chapter eight Inflation Budgeting
- A. Premchand
- Published Date:
- March 1989
Never imagine yourself not to be otherwise than what it might appear to others that what you were or might have been was not otherwise than what you had been would have appeared to them to be otherwise.
LEWIS CARROLL, Alice's Adventures in Wonderland
Inflation became a worldwide phenomenon during the late 1970s. If, during the 1950s and 1960s, the problem was experienced by a few countries in Latin America and Asia, inflation became more rampant and chronic and persisted at high rates during the 1970s in a variety of countries. Some expect that the trend will continue during most of the 1980s. Others predict that there will be a massive downturn and that the problem faced by policymakers will be, not inflation, but a deepening recession, and that concern will be on the ways in which the downturn could be avoided. While the prognosis differs, the need for the study of inflation itself requires hardly any defense.
The effects of inflation are far-reaching and affect all aspects of budgeting with varying degrees of severity. If budgeting is to serve the central policy function as it has done during the last three decades, it is essential that the impact of inflation on the budget, as well as the impact of the budget on inflation, are recognized and analyzed in the proper perspective. To the extent that these aspects are not given proper consideration, the budget is formulated on suspect premises and may reflect a wrong reading of signals, as well as giving inappropriate signals to the economy. In several countries, the lack of recognition of inflation may contribute to the worsening of the government's fiscal situation as its slowness to react to taxation and to introduce tariff changes may cause reduced revenues, which, with increasing expenditures, will lead to higher deficits. Although these aspects are felt and experienced by every policymaker, little attention has been given to this aspect in the literature. A good deal of the available literature deals with the individual's expectation and his behavior in an inflationary environment. Government, admittedly different from an individual in its approaches, received scant attention. Ironically, during recent years and as is evident from the burgeoning literature, considerable attention has been devoted to inflation accounting, largely because of increased tax liability being a concomitant of inflation. The concerns of the budgeteer are, however, real and have grown in their importance along with the persistence of inflation.
Every budget is formulated, either explicitly or implicitly, on a price basis. As prices rise and become relatively unpredictable, the problems of budgeting are felt more keenly. With inflation, one year may appear too long a period to keep costs and budgets the same. Inflation budgeting implies a correct reckoning of current and prospective prices—the prices at the time of the budget formulation and the changes in prices after the budget becomes operational. The greater the rapidity of changes in prices, the greater are the problems encountered in policymaking and for the budget in fulfilling its economic role. What is the nature of inflation? How does it affect the budget? What are the current approaches toward these issues? What are the policy options when the rapidity of price change is greater than estimated? Should the rate of inflation be forecast? What are its implications? How can a budget be formulated to reflect the realities of inflation? Is indexation a partial or complete answer to these issues? What are the implications of inflation for program management? These are only some of the questions that need to be answered. This chapter and the following one consider these issues. This chapter deals with the more general issues of inflation budgeting, while the following chapter is devoted to the role of expenditures in economic management. These chapters do not seek to offer any settled conclusions or guidance on how specific aspects of the budget are to be formulated, but are concerned with a delineation of the issues and a discussion of the state of the art.
Inflation: Causes and Nature
Inflation as a phenomenon is experienced by all members of the community. Some benefited by it but most are affected adversely. While the general perceptions about inflation are fairly common, the factors contributing to it vary. These factors come in various forms and differ in their incidence from economy to economy. However, five factors are enumerated here as those contributing to inflation. First, the pursuit of expansionary fiscal policies by the government has led to higher deficits, easy money policies, and greater liquidity. The commitments to achieve full employment in the western industrial democracies and to use inflation as a source of finance for development plans in developing countries are frequently cited as major contributions to expansionary fiscal policies. Second, either because of the expansionary fiscal policies or because of generous credit facilities to the private sector or a combination of both, there has been an increase in the money supply in the economy and much of the easy liquidity is to be ascribed to the monetary policies of the central banks. Third, economies have been overheated, reflecting specific events of social disturbances such as wars. In industrial countries such overheating occurs when industrial capacity is fully utilized and employment is high and when increased demand causes prices to rise. In developing countries, short-term price increases may take place, not because of full utilization of capacity but because of a sudden drop in crop outturns and consequent commodity shortages. Fourth, there has been a gradual disappearance of money illusion causing wage earners to demand and receive compensatory increases in money incomes to offset actual and anticipated price increases. As a result, wages have risen faster than productivity, pushing up unit costs and leading to price increases. This sets in motion a new chain of activity of wages chasing prices in a spiral and contributing to ongoing inflation. Fifth, the general increases in oil prices since 1973 have changed the supply side and have generally added to the costs. In addition to these factors, it is also suggested that the large balance of payments deficits of the United States through most of the 1960s and the growth of international credit from commercial banking institutions has provided further stimulus to the growth of inflationary pressures in world economy. Depending on the choice of factors, solutions have also been suggested. The fiscal deficit theorists believe that inflation can be cured by balancing the budget and reducing the deficit. The monetary school adherents suggest that controlling the aggregate money supply would be equally effective. Those who subscribe to the overheating theory suggest that as soon as the heat disappears and as demand returns to normal inflation will be reduced; it is believed that a better distribution system will be helpful in reducing commodity shortages. Others argue for tightly implemented wage and incomes policies. However, as wage and incomes policies have not been successful, and as mandatory controls leave much to be desired, emphasis may be placed on a combination of incentives and voluntary action and, therefore, some economists defend tax-based incomes policies. More pragmatic and conventional policymakers prefer a combination of fiscal and monetary policies to fight inflation. The diversity in approaches and solutions is not surprising, given the multidimensional character of inflation. It is evident that inflation is not a one-cause/one-solution problem but is many faceted and requires more organized efforts.
