Article

The Full Employment Budget Surplus Concept as a Tool of Fiscal Analysis in the United States1

Author(s):
International Monetary Fund. Research Dept.
Published Date:
January 1973
Share
  • ShareShare
Show Summary Details

THE FULL EMPLOYMENT BUDGET SURPLUS CONCEPT, which has been developed almost exclusively in the United States, originates from a proposal made in 1947 by the Committee for Economic Development that the budget be designed to “yield a moderate surplus at high-employment national income.”2 The concept found favor during the early 1960s, and the Annual Reports of the Council of Economic Advisers for 1964 and 1965 analyzed the budget program in terms of the full employment surplus. For the next few years the concept drew little official attention, but the present Administration has recently made use of the concept. In particular, the President stated in his Budget Message for 1972:

It [the 1972 Budget] adopts the idea of a “full employment budget,” in which spending does not exceed the revenues the economy could generate under the existing tax system at a time of full employment. In this way, the budget is used as a tool to promote orderly economic expansion, but the impact of the resulting actual deficit is in sharp contrast to the inflationary pressure created by the deficits of the late sixties, which were the result of excessive spending that went far beyond full employment revenues. The full employment budget idea is in the nature of a self-fulfilling prophecy: By operating as if we were at full employment, we will help to bring about that full employment.3

The purpose of this paper is to describe the full employment budget surplus concept and to analyze its usefulness in the particular context of its applicability for fiscal analysis in the United States and other countries. It is intended to concentrate on the use of the concept as a summary measure of budget impact for purposes of both budgetary formulation and evaluation, but attention is also paid to possible uses in formulating budgetary rules such as the ones already outlined. A detailed description of the methods actually applied in practice to measure the size of the full employment surplus is to be found in other studies.4 This paper focuses on the definition of the full employment surplus and the various uses of this budget concept. The reader is also referred to other studies for a systematic quantitative discussion of the accuracy of the concept as an indicator of the expansiveness of fiscal policy in the United States.5

Section I describes the concept of the full employment budget surplus and its method of estimation. Section II identifies and discusses two broad categories of use of the concept: (1) its general use as a summary measure of budget impact for purposes of ex ante policy formulation and ex post policy evaluation, and (2) its use to govern the formulation of budgetary policy. The principal emphasis in this paper is on the first of these two uses of the concept, and Section III examines the economic rationale for its use in this context. This section also looks at the merits of the argument that the full employment surplus permits the separation of the effects on the budget surplus of discretionary action from the automatic effects resulting from fluctuations in the level of gross national product (GNP) and presents a specific comparison of the relative merits of the full employment surplus with the actual budget surplus as a summary measure. Section IV examines the stabilization policy implications of the use of the full employment surplus as a rule in budget formation. Section V presents an overall evaluation of the full employment surplus. Problems of measuring the surplus are considered first and then, using the arguments developed in the preceding sections of the paper, the merits of the full employment surplus vis-à-vis the actual surplus are summarized.

I. Definition and Measurement of the Full Employment Budget Surplus

The full employment budget surplus (or deficit) is an estimate of what the budget result would be with a given expenditure and taxation program if the economy were moving along the path of potential full employment output.6 Thus, it is designed to be independent of fluctuations in the level of economic activity (i.e., deviation from the full employment level of output), and hence the automatic effect of any such fluctuations is removed from the budget result. It is clearly a hypothetical construct designed for the purpose of combining estimates of government revenue at a standardized full employment level of output for any given existing (or assumed) structure of taxation with the existing (or projected) government expenditure.7 Only one component of government expenditure—unemployment benefits—is assumed to vary with the level of economic activity, and with this one exception the concept is designed really to isolate from the budget result any fluctuation in the level of revenue resulting from fluctuations in the level of economic activity. If items of expenditure other than unemployment compensation varied systematically with the level of economic activity, it would also be necessary to include such variations in the measurement of the full employment surplus.8 However, to date, automatic fluctuations in the level of other items of government expenditure have not been allowed in calculations of the size of the surplus in the United States, and in the use of the concept, emphasis has been placed almost exclusively on changes in the level of revenue resulting from fluctuations in the level of output.9

The size of the full employment budget surplus at any point in time will change with any of the following variables: the level of government expenditure, the rates at which taxation is levied, and the structure and rates of payment of unemployment compensation, but not with the level of economic activity. An increase in the rates of personal income taxation, for example, would increase the size of the surplus, while any increase in government expenditure, including the rates of unemployment compensation, or reductions in rates of taxation would reduce it.10 Over time, the size of the surplus will also vary with changes in the potential output that correspond to changes in the productive potential of the economy, provided that there are no countervailing changes in policy instruments.

