chapter seven Expenditure Planning and Forecasting

A. Premchand
Published Date:
March 1989
  • ShareShare
Show Summary Details

Time present and time past Are both perhaps present in time future, And time future contained in time past.

T. S. ELIOT, Burnt Norton

Although the introduction of new forms of budgeting or the adaptation of old forms for new purposes can be achieved, at least in theory, in spectacular jumps, this is best seen as an evolutionary process. One form which has evolved over the years and has attracted both proponents and opponents is expenditure planning and forecasting. Development planning, as noted in the previous chapter, was the framework chosen by the developing world for expenditures to further the objectives of economic growth. The industrial countries, with the major exception of France, adopted systems of expenditure planning and forecasting in the context of growth in expenditures and frequent resort to strategies that in the long run proved to be unsettling. There was, therefore, a search for stability of expenditure over a number of years and for assessment of expenditures in terms of prospective resources and in terms of the implications for the economy. As these objectives were translated into operational tasks, governments had to plan their expenditure strategy as a part of their medium-term economic policy, assess the effect on demand and resource use of individual programs, and look for methods in which expenditures could be used for the growth and management of the economy.

Expenditure planning and forecasting are inevitably intertwined. Planning implies the formulation of the goals and policies, while forecasting is a prediction of the future. Forecasting does not automatically produce a plan, but the process of forecasting provides an opportunity for decision making. Forecasts are sometimes more than mechanical projections, as they reflect a design. Both planning and forecasting, therefore, are parts of the budget and generally take the form of rule-of-thumb projections. During the 1960s and 1970s, however, these two elements became a part of the formal budget process and are now used in a number of industrial countries (Australia, the Federal Republic of Germany, the Netherlands, New Zealand, the United Kingdom, and the United States), while others are initiating efforts to introduce them (Canada, for example). Planning and forecasting have also gained popularity in some parts of the developing world (Kenya, Malawi, Papua New Guinea, and Tanzania) and supplement the development plans.

Expenditure planning, whether undertaken as part of a development plan or as a separate but integral part of the budgetary process, is concerned with the purposes of expenditures, the appropriate rate of growth or decline of expenditures, the implications for the economy and for resource mobilization, and allocations to the sectors that are in need or, alternatively, reductions in expenditures to sectors that can continue with reduced allocations. The distinction between a development plan and formal expenditure planning is that the former represents an organized look into the future taken at a particular time, while the latter is a continuous process of making a forecast and assessing its validity as further progress is made in its implementation. In the development plan, the annual budget provides a corrective. Expenditure plans, however, form the basis for the annual budget and therefore represent updated annual elements of the development plan. Another distinction also merits recognition. Development plans may often be confined, insofar as government outlays are concerned, to new outlays, and may cover the ongoing outlays in broad aggregates. The expenditure plan is more specific and is coterminous with the budget; it includes continuing and new outlays. Expenditure plans are, therefore, continuous, disaggregated, and direct functional inputs into the budget. As such plans are to be viewed as budgets for the next few years, expenditure planning is also described as “multiyear budget planning” or “advance budgeting.” This chapter considers the features of these systems and their technical and operational problems.

Need for Expenditure Planning

Expenditure planning, at either the aggregate or the detailed level, is based on an analysis and assessment of the factors affecting each sector and an attempt to quantify those factors. It involves a three-step process. First, the underlying factors contributing to expenditure increases or reductions need to be defined, identified, measured, and forecast; second, an assessment of the standards of service required, or objectives to be met, needs to be undertaken, and third, the costs of providing services and the time profile of those costs for use in budget formulation must be computed.

These tasks, however, are no different from the usual budgetary, tasks. The additional dimension for expenditure planning is forecasting the three-step process over a number of years. Even this dimension has a checkered record in the history of budgeting. One of the earliest budget forecasts over a period of years was undertaken in the United States in 1946,1 but it did not receive much attention, largely because of the technical problems inherent in any forecasting exercise. By the early 1960s, however, several changes in the nature and magnitude of expenditures led to fresh consideration of the need for systematic expenditure planning. Although other attempts were made at undertaking expenditure forecasts, the need for them to become an integral part of the budget process was specifically recognized by the Plowden Committee in the United Kingdom.2 As much of the impetus for expenditure planning came from this report, it is appropriate to consider here its background and the implications of its recommendations.

The concerns reflected by the Plowden Committee were at that time basically the same as those of the Government, as well as those of the community. From a philosophical standpoint, the major issue was management of the economy with specific reference to the resources of the community, the overall allocation of resources between the public and private sectors and the impact on the balance of payments, rate of price increases, and employment. It was justly felt that private sector operations came to be decided more in terms of a legatee than as a part of a deliberately conceived plan of action. There was also the view that the efficiency of the administration and management of public services, as well as the evaluation of ongoing and new policies, could be greatly improved. Although some efforts were made to strengthen the budgetary mechanisms by improved costing methods, it was recognized that the budgetary process at a technical level also had large gaps. For example, the responsibility for decision on the totality of public expenditure was divided between central or local governments and nationalized industries. There were a large number of decision makers, whose concerns were not always properly captured by the budget coverage and the budgetary process. Decisions were being taken in a piecemeal way, without attention to the linkages with other areas of expenditures or without full awareness of the implications for future resources. The time scale of the fiscal year, within the range of which decisions were taken, was too short in a number of sectors relative to the sizable outlays that were spread over a number of years. Adjustments could not be made in these outlays on an ad hoc basis. At the same time, needed adjustments could not be undertaken because a major part of the resources was already committed and adjustments in new policies require time before they are effective.3 At the departmental or spending agency level, the relationship between the demand for the growing services and outputs of the government, and the estimates of the demands of government departments was not given the recognition it warranted.

