofar as the corresponding liability has been included in the TAXp expressions of the subsection “Financial Flows from the Public Enterprise Sector to Government,” above, but even if it does not appear in the equation, it should be recorded as a memorandum item. A criterion value would be either zero or the ratio of increase in arrears to annual tax liabilities of a comparable segment of the private sector. An unfavorable comparison against the criterion value would cast the public enterprise sector in a burdensome light vis-à-vis the public finances.

AMRtG denotes amortization of government borrowings from the nonfinancial public enterprise sector. In some countries, it comprises redemption of government securities held by nonfinancial public enterprises; in other countries, redemption of public enterprise loans to government follows a more ad hoc pattern (i.e., government pays when, and only when, decision makers perceive the marginal social utility of doing so to be no less than that of any alternative expenditure). if amortization lags behind contractual schedules, the public enterprise sector is being penalized all the more for its close lies to government.

Interest on government borrowings (denoted by INTtG). Here one looks to see whether government is complying with its loan service obligations and what opportunity cost (by virtue of low—in real terms, typically negative—interest rates) government imposes on the public enterprise sector through compulsory borrowings.

Unrequited transfers (grants and subsidies) (denoted by SUBtG). This is a residual category grouping all transfers in cash or kind that do not qualify as payments for goods and services rendered, purchases of equity, or changes in liabilities. Given the heterogeneous nature of the animal, a varied set of analytical questions arises: what proportion, if any, of the transfers can be regarded as compensation for expenses and revenue losses incurred in pursuit of government-imposed noncommercial objectives? What rationale has been applied in allocating transfers to a given set of enterprises between SUBtG and EQItG? In the preceding discussion of the latter component, reservations were expressed about the practice of loading transfers for extraneous purposes onto equity. The converse also applies—namely, that treating as “manna from heaven” what is essentially equity capital understates the asset value on which a positive real return should be earned.

It is, however, the implicit subsidy component of unrequited transfers that deserves the bulk of our analytical attention and to which the paper now turns.

Implicit Subsidies

Three items are identified here, of which the first two are tax subsidies. The first of these constitutes an implicit tax on the population that becomes an implicit subsidy to the public enterprise sector, because government allows the sector to retain it in lieu of handing it over to the treasury. The second implicit subsidy arises when government exempts public enterprises from taxation or levies lower taxes on them than on comparable units in the private sector. Finally, the third implicit subsidy arises because, as already noted, the public enterprise sector enjoys the use of public capital at a price well below its social opportunity cost.

Tax subsidies. The first entry under this heading is SBT1,tG = implicit tax revenue retained by public enterprises, comprising payments to them by their customers—Final consumers or other producers—which, as a result of some governmental policy or act of commission or omission, exceed baseline levels. The baseline here consists of revenues that would enable a public enterprise operating at a “normal” level of efficiency to cover exactly the opportunity cost of its capital—that is, to earn a “normal” profit only. Conversely, revenues—provided they are facilitated by governmental action—that enable such a public enterprise to incur a smaller deficit than would otherwise be the case, or even to earn normal profit or better, are treated as implicit tax receipts.

The justification for this is that the present paper focuses on payments in excess of economic cost that are extracted from the public via exercise of governmental power and are devoted to a governmental objective—namely, strengthening the financial position of public enterprises. Formally, if not politically, government could just as well have mandated the payments in question and directed the enterprises to remit the proceeds to it as conventional indirect tax revenue.

Implicit tax revenues are generated principally by two categories of government policy, the first normally pursued via acts of commission while the second may be associated with either acts of commission or omission. The first category consists of protection of enterprises against foreign and/or domestic competition, the second involves lax control over prices of natural monopolies.

For enterprises producing tradable goods and services, the baseline or zero-implicit-tax case holds where selling prices of output equal unit costs of inputs (which may be protected or taxed well above border price levels) plus unit value added (in the enterprises) reflecting a unit domestic resource cost (DRC) of earning or saving foreign exchange—units of local currency expended on local factor services consumed in producing one dollar of value added at border prices—equal to the shadow exchange rate.

In the case of nontradables, it is appropriate to distinguish goods and services whose production and distribution permit competition and the so-called natural monopolies. With the former, the baseline case corresponds to free competition, ensuring “normal” profits to firms operating at an efficiency level determined, by whatever means are available, to be “average” for the economy in question. Accepting the scope that public utility regulation allows for overinvestment and inefficient operation, the baseline case for natural monopolies is defined as a set of average prices such that production and consumption end up at approximately the level determined by the intersection between the demand curve and an average total cost curve considered by the regulatory authority to reflect reasonable efficiency. Determining such baseline prices is, of course, a standard task for regulatory entities, who have to judge what constitutes reasonable utilization of labor, capital, and purchased inputs to obtain a given level of output. Earning a level of average revenue that meets these criteria does not rule out the use of marginal cost pricing at the margin, since enterprises may still engage in sufficient inframarginal price discrimination to avoid losses or supernormal profits.

Once baseline prices are determined, the implicit tax revenue is given by the area of the rectangle defined by the intersection of the actual price line and the demand curve, and the horizontal line through the baseline price. Graphically, the implicit revenue, which then becomes an implicit subsidy to the public enterprise sector, equals the area bounded by paABpb in Figure 1.

Figure 1
Figure 1

—Collection of Implicit Tax Revenue by Public Enterprises

Explanation: The baseline situation is defined as the establishment, by the regulatory authority, of an average selling price that enables the enterprise to cover the social opportunity cost of all factors, which includes earning a “normal” profit. In the actual situation shown, the regulatory authority is permitting the enterprise to charge a higher average selling price and thus to earn “supernormal” profit. The shaded area represents the implicit tax collected by the enterprise.
Explanation: By levying an indirect tax equivalent to pa-pb on imports of the good or service in question, or through quantitative restrictions limiting the quantum of imports, the government has brought about an actual domestic selling price of pa. As in Part I, the shaded area represents the implicit tax thereby levied on behalf of the public enterprise.

The second tax subsidy applicable here, which we term a “conventional” tax subsidy, is SBT2,tG, representing the value of any preferential lax treatment accorded to public enterprises. In some countries, varying proportions of the sector enjoy exemption from company tax, import duties, and/or other indirect taxes, or lower tax rates compared with the private sector as a whole. As in the computation of implicit tax revenue, it is necessary first to determine a baseline against which to measure the degree of preference involved.

This exercise is subject to two major complications. First, remaining within a partial equilibrium framework, the tax treatment of private companies often varies according to size, economic sector, and other characteristics. One might seek to establish the effective tax rates applied to the majority of private enterprises or to those enterprises accounting for the major share of the base for a given tax. Or one might calculate a mean yield from the application of a particular tax to private sector earnings or transactions.

Second, as one moves into a general equilibrium framework, it becomes necessary to take account of the fact that removal of tax subsidies from public enterprises would reduce the pressure on other points of the tax system, including taxation of private companies (another effect, assuming continuous behavioral functions, would be to stimulate supplemental expenditure). Thus, the baseline company tax against which the implicit subsidies should be measured is, in fact, lower than any that might be computed on the basis of de jure or de facto tax rates.

Any discrepancy between actual taxes paid by a public enterprise and the baseline, however determined, would then be considered an implicit subsidy from government to the enterprise (and would, of course, have to be balanced by expansion of enterprise direct tax payments as a corresponding entry under reverse flows).

It goes without saying that before computing conventional tax subsidies in the case of natural resource-based sectors, any portion of rents which is captured in the form of direct taxes rather than royalties and the like must be deducted from direct tax payments and treated as an implicit royalty.

The capital subsidy to public enterprise. This implicit subsidy, labeled SBCtG, is essentially the difference between the criterion value for payment of returns on public investment and actual payments to government on account of interest and dividends, which was elaborated in the subsection “Payments of return on investment.” As indicated in that discussion, most public enterprises enjoy a substantial subsidy on account of their use of the services of public capital; frequently, it exceeds the total of all other annual operating costs.

The approach followed here and in the subsection “Payments of return on investment” takes the history of government and financial sector capital infusions at face value and looks at the adequacy of the returns they are currently generating as compared with the social opportunity costs of those infusions.

An alternative approach, focusing exclusively on the capital stock’s present earning power, would seek to provide a normative basis for evaluating the macroeconomic performance of the public enterprise sector’s current management. The implicit subsidy calculated in this case measures the shortfall between what the sector is actually earning on its invested capital and what the analysts consider to be its true earning power, with due allowance made for the low or zero productivity of misguided past investment. In other words, with this approach, one seeks to avoid saddling current enterprise management with the responsibility for earning a return consonant with the social opportunity cost of capital on past losses in excess of reasonable startup costs, as well as on politically and socially motivated expenditures of the past. Clearly, the process of netting out such expenditures is subject to a wide margin of uncertainty; but in some of the more blatant cases, reasonable observers might agree on the order of magnitude of the invested capital to be written down.

The question arises as to how adjustment of the “subsidy” entry in Table 1 to incorporate the implicit capital subsidy or attribution of a portion of before-tax profit to implicit tax revenue in the case of monopoly pricing would affect the national income aggregates discussed in “Summary Accounts of the Public Enterprise Sector.”It turns out that these changes do not affect sectoral GDP at market prices, in the first case because introduction of a capital subsidy increases the financing charges/operating surplus pari passu, in the second case because treatment of part of profit/operating surplus as implicit indirect tax revenue merely shifts the sum in question from one component of value added/GDP to the other.

