Abstract
IN THE AREA of fiscal policy, the Netherlands is unique among the developed countries in having attempted over the past decade to pursue a budgetary policy designed to ensure a balanced growth of the government sector consistent with the longer-term structural growth of total output in the economy. Basically, the theory of budgetary policy that has developed consists of a budget margin analysis in which the Government’s revenue and expenditure decisions contained in the budget proposals are tested against a structural standard, which abstracts from cyclical fluctuations in the level of economic activity and embodies the Government’s assumptions about the long-term rate of growth of national output and the built-in revenue elasticity of the tax system. The purpose of this paper is to describe the budget margin analysis and the associated method of fiscal analysis, namely, the budget impulse analysis, and to bring out their implications for fiscal policy. Sections I to IV cover the description, and Sections V to VII examine the implications of the two analyses.
IN THE AREA of fiscal policy, the Netherlands is unique among the developed countries in having attempted over the past decade to pursue a budgetary policy designed to ensure a balanced growth of the government sector consistent with the longer-term structural growth of total output in the economy. Basically, the theory of budgetary policy that has developed consists of a budget margin analysis in which the Government’s revenue and expenditure decisions contained in the budget proposals are tested against a structural standard, which abstracts from cyclical fluctuations in the level of economic activity and embodies the Government’s assumptions about the long-term rate of growth of national output and the built-in revenue elasticity of the tax system. The purpose of this paper is to describe the budget margin analysis and the associated method of fiscal analysis, namely, the budget impulse analysis, and to bring out their implications for fiscal policy. Sections I to IV cover the description, and Sections V to VII examine the implications of the two analyses.
I. The Underlying Budget Philosophy
In Netherlands terminology, the structural standard that forms the basis of medium-term fiscal planning is called the (structural) budget margin, and this nomenclature has been adopted here. The precise procedures for determining the size of the budget margin in any given period are outlined in Section II, but for the present it will suffice to indicate that the available budget margin is a hypothetical construct—calculated by multiplying the trend rate of growth of real output by the income elasticity of the revenue system—that determines the size of the permissible increase in government expenditure.1 In contrast to budgetary philosophy in most countries, it is the hypothetical growth of revenue at the trend level of output and not the actual anticipated growth of revenue that guides expenditure planning in a given period. The actual increase in revenue in any year may exceed, or be less than, the estimate of the budget margin, depending on the stage of cyclical fluctuation in the level of economic activity, but these fluctuations in revenue are irrelevant to expenditure planning. In periods when the actual growth of revenue exceeds the calculated budget margin, the additional revenue is to be used to reduce the actual cash deficit and not to increase the rate of growth of government expenditure, and, conversely, government expenditure is not to be reduced when the actual growth in revenues is less than the budget margin.
Initially, the budget margin was conceived as constraining the rate of growth of government expenditure within the range of the long-term rate of growth of output, the scope available for increasing the relative importance of government vis-à-vis private expenditure being limited to the portion of the margin that is the consequence of a faster automatic growth in revenue than that in real output. Theoretically, this additional revenue resulting from the fact that the income elasticity of total tax revenue exceeds 1 (in Netherlands terminology, the progression factor) could be applied either to increase the relative size of the government sector or to offset the increase in the real burden of taxation resulting from a progressive tax system. In this context, for the purposes of the budget margin testing procedures, a reduction in the tax burden is viewed as equivalent to an increase in expenditure, since both reflect discretionary expenditure decisions of the Government.2 In the latter part of the 1960s the constraint on the growth of expenditure that could be determined by the available margin was relaxed, in that increases in expenditure that exceeded the margin would be permitted if they were financed by discretionary tax increases.3 The thrust of this modified rule is to maintain a structural budget deficit of constant absolute size as the controlling variable in the rate of growth of government expenditure.
This general description is subject to minor modifications in the procedures followed in practice, but the basic budget philosophy is to ensure the relatively steady growth of government expenditure at a rate determined from the projected long-term rate of national income in real terms. However, in periods of inflation, major complications arise in applying this principle both because the size of the budget margin determined in real terms will fluctuate in money terms with the rate of inflation and because inflation has been associated with an increase in the relative cost of providing government services.4
Thus, the definition of what constitutes an increase in government expenditure in real terms becomes complicated. An elaborate system of earmarking has been evolved in practice for the operation of the budget margin procedures in periods of inflation and it is described in the subsection, Adjustment for price increases.5 For present purposes, however, it is relevant that the Netherlands Government considers that, in a progressive tax system, any increase in the real burden of taxation (indicated by the increase in the ratio of tax to income) resulting from inflation should not be allowed to increase the available budget margin but should be offset by discretionary tax reductions. Thus, even when account is taken of the existence of inflation, the budget margin is still determined basically by the automatic growth in revenue that is expected to take place when the economy is moving along its long-term growth path.
The budget as an instrument of fiscal policy
While the emphasis of the budget margin principle is on the balanced growth of government expenditure, it also serves the important function of aiding the balanced growth of real output along its trend path. Professor Stevers, consultant in public finance to the Netherlands Central Planning Bureau, has described this aspect of the budget margin in the following terms:
The equilibrated growth [of the economy] is furthered by an expansion of government finances in accordance with the conditions set by the budget margin, because then the impact of the budget in the national economy is approximately constant. This constancy can be shown by assuming a proportional increase of all government income and outlays. Then, in a demand model, government expenditure as an exogenous variable, remains constant in proportion to national income and the marginal tax quote [sic], as a component of the multiplier does not change either. This constant impact will neither change fundamentally by a change in the composition of taxes and expenditure nor by a shift between both of them, at least if not too large mutations are considered.6
In fact, the objective of the budget margin procedures has been to generate a budget deficit that is constant when the economy is moving along its estimated long-term growth path rather than, as assumed by Professor Stevers, a deficit that is constant as a proportion of output. This hypothetical deficit of constant absolute size is designated here as a structural budget deficit to distinguish it from the actual budget outcome over any given period. In principle, the structural deficit represents a measure of the deficit that is equal at the desired level of full employment to the difference between private savings generated at full employment (including the surplus or deficit in the balance of payments) and the private investment necessary to achieve full employment. That is to say, the structural budget deficit, at the normal level of economic activity that is taken as the basis for the calculation, will be assumed to offset the excess of ex ante private savings over ex ante private investment.7 Thus, the Netherlands procedures are based on the assumption that the movement of the economy along its real growth path is consistent with the maintenance of a structural budget deficit of constant amount over time. This assumption implies an overwhelming reliance on automatic stabilizers in the framing of fiscal policies.
This reliance on automatic stabilizers has been reinforced since 1968. Beginning in that year, items of expenditure related directly to changes in the level of economic activity have been estimated on the basis of normal rather than actual unemployment expected in the coming year.8 Since that date, if unemployment is greater than the level assumed to be normal, the higher expenditure on this count is not tested against the budget margin; thus, the actual budget deficit will exceed the structural deficit by a larger amount than would be true if only actual increases in expenditure were tested against the margin. The resulting increase in the size of the actual deficit in a period of less than full (or normal) employment is an automatic stabilizer, as is the reduction in the size of the actual deficit in the converse situation of overfull employment.
Discretionary variations in the level of expenditure and revenues for purposes of stabilization policy are permitted under the budget margin procedures only if the Government chooses, under a principle first applied in 1968, not to test against the margin (a) certain items of expenditure and (b) changes in tax revenue introduced specifically for purposes of fiscal policy.9 The fact that the original budget margin principles were modified in this way to permit such discretionary changes in budgetary policy implies that the given structural deficit is not entirely adequate for the immediate purposes of short-run stabilization policy. Prior to 1971, only relatively small annual amounts of expenditure (about f. 150 million, approximately 0.1 per cent of gross national product) were exempt from tests against the budget margin for fiscal policy reasons, but more recently tax increases using the regulator have also been excluded from the available margin.10
II. Procedure for Calculating the Budget Margin in Practice
The two components involved in calculating the size of the available budget margin in any given period are (1) the automatic increment in tax revenue that would occur if real output were expanding along the assumed structural growth path and (2) the actual expected increase in nontax revenue in the year under consideration. The automatic increase in tax revenue is calculated using an assumed rate of growth of output measured in real terms determined from past and anticipated experiences in the Netherlands and applying an estimate of the built-in elasticity of the tax system. Over the period 1968 to 1971, for example, an average rate of growth of national income in real terms of 4.8 per cent and (using 1966 as the base period) an income elasticity of 1.25 were applied in calculating the size of the budget margin.11 The resulting automatic increase in tax revenue is 6 per cent per annum, namely, 1.25 times 4.8 per cent per annum.