The reformers, in the form of the Plowden Committee, recognized the tendency of the Government to lose sight of future costs, of the long-term implications of short-term decisions, and of the formulation of tax policies to meet growing expenditures. They demanded, as a basis for action, that explicit and regular surveys of public expenditure be taken over a period of years. The reformers recognized the piecemeal and ad hoc nature of decisions, and they demanded that “decisions involving substantial future expenditure should be taken in the light of these surveys.” Recognizing the economic and resource limitations, they demanded that expenditures be viewed “in relation to prospective resources.” The reformers also recognized the uncertainty of forecasting, over a period of years, and they sought to dispel some of the uncertainty by a continuous examination and re-examination of assumptions and a search for new and improved techniques. These approaches, the Plowden Committee felt, would “help the Government to make good decisions, by providing a better perspective. It will not provide automatic criteria, or create a substitute for the application of the judgment.”4 The Committee intended these surveys for the internal use of government and doubted “whether any government will feel able to place these surveys before Parliament and the Public.”5 The Plowden Committee's recommendations received support, and it was widely believed that the central problem of bringing the growth of expenditure under better control would be fulfilled through their implementation. The new system came to be known as the Public Expenditure Survey Committee (PESC), named after the group of officials who were entrusted with the task of making public expenditure surveys.

The factors implying the need for such public expenditure surveys have an acceptable universality about them. As the system came to be viewed as a viable framework for remedying some of the institutional problems, additional factors justifying the relevance of expenditure planning over a period of years came to be recognized. Such planning leads to improved decision making in specific ways that are recognized by those engaged in budget making more often than by the public. It provides a common base for weighing the cost of both new proposals and continuing policies, and it minimizes the element of “game playing” strategies. An additional factor proved attractive to the finance ministry—namely, that expenditure plans gave it a greater role. Expenditure plans are favored because they provide better policy inputs to central agencies without essentially reducing the operational freedom of spending departments. Most important is the inherent ability of expenditure plans to reflect the concerns of the government and those of the community at large. Depending on the time horizon, however, some of the analyses associated with plans could be speculative.

Expenditure planning and forecasting include the forecast of economic trends and of expenditures of government at various levels, the determination of the requirements of each agency, and the phasing of expenditures. As is true for the budgetary process depicted in Chapter 2, each of these aspects constitutes elements of the iterative process and each one leads to the next phase (Chart 4). Each of the elements depicted in Chart 4 reflects a group of complex activities, and their apparent simplicity is deceptive. The quantification of each aspect and the choices faced at each juncture give rise to several technical issues, not all of which lend themselves to quick, easy, or durable solutions. Before discussing the technical aspects of expenditure planning, it is appropriate to consider the nature of the forecasting and its general implications. In a way, forecasting is the Achilles heel of expenditure planning but, if properly done, it can lead to productive results. Lack of care or more emphasis on subjective aspects could make the forecasts questionable. From the technical point of view, forecasting budgets casts great responsibilities on the decision maker. Much of the accuracy of the forecast is dependent on his ability.

Chart 4.Schematic Presentation of Expenditure Planning Process


In day-to-day life everyone makes forecasts. In some cases, it may be just an intuitive guess, in others, it may be an organized projection based on available knowledge. Reliance may even be placed on the reading of the planetary configurations and their implications for individual or corporate fortunes. But it should be recognized that astrology is a science as well as a technique of forecasting. What gives importance to forecasting is that it is not treated as a single event but is the implicit normative basis for the projection—systematically exploring and selecting policy goals and alternatives so that the isolated events are incorporated into a decision-making process. In this form it has been used for weather forecasting and, since the advent of economic planning, for projecting future economic trends. Notwithstanding this application, the track record of forecasting has been rather problematic. Ascher notes “[a]ll the technological sophistication of our public and private sectors has not produced an enviable record of forecasting future problems and events.”6

In considering the nature of forecasting and its applicability, a distinction should be made between the use of forecasting in the corporate and the government sectors. In particular, three distinguishing features may be noted. First, the goals and operations of corporate planning (even of the multinational conglomerates) are relatively narrow, specific, and well defined. The corporate concerns are with product development and marketing, working generally under the assumption that the goal is to maximize profits. The goals of the government are far more complex, the process includes multiple levels of decision making, and generally the government is concerned with a vast range of goods and services that are not easily measured precisely. Second, decision making in government is largely political; while no organization can be totally devoid of it, politics are less significant in a corporate setting and may not have any immediate impact on the community. If a corporation leaves the market, another will enter, and the inconvenience to the consumer, if any, may be marginal or of short duration. The politics of government, however, extend far beyond the fortunes of a party and may have a lasting impact, in unpredictable ways, on the community. And, third, forecasting in corporations is a private and internal matter, and concerns over competition make it a secret operation. The forecasts of government are intended for public consumption, and the government is accountable not only for the technical accuracy of the plans but also for the results of the forecasts.

Forecasting in government has an added dilemma. Overoptimistic assumptions regarding economic performance lead to unrealistic estimates and thereby erode the credibility of the forecast itself. Specifically, in regard to expenditures, if these are tentatively fixed with reference to overoptimistic assumptions, their downward revision in the light of the changed economic situation becomes a problem that will soon degenerate into a political crisis. On the other hand, pessimistic assumptions are less palatable (although recent political experience suggests that such assumptions could, on occasion, be a strong asset for attaining power) and, in the sensitive financial area, may prove to be a self-fulfilling prophecy.7 It should be recognized, however, that some of these aspects become more apparent on the basis of hindsight. The essential corrective of the forecasting system is that each forecast is not to be seen as immutable but rather as a flexible outline of future policies that must be revised in line with the changes in the economy and must be maintained on a “rolling” basis.