On the other hand, modification of either the subsidy or the indirect tax component affects the transition from market prices to factor incomes—upward adjustment of subsidies increasing GDP at factor cost and higher (implicit) indirect tax revenue decreasing it.

Adding a time dimension to analysis of the capital subsidy received by the public enterprise sector puts its importance, and the corresponding implicit fiscal burden on government, in even bolder relief. The discussion in this paper has hitherto been restricted to flows and stocks pertaining to a particular discrete period, typically one year, during which it is desired to evaluate the public enterprise sector’s performance. However, the provision of public capital to enterprises is customarily a contractual arrangement in which government either binds itself explicitly to accept a specific rate of return over a period of years, sometimes indefinitely, or else renounces any entitlement to a specific return (case of equity capital) and, in many cases, by precedent as well as by current policy, signals to enterprise management that, far from expecting a financial return in the foreseeable future, it will be relieved if the enterprise can survive without periodic operating subsidies.

In either of these cases one can define the present value of an implicit stream of future subsidy payments to which government is committing itself, contractually or de facto. The stream consists of annual differences between the social opportunity cost of capital, as defined in “Payments of return on investment” above, and the interest payments stipulated in the loan agreement, or the expected flow of dividends in the case of equity capital. The social discount rate is applied to calculate present value. The ratio of present values of (1) the subsidy stream and (2) the disbursement stream of the loan or equity investment constitutes the grant element of the instrument in question. A new paper tabulates this grant element on varying assumptions about loan terms, and shows that explicit interest subsidies in Korea’s direct official lending (not only to public enterprises) during the 1970s accounted for a minimum of half the central government budget deficit.17

Financial Flows Within the Public Enterprise Sector

As mentioned previously, intrasectoral flows constitute wash transactions when it comes to measuring the public enterprise sector’s impact vis-à-vis government. However, they are of interest in assessing the fiscal impact of a subset of enterprises constituting on net either a source or an object of flows to or from the rest of the sector. This interest arises from a presumption that, in the case of enterprises serving as a net source of funds to others, the former are relieving government of a burden that would otherwise fall on itself. Conversely, enterprises benefiting from the flows are viewed as diverting resources that could otherwise be channeled to the money market or the exchequer.

Three categories of flows, comprising (1) purchase of equity (EQIt1, with the superscript I symbolizing “intrasectoral”); (2) debt transactions (LNDtI and AMRtI), and (3) payments of return on investment (INTtI and DIVtI), are discussed adequately in the two preceding subsections, and considerations distinguishing intrasectoral flows from those between the public enterprise sector and government do not require further elaboration. Conversely, category (4), cross-subsidies (SUBtI), comprising at once a payment for one set of enterprises and a receipt for another set, applies uniquely to intrasectoral flows.

Perhaps the outstanding current manifestation of cross-subsidies within the public enterprise sector is the supply of energy at below-market prices to energy-intensive manufacturing industries, notably fertilizer, other petrochemicals, and basic metals. Examples of this are found in both developing and industrial countries. The subsidized supply of fuel or feedstocks by state-controlled petroleum and natural gas companies ensures the solvency of many processing industries that could not otherwise meet international competition. In effect, the energy suppliers are relieving the government budget and/or the credit market of at least part of the burden they would otherwise have to bear to keep the enterprises in question afloat. The other side of the coin is that the recipient enterprises are in effect diverting revenues and investable funds that would otherwise accrue to the budget and credit institutions. (Assessing the implicit subsidy becomes subject to greater uncertainty in a situation of production restraint as marginal revenue drops substantially below market price.)

Fiscal Impact of the Public Enterprise Sector: Conclusion

Table 3 summarizes the 17 subheadings and 24 individual flows, explicit and implicit, identified in this subsection.

Table 3.

Parameters of Public Enterprise Sector Fiscal Impact: A Listng

article image

The subscript t (= year) should be understood for all expressions.

The nominal fiscal balance (net flow of resources from the public enterprise sector to government) is the sum of accounting magnitudes under (1) less the sum of those under (2). The real fiscal balance is found by adding the relevant implicit transfers and subsidies to (1) and (2), respectively.

Apart from these overall indicators of fiscal impact, other indicators may be obtained by comparing various components of the flows and subtotals thereof. Potentially significant indicators include the following:


This term represents the difference between the sum of explicit and implicit grants and subsidies from government to the public enterprise sector, and the sum of the sector’s implicit transfers to government. If the first sum exceeds the second, one will most of the time be correct in asserting that the subsidies society is paying to the public enterprise sector exceed the costs that the sector is incurring in order to provide unrequited benefits to society. On the other hand, it would be a riskier proposition to assert the converse—namely, that if the first sum is less than the second, the sector is not recouping its full costs through subsidies, because vehicles of government transfer not included in the first sum—notably, purchase of equity and lending—are frequently tantamount to outright subsidies (the policymakers do not expect significant cash returns on, or redemption of, the investments, and/or effective machinery for realizing the same is lacking).


The two sums comprising this difference represent the nominal flows between government and the public enterprise sector most likely to be estimated in government budget documents. Hence, the expression may be viewed as an approximation to the sector’s budgetary impact determined by accessibility of data. It excludes, of course, the implicit transfers under (1) above, which means all the terms in (1) except SUBG, that is, explicit grants and subsidies. It also excludes tax payments of public enterprises, changes in tax arrears, and changes in arrears of government payments to the sector, which budget documents seldom reveal.

This is the expression used by R.P. Short to measure budgetary “burden” in the following paper.

Anticipating the analysis of the following section on monetary impact, given what is already known in many countries about the implications of the central government’s overall deficit for credit creation and the monetary expansion, comparison of the public enterprise sector’s fiscal deficit with the overall government deficit provides an indicator of the proportion of those phenomena that can be traced back to fiscal pressures from the sector.

Monetary Impact of the Public Enterprise Sector

The focus under this heading is on the net increment in credit associated with operations of nonfinancial public enterprises and its impact on other monetary parameters, such as the money supply, the level of price inflation, the structure of interest rates, and the balance of payments. Financial public enterprises are excluded here, because they function primarily as vehicles, rather than users, of credit; although when a financial public enterprise gets into difficulty, the use of credit or budget resources to bail it out has the same monetary impact as a bailout of a nonfinancial enterprise.

First Level of Monetary Aggregates—Gross Use of Credit

Computation of Credit Accounted for by the Nonfinancial Public Enterprise Sector

The first crude measures of monetary impact are the shares of assets and liabilities of domestic financial institutions accounted for by the public enterprise sector.18 It is convenient to distinguish among the central bank, other banks (whose deposits form part of the money supply), and nonbank financial institutions. Net credit outstanding directly to public enterprises is given by (CcbPDcbP)+(CbPDbP)+(CnbPDnbP) where C represents credit outstanding, D represents deposits, the superscript P refers to the public enterprise sector, and the subscripts represent the central bank (cb), other banks (b), and nonbank financial institutions (nb), respectively. In some countries, either or both of CcbP and DcbP will be negligible or zero, as it is central bank policy not to deal directly with public enterprises. Corresponding to each absolute measure of Cp and Dp will be a proportion of total credit outstanding or deposits in a given category of institution accounted for by public enterprises. These proportions already merit some policy interest—a large public enterprise share on the credit side suggests crowding out of the private sector, a large share in deposits suggests a major public enterprise contribution to financing the rest of the economy.

Apart from the financial sector’s direct lending to public enterprises, one must take into account recourse that government and private nonfinancial enterprises have to financial institutions in order to finance (1) net government loans and transfers to public enterprises and (2) arrears in public enterprise payments to the private nonfinancial sector.19 We denote (1) by CGP and (2) by Ap (A denotes arrears).

In the extreme case where one can assume that a government would reduce its deficit financing pari passu with a reduction in loans or subsidies to public enterprises, the volume of these transfers, identified in the analysis of the fiscal impact (preceding subsection), translates into an equivalent amount of credit. In the more realistic case, where the disutility associated with resort to deficit financing and consequent inflation strengthens the authorities’ resistance to demands A, B, and C for each unavoidable increase in demand D, one estimates a coefficient, say g, less than unity, giving the proportion by which the transfers in question augment the government’s deficit and thus its resort to credit.

The amount of financial sector credit accounted for by the public enterprise sector can now be defined as:


The sector’s proportional share may be denoted XPX; residual shares are accounted for by government and the private sector, direct credit to which is adjusted downward to account for the amounts gCGP and Ap, respectively, which are attributed to public enterprise operations.

Public Enterprise Sector’s Share in Real Credit Expansion

Once absolute credit levels accounted for by public enterprises have been assessed, the next step is to examine the sector’s share in real credit expansion during a given period, ranging from, say, one quarter to a year or more. Expansion of total credit from financial institutions is denoted by XtXt–1 and in real terms by Xt1+p˙Xt1 with p˙ representing the relevant inflation rate during period t. Real expansion of credit accounted for directly and indirectly by public enterprises is given by


The share of public enterprises in total credit expansion from t—1 to t is then given by


One’s policy-attuned ears perk up on hearing that public enterprises in Country A absorbed sp percent of the real increase in credit during year t, which together with direct government use (excluding the share accounted for by gCGP) of sG percent, left only SR percent for the private sector. We are especially disturbed if, by chance, SR turns out to be negative, which is perfectly possible—in other words, while financial sector credit expanded in real terms, credit to the private sector increased less rapidly than the price level. Short of this extreme case, any excess of the public enterprise sector’s marginal share in real credit expansion, sP, over the proportional share XPX may be indicative of growing pressure on the credit markets from the side of public enterprises.