The necessity to fight inflation hardly requires any emphasis in view of the undesirable results that arise from high rates of inflation over prolonged periods. Four effects are easily identifiable. First, as liquidity increases and prices rise, there is a distinct possibility that exports will be discouraged or imports will be encouraged, thus contributing to greater problems for the balance of payments. The experience of the industrial and developing countries conclusively illustrates that strong inflation leads to capital flight, an increased demand for imports, and reduced exports. In due course, pressures will be exerted on the exchange rate and frequent changes in the rate could add to the supply costs of certain commodities. More important, this will lead to the search for financial resources to tide over balance of payments difficulties. Second, inflation has effects on the desire for liquidity in speculative and precautionary motives. Weakening traditional modes of liquidity will encourage consumption and discourage saving and will exert influence on investment and reduce investable resources. As the normal sources of liquidity do not offer protection against the erosion of the value of money, investable resources will move into housing, gold, and other assets that seem to provide a hedge against inflation. Third, there will be an uneven impact on the distribution of incomes. As fixed-income groups struggle to catch up in their incomes, other factor incomes might increase disproportionately to their role, exerting greater stress and strain on the social cohesiveness of society. Fourth, without financial stability, progress in achieving other goals, such as growth, would be slower and would become increasingly difficult. More significantly, the erosion in the confidence of the community as a result of inflation will have a wider, unquantifiable, but enormous impact. Okun graphically described the situation: “when the economy goes wrong, nothing goes right. Its malfunctioning robs us of our self-confidence. It creates distrust. People feel squeezed and cheated and hunt for the villains and oppressors. So when we most need productive partnerships to solve our problems, we are confronted with disunity.”1
These four effects illustrate the need for comprehensive action to restore stabilization and the community's faith in the value of money. To achieve the goal is difficult and requires hard choices. Implementing the policies will involve painful adjustments and, presumably, loss of political support in the short run. If the goal is to receive political support, experience illustrates that inflation could be a potent weapon during the period for which the proverbial memory of the voter is short. If, however, the faith in a statesman is restored, with adequate public discussion and support, implementing these policies may become somewhat easier.2 Policies should not reflect frequent changes from an overriding commitment to reduce inflation to moderate reflation the moment signs of a recession appear. Anti-inflation policies take longer and need steady adherence to the ultimate goal.
Role of Budget
The impact of the budget on the economy and its role in the management of the aggregate demand has been analyzed earlier. The issue, however, is whether the management of the demand has translated itself into a coherent policy for the control of inflation. If it is recognized that the current rate of inflation is not necessarily the result of government deficits, it would then be appropriate for anti-inflation policies to be oriented to other factors, particularly those that have added to costs. Viewed from this point, and based on the experience of both industrial and developing countries, four factors of budget activity merit recognition. First, on the tax side, increases in indirect taxes have a tendency to contribute to higher costs and in due course to increased prices. Second, governments may often increase the money wages of their own employees and, through legislation applicable to the whole economy, the minimum wage. This in turn leads to increased production costs. Third, maintaining price supports for the farming community in industrial countries frequently leads to a cost-price spiral. In developing countries, the incidence of such farm support is uneven and is more prevalent in monoculture economies where marketing boards are required to provide remunerative prices. Fourth, expansion of social regulations, particularly automobile emission controls for ecological purposes, and related health standards frequently lead to increased costs. These aspects show how government budgetary activities could lead to higher costs and higher prices. They have not been given due recognition earlier, partly because of the preoccupation with financing the budget deficit and partly because some of these measures, particularly the impact of regulatory expenditures, general wage increases, and the impact of tax increases on the cost push, are not appropriately analyzed in the budgetary process.
The deficit of the government budget is another important factor that contributes to inflation. Conventionally measured deficits generally are those of central or federal governments and exclude, for lack of statistical data, the deficits of state and local bodies. While in some countries local governments are not permitted to raise loans in the market, in countries like India the deficits of state governments are financed to a major extent by credit from the central bank. The aggregate deficit of all governments may be financed differently in industrial and developing countries. In industrial countries it is financed by the sale of securities to the public and to the commercial banks, which prefer government bonds for reasons of high yield and less risk. In some countries, securities held by commercial banks are rediscounted, while in the Federal Republic of Germany the central bank provides no rediscounting on government bills. In developing countries, government securities are held mostly by the central bank and by other captive funds of government, such as provident and pension funds. Empirical investigations into the financing of debt in selected developing countries in Africa, Asia, the Caribbean, and Latin America for the decade 1967–76 show, for example, that in Botswana, Korea, Swaziland, Trinidad and Tobago, and Tunisia sizable portions of the deficit were financed by external borrowing. (This includes project borrowing from international financial institutions.) In other countries, particularly in Ethiopia, Guyana, Guatemala, Jamaica, Peru, Thailand, and Zaïre, a major part of the budget deficits was financed by the banking system, in particular, by the central bank. While the relative roles of external borrowing and borrowing from the central bank and captive funds have changed over the period, reflecting the growth of other domestic financial institutions, the fact remains that growing deficits contributed to greater liquidity, money supply, and inflation.3
The budget deficit when initially formulated is based on a prognosis of the economy. The perception of the economy itself may contribute, and indeed has contributed, to further inflation. The experience of several industrial countries during the early 1970s suggests that the money illusion has been working heavily with fiscal planners, contributing to more liberal transfers and social security payments that very soon outstripped the available revenues and made resort to bank financing imperative. Further, the prognosis of the economy itself may be made more favorable than warranted. In turn, this has led to situations where expenditure could not be curtailed, resources could not be raised for fear of adverse effects, and greater budget deficits fueled further inflation. When developing countries prepare their budgets, inadequate attention was paid to inflation-induced declines in real revenues and the impact of inflation on expenditure. In practice this meant that the initial budget was formulated on what was considered a reasonable deficit and, as the year progressed, the widening of the deficit loomed large and greater resort was made to inflationary financing. Unwittingly, the budgets, in the course of their implementation, and in terms of their deviation from the original plans, became instruments of inflation.