The assumed rate of growth of real output (4.8 per cent) was chosen in the following way:
In the 1950s and the 1960s the average growth of the national income in real terms was about 4.8%, i.e. the same as that predicted by the Central Planning Office for the 1965-1970 period. Accordingly, he [the Minister of Finance] regards it as reasonable to base his calculations of the budgetary margin on the same figure.12
Determining the appropriate growth rate of real output, while difficult, does not appear to be as complicated as choosing the base year for calculating the size of the margin. As mentioned before, the objective of budgetary policy in the Netherlands is to maintain a structural budget deficit equal in size to the deficit that existed in a base year when the demand for private savings from both the public and private sectors equaled the supply at a normal level of economic activity. In this context, the justification for the choice of 1966 as a base period highlights the difficulties inherent in such a decision.
The tax receipts in 1966 were taken as the basis for calculating the budget margin for the years 1968 to 1971. It is true that 1966 was not a year of balanced growth (inflationary strains were apparent in autumn and were followed by a rapid rise in unemployment) but the total tax receipts in that year were more or less of the order of magnitude that would have been reached if the economy had been balanced.13
Since 1961, three different base periods have been used for calculating the available margin: 1960 for the period 1961–63; 1963 for the period 1964–67; and 1966 for 1968–71. A lower rate of growth of output in real terms (4 per cent) was assumed for the two earlier periods, but the effect of this lower rate on the size of the available margin was partially offset by using a higher income elasticity of the tax system (1⅓ per cent) over this period compared with the figure (1¼ per cent) that is used currently. This reduction in the size of the income elasticity of the tax system reflected primarily the consistent decrease in the share of nonwage income in total income over the relevant period and the lower income elasticity of revenue from wage income.14
The procedure followed concerning the nontax revenue to be included in the budget margin in any year in effect includes the estimated increase in revenue from that source. For the planning period 1968–71, a structural growth of nontax revenue was estimated at f. 150 million, and this amount was automatically included in the budget margin over this period. However, actual increases in excess of this estimate, along with discretionary increases in tax revenue, were included separately as items increasing the size of the available margin in the relevant period. In the budget memorandum for 1970, the following view on the importance of nontax revenue was expressed:
As is evident from [the table], the extra growth of non-tax revenue (i.e. the growth in excess of the structural growth of f150 million already included in the budget margin) enhances very considerably indeed the growth potentialities on the expenditure side.15
For example, in the three years 1968, 1969, and 1970, the budget estimates of the increase in nontax revenue were f. 365 million, f. 455 million, and f. 380 million, respectively, compared with an average structural margin of f. 1,100 million per annum excluding f. 150 million of nontax revenue over this period.16
In calculating the automatic growth in tax revenue to be included in the margin, a four-year average was used over the planning period 1968–71 rather than margins of increasing size in successive years. This is a conscious decision of the Government justified in the following way:
Tax receipts will increase automatically by roughly f4,400 million between 1968 and 1971, i.e. an average of f1,100 million per year…. The budgetary margin, particularly that for 1971, is reduced by taking a 4-year average for calculating the expansion of tax revenue. This reduction of the budgetary margin would seem warranted in view of the policy changes in regard to expenditure that are being studied; some of them have already been put into effect and their impact may be expected to become increasingly noticeable in the years ahead.17
The conservatism inherent in this procedure is highlighted by the further comment that
the Minister [of Finance] is quite aware that … the calculation [of the budget margin] is based on a number of simplified assumptions. But he believes that, by not calculating the margin on the basis of the highest possible growth rate and progression factor and by assuming that the budgetary deficit will remain constant at a nominal figure (not a relative figure, in other words, a constant budgetary deficit in relation to the national income) in the longer term, there is no risk of obtaining results that may later prove to have been too optimistic.18
Adjustment for price increases
The budget margin derived in the manner just outlined is defined in real terms at the price level during the base period, and an adjustment for actual price increases over the relevant period must be made to express the margin in terms of the current price level. In the planning period 1968-71, calculations of the budget margin were based on the 1967 price level; in order to restate the margin in terms of the current price level, the budget margin available in 1970 was increased by f. 110 million and that in 1971 was increased by f. 160 million.19 However, no adjustments for price increases were made in 1969, even though the price level had increased by approximately 3 per cent in 1968.
Such technical adjustments for actual price increases over the planning period solve only a relatively minor problem that arises in periods of inflation. The more significant problems are those that arise because the cost of providing goods and services, including general increases in wages and salaries, increases in periods of inflation at a faster rate than the additional revenues resulting from inflation that are considered by the Government to be available to finance the increased cost of providing government services. In this context, the Government has decided that any revenue resulting from an increase in the ratio of taxation to income that occurs because of inflation should be offset by reductions in tax rates, and should not be used to finance increases in government expenditure. Tax reductions introduced for this reason are exempt from the budget margin procedures that normally treat discretionary tax reductions as being equivalent to increases in relevant expenditure in the sense of reducing the size of the available budget margin in the relevant period. Thus, the increase in tax revenue that is treated as being available to finance increases in the cost of providing government goods and services is proportional to the increase in money income resulting from inflation, and following the Netherlands terminology it is defined here as the proportional increase in actual tax revenue.
In practice, special procedures have been adopted to meet these problems. As shown in Table 1, in recent years the proportional increase in tax revenue as just defined has approximated the general increases in wages and salaries that have occurred over the same period. Reflecting this situation, the Government has adopted the convention of exempting general increases in wages and salaries from the budget margin testing procedures while retaining the definition of the budget margin in real terms except for nontax revenues, which are included in the margin without adjustment for inflation. Increases in the cost of the purchases of goods and other services are accordingly included in expenditure that is tested against the budget margin defined in real terms. This procedure is justified by the Government on the grounds of convenience and really is an arbitrary solution to a most difficult problem, which cannot readily be supported on a priori grounds.20 Provided that the past empirical relationship between general increases in wages and salaries and the proportional increase in revenue resulting from inflation continues to hold in the future, the net effect of the above-mentioned adjustments for inflation is that the size of the budget margin expressed in real terms will vary with the rate of inflation in the prices of goods and other services purchased by the Government.
The Netherlands: General Increases in Wages and Salaries and Proportional Increases in Tax Revenue Owing to Inflation, 1968–70
(In millions of guilders)
The Netherlands: General Increases in Wages and Salaries and Proportional Increases in Tax Revenue Owing to Inflation, 1968–70
(In millions of guilders)
| General Increases in Wages and Salaries (Including Social Welfare) in Central Government and Municipalities |
Proportional Increase in Tax Revenue Due to Inflation |
Difference | ||
|---|---|---|---|---|
| 1968 | 795 | 645 | −150 | |
| 1969 | 1,190 | 1,260 | 70 | |
| 1970 | 710 | 825 | 115 | |
| Total | 2,695 | 2,730 | 35 | |
The Netherlands: General Increases in Wages and Salaries and Proportional Increases in Tax Revenue Owing to Inflation, 1968–70
(In millions of guilders)
| General Increases in Wages and Salaries (Including Social Welfare) in Central Government and Municipalities |
Proportional Increase in Tax Revenue Due to Inflation |
Difference | ||
|---|---|---|---|---|
| 1968 | 795 | 645 | −150 | |
| 1969 | 1,190 | 1,260 | 70 | |
| 1970 | 710 | 825 | 115 | |
| Total | 2,695 | 2,730 | 35 | |
The schema above is summarized in Chart 1, which was provided in the abridged budget memorandum for 1971.21 However, it must be understood that the system of “earmarking” described in this chart is a conceptual one and not a strict pattern of earmarking as it is traditionally defined. In particular, there is no direct correspondence in any given time period between the nominal growth in revenue owing to inflation and the general increases in wages and salaries in the same period. Indeed, the size of general increases in wages and salaries in the government sector in any period is determined primarily by the movement of wages and salaries in the private sector of the economy and not by the proportional growth in tax revenue resulting from inflation. The looseness of the Netherlands earmarking system is possible because the revenues (and corresponding expenditure) resulting from inflation shown in the chart are excluded from the budget margin testing procedures, whether they are equal or not.