Quantification of the concept11

Estimates of tax revenues and government expenditure expressed in money terms can be derived from the estimate of potential output also expressed in money terms at any given point in time. The full employment surplus then represents the difference between the estimated full employment revenues and full employment expenditure. In the United States, the level of potential output consistent with the preferred definition of the full employment of labor is first calculated in real terms and is then converted into money terms by using the implicit GNP deflator as the relevant price index. The value of the GNP price deflator index at the actual level of income—not a hypothetical value of what this index would be if the economy were at full employment—is used in this exercise. That is to say, the “full employment level of income” used in calculations of the size of the full employment surplus does not represent an estimate of what the actual level of full employment income in money terms would be, except when the economy is actually operating at full employment.

Once the full employment level of GNP in money terms has been determined, assumptions about the functional components of full employment income are made for purposes of estimating the revenue. Standard practice in the United States involves the estimation of the shares of the various components of income, such as personal income and corporate profits, for the purpose of applying average effective rates of taxation to determine full employment revenues. In contrast to the procedure whereby the actual GNP price deflator is used to estimate full employment GNP, the actual shares of income in actual current GNP are not used to estimate income shares in full employment GNP, because the relative shares of the various income components tend to vary in a cyclical fashion. For this reason, an attempt is made to estimate what the relative shares of the various components of income would be at full employment for calculating the full employment tax revenue.12 Errors in estimating the size of the full employment surplus will arise if income shares are estimated incorrectly. This problem is discussed in Section V.

On the expenditure side of the budget, except for unemployment compensation it is assumed that the level of expenditure does not vary automatically with the level of economic activity. Thus, the determination of the level of full employment expenditure requires only the additional step of determining the level of payments of unemployment compensation corresponding to the full employment of labor. The full employment surplus is then calculated as the difference between full employment revenues and expenditure.

The treatment of price changes

In measuring the size of the full employment surplus, it is assumed implicitly that the level of expenditure expressed in money terms does not vary automatically with the rate of inflation, but that the level of full employment revenues does vary because potential output expressed in money terms changes with the level of the GNP price deflator used in its calculation. Thus, the size of the full employment surplus is crucially dependent on the price assumptions implicit in its computation.

As mentioned previously, no attempt is made to estimate what the level of prices would be at full employment in calculating potential output in money terms. Accordingly, unless the economy is actually operating at or close to full employment, the estimate of full employment revenues will not, in fact, indicate what the size of actual revenues expressed in money terms would be if the level of unemployment were to change substantially and to result in a change in the anticipated rate of inflation. However, in the view of Okun and Teeters, the difficulties posed by inflation for measuring full employment budget revenues are mitigated because some government expenditure responds more or less automatically to inflation.13 Any such variation in expenditure would serve to offset increases in the size of the surplus that otherwise would be brought about by a change in the rate of inflation, but by definition the size of the calculated full employment surplus increases with the rate of inflation assumed in its calculation.

This point has important implications for both ex post and ex ante evaluations of any budget that uses this concept and is examined in detail later in this paper. It will suffice here to indicate that the anticipated rate of inflation used will significantly influence the absolute magnitude of the calculated surplus. A similar although distinguishable complication for using the concept in periods of inflation also arises for ex post evaluations of any budget, because in any such computations the actual rate of inflation is used in calculating the size of the full employment surplus, even though that rate of inflation may differ from the full employment rate of inflation and may have been influenced by the budgetary action under examination.

II. The Uses of the Full Employment Budget Surplus Concept

The uses of the full employment budget surplus concept are in two broad categories: the first is its general use as a summary measure of the impact of the budget, whereas the second is its use as a rule designed to govern the formulation of budgetary action by the government. The present U.S. Administration is clearly guided by the concept of a full employment surplus in undertaking budgetary action, as can be witnessed by the following statement of the President:

At times the economic situation permits—even calls for—a budget deficit. There is one basic guideline for the budget, however, which we should never violate: except in emergency conditions, expenditures must never be allowed to outrun the revenues that the tax system would produce at reasonably full employment.14

However, later in 1971 there were discretionary reductions in tax rates, which (principally) resulted in changing the original budget, planned to be in balance at full employment, into one with an anticipated full employment deficit of $8 billion. In the budget for the fiscal year 1973, which is expected to be balanced at full employment, the President restated his continued commitment to the rule just given.