Forecasting in government can be unitary or can be an amalgam of components. The former is relevant when dealing with specific aggregates, such as the rate of growth of GNP and population. In expenditure planning, however, forecasting has a component-based approach. Since the attempt is to delineate the interrelationships between the economy and the budget—not as two broad entities but each as a total of significant blocks—forecasting has to be applied at a disaggregated level. Forecasting of expenditures has to facilitate the transition from the aggregates, which are the concerns of the economic policymaker, to the details that are appropriately the responsibility of the program manager in the spending agencies. The forecast of the specifics is based on the forecast of the general trends in the economy, and to that extent, the former has to be an integral part of the latter. Forecasting in terms of the component-based approach should not, however, treat each portion separately, ignoring the linkages, but be more appropriately viewed as a holistic system that permits the detailed analysis of the interaction.

The validity of forecasting is partly dependent upon the remoteness of the target. The longer the period the greater is the possibility that conditions will change. Five-year projections may, therefore, be prone to greater error than the forecast for the next year. The choice of the appropriate period for expenditure planning presents another dilemma. The next fiscal year may be an ideal choice, purely because of the limited time between the forecast and the actual occurrence. But, as noted above, it may be too short to control the endogenous factors of expenditure growth. Also, in some respects, five-year forecasting periods may be considered to be too short.8 In view of the fact that projects take up to a decade from planning to completion, it might be expedient to forecast over periods that synchronize with the activities of government. Research development, major river valley projects, energy, retirement, and education policies are some aspects where periods that are considered appropriate for forecasting may prove too short for planning purposes. Part of this problem is, however, mitigated by the rolling nature of forecasting.

The success of forecasting and its utility to the policymaker are dependent on the comprehensiveness of its coverage and the accuracy of the estimates. The accuracy, in turn, is dependent on the objectivity that the policymaker is able to bring to the assessment of the economy, to the nature of its constraints, and to its operations. Even in the best of times, this is a hazardous area and clearly needs the best attention. In addition, if it is to prove acceptable to the participants, it should be based on a clear understanding of the general political considerations. It should, therefore, include the participation of the ministers and others who are engaged in the budgetary process.

Economic Outlook

The several objectives of government fiscal policy are intertwined, and priorities need to be established. In arriving at choices, however, it is necessary to take into account the outlook for the economy and its implications for expenditure, as well as the autonomous factors contributing to the growth of expenditure and their impact on the economy. These are pursued in two different strands—one dealing with the formulation of the economic outlook, and another dealing with the expenditure requirements in future years. For administrative purposes, however, these strands interact and thus contribute to the formulation of fiscal policy.

The determination of the outlook for the economy involves the assessment of developments in the domestic and external economies, productivity, growth in the economy, the rate of inflation, developments in wage rates, and overall domestic demand and economic activity. Oil price increases or reductions and their impact on the economy would similarly need to be forecast. In countries that are highly dependent on exports, the demand for their domestic products needs to be assessed. Similarly, changing import prices and the future costs of external borrowing need to be identified. The purpose of this exercise in a narrower context is to ensure that the overall implications of fiscal policy in general, and expenditures in particular, are consistent with the macroeconomic plan of the economy. The projections for the economy would establish the broad framework within which to determine the policy guidelines of expenditure forecasts. Without this policy guidance, expenditure forecasts would potentially become shopping lists. By their very nature, however, projections for the economy may not be precise and need modification and if they are seriously in error, it is likely that the number of modifications in expenditures will be considerably greater.

Forecasting the outlook for the economy has reached a high degree of sophistication in recent years, with considerable reliance on economic modeling and related exercises.9 Briefly, however, the formulation of the economic outlook in the current context involves no sectarian view that is exclusively Keynesian or monetarist but often is a pragmatic combination of the two. Following the traditional approach, and after examining previous trends in the economy, judgment is made on the long-term growth potential and outlook in terms of the technological and demographic factors. This is then adjusted for cyclical changes in the economy. At the same time, an estimate is also made of the nominal rate of growth in GDP and there is an interface to reconcile the growth potential with financial constraints. In the light of that analysis GDP growth is formulated. The degree of sophistication involved in the identification and measurement of the interrelationships will determine the practical relevance of the economic outlook.

The specification of the interrelationships and the publication of the outlook have gone through a period in which there was initial reluctance to publish them (in view of the pronounced announcement effects discussed earlier). Over time, however, there has been more willingness to publish them. In the United Kingdom, the publication of medium-term strategy is considered as an important element in the process.10 There is, however, considerable variation in the practices, and in the degree of detail, of those countries that have taken to expenditure planning and forecasting. In the United States, for example, the practice within Government has been to use economic goals, e.g., lower unemployment, as the basis for its five-year projections of revenue and expenditure.11 In the Federal Republic of Germany, the Financial Plan for the period 1979–8312 is based on the assumption that annually the real growth of the economy would be 4 percent and increases in prices 3 percent. The projections of revenue and expenditure are made within this framework and fiscal policy guidelines, as in the United Kingdom, emphasize keeping the annual rate of growth of expenditure below the anticipated rate in the GNP. In the Netherlands, the future of the economy is stated in the Budget Memorandum and is based on the estimates of the Central Planning Bureau.

The outlook and the assessment of the economic trends is less detailed and more aggregate in developing countries. Specifically, in countries that have adopted expenditure forecasting, it appears that the forecast of the economy primarily relates to the rate of growth that is indicated more as a target than as a forecast.