The Netherlands: Schema of Earmarking in the Budget Margin Analysis
Citation: IMF Staff Papers 1972, 003; 10.5089/9781451969290.024.A005
Source: The Netherlands Budget Memorandum, 1971 (Abridged), p. 20.Nonrelevant items of expenditure
Apart from general increases in wages and salaries and discretionary cuts in tax rates to offset increases in the real burden of taxation resulting from inflation, certain other items of expenditure are not tested against the budget margin. These include (a) expenditure that hardly affects the level of aggregate demand, (b) expenditure resulting from virement, (c) expenditure and corresponding revenue disregarded for the sake of simplicity, (d) sporadic participations in the share issues of partly owned state enterprises, and (e) expenditure and changes in tax revenue undertaken for fiscal policy reasons.22
Exclusions in classifications (a) and (c) need little elaboration. Expenditure in category (a) includes transactions with the International Monetary Fund (IMF) and debt redemptions, and these transactions clearly have no direct impact on aggregate demand. The major examples in category (b) are changes in the responsibility for expenditure between the central government and the municipalities associated with corresponding changes in the share of tax revenues of the respective governments, while certain transactions associated with the introduction of the value-added tax in the Netherlands are the most important example of excluded items in category (c). Investment in the share issues of partly owned state enterprises will have a significant impact on the level of aggregate demand unless offset by decreases in the level of investment in the private sector of the economy. The Government’s justification for their exclusion is that such expenditure occurs sporadically and is most appropriately financed through borrowing on the capital market.23 In these transactions, the Government sees its role as effectively that of a financial intermediary and considers that the funds invested in government securities and used for investment in the relevant state enterprises reduce the supply of funds available for private investment, ceteris paribus. Only if the Government were to change its monetary policy so that it could more easily invest without bringing about a reduction in private expenditure could this view be questioned as being inappropriate.
The exclusion from the testing procedures of expenditure and discretionary tax measures undertaken expressly for fiscal policy reasons represents a significant modification of the original budget margin principles, since such measures do directly affect the level of aggregate demand. In the budgets for 1968-71, only relatively small amounts of expenditure were excluded from the budget margin test for fiscal policy reasons. But in 1971 a fiscal regulator allowing the Minister of Finance to raise or decrease the five main taxes by up to 5 per cent was introduced, and revenues raised in this way have been excluded from the discretionary increases in taxation in the budget margin test. Over all, it seems realistic to conclude that this category of nonrelevant items presently serves as a safety valve to help to meet special problems without significantly weakening the constraints on the growth of expenditure inherent in the Netherlands procedure.
Testing expenditure against the margin
All increases in government expenditure other than the aforementioned nonrelevant items are tested against the budget margin available in the relevant period. For this purpose, discretionary changes in taxation other than the reductions in tax rates to offset the increase in the tax burden resulting from inflation are treated as reducing or increasing the amount of relevant expenditure to be tested against the available margin.24 As mentioned previously, this procedure permits expenditure to increase at a rate that is different from the rate of growth of revenue used in calculating the margin, provided that appropriate discretionary tax measures are introduced. One important method used to increase the tax burden in a given year has been to postpone (or cancel) the discretionary tax cuts designed specifically to offset the effects of inflation that do not normally enter into the budget margin calculations. In any event, discretionary tax increases have played a relatively important role in ensuring that expenditure did not exceed the permissible margin by significant amounts. That this has been true in the planning period 1968–71 is evident from Table 2, adapted from the budget memorandum for 1971. This table illuminates the planning of budgetary policy over the relevant period; to aid interpretation it may be mentioned that the Netherlands Government places considerable emphasis on ensuring that expenditure increases in excess of the margin in one year are eliminated against the budget margin of following years.
The Netherlands: Tax Measures and Expenditure in Excess of the Budget Margin, According to Budget Estimates, 1968–71
(In millions of guilders)
The Netherlands: Tax Measures and Expenditure in Excess of the Budget Margin, According to Budget Estimates, 1968–71
(In millions of guilders)
| Expenditure in Excess of the Budget Margin |
Actual Tax Measures (Increase on Balance) |
Expenditure Exceeding Budget Margin or Elimination of Previous Deficit (−) |
||
|---|---|---|---|---|
| 1968 | 628 | 505 | 123 | |
| 1969 | 49 | 88 | −39 | |
| 1970 | 263 | 305 | −42 | |
| 1971 | 226 | 308 | −82 | |
| Total | 1,166 | 1,206 | −40 | |
The Netherlands: Tax Measures and Expenditure in Excess of the Budget Margin, According to Budget Estimates, 1968–71
(In millions of guilders)
| Expenditure in Excess of the Budget Margin |
Actual Tax Measures (Increase on Balance) |
Expenditure Exceeding Budget Margin or Elimination of Previous Deficit (−) |
||
|---|---|---|---|---|
| 1968 | 628 | 505 | 123 | |
| 1969 | 49 | 88 | −39 | |
| 1970 | 263 | 305 | −42 | |
| 1971 | 226 | 308 | −82 | |
| Total | 1,166 | 1,206 | −40 | |
Apart from 1968, the major demand for outlay in excess of the available budget margin was associated with the spending by the municipalities in a period when the most severe pressures for increases in expenditure were experienced by these lower-level governments. Discretionary tax increases occurred in each of the four years covered in the table, and these combined with the built-in growth in revenue resulted in a steady expansion in the share of both tax revenue and government expenditure in total income over this period. For example, total tax revenue as a percentage of net national product at market prices increased from 27.5 per cent in 1968 to an estimated 28.8 per cent in 1971.25
III. The Budget Impulse Analysis
The Ministry of Finance uses the budget margin analysis to determine the levels of expenditure and rates of taxation and then provides in the budget document an “impulse analysis” for the purpose of evaluating the impact of the budget on the economy.26 This analysis is designed to determine the net stimulative effect (that is to say, the impulse) of the actual budget on the level of aggregate demand. An evaluation is then made as to whether this stimulus is appropriate in the existing economic circumstances. While the basic procedures of measuring the budget impulse are firmly established, the interpretation of the significance of this measure—especially whether the most relevant comparison is of the growth in output in real terms or in money terms—has varied from time to time.
Underlying the application of the impulse analysis is the desire to distinguish between the causes and consequences of cyclical fluctuations in the level of output. To this end, the procedures adopted attempt to separate budget changes into automatic and discretionary components, and the discretionary component is identified as the stimulative component of the budgetary action. For this purpose, the Ministry of Finance defines as discretionary measures all changes in relevant expenditure and the increase in tax revenue in excess of the automatic increase that would occur if the income elasticity of the tax system were equal to unity. The procedure of the Ministry of Finance reflects the view that if, in the absence of changes in effective tax rates, revenues increase at a faster rate than income it can be inferred that the Government has made a discretionary decision not to take action to maintain tax revenue as a constant proportion of income.27 Prior to 1971, the Ministry of Finance did not include increases in nontax revenue as part of the discretionary increase in revenue, preferring instead to make specific allowance for increases in nontax revenue in evaluating the significance of the budget impulse in the relevant period. However, in the budget memorandum for 1971, any increase in nontax revenue that results in an increase in the ratio of nontax revenue to income was treated as a discretionary measure.28
As in the budget margin procedures, increases in only relevant expenditure are included in the impulse analysis, but the definition of such expenditure differs in the two procedures. The major difference is that general increases in wages and salaries are included as relevant expenditure in the impulse analysis, although some qualitative allowance is made for this factor in evaluating the significance of the net budget stimulus. As a general principle, the Finance Ministry’s definition of relevant expenditure for the impulse analysis excludes only the expenditure, such as debt redemptions and transactions with the IMF, that has little or no impact on the level of aggregate demand.29 In this particular context, direct government expenditure abroad is included in relevant expenditure on the grounds that such expenditure affects the balance of payments, if not internal aggregate demand.