Impact of Inflation on Budget
The central element in the above behavior is not merely the budget impact on inflation, but the impact of inflation on the budget. The elements of this impact have been studied on an empirical basis and, notwithstanding the diversity of the approaches to these studies, certain broad features of the impact may be stated. For analytical convenience, these are provided in Table 13; these conclusions are not universal and are subject to serious limitations.4 The impact of inflation on revenues and expenditures is based on their respective elasticities and lags with which they react to changes in the economy. If budgeting were to be an automatic exercise, all that would be needed would be to ascertain the elasticities and lags and apply them to the annual forecasts of GNP. Realities are, however, more complex and much is dependent on the revenue system, the operational framework for the collection of taxes, the composition of expenditures, the nature and quality of services provided by the government, and the relative emphasis placed on distributional objectives of government. Specifically, in respect of government revenues and especially in regard to income and corporation taxes, two different trends emerge. In industrial countries with progressive tax systems, with elasticity of more than unity, and with minimum lags in collection, inflation will have the effect of moving people into a higher tax bracket and overall tax revenues will tend to grow faster. In developing countries, however, the effect of inflation may be a reduction in real revenues because of low elasticities of revenue and longer lags in collection. Also, a greater share of revenues in developing countries are derived from indirect taxes—primarily taxes on domestic transactions and international trade—that are often specific and adjustment for changing values is made after long lags. Much is dependent on the rate of inflation and the period for which it lasts. If the rate of inflation is high and if it is not chronic, the elasticity of some income taxes may even overwhelm the lags in collection in developing countries and will have a favorable impact on total revenues. If, however, inflation persists, that advantage would be lost, payment lags would be longer as they would favor the taxpayer, and there would be a decline in real revenue. The effect on nontax revenues is dependent on the rapidity with which interest rates are adjusted and tariffs for government services and products are changed to reflect the impact of inflation. The overall effect is likely to be neutral, however, as government also pays interest on its borrowing and to that extent is adversely affected. Revisions in tariff may permit the continuation of the previous fiscal balance, but this is unlikely to show any improvement. It is more likely, given the political difficulties and the reluctance to change prices frequently, that the overall impact on the budget will be negative. Thus, depending on the nature of the tax system and the rate and duration of inflation, there may be an increase in real revenue and the working of the fiscal drag5 may become more apparent, or the real value of revenue may remain unaffected or may even suffer a reduction.
|Budget Categories||Likely Effects||Remarks|
|Income, net wealth, and inheritance tax||Positive||Inflation is likely to gain more receipts. However, the gains are dependent on techniques of assessment and collection. Government will be adversely affected if inflation prevails over a long period and if the interval between accrual and liability and payment is long|
|indirect taxes: Customs: Taxes on goods and services||Positive/Negative||Impact is dependent on the specific or ad valorem nature of taxes and duties|
|Interest receipts||Positive/Negative||Governments are both creditors and debtors. Government's revenue position will improve if interest rates are adjusted quickly. As an extension of this, its expenditures will also increase|
|Fees: Transfers from public enterprises||Negative||Considerable lags in price adjustments, which generally occur, contribute to government losses|
|Wages and salaries||Positive/Negative||Impact is dependent on the rate and timing of adjustment made in public service salary structure. Automatic adjustments will lead to greater government outlays|
|Interest and debt retirement||Positive/Negative||Impact is dependent on size of debt, level of interest rates, and rate of inflation. If bonds and interest rates are indexed, government outlays will be higher. Delays in adjusting interest rates and absence of indexation contribute to government gains|
|Other Transfer Payments|
|Pensions, etc.||Negative||If indexation or other similar automatic arrangements exist, government expenditures will be higher. Magnitude is dependent on share of transfer payments in total expenditures|
|Subsidies||Negative||During periods of inflation, subsidies tend to become open ended owing to lags in price adjustments|
|Fuel and energy expenditure||Negative||Extent of expenditure increases are dependent on amounts of energy consumption and pattern of subsidies|
|Purchases of equipment and machinery||Generally negative||Private sector price adjustments are quicker and contain margins for expected rate of inflation and delays in payment by government. In countries where the public sector's imports are large, extent of government losses will be higher, depending on the rate of inflation in the exporting country and the type of equipment|
The impact on expenditures is far more diverse and reflects the variety of expenditure. Expenditure on wages and salaries is likely to rise primarily because their revision is often linked to a price index. As the objective of public salaries is not merely to buy a service but to be competitive, to maintain parity, and to maintain a standard of living for its employees, it is possible that expenditures on wages and salaries will rise.6 More often, however, these adjustments are made after a gap and frequently the adjustment is for a past period rather than for current or prospective periods. Where salary levels are pegged to an index, total expenditure will increase with a constant lag but in line with inflation. Other transfer payments—pensions and subsidies—are likely to register sharp increases. Pensions are often related to an index and the volume of subsidies to government-owned enterprises and to other enterprises will rise, depending on the extent of price restraint sought by government in essential goods and the rate of inflation in other traditionally subsidized sectors. The major sector of government expenditures that is adversely affected is the general category of other goods and equipment. A major component of current expenditure is fuel consumption, the price of which tends to fluctuate with inflation. In capital expenditures, purchases of land and existing buildings (the prices of which increase both for speculative and inflation reasons) and equipment reveal that price adjustments are faster in the private sector and the overall impact on government expenditure is negative. In construction, there is usually a provision in the contract allowing the price to be adjusted by reference to wages and material costs paid by the contractor. Because of this escalation clause, government expenditures on construction tend to move contemporaneously with inflation. These factors would be evident from an analysis of expenditures in terms of functional classification. Agencies such as Defense, Public Works, Transport, and Communications register a higher rate of increase in expenditure because equipment and construction is the largest component. Agencies that employ more manpower, such as education, have lower increases in expenditures, as public service salaries lag behind the general rate of inflation.