While some examination of the use of the full employment budget concept in the formulation of budgetary policy is clearly warranted, attention in this paper is focused on the use of the full employment budget surplus as a summary measure of fiscal action—not least because such use can be subjected to detailed theoretical examination.15 Detailed considerations of the use of the full employment surplus as a rule for formulating policy involve the examination of political, historical, and other institutional factors that gave rise to the introduction of the rule and are best treated separately. Certain of the theoretical implications of the use of the full employment budget surplus concept, however, will be examined because of their immediate relevance to the nature of fiscal policy if such rules were to be adopted in any given country.

As a summary measure, the full employment surplus concept has been put to a variety of uses, not all of which are readily supportable on theoretical grounds. The major uses of the concept to be considered here rely for theoretical justification on the judgment that as a summary indicator the full employment surplus is superior to the actual budget surplus. This judgment rests on the argument that the full employment surplus concept separates the effects of the budget on the economy from the effects of the economy on the budget at the standardized full employment level of output. As a consequence of this separation, it is then possible to compare different budgets either for ex ante budget planning purposes at one point in time, or for ex post budgetary evaluation at different points in time. One important instance of its use as a summary measure of the overall restrictiveness of the budget was the emphasis that the proponents of the concept placed on the magnitude of the full employment surplus early in the 1960s when the economy had been operating at levels of actual output that were significantly less than potential output.16

III. The Economic Rationale of the Full Employment Budget Surplus Concept

For any budgeted level of government expenditure it is possible to envisage a range of actual budget outcomes depending on the level of GNP and on corresponding revenues achieved in any given period.17 On this basis, for any given budget, a hypothetical schedule relating the size of the budget surplus to the level of GNP can be constructed, and the full employment surplus and the actual surplus will represent two points on this schedule.18 A movement along the function would indicate a change in the size of the budget result consequent upon the change in GNP, while a shift in the curve would indicate a change in budgetary policy. A shift in the function can result from a change in either the level of expenditure (except for unemployment compensation) or in the rates and forms of levying taxation. Any discretionary change in the level of expenditure will shift the function vertically without changing its slope, whereas discretionary changes in taxation may result in a vertical shift of the curve, a change in its slope, or some combination of these two. That is to say, discretionary changes in taxation may affect both the level of revenue collections at any given level of output and the income elasticity of the revenue system, and these would be reflected in the shape of the curve relating the size of the budget surplus to the level of GNP.

In Chart 1, AA is a schedule (assumed to be linear for expositional convenience) relating the size of the actual budget surplus to GNP for a given budget.19 The schedules BB and CC have been derived from AA by assuming discretionary changes in the level of expenditure (BB) and in taxation (CC), and the vertical distances ab and ac measure (at the assumed level of GNP corresponding to full employment (YF) the extent of the discretionary action that caused the movement of the curve AA. The net discretionary changes in revenue and expenditure could, however, also be measured at any other level of income, but when the slope of the function changes, as from AA to CC, the size of the estimate changes with the level of income at which it is measured.20 In the situation depicted in Chart 1, the estimated discretionary increase in tax revenue is larger when it is measured at full employment (i.e., at YF) than at levels of GNP less than YF. Chart 1 depicts the static situation where three alternative budgets are compared at any given point in time, and in this context the measurement of size of discretionary measures would pose few problems other than the estimation of the position of the curves, at least in the relevant ranges of income at which it is desired to measure the discretionary changes.

Chart 1.Comparison of Three Hypothetical Schedules of the Relationship Between the Actual Budget Surplus and the Level of Gross National Product (GNP)

Chart 1 can also be interpreted in a dynamic context by assuming that AA, BB, and CC are schedules corresponding to budgets at three successive periods of time. BB, derived from the second-period budget, would reflect changes in expenditure from the first-period budget (AA), discretionary changes in taxation that do not affect the slope of the schedule, or some combination of these two actions. In the third period, however, discretionary changes in taxation altering the income elasticity of the tax system must have been introduced to change the slope of CC. The size of the full employment surplus in the successive periods will also be influenced by the automatic growth of revenue corresponding to the growth of full employment income over this period. A separation of this automatic growth in revenue is then required if judgment is to be passed on discretionary budgetary policy (however defined) using the full employment surplus. This is in contrast to the situation at any given point in time when automatic changes in the size of the full employment surplus cannot occur.