The net budget impulse is then derived as the difference between the increases in relevant expenditure and increases in revenues as defined above. That is to say, the budget impulse is defined as the difference between the sum of (a) the increase in tax revenue resulting from direct government action, (b) the increase in tax revenue in excess of what would accrue if the income elasticity were equal to unity, and (c) the actual increase in nontax revenue, and the increase in relevant expenditure in the given period. Under the procedures of the Ministry of Finance, the net impulse expressed as a percentage of the preceding year’s relevant expenditure is designated as the net stimulus to the economy, and an interpretation of the significance of this measure of the net stimulus is then undertaken by direct comparison with the trend rate of growth of income and the actual growth in national income expressed in both real and nominal terms.30
Only general observations of the fiscal impact of the budget using the net stimulus concept are presented in the annual budget documents, and the impulse analysis is presented as a supplement to the budget margin analysis, which embodies the underlying philosophy of budget revenue and expenditure. Normally, the size of the net stimulus in a given year is compared with that in preceding years as well as with both the assumed structural and the actual increase in national income anticipated over the relevant period. The comparison with the stimulus in preceding years is a direct one, which is evaluated in the context of the current level of aggregate demand, whereas the comparison with the growth in national income is complicated in periods of inflation by the problems of ascertaining whether the most relevant comparison is with the growth in income expressed in real terms or in money terms. In periods of inflation, the choice facing the Government is whether to compare the net stimulus with the increase in national income expressed in real terms, or to make some allowance for the actual or anticipated rate of inflation.
In practice, the comparison is made with the increase in national income in both real and money terms, not least of all because the comparison of the net stimulus with either the estimated actual or the structural increase in real national income in periods of inflation would indicate a more expansionary budgetary policy than comparisons in money terms. In the comparison with the growth in income in real terms, specific reference is made to the extent to which increases in government wages and salaries affect the size of the net stimulus. This procedure effectively recognizes that any comparison solely with the growth in income in real terms is a harsh one, while at the same time it accords with the special treatment of general increases in wages and salaries in the budget margin procedures. Since the manner of evaluating the net stimulus concept is essentially a qualitative one, it is illuminating to reproduce both the calculation and the evaluation of the net stimulus concept in the abridged budget memorandum for 1971.31 The method of calculating the net stimulus for the years 1969 to 1971 is shown in Table 3.
The Netherlands: Calculation of the Net Stimulus, 1969–71
Increase in revenue resulting from the income elasticity of tax revenue exceeding 1.
The Netherlands: Calculation of the Net Stimulus, 1969–71
| 1969 | 1970 | 1971 | 1969 | 1970 | 1971 | |||
|---|---|---|---|---|---|---|---|---|
| Millions of guilders | As percentage of relevant expenditure in preceding year | |||||||
| Increase in relevant expenditure | 2,920 | 2,785 | 2,715 | 12.7 | 10.7 | 9.4 | ||
| Less discretionary increase in taxes | 655 | 590 | 665 | 2.8 | 2.3 | 2.3 | ||
| Of which | ||||||||
| Effect of change in tax rates | 625 | −75 | 25 | 2.7 | −0.3 | 0.1 | ||
| Higher revenue owing to progression 1 | 30 | 665 | 640 | 0.1 | 2.6 | 2.2 | ||
| Less increase in nontax revenue | 180 | 165 | 350 | 0.8 | 0.6 | 1.2 | ||
| Net impulse stimulus | 2,085 | 2,030 | 1,700 | 9.1 | 7.8 | 5.9 | ||
| Percentage rise of national income at market prices | ||||||||
| In current prices | 11.1 | 10.7 | 11.4 | |||||
| In constant prices | 4.9 | 5.6 | 4.3 | |||||
Increase in revenue resulting from the income elasticity of tax revenue exceeding 1.
The Netherlands: Calculation of the Net Stimulus, 1969–71
| 1969 | 1970 | 1971 | 1969 | 1970 | 1971 | |||
|---|---|---|---|---|---|---|---|---|
| Millions of guilders | As percentage of relevant expenditure in preceding year | |||||||
| Increase in relevant expenditure | 2,920 | 2,785 | 2,715 | 12.7 | 10.7 | 9.4 | ||
| Less discretionary increase in taxes | 655 | 590 | 665 | 2.8 | 2.3 | 2.3 | ||
| Of which | ||||||||
| Effect of change in tax rates | 625 | −75 | 25 | 2.7 | −0.3 | 0.1 | ||
| Higher revenue owing to progression 1 | 30 | 665 | 640 | 0.1 | 2.6 | 2.2 | ||
| Less increase in nontax revenue | 180 | 165 | 350 | 0.8 | 0.6 | 1.2 | ||
| Net impulse stimulus | 2,085 | 2,030 | 1,700 | 9.1 | 7.8 | 5.9 | ||
| Percentage rise of national income at market prices | ||||||||
| In current prices | 11.1 | 10.7 | 11.4 | |||||
| In constant prices | 4.9 | 5.6 | 4.3 | |||||
Increase in revenue resulting from the income elasticity of tax revenue exceeding 1.
The significance of the measures of net stimulus derived in the above way is evaluated as follows:
The stimuli in 1969 and 1970 calculated in this manner work out at roughly 9% and almost 8% respectively, i.e. in both cases at appreciably more than the structural real increase in national income of just under 5%, which, viewed in the light of economic stresses, is on the high side. However, we must not lose sight of the fact that the rise in wages and salaries has contributed very greatly to the growth in expenditure, accounting as it does for 4.2% and 3% of the calculated figures of 12.7% and 10.7% respectively. This nominal aspect is not included in the yardstick used for the comparison, viz. the real long-term growth in national income. Consequently, the stimulus assumes more favourable proportions when compared with the nominal growth of the national income (11.1% in 1969 and 10.7% in 1970), notwithstanding the fact that in 1970 the actual real growth exceeds the structural growth. Part of the extra stimulus in 1970 compared with that anticipated in the budget memorandum is also ascribable to the larger increase in wages and salaries.
Calculated on the basis of the draft budget, i.e. disregarding the additional economic measures, the stimulus in 1971 is less than that in the two preceding years. The rise in expenditure is smaller, although only slightly so. The increase again includes a substantial rise in wages and salaries, viz. f1,000 million or over 3½%. Moreover, a relatively large proportion of the increase in expenditure in 1971 is covered within the structural budgetary policy, being compensated for by the growth of non-tax revenue. This is also reflected in the greater negative effect of these resources [i.e., nontax revenues] compared with the two preceding years.32
From this argument, it is apparent that while no single standard is used to evaluate the fiscal impact of the budget, the measure of net stimulus is used to pass judgment on the expansiveness of budgetary policy. Further, officials of the Ministry of Finance have stated that, in a cyclical context, if the complication of price and wage increases is ignored budgetary policy can be considered neutral when the net stimulus equals the structural long-term rate of growth of national income, and that judgments on the correct direction of budgetary policy in the context of any given economic forecasts can be made by comparing the size of the net stimulus with the structural growth of national income.