Anatomy of Expenditure Plans

Formulation of the expenditure forecasts and determination of the related policy design require that their principal ingredients, such as the totality of expenditure, be clearly recognized in the system evolved for the purpose. The frame of reference of expenditure forecasts and the practices may conveniently be considered in terms of the coverage, period, content, and basis of the forecast, and the form of presentation of expenditure estimates. These aspects are considered below.


The purpose of expenditure planning is to bring together the operations of those entities that have an impact on the economy regardless of legal and other differences in their operations. This indicates that, in addition to outlays by the central government, the expenditure plans should ideally include the transaction of other branches of government including public enterprises. Consideration of the totality of expenditures poses more problems in a federal arrangement than in a unitary form of government. Integrated plans are not obviously feasible in the classic federations such as Australia, Canada, India, and the United States in view of the specific constitutional rights of provinces and states for raising revenue and incurring expenditures. In countries such as the United Kingdom where no state governments exist, the grants and other capital expenditure to the local authorities are included in the expenditure plan. In the Federal Republic of Germany, state governments are obliged to formulate their own financial plans and coordinate with federal plans. In Tanzania, however, local governments have no revenue-raising powers and are dependent on the Central Government. As such, their transactions form part of the total expenditure plan.

Target period

Just as there are various ways of assessing the economy, or measuring the budget impact, there are also various periods for which forecasts can and are made. In some countries, the practices have varied over the years. For example, in the United Kingdom, the initial practice was to make the forecast for five years. However, as the estimates revealed a greater margin of error for the later years, a transition was made to the “focus years,” or the period in which the working of the policies could be envisaged with greater certainty. Under the procedures now in vogue, the U.K. expenditure plans provide estimates for three years in addition to the next fiscal year (3 + 1). A similar procedure is followed in Australia, New Zealand, and Kenya. In the United States the forecasts are divided into two groups: (1) a three-year framework in which budgetary decision making takes place, and (2) an additional two-year period containing the extrapolations of the budgetary effects of current policies. In the Federal Republic of Germany, forecasts are made for four years, in addition to the fiscal year (4 + 1). In Tanzania, however, they are made for only one year after the fiscal year for which the budget is presented; however, in the event, the period is too limited for making meaningful assessment of the policy directions and related adjustments. Although the choice of the period is influenced partly by administrative convenience and partly by actual use of the forecasts for budget policies, it appears advantageous that the period should be longer than two years. On the basis of experience, three years seem more popular.

Content of expenditure plans

A feature distinguishing expenditure plans from development plans is that the latter are usually structured to deal with new outlays rather than with existing policies. Expenditure plans, however, involve the future implications of the existing policies as well as new outlays. In terms of the PESC practices in the United Kingdom, outlays on existing policies and new services are dealt with in a pragmatic way. Ideally, expenditure forecasts could be formulated to show (a) the present level of expenditure, (b) additional expenditure to provide the same service in future, for example, to maintain a pupil/teacher ratio in educational institutions, and (c) additional expenditure if the service is to be changed. Elements (a) and (b) can be described as existing policy, while (c) can be defined as new policy. In reality, however, the distinctions between existing and new policies may not be clear and may be blurred depending on the nature of the sector and the statistical data available. From the point of view of spending ministries, changes in the policies may not always imply radical departures from the existing ones and are, at best, minor modifications to prevent what the ministries perceive to be a deterioration in the services. Moreover, for many activities of government the type of objective tests of requirements or of measurement of productivity may not be available. Further, the development of such information presupposes the existence of a performance budgeting system. The practice in the United Kingdom is that for each year of forecast the previous year's level of service is taken as the base, this is then modified by additions representing new policies and subtractions reflecting the closure of activities that have already been agreed to during the year. Proposals over and above the levels so indicated constitute new policy proposals. This procedure implies that forecasts become effectively the annual budgets. In other countries, particularly in the United States, the eventual budget is made by the legislature and forecasts of government are illustrative of the policy implications. In the U.K. system, division of expenditures into existing and new outlays provides a firm basis for the decision maker, particularly at the ministerial level, to take a view on the programs as well as on the aggregate. If a trade-off is to be made between the existing and new outlays within specified resources, the implications will be clear to the decision maker. In the developing countries forecasting is done at a more aggregate level and basically involves the application of percentage growth rates to existing and new outlays. A technical issue that could often contribute to divergence between the content of the expenditure plans and the actual spending is the use of different bases. For example, in the United States, planning is done in terms of the obligational authority but the outlays are not controlled by it. Although certain stable relationships between authority and actual spending are assumed, misjudgment of exact timing could, and does, lead to significant divergences between the forecast and the actual amount of spending. Congruence between these two is essential for the success of the forecast.

Price basis

One of the major problems encountered in formulating expenditure forecasts is the choice of the proper price basis. The essence of medium-term forecasting and allocation of resources is to ensure that the decisions involving commitments of economic resources for the future are taken according to a rational pattern of priorities. Planning, therefore, involves the measurement of the cost of services in real and monetary terms, as well as the value of resources in terms of the relative claims on GDP. Formulation of expenditure estimates over time involves the choice of using constant or current prices. As the approach of the Plowden Committee emphasized, the establishment and attainment of real expenditure targets both by each functional category and by total meant that forecasts had to be analyzed in terms of the relation of the expenditure targets to the projected growth in physical resources as measured by GDP at constant prices. But the approach of constant prices had several conceptual and operational implications that could not be ignored and if constant prices were chosen, compensating devices had to be utilized.