Nevertheless, a larger full employment surplus will in all circumstances indicate that the budget surplus is larger in one situation than in another, at least in the ranges of income close to full employment. This information is relevant for evaluating the effect of the budget on aggregate demand at these levels of income, but, as will be shown, a higher full employment surplus does not necessarily indicate the relative effect on aggregate demand of these alternative budgets at significantly different levels of income. Chart 2 compares two budgets by showing the relationship between the relevant schedules relating the actual budget result and the level of GNP. It is assumed that the budget corresponding to BB is one that is designed to yield a small surplus at full employment (i.e., at the level of GNP, YF), whereas the budget corresponding to AA would be in considerable surplus at this level of income. The latter budget is clearly more restrictive than the former if the full employment level of income is taken as the reference point, but when the comparison is made at levels of income less than YA in the chart, the opposite conclusion may be reached. The implication of this is that the full employment level of income cannot be used unequivocally to pass judgment on the expansiveness of budgetary policy, even when attention is not confined solely to discretionary policy. This would not preclude the use of the full employment surplus as a guide for current policy at levels of income less than full employment, except when changes in the tax structure have occurred and when it is possible that schedules similar to those depicted in Chart 2 may cross in the relevant ranges of income. The technical conditions under which it is possible for schedules to cross have been described elsewhere.21 For the curves to cross requires that the rates and forms of levying taxation change substantially. Thus, only when significant changes in the structure of tax schedules have taken place would it be necessary to consider whether the evaluation of budgets at full employment yields quantitative results consistent with those that would be derived at levels of income at less than full employment.

Chart 2.Comparison of Two Hypothetical Schedules of the Relationship Between the Actual Budget Surplus and the Level of Gross National Product (GNP) in Which the Schedules Gross Before Full Employment

Comparison with the actual surplus

The consensus of opinion of users of the full employment budget surplus concept would be with Okun and Teeters, who state that the full employment surplus is not and was never meant to be a precise measure of the fiscal impact of the budget for use by the expert.22 In their opinion, the concept can be considered as a simple one-parameter description that is a superior alternative to the actual budget surplus as an indicator of fiscal policy. Its superiority arises, in their view, from the fact that the full employment surplus permits the separation of the effects of the budget on the economy from the effects of the economy on the budget, that is, the full employment surplus permits the separation of the discretionary components of government fiscal action. Even though this opinion is theoretically correct only in the static context when the level of potential output does not change, the relative merits of the full employment surplus and the actual surplus as summary indicators of budgetary policy are considered in some detail here.

That the full employment surplus permits the isolation of discretionary changes in budgetary action in only a static context should be apparent from the preceding discussion. For if the level of potential output changes, the resulting automatic growth in revenue will be reflected in the change of the full employment surplus over time. In the normal dynamic context of an economy when potential output in money terms increases with both the supply of factors of production and inflation, additional information on the growth of revenue associated with this increase in potential output is necessary if the discretionary component of government fiscal action is to be identified. Solomon has proposed a method for separating the automatic components of the change in the full employment surplus,23 but this method is only an approximate one that in any case does not differ in principle from a similar method of determining the extent of the automatic change in the size of the actual budget surplus. Thus, the full employment surplus can be superior to the actual surplus as a summary indicator for the reasons provided by Okun and Teeters only in the following conditions: (1) in a static context when ex ante comparisons are being made of alternative budget strategies, and (2) in a dynamic context in which one is willing to assume that a budget surplus that includes the effect on revenue of the growth over time of the full employment level of income is conceptually superior to one that is affected by automatic fluctuations in revenue resulting from changes in the level of income. Also, as mentioned previously, it is possible that the full employment surplus is an erroneous indicator of the restrictiveness of alternative budgets at the current level of income if significant discretionary changes in taxation have occurred.24 For this reason, the preference of the full employment surplus over the actual surplus requires an additional assumption.