The budget for 1972 contains a similar statement that the trend rate of growth of income is to be preferred to the actual growth as the yardstick for defining a “neutral” growth in the level of expenditure.33 Any increase in expenditure in excess of the structural growth rate of income is considered neutral only if it is offset by discretionary increases in taxation. While the actual rate of growth of revenue is used in practice to calculate the discretionary tax increase, it is stated that the use of the structural rate of growth of income in this calculation would be justified.34 If the revenue growth resulting from the trend rate of growth of income were so used in the impulse analysis, it would imply that a budget would be neutral if the rate of growth of expenditure (adjusted for discretionary changes in taxation) equaled the trend rate of growth of income and the implied rate of automatic growth of revenues. That is to say, the budget would be neutral if the structural budget deficit remained as a constant proportion of potential national income. This essentially is the link with the philosophy underlying the budget margin procedures.35
IV. Relationship Between the Two Analyses in Practice
Since the Ministry of Finance uses both the impulse analysis and the budget margin analysis, it is interesting to examine the relationship between the two analyses. It is readily apparent that the two are designed for essentially different purposes: the budget margin analysis guides the long-run growth of government expenditure, while the impulse analysis serves as an indicator of fiscal impact in a cyclical context. Thus, the impulse analysis considers the actual revenue growth expected over the budgetary period and not, as in the budget margin analysis, the revenues that would accrue if the economy were moving along its long-term structural growth path. As already described, the definitions of relevant expenditure in the two analyses also vary, but, with the exception of certain expenditure excluded from the budget margin analysis for fiscal policy reasons, these differences can be explained by the particular method chosen to deal with the problems posed by inflation in the budget margin analysis.
In the two analyses, the comparisons with the budgets of preceding years are made from slightly different viewpoints. The impulse analysis considers only the actual or probable outcome of the budget in the preceding period, whereas the budget margin analysis takes the original estimates of the preceding year as the starting point. This difference is introduced because the impulse analysis is concerned with the effect on the economy of the actual budget in the preceding period, while the budget margin analysis requires that departures from the available margin in previous periods be rectified in following periods.36
If it is valid to assume that in any chosen period general increases in wages and salaries approximate the proportional increase in tax revenue resulting from inflation, then differences in the definition of relevant expenditure in the two procedures can reasonably be ignored on the grounds that an equivalent amount of revenues and expenditure are excluded from the budget margin analysis.37 On the basis of this assumption, the impulse analysis differs from the budget margin analysis in that the actual growth of tax revenue is used in the former while the hypothetical trend growth of revenue is used in the latter. Beginning with the 1971 budget, nontax revenues have been treated in the same way in the two analyses. Accordingly, revenues used in the budget margin analysis will differ from those used in the impulse analysis—both if the actual growth of income in real terms differs from the assumed long-term trend rate of growth of income and if the actual income elasticity of the tax system varies over the business cycle from its long-term trend value used in the budget margin calculations.38
V. Stabilization Policy Implicit in the Budget Margin Concept
The stabilization policy rules implicit in the budget margin analysis are as follows:
(1) The additional revenues available to finance the increase in expenditure or to reduce the burden of taxation in any given budgetary period are determined by the structural trend rate of growth of the economy and not by the actual increase in revenue estimated in the period.
(2) Any increase in expenditure greater than the structural increase in tax revenue must be financed either by discretionary tax increases or increased nontax revenues, in order to maintain an unchanged structural budget deficit over time.
(3) Certain items defined to be nonrelevant, including expenditure and discretionary changes in taxation for the expressed purpose of fiscal policy, are not subject to these rules.
The net effect of the first two rules is to ensure a preponderant reliance on automatic stabilization for government fiscal policy in the Netherlands. Thus, subject to the flexibility of fiscal action that is possible under the third rule, the efficacy of stabilization policy will then depend principally on the suitability of the automatic stabilizers to achieve the desired policy objectives of the Government.
Under the first two principles, these stabilizers operate in the following way. Independently of the level of economic activity, government expenditure (or private expenditure consequent upon discretionary reductions in taxation) will provide a constant net addition to the level of aggregate demand. The actual growth of the economy consequent upon changes in the level of government expenditure and other exogenous variables will determine the size of the increase in tax revenue and the resulting change in the size of the actual cash deficit. Provided that the actual income elasticity of the revenue system is approximately equal to that assumed in the budget margin calculations,39 then a rate of growth of income equal to the structural norm would result in no change in the size of the actual budget deficit. A rate of growth of real income slower than the assumed norm would result in an increase in the size of the actual budget deficit, while a rate of increase faster than the norm would result in a reduction of the size of the actual deficit. While the magnitudes of these changes may be insufficient to return the economy to the desired equilibrium growth path, clearly they will be in the right direction.
The stabilizing features of this system compare favorably with the conventionally assumed textbook case where the stabilizing properties of the system depend solely on the automatic variation in tax revenue and (possibly) in unemployment compensation over the business cycle.40 In this case, the possibility of fiscal drag, that is to say, of revenues rising faster than expenditure, in any business upturn may well result in destabilizing automatic fiscal action, which will slow down the increase in income when income is increasing at less than a desired rate.41 The superiority of the Netherlands system arises from the fact that the planned growth in expenditure effectively neutralizes any fiscal drag, because when growth is at less than the desired trend rate, expenditure will increase more rapidly than revenues and will neutralize any incipient fiscal drag. It is inherent in this system that the Government is constrained by the budget margin principles from taking discretionary fiscal action that may be more appropriate when, for example, excessive aggregate demand, not fiscal drag, is the major stabilization policy problem in any particular period.
That is to say, there is a counterpart of the advantage of the Netherlands system in avoiding the problem of fiscal drag. It is that the planned steady growth of government expenditure at a faster rate than the trend growth path of the economy (or potential output) implies both a relative inflexibility of government expenditure and a continually expanding relative size of the government sector in the economy unless, as was envisaged originally, a significant part of the permissible budget margin is used for discretionary tax reductions. Furthermore, the steady growth of government expenditure implies that the fluctuations in the use of resources by the private sector of the economy are greater than those by the government sector. The exclusion of expenditure or discretionary taxation for the expressed purposes of fiscal policy from the budget margin test also adds a further bias by permitting only increases in taxation, not reductions in the level of government expenditure, in periods of excess demand.42 In sum, the steady increase of government expenditure is thus at the same time a major strength (in avoiding fiscal drag) and a major weakness (in the inability to vary government expenditure in countercyclical fashion) inherent in the budget margin concept. The power granted to the Minister of Finance to raise or lower the five major taxes by up to 5 per cent may, however, imply a lesser reliance on automatic stabilizers in the future.
The problem of rising prices
In the preceding section, the serious problems of procedure for the budget margin concept raised by inflation were ignored. In actual practice, tax revenue included in the budget margin is determined from a hypothetical growth in income expressed in real terms, while nontax revenue is determined from the actual growth anticipated in such revenue in money terms. On the expenditure side, the largest part of the increase in government expenditure owing to inflation—namely, increases in wages and salaries—is not tested against the budget margin, whereas increases in the costs of goods and other services (that is, excluding services of employees) purchased by the Government are so tested. In periods of inflation, it would be clearly preferable for the budget margin test to be undertaken in money terms rather than in real terms while retaining the present elaborate earmarking system for certain components of revenue. This approach would have the advantage of ensuring that all expenditure or reductions in the level of taxation defined in money terms that increase the level of aggregate demand are in fact tested against the budget margin. It would also result in a consistent treatment of increases in tax and nontax revenue, whereas at the present time tax revenues are defined in real terms and nontax revenues in money terms.
However, since the procedure followed in the Netherlands differs from this, it is rewarding to examine the important fiscal policy implications of the present system. Increases in the prices of goods purchased by the Government are tested against the available budget margin defined in real terms, while general increases in wages and salaries are excluded as nonrelevant items of expenditure. In much the same way, increases in tax revenue resulting from inflation are treated as irrelevant for the budget margin calculations, since they are earmarked to finance tax reductions and expenditure on general increases in wages and salaries.43 In the Netherlands, general increases in wages and salaries in the government sector are linked, with a short time lag, to the movement of wages and salaries in the private sector of the economy. Provided that wage increases in the private sector fully reflect past price changes and increases in productivity and that productivity in the government sector increases less rapidly than in the private sector, wage increases in the government sector will result in a faster rate of increase in the cost of providing any given standard of services than the rate of increase in the general price level. An increase in the relative cost of providing government services has, for this reason, been a normal occurrence in the Netherlands. Further, while the Government may influence the prices of the goods that it purchases, it is unlikely that the prices of these goods will change in a manner significantly different from the general price level. Thus, since the general percentage increase in wages and salaries of providing a given standard of services will exceed the rate of inflation in the normal course of events, the general percentage increase in total government expenditure owing to inflation will normally exceed the actual rate of inflation over any given period. This fact, together with the action taken to ensure that the real tax burden is not increased as the result of inflation, ensures that the available proportional revenue growth attributable to inflation will be less than the increase in the costs of providing the given level of goods and services.