The choice between constant and current prices is not in any event an easy one. The United Kingdom, which pioneered the expenditure forecasting system, resorted to constant prices in its price surveys, while other countries utilized current prices for formulating their forecasts. The adoption of constant prices had, however, been continually riddled with problems, and after approximately two decades of use these have recently been abandoned by the U.K. authorities. The experience with constant prices fully illustrates the complexity of problems involved in the choice and it is, therefore, instructive to deal with the experience of the United Kingdom in some detail.

Constant prices

The use of constant prices gives rise to difficulty in comparing actual expenditures (which are in current terms) with the provisions shown in the forecasts. It also obscures the real factors contributing to increases in costs. This was particularly true in respect of cost increases caused by increases in the salaries of government employees, which, in terms of the national accounting conventions, are not accompanied by any increase in productivity. The increase in the costs of services caused by pay increases are as important as those caused by expansion of public services, and anomalies of public policies may arise when one is differently treated from others.13 Moreover, the constant price presentation will constantly underestimate the future weight of public services in relation to the economy as a whole because of the zero productivity convention. If salaries of government employees increase at the same rate as those in the private sector, it will mean that the cost of providing the services rises more than the average prices for the economy. This, however, is not captured in constant prices and, if undue reliance is placed on that approach, the actual experience may turn out to be vastly different from that estimated. It could also mean that, if constant prices are to be trusted, the community is either getting increased and better services or, alternatively, the same services are provided by fewer staff. More significantly, by understating the cost, the opportunity cost of the use of resources is not fully revealed. In the narrower context of budgeting, estimates in real and constant terms imply that cash has to be provided, regardless of the price rise, in order to maintain the agreed levels of services in real terms.14 This had the potential of depriving the government of control over cash expenditure exactly when it needs this most.

In practical terms, in the United Kingdom the forecast of expenditure was given in volume terms or in constant prices.15 Under the practices in vogue until the early 1980s, the base dates for constant prices varied. For most expenditures on goods and services constant prices were those that obtained in the last year. For computing transfer expenditures, the average of the current year's prices was used as a basis. However, because the use of constant prices did not fully reveal the value of the services, and reflected not the quantity of services produced but the quantity of manpower and other resources that were used in supplying those services, the concept of relative price effect was used. The term refers to the change in the prices of goods and services bought by government relative to the average price of national output and measures the extent to which the resource used by the public sector becomes more or less valuable than those used in the national economy. In short, the relative price effect is the difference between the estimate in current cost terms and in volume or constant terms. In addition to the forecast made on the above basis, the plans also provided for a contingency reserve to cover requirements for increased expenditures that were not foreseen or that could not be properly quantified. The amount so set apart varied depending on the uncertainty.

The configuration of prices, therefore, included three approaches: (a) in constant and volume terms, (b) in cost terms, including the operation of the relative price effect, and (c) in outturn prices or current prices in terms of which the actual expenditures were incurred. The use of this three-tier structure, and the problems inherent in making an accurate forecast, contributed to annoying issues that became more prominent after 1973. Those that merit consideration are the implications of the three-tier price structure, the computation of the relative price effect, and the fear of the loss of financial control. The use of three different prices obviously generated considerable confusion among the spending agencies. Expenditures in constant prices were not the ones in which payments were made, nor did they represent the figures recorded by the accountants. Comparing ongoing expenses with those of previous years was not meaningful, unless they were deflated and made uniform—an aspect in which the lay administrator did not feel proficient. Nor did this give to the public any better perception of the role of public sector. At the same time, the revaluation of outlays that had to be undertaken at the end of the year caused additional burdens. The impact of the new language of constant prices can be seen by the epithet given to them—funny money—or money in which no transactions were carried out but which was used for policy purposes. While the severity of the problem may have been reduced over the years, there is no gainsaying that it contributed to a language interpreted by participants—each in his own way—with no common wavelength.

Estimating the relative price effect is also difficult. After the general price increase of 1973 and reflecting some speculative activities, the prices of land and houses moved very sharply upward,16 and the catching up in the pay awards for civil servants contributed to higher expenditures. These factors, associated with others, in particular the rate of inflation, contributed to expenditures that were substantially higher than those shown in expenditure forecasts, in turn, implying that the price effect was not fully estimated. This rising discrepancy between the cool rationalism of the expenditure White Papers and the more sordid actual expenditures cast aspersions on the efficiency of the expenditure control mechanism. Some felt that in paying more attention to planning expenditures, the use of conventional financial control had been relegated to the background, leading to more expenditures than planned.17 The Expenditure Committee also reported “we feel bound to conclude … the Treasury's present methods of controlling public expenditure are inadequate in the sense that money can be spent on a scale which was not contemplated when the relevant policies were decided upon.”18 The Chancellor of the Exchequer himself felt that the time had come “to reflect seriously on the wisdom of planning public expenditure solely on the basis of constant prices.”19 Another Chancellor admitted, albeit half a decade later “there is something inherently unreal in trying to plan and measure things in terms of what is rightly described as ‘funny money’.”20 The recognition of the problem did not lead to any immediate changes in the basis, but a compensating device known as cash limits (discussed in the following chapter) was introduced to contain expenditures.

The dangers of formulating expenditure plans in constant terms has persisted and there has been growing disillusionment. The final break from constant terms was announced in 1981 and is operational from fiscal year 1982/83. Announcing this change, the Chancellor stated that the examination of expenditures would be conducted in terms of cash (in current prices) that would be available for that year. It was recognized that the new system could be difficult for those in the spending agencies who will be obliged “from the outset to form a view as to what their money will buy.”21 This change involving the use of current prices or “money of the day” implies a substantial modification of the PESC system but is in line with the broad evolution of the control and administration of public spending since the mid-1970s. The procedures relating to the use of prices in current or cash terms are yet to acquire extended experience and their success depends on the care taken to project future prices. (This aspect is considered in the following chapter.) For example, in the Federal Republic of Germany, there are three types of expenditure in the formulation of the plan: (a) expenditure on wages and salaries that take into account future increases; (b) index-linked social expenditures that are based on future price rises; and (c) other expenditures that also allow for future price increases. The financial plan itself is expressed in terms of expected prices. The Netherlands has a similar practice, in which expenditure plans are expressed in outturn prices. In two developing countries in Africa (Kenya and Tanzania) estimates are expressed in outturn prices but, in computing them, it is expected that the rate of expenditure growth stipulated for various programs will also include the rate of price increase.