Unless the normal growth in expenditure matches the automatic growth in revenues with unchanged tax rates resulting from the growth in full employment income, the full employment concept in a dynamic context is subject to the same defect as the actual budget result in reflecting the effects of automatic changes in the budget result, owing to changes in the level of income as well as of discretionary government action. Expressing the size of the full employment surplus as a percentage of potential output and the size of the actual budget surplus as a percentage of the actual level of income will provide a more meaningful basis for comparisons over time, but it will not serve to identify whether any growth in the size of the surplus relative to actual or potential output is the result of automatic forces or of policy action by the government.25 There is one situation, however, in which the full employment surplus is the superior indicator of the direction of discretionary fiscal policy even in a dynamic context, viz., when the size of the full employment surplus falls between two successive periods of time. In an economy in which potential output is increasing, and in which the general price level does not fall between the successive periods of time, a reduction in the absolute size of the full employment surplus over time must indicate that discretionary fiscal policy has become less restrictive, because the automatic increase in the size of full employment revenues must always be a positive magnitude. By contrast, a reduction in the size of the actual budget surplus over the same period may reflect either a decline in the level of economic activity or stimulative discretionary fiscal action. Indeed, it is well recognized that stimulative discretionary fiscal action may increase the size of the actual budget surplus over time, but it is conceptually not possible for the full employment surplus to be so affected.

If comparisons of the size of the full employment surpluses relative to potential output are made over time, this conclusion must be modified, because an increase in the size of the full employment surplus in absolute terms is consistent with a fall in the size of the surplus relative to potential output. Thus, a decrease in the size of the full employment surplus relative to potential output does not necessarily indicate a less restrictive discretionary fiscal policy in the later period of time; it may well reflect a revenue system that is relatively income inelastic at full employment. Nevertheless, if, as in the United States, the built-in elasticity of the revenue system at full employment is equal to or greater than unity, a reduction in the size of the full employment surplus relative to potential output will indicate unequivocally a less restrictive discretionary fiscal policy at successive points in time. Naturally, this advantage of the full employment surplus over the actual surplus in indicating the nature of discretionary fiscal policy is asymmetrical and does not apply to situations where the size of the full employment surplus relative to potential output increases.

Both the full employment surplus and the actual budget surplus are affected in much the same way by assumed or actual changes in the price level in any period of time. In particular, the absolute size of both surpluses will be increased for any given level of expenditure (defined in money terms) if the rate of inflation used in calculating the size of the surplus is increased. However, when actual output is less than potential output, the level of expenditure assumed in the calculation of the full employment surplus is equal to or less than the level of expenditure included in the calculation of the actual surplus, and the increase in the absolute size of the full employment surplus will exceed that in the size of the actual surplus for any increase in the assumed or actual rate of inflation used in the budget calculations. If the actual rate of inflation is greater than that assumed in ex ante budget calculations when actual output is less than potential output, the size of both the ex post actual surplus and the ex post full employment surplus will be greater than the ex ante estimates of the size of these surpluses. It then follows that the ex post judgment on the expansiveness of fiscal policy using the full employment surplus will be subject to complications similar to those inherent in using the actual budget surplus for the same purposes when the actual rate of inflation differs from that used in budget planning estimates.

This proposition might suggest that the size of the full employment surplus might be measured at some standardized or invariant rate of inflation rather than at the actual price level and thus avoid the complication posed by inflation for the evaluation of fiscal policy using the actual surplus. The problem of dealing with inflation when using the full employment surplus has been considered in detail by Okun and Teeters,26 and from that analysis it is readily apparent that such a proposition cannot be considered as a practical one in periods of high rates of inflation, because the relationship between actual GNP in money terms and potential GNP at the hypothetical price level would depart increasingly from that between actual and potential GNP at current market prices. The confusing situation where the size of the actual surplus exceeds the full employment surplus at levels of output at less than full employment might, for example, result quickly under such a procedure. In short, the full employment surplus concept does not permit a resolution of the problems posed by the existence of inflation for the evaluation of fiscal policy using the actual surplus.

The question of weighting

At the theoretical level, the choice of the most suitable summary indicator of the budget impact involves a decision on whether the various components of the budget should be weighted to reflect their differing impacts (if any) on the level of aggregate demand. The preceding discussion has been based on the implicit assumption that all components of revenue and expenditure should be given an equal, although opposite in sign, weighting, whereas in practice it may well be more realistic to assign different weights to various components of revenue and expenditure if the first round or total effects on the level of aggregate demand of different budgets are the subject of comparison. Nevertheless, the merits of the use of a weighted full employment surplus vis-à-vis that of the unweighted surplus concept will not be examined here, since the relevant considerations are the same ones that exist for comparing the relative merits of the use of the weighted and unweighted actual budget surplus for fiscal analysis.27 The statements made with respect to the comparison of the actual budget surplus with the full employment surplus are, with one exception, equally valid for comparisons of the weighted full employment surplus with the use of the weighted actual surplus as summary indicators. The exception is that the conditions previously outlined, under which the full employment surplus is superior to the actual surplus as an indicator of fiscal policy in a dynamic context, are no longer valid when weights of different size are applied to separate components of revenue and expenditure.