A major finding of empirical investigation has been that expenditures in developing countries adjust faster than revenues to inflation.7 This conclusion is debatable for both assumptions made in investigations, as well as for more practical considerations. It was assumed in the formulation of the model that, in the short run, it would be difficult to reduce government commitments and expenditures and, therefore, these are automatically adjusted to keep pace with inflation. This implies a pure mechanistic approach by the policymaker and assumes that he has no discretion. Experience does not lend support to this argument. The use of data is also weakened by the fact that they do not distinguish between volume and price changes and do not account separately for increases in expenditures following a depreciation in the exchange rate. Also, the nature of results changes depending on the deflator used. In fact, as pointed out in Chapter 5, salaries in public services have often lagged behind and, as shown in the following chapter, the reaction to inflation has been to reduce the expenditures at any cost. What emerges from the experience is that the overall impact of inflation in the budget is indeterminate. In countries with highly elastic tax systems the impact in the short run may be beneficial. In countries that have less elastic tax systems, the impact on expenditure being greater, inflation contributes, for technical and administrative reasons, to increased deficits in the short run.
Approaches to Budgeting
The approaches to budgeting in an inflationary context reveal several weaknesses that affect the overall budgetary process. One common approach that had a good deal of acceptance at a practical level, and that is still to be found, is one of wishing the problem away. Support for this approach may be considered from several angles. For example, it is suggested that ignoring the problem could eliminate the tiresome business of deciding how much to put in the budget estimates to reflect growing costs. Similarly, it is felt that explicit provision of increased pay and expenditures in the budget estimates may attract the critical eye of the legislature and of the public as well. More important, inadequate estimates may intrinsically be a good thing, as they may provide a powerful incentive to restrain spending within the budget by seeking compensating reductions in some areas. Explicit provision for inflation is regarded as a self-fulfilling prophecy.
Operationally, the above approaches converted themselves into budgeting based on prices prevailing at the time of the preparation of the budget. This, however, had an inconsistent application in that the likely increases in prices were used for estimating revenues and not for expenditures, and budgetary deficits with a downward trend were achieved. To the extent that there was any increase in prices affecting expenditures, it was expected that such increases would be offset by increased productivity in government services. These approaches did not have any discernible consequences, as long as the rate of inflation was low and increases in expenditure were compensated for by increases in revenue, which, despite the recognition of the price impact, had an adequate margin. But with increases in the rate of inflation, it was recognized that these approaches had limitations. Productivity increases could not match up with the rate of inflation. The effort of using the prices prevailing at the time of the budget preparation proved a liability, for there was a considerable gap between the compilation of estimates by agencies and the implementation of the budget. This approach had a twofold effect. First, there was difficulty in interpreting expenditure proposals and assessing their financial implications and in evaluating the impact of budget on inflation. Second, the agencies went through a scissors effect of reduced allocations as a part of anti-inflation strategy and erosion in real terms in its expenditures. To a certain extent this may have been a balancing mechanism in the economy as a whole. Fellner argued, for example, in his treatise on wartime inflation, that a decision that seeks to allocate more than 100 percent of available resources to the various claimants cannot be executed,8 and that inflation functions as a mechanism for revising the supposed decision in order to make final shares within 100 percent. In normal administrative life, however, budget formulation and management became difficult. It became a series of crises and allocation of funds came to be made for meeting dramatic crises, such as a transport strike, a doctors' strike. As each case threatened public convenience, funds had to be allocated to them, although this meant, in the process of decision making, a neglect, sometimes benign but more frequently deliberate, of other needs and claims. In countries that had medium-term plans, the budgets ceased to fulfill their annual role as the direction of the budget and the economy and their mutual interface became difficult to identify in the welter of ad hoc approaches.
The experience clearly indicates that inflation cannot be ignored or compensated for by balancing mechanisms between revenues and expenditures or productivity gains, but must be explicitly recognized so that for coherent fiscal policy, budget estimates can be properly analyzed, reviewed, and improved. There are several ways in which this objective could be attained. Essentially, and on the basis of the experience of several countries, three different approaches may be considered.
The first is to anticipate the rate of inflation and provide for estimates that reflect increasing wages and prices. The rate of inflation itself is influenced by several exogenous factors, including the fiscal position of the public sector itself. The rate is then used for explicit provision in the detailed budget estimates. In the early 1970s, Australia, the United States, and some developing countries resorted to a “lump sum provision” for the purpose, which would then be allotted to different agencies. This lump sum was intended in the United States to cover the costs of procurement and other expenditures and, as the rate of inflation grew to double digits, it was considered more convenient to make detailed provisions. This approach implied that the effect of inflation would be fully taken into account and that expenditure programs would not be crippled. A variant of this is adopted in developing countries. Under this approach, no provision is made for inflation in the initial estimates and, as inflation is experienced, supplementary provisions are made. But this, however, is more of a crisis management approach.
A second approach, which is applicable mostly to the United Kingdom, is to evolve medium-term plans in real or constant terms and to make adjustments at the beginning of the fiscal year for the rate of inflation and to provide needed amounts. In both approaches, the key element is the rate of inflation forecast for the next year; like all forecasts, there is a speculative element in the prediction of this rate. The rate has its own impact on other aspects, particularly wage bargaining, and on the behavior of the agencies. The problem is, however, that if the actual rate of inflation is higher or lower than the estimated rate, there would be expenditure excesses or, alternatively, shortfalls.9 Some elements of expenditure are particularly difficult to forecast and are dependent on a larger group of exogenous factors. For example, subsidies are demand related and difficulties are experienced in forecasting them. Much the same could be said about transfer payments, which may increase or decrease depending on the state of the economy.10 The methods by which these overruns (or unanticipated inflation) are accommodated lead to the consideration of other approaches to inflation budgeting.
The third approach, which was used in Canada during 1977–79, was to organize a “reserve against statutory overruns.” The problem faced by Canada and a number of other countries was that expenditures under entitlement programs that specify conditions of payments to the beneficiaries in law tended to exceed estimates owing to unanticipated changes in the economy. These expenditures, which were called statutory in Canada, were to be contained within the limits of the reserve. The concern of the reserve was with expenditure overruns, while shortfalls were considered to be more manageable.