The difficulties of choosing the proper technique and of making it credible over a period illustrate the strengths and weaknesses of an expenditure forecasting system and the reason for its relatively slow growth and spread.

Form of presentation

The form in which expenditure plans are formulated and presented should, as far as possible, be close to the budget categories. Without such congruence, there will be problems of coordination. Also, if expenditures are to be monitored and controlled in terms of the plans, then they have to be essentially the same as the budget groups. In recognition of this obvious need, the U.K. authorities have made improvements both in plans and budgets and now the same functional, economic, and spending authority groupings are used by them. In the Federal Republic of Germany relatively broad groupings are used for financial plans and detailed categories are used in the budget. In developing countries, the categories reflect primarily the spending authority groupings and the forecasts are very broad and lack detail. Budgets, however, are more detailed, necessarily so for the purpose of obtaining appropriation authority.

Institutional Aspects

In order to be effective, the expenditure planning and forecasting system has to become a part of the annual budget cycle. This fusion has been attained in the United Kingdom. The cycle of operations, like that of the budget takes approximately a year. At the end of each calendar year instructions are issued by the U.K. Treasury regarding the procedures of the expenditure survey to the spending agencies, which then furnish the relevant information and estimates by the end of February. The estimates are then discussed by the Treasury and the spending departments and agreement is reached over two months. A draft report is then prepared by the Treasury and, after approval by the Public Expenditure Survey Committee, is submitted by midyear to ministers, who then take decisions on the aggregate as well as on the composition of public expenditures. The estimates are then published. Three aspects of these procedures merit particular attention. First, formal consultative machinery has evolved for decision making on public expenditures. The proposals for expenditure increases and reductions, and the assessment of their costing, are reviewed by the Principal Finance Officers, who work primarily as a part of the administrative department but represent, as it were, the financial conscience of the Treasury. These Finance Officers represent an inner Public Expenditure Survey Committee and their approaches have a decisive influence on the final outcome. The approval by the Survey Committee is more formal, and during recent years its deliberations have not taken more than an hour. The growth of the consultative machinery has been largely instrumental in providing a degree of stability to the process and in reducing the distrust that spending agencies traditionally have toward finance ministries. Second, the involvement of ministers has added the requisite political dimensions to the process. Third, the surveys so made reflect not merely multiyear planning but, for the following year and some of the focus years, multiyear budgeting as well. Similar but less formal procedures are observed in the Federal Republic of Germany. A new financial plan is prepared each year, together with the Government's draft budget, and is then finalized by the Finance Ministry in close cooperation with expenditure departments. After approval by the Cabinet the plan is presented to the legislature for information. Australia and New Zealand have installed procedures that are closely modeled on the practices of the United Kingdom.22 In the United States the system is used primarily as a management tool to ensure that the long-term consequences of present policies are identified and recognized in the budgetary decision-making process. In the few developing countries where the system is prevalent in one form or another, it appears that there is neither formal consultative machinery nor any attempt to integrate the forecasting system with the budget cycle. The spending departments are asked to furnish proposals but these are substantially changed by the planning agencies (in Kenya and Tanzania the forward budgets are organized by the planning branches of the finance ministries) and the final estimates are shown only in the aggregate.

Country Experiences

The experience of expenditure planning and forecasting systems, as has been repeatedly pointed out above, is restricted to a few countries. The longest experience is that of the United Kingdom, which initiated the experiment in the early 1960s. The last two decades have revealed attitudes of probing, strengthening, further probing, and undertaking necessary adjustments. There was a willingness to change, although the process of change itself took time and, on the whole, the system showed greater resilience and responsiveness. Although the beginning was hesitant, more arduous tasks in specifying the outlook for the economy were undertaken. The system itself was buffeted during 1965–68 by economic crosscurrents and was caught between the optimistic assessments of the economy, on the one hand, and the necessity to control expenditures in the context of a gloomier economy, on the other. The developments in the mid-1970s proved conclusively the difficulties in the use of constant prices and the correct calculation of the relative price effect. While these aspects may lead to the opinion that there is no simple way in which forecasts can be made, it also illustrates that more faith than warranted may have been placed on the forecasting techniques.

The German experience, which started in 1967, has had relatively more stability and coherence. This may be due partly to the dread that the Federal Republic has for budget deficits and, therefore, the general readiness to ensure containable growth rates in expenditure, and partly to the state of the economy itself. Australia and New Zealand established the system of expenditure planning and forecasting, and installed sophisticated information systems for keeping track of value and price changes. Progress in both countries is considerable and the preparation of expenditure forecasts has become a part of budget folklore.