Thus, in periods of inflation, some part of the increase in government revenue measured in real terms will be required to meet the additional costs of providing government services. The excess of the percentage increase in general wages and salaries over the rate of inflation and the percentage increase in the average productivity of the provision of government services will determine the extent to which inflation reduces the capacity of the Government to provide additional real expenditure within the budget margin. If the experience of the period 1968-70 is any guide,44 it appears that the budget margin defined in real terms should fully cover the increase in the cost of goods purchased by the Government, because general increases in wages and salaries have approximated the proportional increase in revenue attributable to inflation. This is to imply that the stimulus to the economy measured in real terms varies with the rate of inflation and as such represents a departure from the theoretical underpinnings of the system as outlined by Professor Stevers, wherein the objective of the budget margin procedures is to provide a constant stimulus in real terms to the level of aggregate demand. That Professor Stevers sees the treatment of inflation as a major problem can be readily understood in the light of this view of the rationale underlying the budget margin philosophy.45
A general evaluation of Professor Stevers’s model will not be attempted here, but it is apparent that the choice of a suitable base year with a structural budget deficit equal to the excess of private savings over private investment, including the overall balance of payments, is crucial to the realism of applying such a model in practice. Even then, while discretionary fiscal policy may well be destabilizing because of imperfect knowledge or time lags, it must surely be reasonable to expect that governments can improve their fiscal performance through discretionary action. Indeed, Professor Stevers clearly recognizes that in practice discretionary fiscal action is necessary in certain circumstances.46
VI. The Net Stimulus as a Summary Indicator of Budget Impact
In the impulse (and budget margin) analysis it is assumed that the size of the balanced budget multiplier is zero, or sufficiently small to be negligible, as is true when the actual budget surplus is used as a summary measure of the fiscal impact of the budget. That is to say, no attempt is made in either summary indicator to weight the different components of government revenues and expenditure according to the differing impacts that such components may have on the level of aggregate demand, except perhaps for the specific exclusion of a relatively few items from relevant expenditure in the impulse analysis. Thus, even though expenditure abroad that has no impact on the level of domestic aggregate demand is included in the impulse analysis, this procedure in no way differs from conventional budgetary analysis when weighting procedures are not applied.
Therefore, it is meaningful to examine the usefulness of the impulse concept as a summary indicator of the impact of the budget on the economy in terms of whether or not it is superior to the budget surplus concept when the latter is also used for the same purpose. The important difference between the two indicators is the separation of increases in revenues into automatic and discretionary components under the impulse analysis and the absence of such separation in the analyses based on changes in the size of the budget balance. For the purposes of the present comparison, two alternative definitions of a neutral budget will be taken as a standard: (a) one in which there is no change in the size of the actual budget balance, or (b) one in which there is no change in the size of the budget balance in relationship to national income. In this context, a budget will be considered as expansionary (contractionary) if the size of the deficit in the respective definitions increases (decreases) over the relevant period.
In essence, the difference between the measure of the net stimulus and the measure of the change in the budget balance is that the net stimulus concept considers only those increases in expenditure and revenues that are defined as discretionary, while all increases in expenditure and revenues are included in determining changes in the size of the budget balance. The size of the impulse in absolute terms thus differs from that of the change in the budget balance by the amount of the automatic increase in tax revenue not considered in determining the size of the budget impulse. The chosen definition of automatic revenue growth is the increase in tax revenue that is proportional to the growth in actual income. It follows that the actual rate of growth of income can serve as a proxy for the percentage change in tax revenue that is defined as automatic. Further, the net stimulus expressed as a percentage of relevant expenditure in the preceding period is equivalent to the percentage increase in expenditure after allowing for discretionary increases in taxation. Thus, defining a budget as neutral when the size of the net stimulus (expressed in percentage terms) equals the actual rate of growth of income is equivalent to defining a budget as neutral when the size of the budget balance in relation to national income does not change, because here the automatic growth in revenue expressed in percentage terms will equal the (adjusted) percentage increase in expenditure, implying no change in the size of the budget balance expressed as a percentage of national income.
On the other hand, defining a budget as neutral when the size of the budget balance in absolute terms does not change is not consistent with defining a budget as neutral in which the size of the net stimulus equals the actual rate of growth of income except when the budget was balanced in the initial period. This one special case arises because the automatic increase in tax revenue in this circumstance would equal the increase in expenditure less the tax increase defined as discretionary. But, in situations where the budget was in deficit (surplus) at the beginning of the period, the increase in expenditure would exceed (be less than) the increase in tax revenue, resulting in a change in the size of the budget balance when the net stimulus equals the actual rate of growth of national income. Since the overall budget of the Netherlands is presently in deficit, the normal tendency would be for the size of the actual budget deficit to increase when the net stimulus equals the actual rate of growth of income.
When the size of the net stimulus is compared with other standards, such as the trend rate of growth of income or the actual rate of growth of income expressed in real terms, the analysis cannot be compared directly with one based on changes in the size of the actual budget balance. For, unless the actual rate of growth of income equals, say, the trend rate of growth of income, the automatic percentage increase in revenue not examined in the impulse analysis will be greater (less) than the rate of growth taken as the standard when the actual growth in income is greater (less) than the standard. There is thus no direct relationship between the change in the size of the budget balance and any evaluation of the fiscal impact of the budget on this standard. In comparisons with standards other than the actual growth in money income, no attention is given to the size of the actual increase in revenue that is defined as automatic. A budget that is defined as neutral when the net stimulus equals the assumed trend rate of growth of income may be highly contractionary when evaluated in terms of changes in the size of the budget deficit in the given period. This would, for example, be true when the actual rate of growth of income exceeded the trend rate of growth of income by a significant margin. For purposes of short-term analysis, it must surely be relevant for evaluating the effect of the budget on the economy that a substantial part of the automatic increase in revenue be reflected in a reduction in the size of the budget deficit. Any comparison of the net stimulus with standards other than the actual rate of growth of income, such as the structural rate of growth, must for this reason be interpreted as being more pertinent to ascertaining whether the objective of the budget margin philosophy, interpreted as providing a constant proportionate discretionary net stimulus to the level of aggregate demand independently of fluctuations in that level, is being achieved.
VII. Summary and Conclusion
The Netherlands budgetary philosophy must be considered as unique in the emphasis that is placed on the balanced growth of government expenditure in relation to the anticipated long-term structural growth of the economy. Under this philosophy, the growth in government expenditure is governed by the size of an available budget margin determined in a dynamic context through the growth in revenues that would take place when the economy is moving along its long-term growth path. While the size of the budget margin can be increased through discretionary tax increases, in the past the available margin has been determined principally as the sum of (a) the product of the income elasticity of the tax system and the anticipated trend rate of growth of income and (b) the expected actual increase in nontax revenues over the budget planning period. Except for this nontax revenue component, the emphasis in expenditure planning is placed on the long-term trend rate of growth of revenues, not on the anticipated growth of revenues.
Not all items of government expenditure are subject to the budget margin testing procedure; the exclusions are in three major categories: (1) items that have little or no impact on the level of aggregate demand, (2) items resulting from inflation, which are excluded on the assumption that such expenditure approximates the additional revenue resulting from inflation that is not included in the size of the available margin, and (3) certain items designated for the express purpose of discretionary fiscal action. The exclusions in category (2) represent an essentially arbitrary solution to the difficult problems raised by the existence of inflation for the practical application of the budget margin philosophy. To date, exclusions under category (3) have been relatively small, and the possibility of excluding expenditure and discretionary tax action for fiscal policy purposes can be viewed as an attempt to introduce a modest degree of flexibility of fiscal action into a system that otherwise relies exclusively on automatic stabilizers for fiscal policy purposes.