Classified Loans to the Public Enterprise Sector

Another set of parameters of policy interest are those relating to amounts and proportions of loans outstanding to public enterprises that are “classified” or “graded” by management or outside examiners as being of doubtful collectibility. Monetary complications arise if public enterprises are threatening the solvency of financial institutions or merely constraining their flexibility while government delays making good on explicit or implicit guarantees.20

The Impact of the Public Enterprise Sector on Financial Institutions’ Assets and Liabilities

Turning from the gross parameters to the sector’s net impact on financial institutions’ assets and liabilities, estimating the latter becomes an ambiguous exercise. Short of general equilibrium modeling, the neutral course is to assume a break-even baseline, with capital costs covered, and to focus attention on operating deficits. An increment of ΔA in the aggregate operating deficit may inflate XP by anywhere between zero and ΔA. The extreme value of zero would prevail if monetary authorities and financial institutions collaborated to hold the nonfinancial public enterprise sector, the government as one of its creditors, and the sector’s private suppliers to a firm credit ceiling such that any increase in sectoral demand for credit occasioned by the incremental deficit was exactly offset by a decrease in the allocation of credit for public enterprise investment, working capital, etc. This implies that banks and other financial institutions would hold the line against reducing their loans to other borrowers apart from public enterprise creditors.

At the opposite extreme, the increment in the operating deficit induces some combination, equal to ΔA, of new deposits, credit diversion, and credit creation. New deposits would likely account for only a small fraction of AA, and that only if the money market were sufficiently free so that an increase in the effective demand for credit—assuming that public enterprises, with their government backing, constitute such demand—could evoke a rise in the expected return from holding financial assets.

Any of the three categories of financial institutions can divert credit from their other borrowers in order to increase the relevant CP term. The central bank can raise CcbP at will through credit creation. The most plausible assumption about banks and other institutions is that they will lend to the maximum permitted by the monetary authorities in the absence of the incremental deficit. Thus CnbP can rise only through additional deposits or credit diversion, while CbP can increase through either of those two mechanisms or, to the extent the central bank rediscounts the public enterprise paper, through credit creation.

In conclusion, only by collecting evidence at the microeconomic level regarding the behavior of the group of public enterprises accounting for the bulk of the aggregate operating deficit, together with that of their creditors; or, alternatively, via econometric modeling from time series with public enterprise deficits as the exogenous variable of primary interest;21 or, even better, using a combination of the two approaches, can one estimate the impact of sectoral operating deficits on total credit outstanding and its allocation between public enterprises and other borrowers.

Second Level of Monetary Aggregates—Money Supply, Level of Inflation, and Interest Rate Structure

Policy interest of a whole new order of magnitude attaches to appraisal of the nonfinancial public enterprise sector’s impact on these parameters. There are, of course, no direct measures of sectoral impact as there are of the sector’s participation in credit outstanding. Hence, one cannot avoid modeling (at least implicitly) the baseline alternative(s) and, as per the subsection, “First Level of Monetary Aggregates: Gross Use of Credit” above, attributing the credit by which deficits are financed among (1) new deposits, (2) credit diversion, and (3) credit creation. New deposits (ΔD), unless neutralized by action of the monetary authority, lead to expansion of the reserve base, and thus of the money supply, according to the familiar money multiplier (reciprocal of the reserve ratio), k,22 Credit diversion, on the other hand, has no direct impact on the money supply; in the medium and long term, repercussions may follow from its impact on resource allocation if relatively efficient activities are deprived of credit in order to finance public enterprise operating deficits.

Finally, credit creation arises from the monetary authority’s action in financing public enterprise deficits with high-powered money (ΔH)—that is, infusions leading to a one-for-one increase in bank reserves—whether via government or the banking system. Here again, the money multiplier comes into action.

The increment in money supply (ΔM) caused by a unit increment in public enterprise operating deficits is now seen to equal kD + ΔH). Of the two terms in parentheses, ΔH will almost invariably be dominant.

Turning to the level of inflation, the price index of output produced by the public enterprises themselves can be computed as an initial, essentially trivial, measure of sector impact. Such indices are readily available for individual enterprises, especially public utilities and major transport enterprises such as railroads, and the managements of nearly all public enterprises will (or should) have a good idea as to the trend of their output prices. The measure is trivial, however, because the force of the sector’s impact on the overall price level comes via the impact of its operating deficits on money supply, not the weight of its output in any general price index.

The impact on prices of any increase in money supply is heavily influenced by the degree of openness of the economy. At one extreme, in an economy with a high propensity to import that is not frustrated by quantitative restrictions, the purchasing power generated by the credit expansion quickly evaporates in the form of imports, and the mean additional demand in effect during any given period of time is considerably less than it would be in an economy where the authorities have imposed an absolute ceiling on imports in accordance with projections of available foreign exchange. The extreme of this latter case is represented by countries that maintain constant exchange rates during periods when their inflation rates greatly exceed the rise of prices (expressed in foreign currency units) of their imports, devaluation being one means of accelerating the evaporation of purchasing power without a commensurate rise in import ceilings, (The effect of the devaluation-induced rise in nominal c.i.f. values by way of offsetting the anti-inflationary impact of the leakage is normally trivial, since the imports already carry scarcity values far in excess of their c.i.f. prices.)

Computing Increment in Mean Level of Money Supply

Of interest in analyzing inflationary impact is the relationship between an injection of money (ΔM) at a given point in time and the mean level of the money supply prevailing during the subsequent period whose incremental inflation is to be measured. With instantaneous leakage of ΔM into imports, the level of the money supply that would have prevailed without the injection in question, which is denoted by M, continues to prevail in spite of the injection. Conversely, with zero leakage the mean level of money (M¯) during the period following the injection is equal to M + ΔM. The more general case features incremental imports of ΔI, which can also be interpreted to contain an element of capital outflow on the assumption that income receivers have a propensity to invest abroad out of incremental income. At the end of the period, money supply is given by M + ΔM – ΔI. On the simplifying assumption that the leakage into imports and capital outflow is spread evenly over the period, the mean level of money during the period would be given by


As a standard pattern of financing public enterprise deficits it is convenient to assume, for purposes of monetary analysis, a total injection of one unit of credit (ΔXP) during a given year, spread equally among v “income periods,” v representing the income velocity of money.23 Thus, 1/v amount of credit is injected at the out set of each period. By virtue of the definition of an income period, the amount of income generated in the first period by this infusion equals the amount of the infusion, less the portion of it leaking out via imports (and possibly also capital outflow—henceforward “imports” will be used as a proxy for both). Incremental imports are given by a constant marginal propensity to import (m) times incremental income, which is now seen to equal Δy = 1/vmΔy. Thus Δy(1+m)=1ν,ΔI=mν(1+m), and the increment in money as of the end of the period, following leakage of ΔI worth of imports, is 1/νmν(1+m)=1ν(1m1+m)=1ν(1+m)1

It can easily be shown that the increment in money at the end of period t following infusion of 1/v amount of credit at the beginning of a period defined as No. 1, assuming no further infusions in subsequent periods, is 1ν(1+m)t. Taking into account the infusions in succeeding periods, the increment in money at the end of each period t. consists of the residual—whatever portion has not yet leaked out via imports—of that period’s and each preceding period’s infusions. These residuals comprise a geometric progression whose sum can be written:


The next step is to calculate the increment in average money supply during the year, arising from credit creation to finance public enterprise deficits. For this purpose one first seeks the sum of successive money increments at the end of all v periods in the year; this is done by summing expression (7) over v, yielding another geometric progression, which can be written:


The best measure of the increment in average money supply is then the overall mean of the means of money supply values at the beginning and end of successive income periods, which may be written thus:


In other words intrayear values receive double weights as compared with successive year-end values, namely, ΔM0 and ΔMv.

The term ΔM0 is zero, since the relevant credit infusions begin only after the end of the preceding year. Thus the only adjustment that need be made in expression (8) in order to obtain (9), apart from dividing by v, is to subtract half of ΔMv, namely, 1(1+m)υ2m The increment in average money supply during year t attributable to public enterprise credit infusions is thus given by:


The asterisk atop the Δ in expression (10) denotes an incremental value attributable to expansion solely of credit to public enterprises; this notation will be used for other variables similarly affected by the credit expansion in question.

Computing Increment in Prices

A simple quantity theory model will be applied in analyzing the impact of public enterprise sector credit expansion on inflation. The model is based on the identity M¯ν =P¯Q, where M¯ denotes the average level of money supply during, a year, v the income velocity of money, P¯ the average price index during the year, and Q the year’s GDP at prices of the year serving as the base for the price index.

The model’s key assumption is that v remains constant—in other words, one abstracts from dynamic effects that might cause the marginal income velocity applicable to ΔM to differ from the average value, v, applicable to M.