The experience of the United States reveals a different dimension and illustrates the difficulties inherent in an environment in which planning is considered in emotional terms and as an anathema. Notwithstanding the occasional projections of expenditure at the federal level from 1945 to 1961, the real impetus for medium-range or long-range expenditure planning came into formal existence only as a part of the program budgeting introduced in the Department of Defense after 1961. This was later extended to the whole Federal Government as part of the Planning, Programming, Budgeting System (PPBS). One of its major features was “projection of government activities over an adequate horizon.” The formal shape for this was given as program and financial plans (PFPs), under which the spending agencies were required to provide a multiyear summary of the programs over a planning period covering the budget year and four future years or over longer periods wherever considered essential.23 But the implementation of this aspect, as with others, revealed several difficulties, all of which owe their origin to the lack of an adequate theory of requirements.24 In the preliminary stages of the implementation of the PPBS, no specific direction was given to spending agencies on how expenditure planning for future years was to be done. They were merely asked to indicate how much and in what way they would spend money in the future. The results of the approach were predictable; there were lengthy “wish” lists of what the agencies would like to spend on their programs if no fiscal constraints were observed.25 The idea that the computation of future cost implications of current decisions would drive the “thin edge of the budget wedge” into the program did not materialize. Later, the scope of the PFP was changed to include only those outlays to which the Government was committed by legal or moral (emphasis added) obligations resulting from past decisions or required by current decisions. The definition of the “commitment” lacked precision and agencies tended to use different definitions and resource limitations. If at earlier stages lack of guidance contributed to problems, it was lack of specificity that compounded these problems at later stages. The exercise itself has not been without benefits for it clearly enabled many agencies, depending on the competence of their internal financial management, to gain perspectives on funding requirements in the future. But the submission of PFPs by agencies was quietly dropped and is no longer required as a part of the budget process.

These practices were reviewed in 1977, as a part of the Zero-Base Budgeting (ZBB) system. The instructions relating to the introduction of ZBB specified that the resource requirements for four years beyond the budget year were to be submitted along with other materials.26 Another development was that the legislature required a five-year projection for mandatory or entitlement outlays. During the 1960s and most of the 1970s, the agencies gained much experience in undertaking future-year expenditure planning, but for several years these were primarily internal or selective exercises that were generally in the nature of projections. Multiyear advance expenditure planning became an integral part of the budget process only in 1979 and the Government's planning horizon was formerly extended to three years, with projections for another two years, all of which are published as part of the budget documentation. The actual impact of these developments is still to be felt.


The experience is both limited but paradoxically adequate to reveal the practical benefits and problems of the system. It is limited because it is primarily the experience of selected industrial countries that offers us the relevant evidence. But it is adequate because the experience extends beyond a decade. To consider the benefits first, it forces the agencies, the public, and the legislature to consider the future implications of current decisions and their impact on the economy. It affords an opportunity to consider the national initiatives in a broader setting and in a phased way, thus minimizing the need for abrupt movements. Within the Government, it changes the value structure that has dominated budgetary thinking and financial controls for more than a century. The explicit and institutionalized (as distinct from ad hoc) consideration of the economy helps to form a clearer policy and, more significantly, the holistic approach of bringing diverse elements together permits a better recognition of the linkages. The planning system is not a single cosmic operation but can serve the complex requirements of current economic management better than the traditional systems. Each year the past as well as the future can be reviewed and this constant monitoring offers a substantial degree of flexibility.

But these very benefits, depending on the skills with which they are used, could, and indeed have, become problems, which may be examined in terms of those relating to the preparation of spending plans in conjunction with prospective resources, in forecasting techniques, and in the implications of the system for expenditure control. The experience described above suggests that the specification of the future of the economy has been a matter of primary concern. Given this uncertainty, formulation of spending plans contributed to a planning paradox that, if the actual growth of GNP were less than forecast and assuming that expenditures were proceeding as forecast, a higher ratio of expenditures to GNP would emerge.27 There are also factors that do not lend themselves to any meaningful forecast, such as natural disasters. It is also evident that some of the national accounting techniques (e.g., constant prices) have contributed to avoidable problems and clearly illustrate the limitations of economic principles in governmental policymaking. Over the years this has led to considerable skepticism and to the claim by political scientists that some of these approaches may not provide for a viable public decision making. The alternatives for this clearly cannot be decision making by faith or on an ad hoc basis, but must be by assiduous efforts at improving the system. Indeed, viewed over the long term, some of these may be considered as problems of the initial or trial period.

The implications for control are more diversified. The critics of the system argue that expenditures once included in the plan become a. “right” for the spending departments and a commitment on the part of the government.28 So considered, the existing policies will become sacrosanct, will resist innovation, and may create a bias for perpetuation. Also, in considering the new policy prospects, the spending departments may try to “beat the system” by having proposals considered outside the normal expenditure planning system. Admittedly, if forecasts are considered as firm commitments by government, the system is liable to break down sooner or later. Although pressures, both internal and external, will develop for the continuation of policies, it is only appropriate that estimates be considered as forecasts and not definitive financial plans.

Another problem that is rather pronounced in the United Kingdom is in regard to the phasing of expenditure. The experience suggests that there is a bow-wave in expenditures implying higher expenditures for the immediate fiscal year and tapering outlays for future years. Cynically known as “fairy gold” it implies that the spending departments trade cuts in future years in order to maintain the present amounts.29 Also, in some countries, the implicit pact between the spending departments and the central agencies was not maintained and there were spurts in expenditures.30 This implies the need for Strengthening the expenditure control mechanism, in particular, the need for detailed review of the estimate bases, A fundamental weakness of the system is that it is based only on inputs and does not emphasize the outputs. It also provides information on existing policies but does not consider alternative policies. Although part of the problem is reduced by undertaking separate policy reviews, it is important to relate the results of such studies to the budget process.

The system requires skills different from those that are found in traditional bureaucracies and it demands the provision of varied information related to expenditure plans, in particular, information on the costing of projects and programs. The slow progress in this area has been a reason for the system not making much headway in developing countries. Even where it has been initiated in some countries, because of a lack of information and of explicit commitment or attention by policymakers, it remained more as a supplemental instrument to the traditional machinery.31 There is more to be done if this system is to become a regular instrument of government financial management.