IV. The Use of the Full Employment Budget Surplus Concept as a Rule for Budgetary Policy

In theory, the full employment budget surplus indicates the differences between government revenue and government expenditure at the full employment level of income. The consistency of the full employment surplus associated with any budget with the estimated full employment level of private savings and private investment (including net foreign investment) for maintaining an equilibrium full employment level of income can be evaluated. For any given level of private savings and private investment at full employment determined on the basis of chosen monetary assumptions, the achievement of a full employment surplus consistent with the fulfillment of the equilibrium conditions necessary to achieve full employment cannot by itself ensure that the full employment level of income will, in fact, be achieved. For, the automatic stabilizers implicit in the given budget structure may be insufficient to ensure the desired correction in the level of private savings and investment to their respective full employment levels, which is required to achieve full employment. This is not to deny that appropriate adjustments in monetary policy could result in the achievement of full employment when no changes in budgetary policy are made, but this would imply an exclusive reliance on automatic stabilizers for fiscal policy purposes.

The rule espoused by the present U.S. Administration that budgetary policy be designed to ensure a balanced budget at full employment is justified on the grounds both that it provides a guide for long-run expenditure policy by imposing a discipline of an upper limit on outlays and that it helps to stabilize the level of income by not increasing the level of expenditure above the long-term trend rate in periods of boom and by not reducing it below the trend rate in periods of slack.28 Apart from the practical difficulties inherent in calculating the trend rate of growth of government revenues at a hypothetical full employment level of income, there would appear to be no major obstacles for a government to base its longer-term expenditure policy on the calculated level of full employment income.29 The pursuit of such a policy, however, does imply that variations in the rate of growth of expenditure from the trend rate of growth of expenditure can occur only if accompanied by discretionary changes in taxation with revenue yields at the full employment level of income equal to the absolute size of the faster, or slower, than trend growth in expenditure. A rigid application of the balanced-budget-at-full-employment rule would then necessitate a discretionary fiscal policy that could vary expenditure and full employment revenues by equal amounts. Thus, unless the balanced budget multiplier is quantitatively significant or unless the full employment estimate of the revenue yield of discretionary tax policy differs substantially from the actual revenue yield of these measures in a given period, the rigid application of the balanced-budget-at-full-employment rule severely constrains the use of discretionary fiscal policy for stabilization policy in the United States. There is also the additional problem, for demand management purposes, that a budget that is balanced at full employment may not be consistent with maintaining an equilibrium full employment level of income, given the savings and investment propensities in the private sector of the economy. Unless assumptions are made that the required equilibrium conditions are to be brought about exclusively by monetary policy, it would not be possible except perhaps in extremely fortunate situations to eschew the use of discretionary policy in obtaining full employment.

V. Evaluation, Summary, and Conclusions

Problems of measurement

As mentioned previously, the full employment budget surplus as it is defined and measured in the United States does not indicate what the actual budget surplus would be if the level of income corresponding to the full employment of labor were to be achieved. For, the actual price level at the existing level of output—not an estimate of what the price level would be at full employment—is used to convert potential output, which is first estimated in real terms, into money terms. If the anticipated price level used in ex ante estimates of the size of the full employment surplus differs from the actual realized price level in the relevant period, the ex post size of the full employment surplus will reflect this variation in the rate of inflation as well as any discretionary policy changes made in the ensuing period. Allowance for this possibility would be necessary if ex ante and ex post evaluations of any given budget are being undertaken, but apart from this the problems posed by inflation in measuring the size of the full employment surplus are not major ones, provided that the actual price level and not a hypothetical full employment level is used for this purpose. In addition, comparing the size of the actual budget surplus with that of the full employment surplus is facilitated when both are determined at the same price level.