With v constant, the following identity holds for the case where an increment in public enterprise sector credit has had the indicated effects (Δ* terms) on average money supply, the average price index and GDP:


This identity is then divided by the analogous one applying in the absence of the increment in public enterprise sector credit, namely, M¯tν =P¯tQt; the v’s cancel out, as do the unity terms,


on both sides of the new equation, which now reads


It is clear on inspection of equation (12) that, should Δ*Qt equal zero—that is, expansion of public enterprise sector credit brings no change in real output—then the proportional increment in average prices is identical to that in average money supply. It may be argued that some enterprises will have to cease production and discharge their labor forces if their deficits are not financed. Conversely, in many countries improved trade and incentive policies would facilitate re-employment of the public enterprise labor and pick up the slack in production within a short time.

In order to compute the impact of the public enterprise sector credit expansion on the price level over the course of a full year, the term representing the increment in the average price level must be modified accordingly. It would be consistent with the assumption of equal amounts of credit creation in each income period to treat Δ*P¯t as simply half of the full-year price level increment, Δ*Pt.

Accordingly equation (12) can be written:


Useful indicators of inflationary impact are (1) the ratio of Δ*Pt to pt-1, showing the proportional increase over the end-of-last-year index that can be attributed to the public enterprise sector credit expansion and (2) Δ*Pt, as a ratio to year PtPt1 +Δ*Pt, showing the proportion of the year’s inflation that can be so attributed.24

In many countries, more sophisticated models have been applied, using time series data to relate public enterprise deficits to money supply and inflation, or simply to relate money supply to inflation. Such models can be adapted to the pattern of inflationary expectations, lag structures, and other dynamic characteristics of a particular economy to yield more reliable estimates than the naive predictions of a simple quantity theory model.

One such exercise that is of more than passing interest here—because it examines the impact of price controls on operating deficits, credit creation, and ultimately inflation with respect to both public enterprises and private firms—was reported in an article on Argentina by Chu and Feltenstein.25 Government transfers to public enterprises translated one-for-one into high-powered money through central bank financing of the government deficit, while private losses were financed by the commercial banks and were not, as a general rule, rediscounted by the central bank. Hence, the Chu and Feltenstein data suggest that the public enterprise transfers were proportionately ten times as inflationary per unit of transfer as the private losses; however, since the latter were so much larger in aggregate, they contributed 2.3 times as much as the former to Argentina’s 15 percent average quarterly inflation rate during 1965–76.26

Impact on Interest Rate Structure

In conclusion, the possible impact of operating deficits of public enterprises on interest rate structure is worthy of passing mention. In the event—more common in industrial than in developing countries—that nominal rates are allowed to respond to modest shifts in demand and supply schedules in the formal money market, the occurrence of public enterprise deficits and the corresponding upward shift of the credit demand curve would force the rates upward. To estimate this effect, one would look primarily at cross—section data on the elasticities of credit supply and demand with respect to perceived financial returns. Over the medium to long term, one would have to take into account feedback from inflation fueled by the public enterprise deficits.

Balance of Payments

A naive approach to estimating the impact of the public enterprise sector on the balance of payments would be to attempt to draw up a balance of credits and debits on current and capital accounts for which the components of the sector are directly responsible. Apart from being an extremely tedious assignment this would lead the analyst into a maze of classification problems. What to do about goods and services purchased from a private importer? What about goods purchased from a local private assembler who embellishes imported inputs with negative value added? What about production of tradables that happen to be consumed domestically in one year and exported the next?

These issues illustrate why attempting to draw such a foreign exchange balance for an industry or a sector classified along other lines is not an economically meaningful exercise. A country’s most efficient public enterprises and/or those whose expansion would yield the highest social benefit may be public utilities earning no foreign exchange on their own account while expending large sums on imported inputs. Instead, what is of interest is whether an industry or sector yields sufficient benefits to cover its social opportunity costs, with tradable inputs and outputs valued at border prices. That issue was already covered in “Summary Accounts of the Public Enterprise Sector” in connection with the public enterprise sector’s contribution to GDP.

Balance on Capital Account

The capital account is arguably an exception to the stricture against attempting to compute a public enterprise sector balance of payments. It will often be of interest to determine whether the sector is contributing a net inflow or outflow of capital. Inflow arises primarily from borrowing, typically medium-to-long-term from official lending agencies, commercial banks, and suppliers, normally backed by host government guarantees. Additionally, petroleum and minerals exporters, power companies, and public enterprises generally in countries regarded as particularly creditworthy have also borrowed internationally at the short end of the spectrum. Foreign equity investment is effectively limited to enterprises with a reasonably assured export market, or situations where the foreign partner derives a satisfactory return from supplying goods and services to the enterprise, and hence can afford to take a nominal equity whose return is uncertain.

Capital outflow arises primarily from repayment of borrowings, though again petroleum and minerals exporters and a few other well-positioned enterprises have transferred funds either to take advantage of interest rate differentials and other short-term earnings opportunities, or even as direct foreign investment. Notwithstanding government’s nominal control over the sector, public enterprises engaging in such operations often find it advantageous to disguise them via transfer pricing, and measurement of the flows by outsiders is subject to considerable uncertainty.

Nevertheless, where public enterprises are sufficiently active in the capital market to affect the outcome of government stabilization policies, an attempt should be made to determine the direction of the net flow and measure it.

Focusing on medium-to-long-term debt capital, a finding that, for example, the public enterprise sector had switched from a position of net inflow to one where amortization payments strongly outweighed loan disbursements, with a concomitant negative impact on the balance of payments, might reflect a relatively stagnant position calling for reform measures. Yet such a conclusion is by no means automatic, since the imbalance in question could also reflect a process of the state divesting itself of those enterprises strong enough to continue borrowing, or a policy of suspending new borrowing to allow time to digest a spate of ill-advised investments.

Import Generation from Credit Created to Finance Public Enterprise Deficits

In many countries the most significant element in the public enterprise sector’s balance of payments impact is the flow of payments associated with the creation of credit to finance operating deficits, if any. The subsection “Second Level of Monetary Aggregates—Money Supply, Level of Inflation, and Interest Rate Structure,” above, examined the case of a strict regime of import and exchange controls under which the government prevents incremental credit from leaking into international payments. The converse situation of an open economy lends itself to analysis via the Polak model of money, income, and the balance of payments, which adds the assumption of a constant propensity to import (marginal = average) to the simple quantity theory’s assumption of constant income velocity.

The rate of generation of imports resulting from a unit expansion of credit at the beginning of a period is determined by a coefficient mv, equal to the product of the propensity to import and income velocity of money, which in turn is equal to imports divided by money supply (the GDP expression in the denominator of the import propensity and the numerator of the velocity term cancel out). For example, import generation during the first year following the injection equals mν1+mmν.27

For analysis of the impact of deficit financing measured over a period of one year, which will be the normal application of this approach, it is useful to derive coefficients giving import generation during the same year (while the deficit in question is also being financed) and then during subsequent years (without taking into account financing of new deficits in those later years). This is done on the basis of the simplifying (but reasonable) assumption that equal increments of credit/money are created in each of v “income periods” during the given year, summing to one unit. The relevant formulas are shown in Table 4.

Table 4.

Formulas for Polak Import Generation Coefficients1

article image
Source: J. J Polak and Loretle Boissonneault. “Monetary Analysis of Income and Imports and its Statistical Application,” International Monetary Fund, Staff Papers. Vol. 7 (April 1960). p 356.

The formula for each year gives the coefficient of import generation for that year caused by creation of one unit of credit/money during year 0 distributed equally among v income periods of year 0. propensity to import and income velocity of money supply being denoted by m and v. respectively.

With increasing values of n, the summation of the formulas for years 0 through n rapidly approaches 1.0. For typical developing country values of mv, additional imports/reserve loss cumulate to at least 82 percent of AM after two years (that is, by the end of year 1) and 94 percent after three years.28


Assessing the Impact of Public Enterprise on Attainment of Noncommercial Objectives

Whatever measurements made pursuant to Section V may tell one about the impact of Country A’s public enterprise sector on economic aggregates, including particularly fiscal and monetary magnitudes, even if one succeeds in conducting a general equilibrium analysis that compares observed parameters with those of a baseline alternative, the elements required for a definitive policy prescription will not be complete until the sector’s impact on attainment of government’s noncommercial objectives has also been assessed. Actual policy decisions, even apparent nondecisions that result in maintenance of the status quo, are the outcome of a (normally more implicit than explicit) process of weighing costs and benefits associated with the impact of alternative courses of action on a wide range of objectives. Depending on the structure of the objective function (s) of Country A’s decision makers (i.e., the value of the weights attached to some unit of advancement of each objective), perfectly rational behavior on their part may be to choose alternatives that appear to conflict with standard goals of fiscal and monetary policy.

There can be no substitutes for adequately trained local officials in the task of analyzing implicit objective functions and measuring the public enterprise sector’s impact on their fulfillment if such work is ultimately to influence the future choices of a country’s policymakers. On the other hand, outsiders who aspire to conduct the more straightforward computations of Section V, and thereby to create presumptive evidence for the desirability of certain policy changes, perhaps reinforced by conditionality of external assistance, should be aware of these more subjective elements in the policy equation and should be prepared to exhort their local counterparts to undertake the necessary work on them.