The need for containment of expenditures is seen in these approaches as a key element to restrain the budget deficit, otherwise the increase in the deficit would generate more inflation. In countries such as the United States and the United Kingdom, where the maintenance of estimated deficit level is an important policy element, an option is to maintain all expenditures other than the entitlement type and to permit the movement of deficit within reasonable limits, provided that change is totally ascribable to entitlement payments. A variation of this approach is to accept the amount of deficit as given and to mobilize additional revenues for meeting the impact of unanticipated inflation. Yet another option is to maintain the deficit as planned and to undertake expenditure reductions or adopt a combination of expenditure reductions and increase in tax rates. These options illustrate the imperative need for estimating prices in an appropriate way.
In budgeting for inflation or, more appropriately, in providing for an explicit recognition, measurement, and linking of price inflation for both revenues and expenditures, distinction has to be made between medium-term planning purposes and short-term budgetary needs. The concern here is with shorter-term purposes of the preparation of the annual budget. In formulating a budget for meeting and reflecting on the inflationary content in the economy, it is assumed that the sources of inflation and its nature have been identified and that necessary policy options have been made. Within that policy framework, two approaches to budgeting may be distinguished—the economist's and the budgeteer's. These two functions do, in practice, overlap or may be combined in a single office in a government. Notwithstanding such a combination, there is a difference in the approaches. The concerns of the economist are with the aggregates and the balance in the division of resources between the public and private sectors and how this balance, particularly the opportunity cost of expenditure, will be affected by inflation. The tasks of the budgeteer, while working within that framework, is more at the disaggregated level as he ensures the proper basis for estimating revenues and allocating expenditures. The economist's choice, reflecting his concerns, is to use national accounts, which provide, in constant prices, the patterns of public sector demands over time and on the acquisition of the real resources of the community. These aggregate accounts do not provide the same facility to the budgeteer, who has to consider the differences in the nature of the resource acquisition among agencies and to make appropriate procurement arrangements. While there is no easy compromise between these approaches, as each one of them has a vital purpose and complement each other, it also appears that they have influenced the choice of instruments over the years, thus causing a degree of complexity. It is essential to recognize that these paths are not independent and do converge at various points in the budgetary process.
The preparation of revenue estimates in an inflationary context is less complex than the preparation of expenditures. For income and corporation taxes, the elasticity of the tax system, the lags in collection, and the accepted assumptions for price increases during the period provide a reliable guide for estimation. Such estimates should take into account the extent of fiscal drag and the adjustments that are considered necessary to reduce the drag. Taxes on domestic transactions and taxes on international trade, to the extent that they are specific, would be based on the projected levels of production or imports. As for fees and other rates charged by the government for its products and services, it has now become customary in government to revise them with a lag rather than in anticipation of the increased rate of inflation.
Government expenditures offer a wider range of issues reflecting the different tasks of each agency. Expenditures are incurred for personnel and also for the procurement of sophisticated technology to produce atomic energy. These two items illustrate the range of services available. What prices should be used by government to forecast expenditures in an inflationary environment? Economists tend to support the use of a single index for preparing forecasts, while the budgetary experience of industrial and developing countries shows that agencies prefer detailed prices reflecting the plant and machinery used by them. A traditional approach has been to use, at the aggregate level, the GDP deflator. 11 The GDP deflators, it must be admitted, are difficult to forecast, and refer to a basket of commonly purchased items at some convenient point of time that do not always reflect the nuances of the variety of public expenditures. Furthermore, these deflators may be influenced by special factors such as commodity price increases overseas that may not altogether be relevant to public sector management. There is, however, a dilemma that confronts the policymakers in this respect; for measuring the balance in the economy a policymaker requires a general price index, while for budgeting more detailed indices are needed.12 If the former is used for budgeting, there is a possibility that it will understate the requirements of the public sector. But the use of more detailed indices within the government may require the use of similar price indices for the private sector, so that the demand for resources for both sectors may be denominated in a comparable fashion. The development of such indices would involve greater investment of resources, which, apart from being difficult, may not be attainable in the short run. Therefore, reliance has to be placed on expedient approaches; if detailed indices are to be used within the government they would have to be adjusted to be compared with the private sector. Such adjustments may involve arbitrary elements but these will be no more than are now in use in the compilation of national accounts. This approach has the advantage, in addition to being realistic, of minimizing the wider margins of error inherent in the use of a single index.
Specifically, the three main categories of expenditure that vary with inflation are expenditure on wages and salaries, transfer payments, and expenditures on capital formation. The use of GDP deflators on wages and salaries, as noted in Chapter 7, tend to overstate the share of public spending in GDP. Determination of annual manpower requirements of government is based, in any event, on the more practical measurement of increase in work load, additional tasks undertaken, and related factors. Therefore, a more appropriate base for predicting wage and salary expenditures by the government would be to formulate it with reference to the expected increase in the national average and the lag with which government salaries are revised, as is done in Japan and Sweden. In developing countries, some types of emoluments are revised on the basis of the consumer price index and projections of that index would then form the basis for the government's own expenditure. Similarly, transfer expenditures, particularly pensions based on price indices, are also adjusted. A thorny issue arises in regard to subsidies to specific industries and to government purchases of land, buildings, and capital machinery and equipment. In this respect, the use of sale prices of broad sectors of industry relevant to general and government purchases would be more appropriate.
The use of these indices does not and should not imply that government revenues and expenditures are automatically linked to selective indices and that allocations would be made to maintain the real levels. On the other hand, the use of these indices has two basic purposes—first, to ensure that there is a proper reckoning of prices and that assumptions regarding future price increases, nominal budget amounts, and the real levels of programs are all mutually consistent, and second, to ascertain the adjustments needed either by mobilizing additional resources or by reducing expenditures in real terms if the budget deficit is to be maintained at the amount forecast. The absence of these implies a money illusion and a budgetary policy that lacks purposiveness.