Estimating government revenues at the assumed level of potential output, however, is more complicated than that at the actual level of output, because in practice it requires a determination of what the functional shares in income would be at full employment. In this calculation, unless the actual level of income approximates the full employment level of income, the actual or anticipated functional shares in the level of income may not serve as an accurate guide to the likely shares at the full employment level of income. In such circumstances, sophisticated analysis may be necessary to determine the most likely functional distribution of income if the full employment level of income were to be achieved, and it may well prove difficult to make informed judgments of the accuracy of the calculated shares in the assumed circumstances. Inaccuracies in estimating the distribution of income in full employment income will be quantitatively important when there are considerable differences in the effective marginal rates of taxation on the various components of income. This, for example, is true in the United States, where company profits are subject to higher effective rates of taxation than personal income. But, nevertheless, any such inaccuracies will affect the absolute size of the full employment surplus and not the relative size of a series of full employment surpluses calculated under the same assumptions. Thus, the problem of accurate revenue estimation at full employment is more relevant to uses of the full employment surplus that concentrate on its absolute size, such as the balanced-budget-at-full-employment rule, which is used by the present Administration.

Overall evaluation

Any overall assessment of the full employment surplus concept as a tool for fiscal analysis is dependent on the uses for which the concept is considered, because the full employment surplus is not superior in all its uses to other tools of fiscal analysis, including the actual budget surplus. There is, however, one general feature of the full employment surplus concept that is of particular interest and that may explain in large part its popularity in periods of relatively high unemployment in the United States, namely, the emphasis that it places on the shortfall in government receipts from the level that would be obtained were the objective of full employment of labor to be achieved. By the mere fact that the full employment surplus exceeds the actual surplus, attention is drawn to the fact that the objective of full employment of labor has not been achieved, and this may facilitate the pursuit of a discretionary fiscal policy that is not actually destabilizing by further reducing the level of employment. If the government’s stated objective of full employment is not fully compatible with other stated objectives, such as the level of inflation to be tolerated and the desired balance of payments, the government may decide on occasion to modify its full employment objective in order to concentrate on its other stated objectives, and discretionary fiscal policy that serves to reduce the level of employment may be introduced in such a situation. Be this as it may, the relationship between the size of the full employment surplus and the actual surplus will still emphasize the departure of actual output from potential output, even though the government’s objectives at a given point in time need not include the immediate achievement of potential output.30

Comparison with the actual surplus

In the United States, the full employment surplus concept has been viewed principally as a substitute for the actual budget surplus in its use as a measure of budget impact and a guide for budgetary policy, and it is thus relevant to examine the merits of the concept for both these uses. The differences in the absolute size of the full employment surplus and the actual surplus arise for two reasons: (1) The differences in the level of revenues and expenditure that result from the divergence in the actual level of income from potential output, both measured at the same price level; (2) the differences in the level of revenue collections that result from differences in the functional distribution of income at the two levels of income.

The latter difference may be important in certain circumstances if variations in the functional distribution of income with the level of unemployment produce substantial variations in the income elasticity of any given revenue system over the business cycle. In such a situation, in contrast to changes in the size of the actual deficit, changes in the size of the full employment surplus over time will not reflect any variations in the actual income elasticity of the revenue system over time. The relevance of this point for policy purposes depends on whether the full employment surplus is to be used to measure the fiscal drag, which, in the absence of increased expenditure, will serve as an automatic influence on the level of economic activity that makes the attainment of full employment more difficult in periods when income is increasing at levels of output less than potential output. For such a purpose, fiscal policy in the given period would be more appropriately guided by anticipated or actual fluctuations in revenue collections that reflect actual or anticipated fluctuations in the functional distribution of income. Thus, in this particular context, the full employment surplus concept is subject to the disadvantage that the trend of revenues used in its calculation may reflect the extent of any incipient fiscal drag at any point in time less accurately than the trend in actual revenues used in calculating the size of the actual surplus.

For other possible uses of the full employment surplus, it has been demonstrated previously that in a dynamic context neither the full employment surplus nor the actual surplus permits the separation of the effects of discretionary policy action on the budget from automatic variations in the size of revenue and expenditure.31 But, nevertheless, under certain limiting assumptions, the full employment surplus concept may permit a qualitative but not quantitative evaluation of discretionary fiscal policy. This is the direct consequence of the fact that the automatic component of the growth in revenue included in calculating the size of full employment surplus is related directly to the assumed normal growth in potential output over time when the productive resources of the economy are increasing. In contrast to the actual budget surplus when the automatic component of the growth in revenue may be positive or negative, depending on the actual growth performance of the economy, the automatic component of the increase in revenue included in the full employment surplus will always be positive in magnitude. This difference permits certain limited conclusions to be drawn about changes in the size of the full employment surplus that, in the absence of additional information, cannot be drawn from an analysis based on the actual surplus.32 Two important cases where it is possible to make judgments on discretionary fiscal policy using the full employment surplus are as follows: (1) If the size of the full employment surplus decreases from one period to the next, this indicates an expansionary discretionary fiscal policy; (2) if the income elasticity of the revenue system is equal to or greater than unity, a decrease in the size of the full employment surplus as a percentage of potential output over time will also indicate an expansionary discretionary fiscal policy.