In this spirit, the present appendix sets forth a taxonomy of noncommercial objectives; indicates in what ways public enterprises are considered by many governments to be effective vehicles for promoting them; suggests conditions under which some objectives might be as well (or even better) served by an alternative industrial organization; and, in some cases, touches on a possible approach to assessing the public enterprise sector’s impact on the objectives. Objectives are organized under five headings: (1) economic stabilization, (2) economic growth, (3) income redistribution, (4) localization/indigenization, and (5) miscellaneous.

Economic Stabilization

Control of Inflation

Policymakers in many countries believe that government’s more direct control over the pricing decisions of public enterprises gives it a better handle on controlling inflation than is available to it through regulation of the private sector. While admitting the inflationary impact of financing, with high-powered money, public enterprise operating deficits that result from price control, advocates of this position argue that quiet restraint of public enterprise pricing, avoiding the fanfare and speculation that accompany broad-based price ceilings, retards inflationary expectations, thus limiting the dynamics of the process. If such an effect is, indeed, present, it will be difficult to spot through partial equilibrium comparison of public enterprise aggregates with criterion values, though it will still come out in an appropriate general equilibrium model.

On the other hand, the Chu and Feltenstein study of Argentina cited in the subsection “Monetary Impact of the Public Enterprise Sector” illustrates how the initial objectives of price control can be achieved through the private sector, as long as compensatory financing is available from the banking system (to be sure, ultimately undermining the anti-inflation objective); moreover, if the supply of commodities is insufficient to meet the demand at controlled prices and unless government’s enforcement network can monitor each point of sale to final users, exchange will eventually take place at scarcity-determined prices, regardless of what role public enterprises have played in production and prior stages of distribution.

Subject to these observations, analysis of the public enterprise sector’s impact on inflation flows from the aggregates discussed in “Monetary Impact of the Public Enterprise Sector.”

Food Security

A prime objective of virtually every government is to ensure that the demand of the population for basic foodstuffs, at price levels perceived as politically acceptable, is met. Political pressures from the producer side counterbalance, to some extent, those emanating from the consumers, and some policymakers are aware of a positive functional relationship between producer prices, on the one hand, and the domestic supply (net of imports), on the other. These conflicting pressures lead many governments into price control measures and other marketing policies that put a financial squeeze on one or more components of the production/marketing system—sellers of inputs, producers using significant quantities of purchased inputs, transporters, processing units, marketing boards, wholesalers, retailers, and/or other categories of intermediaries between producers and final consumers.

Food security issues may be regarded as a subset of those falling within the ambit of anti-inflation policy; however, they merit separate identification in the present context, both by virtue of special characteristics of the public enterprises (notably central agencies for input supply and/or commodity purchasing and marketing) that bear the financial brunt of government policies in the sector and because of the emphasis which even short-term policies normally place on the supply side, drawing on imports to close the gap. By contrast, price control policy in other sectors traditionally pays less heed, at least in the short run, to ensuring supplies that will clear the market at the target price, with the result that controls invariably engender nonprice rationing.

Some governments meet their food price and supply objectives by operating through private, rather than public, enterprises. The approach is normally to (1) prevent monopolization or the operation of domestic cartels in importation and exportation of foodstuffs, and (2) subsidize imports or exports (depending on the domestic demand/supply balance in a given period) with as much cash on the barrel as is needed to maintain the desired retail price ceiling or producer price floor. Reports on some countries by outside observers have taken the position that such an arrangement would evoke more production and reduce effective consumer prices with little, if any, additional fiscal burden as compared with the present situation of intervention by state-controlled marketing agencies.

Noting the opposite stand taken by other observers who attribute certain recorded supply disturbances to unbridled speculation by private agents, the present paper takes no position on the merits of the case. Rather, it contents itself with noting that the issue of food security does not necessarily overlap with that of public enterprise; conversely, where such overlap occurs, the crucial question is the comparative efficiency of the public sector as entrepreneur and manager in this particular activity.

Dampening Economic Downturns

The principle underlying this objective is that by virtue of its control over public enterprises, government can direct them into countercyclical measures consistent with its efforts to maintain output and employment in times of recession. In other words, public enterprises can be relied on more consistently than private firms, notwithstanding the regulation to which the latter can be subjected, to maintain production, build up inventories, and/or, regardless of the course of their production, keep staff on the payroll in the face of slack demand.

As with virtually all noncommercial objectives, this one worsens the enterprises’ financial position and aggravates the sector’s fiscal and monetary impact. This becomes especially clear when one examines cases where a failing private enterprise has been nationalized to prevent it from closing its doors and releasing its labor force. A particular case in point is the Bolivian tin industry, where the continuing deficits of such enterprises pose major fiscal and monetary burdens.

To assess the public enterprise sector’s performance in preserving jobs, which is clearly its prime counterrecessionary contribution, one estimates the fluctuation in employment that would have characterized the units in question had they been under private control. (For this purpose one may apply a coefficient derived from observing the modern private sector’s employment response to business downturns.)

The culmination of this evaluation is a general equilibrium model in which a coefficient of benefit (in monetary terms) is multiplied by person-months of employment preserved by the public enterprise sector. Determining the coefficient is tantamount to a wild exercise of value judgment in which one estimates the probability that unemployment beyond a certain level will cause political disruption and multiplies it by an estimate of the economic cost of the disruption. The opposite side of the balance sheet will contain cost parameters corresponding to the fiscal and monetary impact of the deficits incurred by virtue of preserving the jobs. (In some situations, one might argue that it should also account for output sacrificed because the employment policy in question raises the cost of labor to other producers.)

Whatever the balance of economic costs and benefits, situations will be encountered where the national political leadership weighs the avoidance of political disruption so heavily as to offset virtually any conceivable increment to the budget deficit and inflation associated with preserving modern-sector jobs.

Economic Growth

Increasing Investment, Output, Exports, Income, and Employment

For this subset of objectives, emphasis is placed on the absolute levels of the parameters in question rather than on the locus of control of economic activity or its distributional impact.

The possibility that greater reliance on public enterprise would increase output, income, and employment, as compared with a baseline alternative, depends on whether such a policy would (1) raise investment; (2) affect the sectoral composition of output so as to maximize backward and forward “linkage” effects; and/or (3) utilize capital stock more efficiently and/or employ more workers per unit of investment.

Taking the third possibility first, it turns out that relatively efficient utilization of capital stock has not been a hallmark of public enterprise in most mixed economies, and this has been reflected in the operating deficits and implicit capital subsidies discussed in the subsection on the “Fiscal Impact of the Public Enterprise Sector.” Some public enterprises have attained higher employee/capital ratios than is characteristic of some private firms in the same industries, but more often than not this has been the result of political patronage, which is, again, the antithesis of efficiency. On the other hand, many instances have been documented where, precisely because of their access to subsidized capital, public enterprises have chosen more capital-intensive technology than was suitable for the country’s factor proportions. Whether or not these enterprises were then obliged to employ redundant workers, forcing up the employee/capital ratio, efficiency was impaired.

Turning to linkages, many governments have acted on the supposition that these are enhanced by promoting basic industries, such as iron and steel and heavy chemicals, as well as industries that process the output of the country’s primary producing sectors—notably, agriculture, livestock, forestry, hard minerals, and hydrocarbons—over and above the levels of investment that would occur without direct government intervention. On the other hand, some countries have found more effective, largely skill-related linkages in expon-oriented light manufacture utilizing imported inputs where ability to respond more rapidly and flexibly to changing market opportunities appears to give private enterprise a comparative advantage.

A policy favoring public enterprise may raise the investment rate in one of two ways—first, it may steer into investment domestic savings that, for want of private investment demand (lack of entrepreneurs, unfavorable business outlook, or what not), would not translate into a similar investment rate through the private sector. In other words, if steered into private hands by public credit institutions or other intermediaries, the given volume of resources would largely displace the financing of private investment through other resources, and the economy would end up with a higher consumption/savings ratio than under a public enterprise orientation. Second, the policy may open up channels for a larger inflow of foreign capital titan the sources of such capital would choose to make available for on-lending to private investors. Whether the stepped-up investment rate makes output, income, and employment over the long run higher than they would have been in the baseline case depends on the efficiency of the investment as well as on the debt service burden generated by the increased capital inflow.

To evaluate the impact of public enterprise under these headings without recourse to a general equilibrium model, one may compare (1) social rates of return, with nominal values of tradable goods and services being adjusted for border prices, along with (2) capital/value added, and (3) employment/capital ratios between public enterprise and private firms that are analogous in other respects. Also relevant are characteristics of the currently unsatisfied private demand for credit, and the willingness of foreign capital sources to channel funds to the local private sector through financial institutions, as opposed to lending to public enterprises.

Accelerating Industrialization

As a rationale for the establishment of public enterprises in the manufacturing sector, of greater importance than any ideological preference for state versus private ownership has been, in many countries, the perception that private capital and entrepreneurship were not mobilizing as rapidly as desirable, from the viewpoint of social objectives—and indeed, given various and sundry constraints, could not be expected to do so—to exploit socially profitable investment and production opportunities, especially in the field of import substitution.

By international standards, many of the facilities established in this area by the public sector have been too small to exploit the economies of scale associated with modern technology, and many have operated at only a small fraction of even the limited capacity with which they were established. High average production costs resulting from these (and other) circumstances have engendered financial deficits for many such enterprises, creating yet another source of pressure on government budgets and credit markets.