Inflation budgeting adds additional responsibilities to budgeting and planning agencies at the central level. It is appropriate that the central agencies, which have the responsibility for national economic management, also provide guidance on the price assumptions to be used by different agencies. Provision of such guidance implies a detailed framework of economic planning. In many developing countries, as noted earlier, planning continues to be largely an exercise in allocating resources for investment purposes and there has not been rapid progress in the anticipation and measurement of inflation. Relegating this responsibility to the agencies would lead to confusion, disparate assumptions, and lack of consistency in the policy framework. Central guidance would also be needed for the selection and utilization of sector-oriented indices. These additional responsibilities are necessary inconveniences in an inflationary era and the tasks of policy management would be considerably eased if adequate attention were paid to these aspects. Will these problems be less if there is selective indexation of revenues, expenditures, and debt? Is indexation a better alternative to discretionary action? What are the relative advantages and disadvantages?
Aspects of Indexation
Reflecting the rapid increases in prices and the difficulties in adjusting categories of the budget to changing prices, selective indexation of revenues, expenditure, and debt has been undertaken by a number of industrial countries and, to a lesser extent, by developing countries.13 Although indexation as a phenomenon received wider attention during the early 1970s, the concept itself has a good deal of history.14 As early as 1887 Alfred Marshall observed in a memorandum to the Royal Commission on the Values of Gold and Silver “that Government should provide a tabular standard of value for optional use within the United Kingdom in all transactions which extend over a long period of time. … A theoretically perfect standard of purchasing power is unattainable. … But the index numbers with which we are already familiar would give ten times better standard of value for optional use within the country in long-standing contracts than even a true bimetallic currency.” Since then, indexation of one type or another has been used in government transactions. The use of the term itself has, however, introduced an element of confusion into the discussion of the subject. It has come to be considered as being restricted to the adjustment of investments, savings, or bonds, generally by an index assumed to reflect the change in the general purchasing power of money. In practice, however, it is appropriate that usage of the term should include the changes in contractors' prices, as well as changes in a number of other areas.15 In this larger sense, some kind of indexation is observable, even if on a smaller scale, in government transactions.
Over the years the issue of indexation has generated a good deal of debate, and opinion is evenly divided between those who see it as an imperative need and others who view it as virtually giving up the fight against inflation. The need for indexation is argued for several reasons. First, when inflation persists at a high rate over a prolonged period, it becomes desirable to introduce indexation to preserve the basis of economic life and to mitigate the adverse influences of inflation on saving and investment. Second, the impact of inflation is uneven and unfair to those sections of the community where pay or receipts cannot be adjusted and, therefore, indexation is needed to insulate real incomes. Third, indexation of loans would protect the investor and promote greater mobilization of financial resources for the government. Fourth, that it will not have the lags and other drawbacks that are inherent in any system in which corrective action is sought through discretionary action. Those who are disinclined to indexation refute these arguments.16 The applicability of these arguments differs, since wage settlements, as distinct from other settlements, evoke different responses. The relevance of these purposes and arguments is to be seen more in terms of specific budget categories.
Indexation has been applied for income taxes in a number of countries of the Organization for Economic Cooperation and Development (OECD). This was done primarily to mitigate the inequitable impact of a progressive tax system with high elasticity, which, through the working of fiscal drag, obtains more revenues for government. While normally an increase in tax rates during periods of inflation would win approval as a policy choice, it is the inequitable and arbitrary nature of the revenue increases that is questioned. Furthermore, the full operation of fiscal drag could in due course, prove to be deflationary. Indexation, on the other hand, will maintain the tax equity and will help stabilize real output by sustaining real disposable income in the context at rapid price increases. In practice, however, indexation is not applied in its entirety to the tax system—the most notable exceptions being excise duties and other taxes on sales and related transactions—and, when applied, is done retroactively through administrative action. Several countries have a system of bracket indexation, while a few others seek to minimize the impact of inflation on company taxation by frequent but ad hoc reliefs. Governments have been, by and large, reluctant to index systems lest the buoyant revenues that enabled them to hold down deficits would be lost and they would be plunged into greater deficit, with attendant adverse impacts on the economy.
The indexation of expenditures is relatively more widespread, although there are noticeable differences between industrial and developing countries; in both, indexation of wages is widely prevalent. In most countries, wages are linked to a cost of living index.17 In Brazil and Israel, wages move along with changes in the index, while in several other developing countries they are adjusted after a lag through deliberate government action. Social security payments, particularly unemployment benefits and pensions, are generally indexed in OECD countries and in a number of Latin American countries. Certain types of income maintenance programs, for example, food stamps, and nutrition programs are similarly linked to a price index. An additional feature of benefit indexation is that the adjustment is generally automatic and contemporaneous rather than lagged. Capital expenditures, except for specific contracts, are not pegged to any index per se, although escalation in costs is recognized and compensated for.
The relevance of indexation for debt has received additional stimulus since the mid-1950s, in both industrial and developing countries. If the concern in industrial countries is to restore the original purchasing power of government bonds, the additional feature in developing countries is the need to promote small savings by the public and to channel them into government finances.18 Unless appropriate protection is offered against erosion by inflation, it is felt that the public will not be interested in buying government bonds. Those who favor indexation argue that it is the moral duty of the state to refrain from “defrauding” bondholders via inflation, that indexation will actually prove to be helpful in promoting saving, and that, on the whole, it will bring in the needed resources to the government. These considerations can be and have been subjected to a variety of considerations, including the fact that the issue of a guaranteed purchasing power bond would encourage inflation by removing a fixed-sum asset holding group from among the ranks of anti-inflationary pressure groups that influence government policy.19 It is also suggested, and evidence supports it, that the increase in savings need not necessarily contribute to reducing inflation, as many offsetting increases in expenditure will neutralize that. Also, to the extent that the bond issue is supported by the central bank, it will contribute to greater liquidity. Furthermore, when debts are repaid, there will be increased outlays that may aggravate inflationary pressures. Much is also dependent on the actual use of resources raised through bonds, which, when utilized for purposes other than the expansion of real capital, can be saddled with a debt burden that cannot adequately be serviced by the growth in real output.20 An argument implicit in the indexation approach is that the real interest rate under inflation should be the same as the rate under conditions of monetary stability. This may not in fact be so, as the interest rate that a borrower would be willing to pay depends on the type of inflation. The inherent rates tend to have a premium during periods of inflation which should be adequate for attracting bond purchasers. Specifically in regard to developing countries, it is agreed in a context where a major part of the debt is held by foreigners, who may insist on exchange rate guarantees, governments may not be enthusiastic about this. Also, once there is indexation with reference to a price index that includes indirect taxes, governments may not be willing to raise revenues by revisions in the rates of indirect taxes, for each such revision would also imply higher debt service charges.21 Experience shows that indexation in industrial countries is primarily on long-term loans to protect the small saver. In developing countries, excluding Brazil, greater reliance is placed on the utilization of captive funds (provident funds, trust funds) for raising debt capital. There has been, however, a growing consensus for a selective and flexible indexation approach within the overall policy framework.