Occasional forecasts of government expenditures have been undertaken by economists. Specific mention should be made of Otto Eckstein (1959), and Gerhard Colm and Peter Wagner (1966). Similar attempts were made in the reports of a symposium. Public Finances: Need, Sources and Utilization (New Jersey: Princeton University Press, 1961).

United Kingdom, Control of Public Expenditure (1961).

There is considerable official and other literature on the U.K. experience. In the former category, specific mention may be made of Public Expenditure White Papers: Handbook on Methodology (1972a); Public Expenditure: Planning and Control, Cmnd. 2915 (1966). Public Expenditure: A New Presentation, Cmnd. 4017 (1969a); and “The Management of Public Expenditure” (1979b). All these are published by Her Majesty's Stationery Office, London. In the other literature, the following provide a historical and critical account: Sir Richard Clarke (1964); Sir Samuel Goldman (1973); Hugh Heclo and Aaton Wildavsky (1974); and Christopher Pollitt (1977), pp. 127–12.

United Kingdom, Control of Public Expenditure (1961), p. 8.

Notwithstanding this approach, pressures soon led to the public announcement of the U.K. Government's public expenditure plans—a tradition that is also continued and adopted in other countries. In Australia, however, expenditure plans are largely for internal purposes and are not published.

Ascher maintains that the importance of self-fulfilling forecasts has been exaggerated. While the decision makers do not control the pace of events involved, the fact remains that the public's reactions and expectations play a crucial role in the success of financial policies. The announcements of government policies have significant effects in themselves (Ascher, 1978, p. 5).

See, for example, the analysis of the U.S. Government on population change and long-range effects on the budget, which cover a decade or more, U.S. Government Budget: 1981.

One of the more detailed macroeconomic models is the one utilized by the U.K. Treasury, which has more than 250 equations. For a discussion of these aspects, see United Kingdom, Treasury Macroeconomic Model: Equation and Variable Listing (1980c). Similarly, there is an impact program in Australia, which is devoted to the development of a medium-term to long-term, economywide, economic-demographic model. It is an integration of three modules: (i) Macro, concerned with the fiscal and monetary aggregates; (ii) BACHUROO, which is a module designed to keep track of the evolving population and work force; and (iii) ORANI, which is an interindustry module. For a descriptive account of these models, see Report of the Royal Commission on Australian Government Administration, Vol. 4 (1976), Appendix, pp. 11–22.

In the United Kingdom economic forecasts are required to be published twice yearly under the Industry Act, 1975. The medium-term strategy is announced in the annual White Paper on Government Expenditure Plans.

In the United States this has been the practice of the Government. The Congressional Budget Office (CBO)—a branch of the legislature—however, presents two sets of projections assuming different rates of growth in the economy. The economic assumptions of CBO projections do not reflect the probable economic conditions of policy targets and are more illustrative in nature. The projections of the Office of Management and Budget could lead to higher surpluses because of the working of the fiscal drag.

Germany (1980), p. 2. Only an abridged English translation of the Plan is available.

These problems were recognized by the U.K. Government. See United Kingdom, Public Expenditure: Planning and Control (1966), p. 12. Set also United Kingdom, Public Expenditure: A New Presentation (1969a), pp. 22–23.

This was foreseen by Enoch Powell. See extracts from his letter in David Galloway (1976), pp. 154–55.

Volume term connotes that physical yardsticks are used for measuring the services and their requirements. In practice, however, this has not been possible for a large area of government and constant prices have been used instead.

See Maurice Wright (1979), pp. 143–69; and R. W. R. Price (1979), pp. 68–76. The measurement of the actual impact of the relative price effect requires detailed calculations and is not obvious from the White Papers which show only a total figure for public expenditure as a whole.

United Kingdom, The Financing of Public Expenditure (1975b), pp. VI–VII.

Hansard (H.M. Stationery Office: London, April 1975), col. 279.

Budget Speech,” The Times, March 11, 1981, p. 6.

The Chancellor observed: “This is bound to be less easy than just deciding what they want—and then simply looking forward to receiving all the money necessary to pay for it,” Budget Speech, ibid. The new arrangements also envisage a change in the use of the contingency reserve, which hereafter will be also used for meeting unforeseen price increases. These changes, which have some presentational and administrative convenience have also been criticized on the count that undue reliance on cash or current prices may lead to neglect of volume and output aspects.

In New Zealand, the Public Expenditure Survey Committee is known as the Committee of Public Officials for Public Expenditure (COPE).

United Stares, Guidance for PPBS (1968).

United States, Zero-Base Budgeting (1977).

By 1979 prevention of this phenomenon became a major government objective. The U.K. authorities noted “To plan more public expenditure before the required output is available to support it would ensure that, in the event, that growth of output does not take place. Higher public expenditure cannot any longer be allowed to precede, and thus prevent, growth in the private sector,” United Kingdom, The Government's Expenditure Plans, 1980–81 (1979a), p. 2.

Such bow-wave phenomenon is not evident in the German financial plans. The experience of developing countries shows a reversal of the bow-wave by the fact that expenditures increase gradually and are sizable in the later years of the plan.

Ireland gave up advance expenditure planning when the spending departments did not comply with the resource limitations.

In India, when rolling expenditure planning was tried to be introduced in 1979, it was viewed as an attempt not to strengthen the five-year plans but to replace them. It, therefore, never became operational, except that future outlays on major projects are shown in performance budget documents.

    Other Resources Citing This Publication