On the other hand, the problem is not generic to public enterprises. Government policy has drawn large sums of private capital into inefficient import-substitution industries, which have not always received, ex postfacto, the degree of protection required to support selling prices that would cover costs.29

Before the problem is made to took too much like that faced under the earlier topic of price control, it should lie noted that the economist’s optimizing concern here approaches the price issue from the opposite direction, seeking to avoid the distortions that arise where resources are allocated to high-cost activities when improved policies would allocate them to efficient uses and thereby accelerate the growth of output and employment. It is this concern that lends one to reject additional protection as a necessarily desirable solution to the financial strains encountered by public and/or private enterprises in this area.

Likewise, the absence of such strains in import-substituting public enterprises, if achieved by high effective protection that distorts the economy’s cost structure and discourages export-led growth, should not be taken to mean that the enterprises in question pose no problem from the viewpoint of macroeconomic policy. A de facto policy of ensuring financial viability for a wide range of industries with very high unit domestic resource costs, even negative value added at border prices, certainly qualifies as a macroeconomic policy, and one highly counterproductive from the viewpoint of long-term economic and political stability.

Income Redistribution

Promotion of Small-Scale Producers Through Credit

This objective comes first under the heading of income redistribution because it has loomed large in the financial position of public credit institutions in many countries. The problem arises when governments—often aided and abetted by donor agencies—direct credit institutions to finance provision of capital equipment, current inputs and/or working capital to farmers, artisans, and other small-scale producers whose collateral, if any, cannot be readily converted into cash. The borrowers are not oblivious to the difficulties, whether economic or political, that creditors would face in trying to foreclose on and sell their land or movable assets, and except where management conveys a credible threat of cutting off future credit (which has often proven an elusive goal) and, moreover, the net present value of the future credit exceeds that of the current amounts due, the institutions in question have encountered serious problems of arrears and defaults.30

Insofar as an initial government subvention leaks out, is not replenished, and the program winds down, continuing fiscal implications are absent. But where the government faces political pressure to continue lending, as in the case of agricultural credit that is cast as seasonal crop finance, the repayment shortfall may engender ongoing fiscal pressure.

In many countries, the problem is assailed as a reflection of weak management of public credit institutions and perhaps also as a slackening of political discipline when governments hesitate to crack down on a large mass of low-income defaulters. The policy discussion revolves around ways and means of improving management and stiffening politicians’ backbones. However, even if implementation of the policy were entrusted to the private money market, institutions in the formal sector would ordinarily seek government guarantees before intervening in markets with dubious collateral, and continuation of the program would involve a fiscal burden.

Once more the quintessential public enterprise issue is whether social objectives are enhanced by entrusting the activity to public institutions as opposed to working out some cost reimbursement and loan guarantee arrangement with private agencies.31

Other Modes of Vertical Income Redistribution

The public enterprise mode of industrial organization can contribute to achievement of income distribution goals because sectors of investment can be chosen and employment configurations designed so that disposable incorile flows are targeted at particular segments of the population.

On the other hand, critiques of public enterprise performance in some countries have alleged that, by appropriating an undue share of the nation’s fiscal and credit resources for a small set of highly capital-intensive, corruptly installed, and/or badly managed public enterprises, governments, while undoubtedly raising the disposable incomes of a small privileged class of managers and workers, have retarded the creation and expansion of a much bigger set of small- and medium-sized private businesses that would have drawn a larger segment of the labor force into modern economic activity and thus would, in the final analysis, have made for a lower Gini coefficient of income inequality. While these allegations certainly do not apply to every country, they raise a set of issues that must be addressed in measuring the distributional impact of public enterprise. Given the extensive overlap between the issues of employment creation and income distribution, the parameters to be evaluated under the present heading include those listed above under “Economic Growth,” notably comparative capital/value added and employment/capital ratios, and characteristics of the unsatisfied private demand for credit.

Geographical Redistribution (Promoting Regional Equity)

Once again, a public enterprise-oriented policy confers direct control over the locations whose populations are to benefit from an expansion of economic activity.

The appropriate baseline with which to compare the contribution of public enterprise toward regional equity is a program of subsidies and other incentives by which government would seek to attract private investment, domestic or foreign, into less developed regions. Public enterprise will come out ahead in the comparison if its fiscal impact is less burdensome than the incentives government would have had to offer private firms to generate the same disposable income. In the absence of a negotiating record that shows just what private investors asked in order to locate in a region, this exercise will unavoidably involve a greater or lesser degree of speculation.

In the event public enterprise is shown to have pumped X amount of disposable income into a region at lower social cost than might have been expected from the baseline alternative, the question of negative, zero, or positive net present value will still remain. An approach that has won considerable favor in recent years is the revealed preference method expounded in the United Nations Industrial Development Organization (UNIDO) project evaluation guidelines, whereby the analyst assesses the implicit weight given by policymakers to an increment of disposable income in Region A as opposed to Region B by examining their past or present choices among investment alternatives.32 This paper will not tarry on this point, which is tangential to the intrinsic social efficiency of public enterprises.


Supply Sources

The principle here is to ensure production locally of certain goods and services that are regarded as “strategic” to the economy, the defense effort, or whatever, and in which the private sector evinces little interest or demands excessive subsidies. The key question in evaluation: have the state-controlled producers succeeded in generating sufficient local value added, measured at border prices, to make the country’s dependence on imported inputs less of a problem than its previous dependence on imports of the final product was perceived to be?

Asset Ownership and Control

Here the objective is to displace or forestall foreign ownership and control of existing and potential productive facilities. The current government’s intention may be to leave the facilities in state hands indefinitely or to turn them over to local private ownership as and when the government considers the private sector “ready” to receive them. (“Readiness” may be measured in the proportion of share value, on whatever basis estimated, that local private investors are able and willing to put up out of their own resources, which should not exceed a “reasonable” amount of debt capital; it may also relate to government’s perception of the availability of private managerial talent.) Among criteria for performance evaluation: does government exercise de facto control over the facilities in question, or is it effectively at the mercy of foreign managers and suppliers, its nominal partners in a joint venture? What is the record of divestiture of state-owned facilities to local private investors? To what extent are the latter merely “fronting” for foreigners?


Some governments consider labor legislation and immigration control over employment of foreign manpower by private firms to be an inadequate mechanism for ensuring citizens access to skilled and white-collar positions. Accordingly, public enterprise is viewed as a vehicle for accelerated training and promotion of citizen labor. Partial evaluation of its performance in this regard may be achieved by comparing the relevant cadres of analogous enterprises under public and private (domestic or foreign) control. The far more complex general equilibrium issue is: which mode of industrial organization ultimately prepares more citizens for effective economic participation at middle and upper levels? Should the allegation prove correct that public enterprise in some countries crowds out more value added under the aegis of small- and medium-sized private business than it generates itself, the case for the public enterprise regime as a superior contributor to human capital formation in those countries would be weakened.

National Policymaking

The fact that this is the most nebulous of the subobjectives under localization does not gainsay its importance as a rationale for resort to public enterprise, based on the perception that concentration in the foreign headquarters of multinational firms of authority over investment, employment, procurement, production, pricing, and marketing decisions of major local enterprises heavily circumscribes the ability of the government to determine macroeconomic policies in these areas. This concern is cited when certain areas of production are closed to foreign investment, and it also partly motivates a policy on the part of some governments to resist the establishment of branches and wholly owned subsidiaries of foreign companies in favor of joint ventures with sufficient public sector participation to qualify as public enterprises. To evaluate achievement of the objective, one reviews the performance of the enterprises in whose establishment this concern has been a motivating factor and considers how they might have acted differently in the areas listed above, had they responded solely to interests of multinationals.

The mere fact that an enterprise qualifies as public in no way guarantees that its performance adheres more closely to government policy guidelines (with the degree of ambiguity that often attaches to identifying what these are) than would have that of a multinational subsidiary. An extensive and rapidly growing literature dealing with the behavior of public enterprise managers points to many dichotomies between their interests and government’s policy concerns.

Miscellaneous Objectives

Concluding the discussion of noncommercial objectives supposedly served by the establishment of public enterprises are two that figure less often in explicit rationalizations, while carrying weight from time to time in some countries.

Ensuring Government Access to Market Information for Regulatory Purposes

The principle here is that government, by entering certain lines of production on its own account, acquires hard data about costs and selling prices with which to design more enlightened price control measures and other regulatory policies vis-à-vis the remaining enterprises in an industry. For purposes of evaluation, the analyst would require access to the flow of information from the public enterprises in question to the regulatory authorities and would have to judge the effectiveness of its use by the latter. Once again the fact that a public enterprise may have access to “inside” information pertaining to its industry does not guarantee that management will consider it in its own interest to reveal the complete picture to government regulators.

Improving Working Conditions

With this variant of subobjectives relating to employment, a government seeks to provide better working conditions for the labor force in a given line of production than those prevailing under private management or thought likely to prevail should an industry be left to develop under private auspices. To be sure, such magnanimity vis-à-vis the proletariat is likely to bear a fiscal, and even a monetary, cost. Evaluation of performance is a relatively simple matter of comparing terms and conditions of service and work environment of the public enterprise in question with a sample of private firms of comparable size. Intensified regulation of the private sector in this area should not be overlooked as a baseline alternative.

Concluding Note to Appendix I

It is virtually axiomatic that a given number of economists will produce at least as many different typologies of public enterprise noncommercial objectives. One prepared some seven years before the present paper is of particular interest because its author went to the trouble of drawing up a matrix of information requirements associated with assessing attainment of the objectives.33 The matrix is reproduced on pages 102–103 as Table 5.