The impulse analysis, which—as a supplement to the budget margin analysis—concerns itself with cyclical fluctuations in the level of income, represents an attempt to separate government fiscal action into separate automatic and discretionary components for direct use for policy purposes. This separation is achieved mainly by excluding the growth in revenues defined as automatic by the Netherlands Government. The chosen definition differs from the conventional one in that it includes only the growth in revenues resulting from a unitary elastic revenue system. The net budget (discretionary) impulse is derived as the difference between the increase in relevant expenditure and the discretionary increase in revenue. A measure of the net stimulatory effect of the budget is computed by expressing the budget impulse as a percentage of the relevant expenditure of the past year. A comparison of this measure with the rate of growth of actual income (and, hence, implicitly with the automatic rate of growth of revenue) and with the trend rate of growth of income is used to indicate the expansionary or deflationary effect of discretionary budget policy.
The budget margin procedure can be viewed both as a rule governing the rate of growth of the size of the public sector and as a rule determining the proper budgetary stance in the context of the long-term growth of the economy. This budgetary stance entails a heavy reliance on automatic stabilizers for fiscal action, because under the budget margin procedure the size of the actual budget deficit increases when the actual rate of growth of income is less than the trend rate of growth of income, and conversely. In periods of inflation there is a departure from the principles of the budget margin philosophy, because the procedures adopted for dealing with price increases result in a decrease in the size of the budget margin defined in real terms when increases in the costs of goods and services (other than the salaries and wages of government employees resulting from inflation) are tested against the margin. In the impulse analysis, it is assumed that the size of the balanced budget multiplier is zero, as is true when the actual budget balance is used as a summary measure of the fiscal impact of the budget. As a consequence of the chosen definition of discretionary increases in taxation that is used in the impulse analysis, defining a budget as neutral when the size of the net stimulus (expressed in percentage terms) equals the actual rate of growth of income is equivalent to defining a budget as neutral when the size of the deficit expressed as a percentage of national income does not change. If the comparison of the net stimulus is with the trend rate of growth of income, there is no relationship between this analysis and one based on the actual budget deficit.
Techniques d’analyse budgétaire aux Pays-Bas
Résumé
Le présent document étudie les implications en matière de politique budgétaire de deux techniques d’analyse financière utilisées aux Pays- Bas à des fins totalement différentes: l’analyse dite de “la marge budgétaire” qui sert à diriger la croissance à long terme des dépenses publiques, et l’analyse dite de “l’impulsion budgétaire” qui joue le rôle d’un indicateur de conjoncture pour la politique budgétaire.
Contrairement à la plupart des autres pays, les Pays-Bas se fondent pour planifier les dépenses sur un accroissement théorique des recettes en partant des tendances de la production et non pas de la croissance projetée des recettes elles-mêmes. La “marge budgétaire” correspond à cette augmentation théorique des recettes et peut, en principe, être utilisée soit pour financer des dépenses publiques supplémentaires, soit pour permettre un allégement discrétionnaire de la fiscalité pendant la période envisagée. Les dépenses ne peuvent dépasser cette marge que si elles sont accompagnées de mesures fiscales discrétionnaires d’un montant de recettes équivalent. Cependant, les dépenses ne sont pas toutes rapportées à la marge budgétaire. Deux exceptions notoires sont les hausses de traitements et salaires dans la fonction publique résultant de l’inflation et les dépenses ou les mesures fiscales discrétionnaires utilisées spécifiquement pour des motifs de politique budgétaire. Les hausses de traitements et salaires de la fonction publique sont examinées à part dans le cadre des problèmes que l’inflation pose à l’analyse de la marge budgétaire, tandis que l’exclusion (depuis 1968) des mesures de politique budgétaire ne laisse qu’une liberté de manoeuvre limitée, dans ce domaine. Les problèmes posés par l’inflation à l’analyse de la marge budgétaire et les conséquences financières d’une croissance soutenue des dépenses publiques indépendamment de la croissance réelle des recettes font l’objet d’une analyse relativement approfondie. La conclusion générale qu’on en tire est que la croissance soutenue des dépenses publiques est à la fois un élément de force qui contribue à éviter le “freinage budgétaire” et un élément de faiblesse en ce sens qu’elle suppose, pour la politique budgétaire des Pays-Bas, une dépendance étroite vis-à-vis de stabilisateurs automatiques. Le pouvoir accordé en 1971 au Ministre des Finances, de majorer ou d’abaisser les cinq principaux impôts dans la limite de 5 pour cent peut, cependant, laisser supposer pour l’avenir une moindre dépendance vis-à-vis de ces stabilisateurs automatiques.
L’analyse de l’impulsion budgétaire a pour objet de mesurer la portée des mesures fiscales discrétionnaires du Gouvernement pendant une période donnée, mais elle adopte pour ces mesures une définition différente de la définition traditionnelle. Aux Pays-Bas, toute augmentation des recettes fiscales excédant le montant qui proviedrait d’un système fiscal d’élasticité égale à 1, est considérée comme le résultat d’une mesure discrétionnaire destinée à accroître la pression fiscale. Ainsi calculée, l’augmentation discrétionnaire des recettes fiscales et non fiscales est déduite de l’accroissement discrétionnaire des dépenses publiques puis exprimée en pourcentage des dépenses de la période précédente. Cette mesure de l’impulsion discrétionnaire communiquée à l’économie par le budget est comparée à l’accroissement survenu au cours des années antérieures ainsi qu’aux taux réels et projetés de la croissance des revenus et sert de base pour évaluer le rôle de la politique budgétaire. Le document compare l’analyse de l’impulsion budgétaire à celle de l’excédent budgétaire réel comme autre indicateur de ce rôle.
Técnicas de análisis fiscal en los Países Bajos
Resumen
En este artículo se examinan las consecuencias que tienen en la política fiscal dos técnicas de análisis fiscal utilizadas en los Países Bajos con fines esencialmente distintos: el análisis del margen presupuestario, que indica el crecimiento del gasto público a largo plazo, y el análisis del impulso, utilizado como indicador de la política fiscal en un contexto cíclico.
A diferencia de la mayoría de los países, los Países Bajos utilizan el crecimiento hipotético de la renta pública al nivel de tendencia de la producción, en vez del aumento previsto en la renta, para orientar la planificación del gasto. El margen presupuestario es este crecimento hipotético calculado de la renta que, en principio, puede utilizarse para financiar un aumento del gasto público o una reducción discrecional de los impuestos durante el período de planificación. Sólo se permite que el gasto sea mayor que el margen si al propio tiempo se modifica discrecionalmente la tributación de modo que se obtenga un aumento equivalente de la renta. Sin embargo no todos los gastos se comparan con el margen presupuestario. Dos importantes excepciones son los aumentos de sueldos y salarios de los funcionarios públicos debidos a la inflación y los gastos o las medidas tributarias discrecionales determinados expresamente por motivos de política fiscal. Los aumentos de sueldos y salarios de los funcionarios públicos se excluyen como solución de los problemas que plantea la inflación para el análisis del margen presupuestario, mientras que la exclusión (a partir de 1968) de las medidas adoptadas con fines de política fiscal dan cierta flexibilidad a la política fiscal discrecional. Se analizan con algún detalle los problemas que supone la inflación para el análisis del margen presupuestario y las consecuencias fiscales de un crecimiento sostenido del gasto público, independientemente del aumento real de la renta. La conclusión general es que en los Países Bajos el crecimiento sostenido del gasto público es a la vez un importante factor positivo para evitar el efecto de contracción ocasionado por un aumento automático de la renta pública no gastado, y un importante factor negativo que exige el principal recurso a estabilizadores automáticos en la política fiscal. Sin embargo, las facultades otorgadas al Ministro de Hacienda en 1971 para aumentar o reducir hasta un 5 por ciento los cinco principales impuestos significan que en el futuro se recurrirá menos a los estabilizadores automáticos.