Issues in Indexation
The approach of indexation is considered an implicit defeatist attitude toward the problem of inflation, that it is a shelter from the effects of taxation, and that once initiated it will prevent consideration of other alternatives and will in due course generate pressures for the extension of the system to all financial transactions.22 Indexation of wages implies an automatic movement that may not involve negotiated settlements. Those who support indexation argue that lagged adjustments to prices lead to welfare losses and that it is naive to suggest that negotiated wage settlements would be any less than automatic ones. The whole issue is dependent on the impact of indexation on the stabilization of real output in the context of changes in demand and ability to control inflation. On both these counts, the debate is inconclusive. It is also clear that indexation does not offer a cure to the hydra-headed nature of inflation faced during recent years. At a practical and operational level also, indexation raises a number of issues. First, timing is of crucial importance. Efforts at maintaining the levels of taxation and expenditures as of a particular date imply a judgment that the level is appropriate for the economy. There is a probability, however, that indexation might take place at a wrong point in the business cycle. Second, there is the question of the appropriate index to be adopted for the purpose.23 In some countries, indexation was introduced with reference to the consumer price index, while in some OECD member countries, it had reference to wages. The consumer price index may, however, have an upward trend even when the prices of manufactured goods are stable, and at times the increases in the price index may be so large that for economic reasons there is no need for offsetting adjustments. If indirect taxes and increased oil prices are included in the index (as they are), it implies that by indexation, increases in excises and sales taxes and oil prices will automatically lead to indirect reimbursement by the government. In some cases, the index could be distorted by administered prices and subsidies. The issue also arises whether certain sections of society should be given preferential treatment. Beneficiaries of social security systems fare better than average wage earners in some countries, as their benefits are based on a price index.24 Other issues of major importance include the frequency of adjustment (once or twice during a year) and whether the benefits should be adjusted to past or prospective rates of inflation. In the process, the computation of the index itself has become a major contending issue—should it be an all-embracing single index or a series of selective indices?25 Prevailing practices reveal, for obvious reasons, preference for several indices.
In the final analysis, the choice of indexation is dependent on the old debate about discretionary action versus indexation, and about medium-term planning, which has built-in factors for annual adjustments for changes in the rate of inflation. The experience of indexation, despite extensive debate, has been rather limited. Even in countries like Brazil and Israel with experience of indexation, it is limited to wages, pensions, and government long-term debt instruments. To that extent there has been greater preference for discretionary action. While such action has the potential for some lags, it also enables government to formulate policies for changing situations, rather than leaving them to automatic adjustments. Medium-term planning of the rolling expenditure forecasting type has also been helpful in anticipating problems and has had as many facilities as indexation. The crucial problem area has been one of forecasting the rate of inflation. If annual correctives are made in medium-term expenditure plans, the role of indexation becomes academic.
Inflation and Program Management
The impact of inflation on program management may be analyzed in terms of the effects on central agencies and on spending departments. The central agencies, which have the responsibility for managing the budget, are confronted with difficult tasks, particularly when a major part of expenditures is for transfers to other levels of government, subsidies to productive sectors, and transfers to households. As noted above, a feature of the industrial countries is the extensive prevalence of social security systems under which benefits are indexed. In a context of high rates of inflation these expenditures, as noted in Chapter 1, grow along with the cycle rather than being countercyclical. As a consequence, the central policymakers have lost a good deal of control and are hemmed in on all sides by commitments that have already been made and over which they have no discretion.
At a program level, inflation imposes greater burdens on managers. Such burdens accentuate when programs have to be implemented within specified resource ceilings. Inflation demands that the program manager be aware of the differential impact of price changes on his program content, that he is able to forecast the rate of inflation, and can contain the outlays when the actual rate of inflation is different from the estimated rate. This requires, in turn, a cost awareness, an. information system that provides data on the progress of the programs, as well as changes in the prices of goods and services used by them, and a decision-making framework for making adjustments. Experience of both industrial and developing countries suggests that these are areas where more progress is needed. Specifically, there is a need for a mechanism that would “cap” the outlays when the actual rate of inflation is higher than the estimated one. The United Kingdom has introduced a system of cash limits to meet this problem.
Cash Limits in United Kingdom
The system of cash limits, which was selectively introduced in the United Kingdom in 1974/75, was extended to a major part of the Government in 1976/77. The limits were essentially indications to the spending departments of the amounts that would be available to them. The limits incorporated prespecified allowance for inflation and were intended to convey a financial discipline by making clear the amount of money that the Government was prepared to pay in the year ahead and some purchases were to be cut to the limits if prices were too high.26 In its operational form, the system indicated the maximum amounts to be spent on blocks of services. Initially, the coverage of cash limits was limited but was extended later to cover central government transfers to local governments and nationalized industries for capital expenditure. As a part of the system, an elaborate reporting framework was also introduced to facilitate the introduction of corrective action.