Table 5.

Nonfinancial Public Enterprises: Objectives and Evaluation

Types of Information Required

article image
Source: Alan A. Tait, “The Fiscal Policy Objectives of Nonfinancial Public Enterprises and Their Information Requirements: A Simple Taxonomy Suggested” (unpublished. International Monetary Fund, 1977).
article image


Computation of Public Enterprise Sector Capital Stock as Determinant of Annualized Opportunity Cost of Capital

Evaluation of the public enterprise sector’s performance in covering the social opportunity cost of its capital, and measurement of the subsidy involved where this cost is not covered, presuppose estimation of the sector’s capital stock.

The first step in deriving such an estimate is to compile a series on annual gross capital formation, revaluing each year’s historical value by the quotient of the relevant price index for the latest year (the year in whose prices capital stock is to be expressed) divided by the index of the investment year in question. If separate price indices are available for components of capital formation (construction, machinery imports, vehicles, etc.), then those should be used for the revaluation. A wholesale price index would be most appropriate for revaluing inventory changes. In some countries, one will have to be content with using a crude consumer price index for the entire exercise.

A series of annual capital depreciation is now needed in order to arrive at a net capital stock. Revaluing the data for each year of the historical series would be extremely cumbersome. A suggested alternative approach is to break down the constant-price capital formation series by broad asset class and to apply a perpetual inventory technique to each class according to its mean economic life. Such an approach has been used by Harberger to calculate the return on national capital stocks of a number of countries.34 A minimum breakdown (used by Harberger) would distinguish between buildings, machinery and equipment, and inventory; data in some countries would permit a more detailed classification, separating out, for example, vehicles. Harberger assumes mean asset lives of 40 years for buildings and 12½ years for machinery and equipment, yielding annual depreciation rates of 2½ percent and 8 percent, respectively.

The trick in applying this approach is to obtain a plausible estimate for the capital stock in some initial year. The longer the investment series, and thus the earlier that year, the less any overestimation or underestimation of capital stock at that point will influence the current estimate, since more of the former will have depreciated out of existence. In some countries, the public enterprise sector has mushroomed to such an extent in recent years that current capital stock is overwhelmingly determined by investment series going back only a short time. Where this is not so and annual investment data are not readily available for earlier years, one can follow Harberger in picking a period of, say, three years when there are reasons for presuming that value added in the public enterprise sector and capital stock of the different asset classes all grew at approximately the same rate.

In this case, the equation It, = (d + r) Kt-1 holds (reflecting investment’s two components, (1) replacement, and (2) augmentation of capital stock), where t denotes the middle year of the period, I denotes annual investment in a given asset class, K denotes the year-end stock of that class, d denotes the relevant depreciation rate, and r denotes the common growth rate of value added and capital stock during the three-year period only (values of r for subsequent periods are irrelevant). To eliminate the effects of random fluctuations in investment, one averages I over the three-year period. Then capital stock of the class in question at the end of year t equals It+1d+r stock at the end of year t + 1 equals (1 – d) times stock at the end of year t plus investment during t + 1, that is,


and values of capital stock emerge readily as functions of d, r (during the initial period), and successive values of I.

The gross capital formation series may require adjustment insofar as it was financed by an inflow of foreign capital, say X foreign currency units, tied so closely to the investment(s) in question that, had those investments not been carried out, the total inflow would have been X units less. (To allow for the possibility that inflows in later periods might have been correspondingly higher to compensate for omission of X in an earlier period, we must convert the comparison to present value terms, discounting foreign capital inflows at the social discount rate.)

In this case, the correct measure of the annualized social opportunity cost of the capital in question is not the product of its constant-price depreciated value and the social discount rate, which costs each unit of capital at the return it would have yielded in the investment opportunity forgone at the margin. Rather, the correct measure of annualized cost is the annual interest or dividend payment made to the investor or provider of funds by way of servicing the capital transfer in question. When the annualized cost of the relevant assets is taken into account in this way, the corresponding asset value must be subtracted from the remaining capital stock by which the discount rate is to be multiplied.

For example, if the transfer comprises a “soft” loan bearing interest well below the recipient country’s social discount rate—and of course the longer the amortization period at any given interest rate meeting this condition, the softer the loan, in other words the larger its grant element—then the annual interest payment is correspondingly well below the product of asset value and the social discount rate, and the procedure outlined here makes for a lower annualized social opportunity cost than would prevail were all investment resources fungible.35


Quality of Investment Decision Making: An Indirect Indicator of Public Enterprise Sector Impact

The need to define this final area of macroeconomic performance of the public enterprise sector arises because reports have been published alleging that declared pursuit of either commercial or noncommercial objectives cannot always be taken at face value.

The discussion in the main body of this paper proceeds on the assumption that policy decisions with regard to the establishment and operation of public enterprises are made in accordance with sincere perceptions of national interest on the part of the authorities. Thus, even where public enterprise operations impose substantial fiscal and monetary burdens, the paper assumes that the sector plays the role it does because the authorities have concluded rationally that the social benefits gained thereby are bona fide and more than offset the costs.

The reports in question allege, on the other hand, that corruption has served, in some situations, as a primary motive in the establishment of certain public enterprises other than public utilities and natural monopolies. In essence, it is charged, establishment of these entities represented a move by one or more politicians and/or officials to tap into a circuit of transactions occasioned by the sale of goods and services and to divert a portion of it to the consumption and aggrandizement of self, family, and friends. It is noted that the initiative for establishment often originated with the private sector in the form of promotors, suppliers, and other interests who have been rewarded with a share of the “take.” Under one frequently cited scenario, foreign suppliers are said to have induced officials to establish public enterprises as customers for their wares, notably factory machinery. In return for a kickback, the officials have supposedly arranged public financing, local and foreign (the tatter on the strength of government guarantees), and enacted measures of protection against domestic and foreign competition. Apart from obvious social costs imposed by such practices it is alleged that they have undermined international competitiveness and choked off industrial growth via balance of payments pressures.

Sociologists focusing on the study of political elites and the bureaucracy build scientific models that predict the response of officials, under specified circumstances, in situations offering such temptations. They add the obvious caveat that corruption is neither limited to the formulation of public enterprise policy, nor more characteristic of it than of other governmental functions.

The present discussion takes no position regarding the truth of the allegations in question. However, should they turn out to be correct in certain situations, it could be argued that economists would mislead by confining their attention to measuring the public enterprise sector’s attainment of putative goals that had little to do with decisions bearing on its organization. In such cases, it would be incumbent on them to seek indicators bearing on the overall quality of the decision-making process with respect to establishment and expansion of public enterprises. In this vein, questions such as the following would have to be addressed.

Choices among alternative uses of resources

How serious an attempt is made to weigh alternative strategies for, and configurations of, the public enterprise sector according to commercial and noncommercial objectives? Are scarce resources expended in collecting relevant data and commissioning studies by bona fide, disinterested experts? Do the authorities appear to give serious weight to the findings of such studies and recommendations by their technical staffs? Conversely, do the authorities’ consultations regarding alternative courses of action appear to be limited to promoters and would-be suppliers? Are feasibility studies at best perfunctory, at worse merely disguised sales pitches?

Design of investments

Once it has been effectively decided to proceed with a certain category of activity under the aegis of public enterprise, how serious is the consideration given to different technologies of production, and different institutional and legal arrangements for implementation? How serious an effort is made to minimize costs for a given level of output? Is competitive tendering provided for? Conversely, do the authorities seem to rush into approval of a design that suits the convenience of a particular supplier rather than offering economies to the host country?

Investment execution

Are standard arrangements made for supervision of procurement, execution of construction contracts, etc.? How severe are the cost overruns and start-up difficulties encountered? Does equipment perform according to specifications and, if not, what is done about it? When the problems encountered are so severe—leading in the extreme to project collapse—as to raise the possibility of fraud in project choice and design, is the matter brought out into the open and investigated, or is it swept under the rug to avoid embarrassment to culpable politicians and officials?


  • 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
  • Gillis, Malcolm, “The Role of State Enterprises in Economic Development,” Social Research (New York), Vol. 47 (Summer 1980), pp. 24889.

    • Search Google Scholar
    • Export Citation
  • Goode, Richard, “Nonfinancial Public Enterprises” (unpublished, International Monetary Fund, May 15, 1980).

  • Harberger, Arnold C., “On Estimating the Rate of Return to Capital in Colombia,” in Project Evaluation: Collected Papers (Chicago: Markham, 1972), pp. 13256.

    • Search Google Scholar
    • Export Citation
  • Harberger, Arnold C., “Perspectives on Capital and Technology in Less-Developed Countries,” in Contemporary Economic Analysis, ed. by M.J. Artis and A.R. Nobay (London: Croom-Helm, 1978).

    • Search Google Scholar
    • Export Citation
  • Harberger, Arnold C., and Daniel L. Wisecarver, “Private and Social Rates of Return to Capital in Uruguay,” in Economic Development and Cultural Change (Chicago), Vol. 25 (April 1977), pp. 41145.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, A Manual on Government Finance Statistics (Draft) (Washington: IMF, 1974).

  • International Monetary Fund, International Financial Statistics Yearbook (Washington: IMF, 1981).