El análisis del impulso tiene por objeto medir el alcance de las medidas fiscales discrecionales del Gobierno en un período dado, pero se basa en una definición distinta de la tradicional de las medidas tributarias discrecionales. En los Países Bajos, todo aumento de la renta fiscal en exceso del que se obtendría mediante un sistema tributario con elasticidad unitaria se considera resultado de una medida discrecional de política para aumentar la carga fiscal. El aumento discrecional de la renta tributaria y no tributaria así calculado se resta del aumento discrecional en el gasto público, y se expresa en porcentaje del gasto durante el período anterior. Esta medida del impulso discrecional que el presupuesto da a la economía se compara con el aumento registrado en años anteriores y con las tasas observadas y de tendencia del crecimiento del ingreso, a fin de evaluar la situación de la política fiscal. En el artículo se compara el empleo del análisis del impulso con el empleo del superávit real del presupuesto como otro indicador de la situación de la política fiscal.
Mr. Dixon, economist in the Fiscal Analysis Division of the Fiscal Affairs Department, is a graduate of the University of Queensland (Australia) and Cambridge University. He formerly taught at the University of Calgary (Alberta, Canada) and the Australian National University.
Over the budget planning period 1968–71, the budget margin was calculated on the basis of an assumed 6 per cent annual growth in government revenues. This figure is the product of the assumed trend growth of income, 4.8 per cent, and the income elasticity of the revenue system, 1.25 per cent.
In earlier years, considerable emphasis was placed on the need to maintain a constant relative size of the government sector through discretionary reductions in taxation, but the inexorable pressures for increased government expenditure in most years resulted in the use of the full margin to increase the level of expenditure. See the budget memoranda for the years 1964 to 1966, and in particular the description of the origin of the structural standard in The Netherlands Budget Memorandum, 1966 (Abridged), pp. 12–17, issued by the Information Department of the Ministry of Finance and hereinafter referred to as the Budget Memorandum. See also Theo A. Stevers, “Some Aspects of the Impact of Social Developments on Public Finance,” in The Budget Today, Semaine de Bruges, Collège d’Europe, 1967, ed. by Rudolf Regul (Bruges, 1968), pp. 53–54.
Compare the description of the budget margin in the Budget Memorandum, 1970, pp. 80-89, with that provided in the budget for 1966 (see footnote 2).
The increase in the relative cost of providing government services is considered further in the subsections, Adjustment for price increases and The problem of rising prices.
See also the Budget Memorandum, 1971, pp. 29–30, for a description of the earmarking procedure.
Stevers, op. cit., pp. 57–58. See also Joergen Lotz, “Techniques of Measuring the Effects of Fiscal Policy,” in OECD Economic Outlook: Occasional Studies, Organization for Economic Cooperation and Development, July 1971, pp. 21–22.
This excess of private savings over private investment at the assumed normal level of activity is consistent with the monetary policy assumption of the Netherlands Bank that as a rule the deficit should not be financed by short-term borrowing or by recourse to the banking system.
The Budget Memorandum, 1970, p. 82.
Ibid., p. 87.
See the subsection, Testing expenditures against the margin.
The Budget Memorandum, 1970, p. 83.
The Budget Memorandum, 1968, p. 20.
The Budget Memorandum, 1970, p. 83.
The actual income elasticity of the tax system is expected to fall below 1¼ per cent in the 1972 budget, reflecting further shifts in the distribution of income. See the Budget Memorandum, 1972, p. 15.
The Budget Memorandum, 1970, p. 65.
The Budget Memorandum, 1968, pp. 20–21.
Ibid., p. 21.
See the Budget Memorandum, 1971, p. 56.
In the opinion of Professor Stevers, the handling of price increases may have already undermined the very idea of a budget margin conceived as a steady increase in real terms in the level of government expenditure. (See Stevers, op. cit., pp. 61–62.) This point is discussed in the subsection, The problem of rising prices.
The Budget Memorandum, 1971, p. 30.
See the Budget Memorandum, 1970, pp. 84–89.
Ibid., pp. 86 and 87. These items of expenditure are included in a supplementary budget, not in the budget estimates.
For a description of the actual testing procedure, see the Budget Memorandum, 1969, pp. 32–33.
The Budget Memorandum, 1971, p. 51. The figures provided in this source have been adjusted to include revenue from death duties.
This section draws heavily on the description and analysis provided in an unpublished working paper of the Organization for Economic Cooperation and Development. For essentially the same purposes, the Central Planning Bureau utilizes a slightly different measure of the budget impact from that of the Ministry of Finance. The significant differences are pointed out in footnotes 27–38.
The Central Planning Bureau, on the other hand, considers that the fact that tax revenue grows at a faster rate than income is the result of endogenous processes. Accordingly, it defines discretionary tax increases in the conventional manner, including only those that result from changes in tax rates or other alterations in the tax system.
The Budget Memorandum, 1971, pp. 81-83. This procedure now accords with the treatment of nontax revenues by the Central Planning Bureau.
The Central Planning Bureau uses basically the same definition of relevant expenditure as the Ministry of Finance; the main difference is that loans to other sectors of the economy are excluded by the Bureau as not representing expenditure of the government sector.
A somewhat different method, however, is followed by the Central Planning Bureau in evaluating the size of the impact. Their measure of the impulse is expressed as a percentage of the preceding year’s national income rather than as a percentage of the preceding year’s relevant expenditure, as in the Ministry’s procedure. The measure so calculated by the Bureau is used to evaluate the effects of budgetary policies of the national economy by applying its short-term econometric model.
The Budget Memorandum, 1971, pp. 82–83.
Ibid.
The Budget Memorandum, 1972, Annex 4, p. 71.
Ibid.
However, the budget margin procedures as applied in practice are designed to achieve a structural budget deficit that is constant in absolute size, not in relative terms. See the subsection, The budget as an instrument of fiscal policy.
See the subsection, Testing expenditure against the margin.
It is here assumed that expenditure excluded from the budget margin analysis for fiscal policy reasons is a relatively small proportion of total expenditure. To date, this has been a realistic assumption.
It is quite possible that the income elasticity of a tax system does fluctuate over the business cycle and is susceptible to secular trends. This possibility complicates any comparison of a system based on a hypothetical growth of revenues, such as the budget margin (or the full-employment budget surplus), with one based on the actual growth in revenues over the period. The argument just presented will be more or less relevant in practice, depending on the extent of the variation in the income elasticity of the revenue system.
This requirement is stated in this form for expositional convenience. If in actual practice the income elasticity of the tax system varies in a cyclical fashion, the manner in which it so varies is of crucial importance to evaluating the efficacy of automatic stabilization in the Netherlands system. To take an unlikely case, if the income elasticity of the system declined as the rate of growth of income increased and conversely increased when the growth rate of income declined, the automatic stabilizing properties of the system discussed here would be reduced correspondingly.
In fact, the Netherlands procedures as outlined previously do encompass changes in unemployment benefits within the automatic stabilizers, because changes in such expenditure from an assumed normal level of expenditure are not tested against the budget margin. See pages 618–19.
Walter Heller defines fiscal drag in the following terms: “Although ‘fiscal drag’ is often used generically to describe the retarding effect of any full-employment surplus, its use in the narrower technical sense refers to the automatic growth in potential revenues arising out of the growth in potential GNP,” Walter W. Heller, New Dimensions of Political Economy (Harvard University Press, 1966), p. 181. The term is used in the less technical sense here, but even under Heller’s technical definition, the Netherlands procedures do eliminate any possibility of fiscal drag.
See the subsection, Nonrelevant items of expenditure. It is relevant, however, that the possibility for introducing discretionary increases in taxation that do not increase the size of the budget margin does decrease the reliance on the automatic stabilization policy just described.
However, if tax reductions are not made to neutralize the effects of inflation on the real tax burden, an equivalent amount of revenue enters the budget margin testing procedures as a discretionary increase in taxation available to finance expenditure increases in excess of the available margin.
See the subsection, Adjustment for price increases.
See also the subsection, The budget as an instrument of fiscal policy, and Stevers, op. cit., p. 61.
Professor Stevers emphasizes, however, that frequent introduction of discretionary policies would seriously weaken the foundations of the budget margin concept. See Stevers, op. cit., p. 58.