The cash limits system is not without its shortcomings. One major limitation is that it does not cover the entitlement outlays, which continue to grow with the rate of inflation. Its success is dependent on the precision with which the actual rate of inflation is forecast and is likely to be fulfilled when the margin of error is minimal. If the actual rate of inflation is substantially higher (as it can be as governments tend to underestimate the initial rate of inflation), then the gap can be so wide that the implementation of cash limits would be rendered doubtful. The experience of the United Kingdom suggests that when the rates of inflation are forecast well, additional expenditures would be needed during periods of recession (1980/81) to alleviate the financial stress on nationalized industries. Apart from undermining the confidence of the system, frequent revision of cash limits to meet exigencies could trigger uncertainty in the spending departments.27 While these problems persist, it seems appropriate that cash limits be viewed as a framework in which benchmarks are set to provide guidance. Such guidance, while obviously useful, also has limitations, and should not be viewed as a complete solution to problems of inflation but rather should be considered as a technique to be used judiciously in conjunction with other techniques.
Arthur M. Okun (1980), p. 1.
Okun compares the acceptance of anti-inflation to an individual going on a diet “overeating is lots of fun and fundamentally enjoyable: Going on a diet is painful, and it brings few results in the short run.… And there is no clear-cut boundary line between normal weight and overweight.” However, as the risk becomes serious “we are tempted to go to the other extreme and adopt a starvation diet. The choices are never easy and they demand a great deal of maturity” (Arthur M. Okun, 1970, pp. 31–32).
For a specific case study of the role of public sector deficits in the United Kingdom during 1973–74, see David Kern (1974), pp. 18–31.
The studies have been rather few. Among these are E. A. Collins (1968), pp. 393–410; Fritz Neumark (1977), pp. 338–58; Vito Tanzi (1977, pp. 154–67, and 1978, pp. 417–51); Bijan B. Aghevli and Mohsin S. Khan (1978), pp. 383–416; and Peter S. Heller (1980), pp. 712–48.
The tendency for tax revenue to rise faster than GNP because of the effect of rising real incomes and inflation in a progressive tax system is called a “fiscal drag.”
In measuring inflation-induced expenditures, it should be noted that measurement in terms of GDP will always show that the rate of spending on wages and salaries increases faster than general prices. This aspect, known as the relative price effect and discussed in Chapter 7, implies that government spending on personnel is invariant with regard to inflation, as it is assumed that there is no gain in the productivity of employees and that increases in the volume of service is achieved only through manpower increases.
Aghevli and Khan (1978), p. 391.
For political acceptability and in the belief that it will have tighter financial control, many governments underestimate the rate of inflation. The experience of the United States and the United Kingdom suggests that when the actual rate of inflation is higher than forecast, it is seen by the public as lack of control and failure on the part of the government.
An analysis of expenditure outturns in the United Kingdom for 1977/78 and 1978/79 reveals that major shortfalls in estimated expenditures occur in subsidies to nationalized industries, capital expenditures, and in procurement of equipment for defense. Part of the problem is due to initial overestimation. An analysis of these aspects is found in Carolyn Jutsum and Graeme Walker (1979); and Valerie Imber (1980).
In the GDP deflator, the average price of a bundle of goods and services representative of the total output of the economy in a base year is designated 100. The average price of a bundle representative of output in year 19–divided by the base year price is the GDP annual deflator for 19–. The same approach is used for extrapolating for a future year.
The Armstrong Committee in the United Kingdom has suggested the use of a general price index to measure the impact of expenditure plans in an inflationary environment. To this Committee, the question of which price index is to be used is unimportant, as long as consideration is given to the issue whether import prices are to be excluded by using the expenditure deflator rather than the GDP deflator and whether the index should exclude the output of public goods and services as this is not to be counted in the measure of opportunity lost, see Armstrong (1980), p. 22. It could be argued that inflation is measured in terms of the changes in general prices and to that extent a price index would be appropriate. When inflation is serious enough to be a big issue, the several general price indices might move together.
In some countries indexation was used as a curb on the excessive enthusiasm of the legislative bodies in providing benefits to certain sections of the community. For example, in the United States social security benefits were indexed, as it was found that frequent revisions in benefits voted by Congress were entailing more expenditures. The benefits were, therefore, indexed. But with rapid increases in inflation rates after the mid-1970s, expenditures also rose.
See, for example, United Kingdom, Inflation Accounting (1975c), pp. 233–34.
This opinion was in evidence in the later periods of the nineteenth century. See Finch (1956), pp. 2–3.
The Brazilian experience indicates that during the mid-1960s wage revision was based on real wages during the previous 24 months and by additional amounts equal to one half of projected inflation. However, as the projection was always on a conservative basis, real wages declined. To avoid this, another formula was devised in 1968 under which the excesses of actual inflation over the projected inflation period, during the previous contract period, were compensated for. A temporary decline in real wages was still possible as wage contracts were usually signed for a year. See Jack D. Guenther (1975), pp. 24–29.
In the early 1950s, numerous government organizations, particularly public utilities, issued bonds that were pegged to their specific products or services, for example, price of a kilowatt hour, railway tickets for travel. In Israel bonds were repayable, up to the mid-1960s, in U.S. dollars. With successive devaluations in currency, this procedure was abandoned and are now linked to the consumer price index.
See Finch (1950), pp. 10–12.
See Benjamin I. Cohen (1966), pp. 449–57.
For a consideration of the issue in the United States in the early 1950s, see United States, Monetary Policy and the Management of Public Debt: Their Role in Achieving Price Stability and High-Level Employment (1952), pp. 888–89
Finch (1956) suggested that in view of the lack of a reliable price index in some countries, it might prove helpful for an international body to advise on the construction of a suitable index (p. 16). Since then there has been considerable progress in the development of data.
In Israel pension payments were indexed not co the cost of living index but to the prevailing nominal basic salary of the rank in the civil service held by the pensioner.
See United Kingdom, Attack on Inflation (1975a), White Paper on Cash Limits on Public Expenditure (1976c), and “The Management of Public Expenditure” (1979b).
The Armstrong Committee considered cash limits as a crude device for controlling public spending and felt that it was not a durable system in situations of wider gaps between estimated and actual rates of inflation (see Armstrong, 1980). Another limitation of the cash limits system is its one-sided stress on budget expenditures.