  • Mali, Republic of, Ministère de Tutelle des Sociétés et Entreprises d’Etat, Analyse de la Situation Financière du Secteur d’Etat au 31 Décembre 1976 (Bamako: October 1977).

    • Search Google Scholar
    • Export Citation
  • Polak, J.J., and Victor Argy, “Credit Policy and the Balance of Payments,” Staff Papers, International Monetary Fund (Washington), Vol. 18 (March 1971), pp. 124.

    • Search Google Scholar
    • Export Citation
  • Polak, J.J., and Victor Argy, and Lorette Boissonneault, “Monetary Analysis of Income and Imports and Its Statistical Application,” Staff Papers, International Monetary Fund (Washington), Vol. 7 (April 1960), pp. 349415.

    • Search Google Scholar
    • Export Citation
  • Tait, Alan A., “The Fiscal Policy Objectives of Nonfinancial Public Enterprises and Their Information Requirements: A Simple Taxonomy Suggested” (unpublished, International Monetary Fund, 1977).

    • Search Google Scholar
    • Export Citation
  • United Nations Industrial Development Organization, Guidelines for Project Evaluation (New York: UN, 1972).

  • United Nations Statistical Office, A System of National Accounts (New York: UN, 1968).

  • Wattleworth, Michael, “Credit Subsidies in Budgetary Lending” (unpublished, International Monetary Fund, May 1983).


Mr. Gray is a Fellow of the Harvard Institute for International Development and consultant to the Fiscal Affairs Department of the International Monetary Fund.


Compare, for example. Malcolm Gillis, “The Role of State Enterprises in Economic Development,” Sanai Research, Vol. 47 (Summer 1980), pp. 248–89.


International Monetary Fund, A Manual on Government Finance Statistics (Draft) (Washington, D.C., 1974). (Hereinafter referred to as the Manual.)


Manual, p. 32.


Ibid., p. 35.


Ibid., p. 29.


Such is the case, for example, with aging heavy industries such as steel and shipbuilding in some industrial countries.


Commercial banks and other financial institutions meeting the definition used in this paper may also act as conduits for such credit, but it is bv no means their sole, or even primary, function.


Calculating and applying the replacement value of capital assets involves a departure from nominal magnitudes; however, it falls short of adjusting the latter for social opportunity cost, which is allocated to stage 3 below.


Methodological considerations associated with netting out the financial impact of pursuit of noncommercial objectives by public enterprise are discussed in Section V in the subsection. “The Fiscal Impact of the Public Enterprise Sector.”


The opposite adjustment applies with respect to the value of construction in progress and investment still in the start-up phase, which should be excluded from capital stock for purposes of applying this criterion.


The format of the financial statements is modeled after Table 2 in a May 15, 1980, office memorandum to staff of the Fund’s Fiscal Affairs Department by then Department head Richard Goode, entitled “Nonfinancial Public Enterprises.” Modifications in that format and the source of the Mali data are described in notes to the table.


Each variant of resource balance described in these two paragraphs is computed on an internationally comparative basis by R.P. Short in the following paper.


All indicators relating values in different time periods are assumed to be computed net of the effects of any discretionary tax rate changes.


Most countries have been the object of multiple efforts to estimate social discount rates, by staff of one or another creditor agency, technical assistance personnel, and local planners. In most cases there should be difficulty in assembling a “consensus range” or mean of alternative estimates, and using such values for rough computations of social opportunity cost.


It appropriate lo enter the caveat here that a fixed interest rate of, say, 12 percent or above on a long-term loan becomes a real burden as the rate of price inflation of the goods and services an enterprise sells to meet its debt service obligations drops toward zero. One supposes that most public enterprises will have access to means of refinancing their debt at lower terms under such circumstances.


Known as a contrat de programme or contrat plan in France and certain Francophone African countries following the French model.


Michael Wattleworth, “Credit Subsidies in Budgetary Lending” (unpublished, International Monetary Fund. May 1983).


Banking statistics in many countries classify all nonfinancial public enterprises, or all but a small selection of them, in the private sector. Collection of this data will thus not always be feasible initially and will often require a reform of current reporting procedures. Implied in the main theses of the present paper is the premise that sound financial management calls for such a reform, and governments should be pressed to institute it.


One might also argue that public enterprises pose an indirect demand Tor credit insofar as private suppliers to (and purchasers from) them utilize credit to finance the corresponding portion of their transactions, quite apart from any requirement posed by payments arrears. By looking only at arrears, one is in effect comparing the existing situation with a baseline need for credit to finance investment and current transactions associated with the given level of output, with or without participation by public enterprises, subject to a “normal” level of arrears.


One does not envisage outsiders sifting through loan portfolios to tabulate these data, but it is among the information that Ministry of Finance control units should be collecting, interpreting, and supplying on an aggregated basis to international creditors. Local bank officials, frustrated by slow turnover of the relevant portions of their loan portfolios, are often eager to collaborate with the authorities in this respect.


In the following paper in this volume. R.P. Short conducts an interesting exercise along these lines, obtaining a regression coefficient of near unity in relating growth in total credit to growth of credit to the public enterprise sector on a cross-country basis. Short interprets this, with reservations, as suggesting that public enterprise credit growth is almost fully passed through into credit creation.


The previous observation that the impact of public enterprise deficits in generating new deposits is likely to be minimal deserves re-emphasis here. Indeed, wherever regulation of the money market is such that formal sector interest rates are not free to respond to higher effective demand tor credit, an increase in public enterprise operating deficits is likely to reduce the demand for real cash balances by aggravating inflationary expectations.


This expository device was utilized in early versions of lhe Polak model of money, income, and imports. See, for example, J.J. Polak and Lorette Boissonneault. “Monetary Analysis of Income and Imports and Its Statistical Application,” International Monetary Fund, Staff Pappers Vol. 7 (April 1960). pp. 349–415.


It will be recalled that pt, represents the price index that would prevail at the end of year t in the absence of the given increment in public enterprise sector credit expansion, hence Δ*Pt bears no relation to the difference between pt, and pt –1 and is supplemental to it.


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.


Ratios computed from the Chu-Feltenstein simulation exercise are reported on page 484 of their article. Reducing simulated quarterly public enterprise transfers of 0.04 billion pesos to zero, ceteris paribus, knocks 2.97 points off the simulated quarterly inflation rate, while a similar elimination of simulated private losses of 0.90 billion pesos knocks off 6.80 points; 2.97/.04 divided by 6.80/.90 equals approximately ten, and 6.80/2.97 = 2.29.


J.J. Polak and Victor Argy, “Credit Policy and the Balance of Payments,” International Monetary Fund, Staff Papers, Vol. 18 (March 1971), p. 3.


Taking the first 27 developing countries (alphabetical order) for which import values in local currency (line 98c) are available in the country pages of the International Monetary Fund’s 1981 International Financial Statistics Yearbook, the ratio of imports to the average level of money (line 34, average of successive end-of-year values) in the latest year for which both figures are given exceeds 1.0 in 20 cases, implying first-year import generation exceeding 40 percent of ΔM and cumulative generation over three years exceeding 90 percent. Most of the countries with ratios below 1.0 fall into special categories, such as large economic scale (Brazil. India), strict exchange controls/high protection (Algeria, Ghana, India), or isolation (the Central African Republic).


A common manifestation of financially inadequate protection ex post facto is that firms are given promises of restrictions on domestic competition that are subsequently not honored, whether for good reasons—desire to reduce costs, create pressure to export, etc.—or bad (bribery by competitors).


The issue considered here does not arise, or assumes considerably less importance, to the extent that geographical or technological factors enable the credit-providing agency to appropriate a substantial share of the borrowers’ output and to deduct loan service payments before passing on the net proceeds.


Some countries have pursued yet another alternative—that of strengthening the informal financial sector to satisfy small producers’ credit needs at less “usurious” interest rates than frequently characterize this market at present. While the fiscal and monetary burdens of such an approach are minimal, its capacity to meet the relevant demand is a subject of debate that cannot be resolved here.


United Nations Industrial Development Organization, Guidelines for Project Evaluation (New York, 1972), See especially Chapters 7, 11, and 12.


Alan A. Tait, “The Fiscal Policy Objectives of Nonfinancial Public Enterprises and Their Information Requirements: A Simple Taxonomy Suggested” (unpublished. International Monetary Fund, 1977).


Arnold C. Harberger, “On Estimating the Rate of Return to Capital in Colombia,” Project Evaluation: Collected Papers (London and Chicago, 1972), pp. 132–56; Arnold C. Harberger and Daniel L. Wisecarver, “Private and Social Rates of Return to Capital in Uruguay,” Economic Development and Cultural Change, Vol. 25 (April 1977), pp. 411–45; Arnold C. Harberger. “Perspectives on Capital and Technology in Less-Developed Countries.” in M.J. Artis and A.R. Nobay (eds.), Contemporary Economic Analysis (Croom-Helm, London, 1978), pp. 15–40. The latter paper applies the procedure to 18 countries, of which 11 (including Greece and Portugal) qualify as less-developed countries.


This discussion should not be interpreted to mean that tied inflows are desirable because their social opportunity cost is less than that of fungible inflows. What matters is the productivity of the projects in which the resources are invested. An inflow tied to the source’s inappropriate equipment often turns out to be very costly in relation to value added.

Some Macroeconomic Aspects