While the full employment surplus does permit the isolation of discretionary fiscal action for the purposes of ex ante budget evaluation, its superiority over the actual surplus in this context is limited to the advantages in avoiding the need to evaluate the effect of the discretionary changes in any given budget program on the actual level of income that is to be used for the purposes of revenue and expenditure forecasting in the actual budget. That is to say, a comparison of two alternative budgets using the actual surplus would necessitate the determination of what the differing effects of these budgets would be on the level of GNP. The use of the full employment surplus would not entail any such calculation but, on the other hand, would permit only a qualitative assessment of the different effects of the two budgets. For the full employment surplus concept permits the measurement of the quantitative size of discretionary fiscal action only at the full employment budget level of income, and this estimate, except for changes in the level of expenditure defined to be exogenous with respect to the level of income, will vary with the level of income at which it is measured.

One of the objections that may be raised against the use of the full employment surplus as a summary indicator for purposes of fiscal analysis can also be validly raised against the use of the actual surplus in the same context. In particular, a valid objection to the use of the full employment surplus for the ex post evaluation of discretionary fiscal policy is that the size of this surplus will vary with the rate of inflation and that the rate of inflation is influenced, inter alia, by discretionary fiscal action. A similar objection to the use of the actual budget surplus for the same purposes is equally valid, for a government may well increase the size of the budget surplus by allowing a higher rate of inflation, at least in the short run when the growth in expenditure is defined in money and not in real terms. Thus, the criticism that the unsatisfactory treatment of price changes is a defect of the full employment surplus concept and may limit its usefulness in certain situations33 is equally applicable to the use of the actual surplus in the same context.

Regarding the use of the full employment surplus concept as a guide to budget formulation, if the actions of governments are to be guided by rigid rules, there will be some advantage in using the full employment budget surplus rather than the actual budget surplus for formulating budgetary policy. A rule that required that the actual budget be balanced in a given period may demand destabilizing discretionary fiscal action that would further reduce the level of income, whereas such an eventuality is not possible if the objective were to balance the budget at the full employment level of income, because the emergence of an actual budget deficit in periods of less than full employment would not immediately require offsetting discretionary action to balance the budget. However, this is not to say that a rule to govern budgetary action using the full employment surplus concept is desirable in practice or justifiable on theoretical grounds, from the viewpoint of undertaking effective fiscal policy, or even determining the long-term growth of public expenditure.

A major disadvantage

Even the strongest proponents of the use of the full employment surplus would admit that this concept is inherently more difficult to quantify than the actual surplus. Indeed, it was mentioned previously that the difficulties of estimating the functional distribution of income at full employment may subject the absolute size of any calculated full employment surplus to some doubt, but that this complication would be less relevant if attention were to be concentrated on the relative size of the surplus, or changes in the size of the surplus, at different points in time. The actual surplus, however, cannot be subject to the same doubts on account of accuracy of measurement, at least as far as ex post analysis is concerned. In addition, the full employment surplus does entail considerably more computational effort than the actual surplus, and it may be argued that the resources devoted to calculating the size of the full employment surplus might be more efficiently applied to the study of different means of achieving the same objective.

To take a relatively important example, if the advantage that most influences the use of the full employment surplus in practice is its superiority in separating discretionary fiscal action from the automatic effects on the budget resulting from changes in the level of income, the identification of the discretionary component of the change in the actual budget result may yield the same benefit with less effort. It might well prove difficult to explain the purpose of such a dissection of the actual surplus (or of changes in the size of the surplus) to laymen, but the information provided would, in fact, be superior to the data that could be gathered from the use of the full employment surplus alone.

SUMMARIES

Financial Integration and Interest Rate Linkages in the Industrial Countries, 1958-711

Victor Argy and Zoran Hodjera

This article focuses on theoretical and policy issues posed by financial integration among industrial countries since the mid-1950s. The problems of measuring international financial integration and of estimating its trend over the years are also explored.

The role of the Euro-dollar market, as a major financial intermediary that channels short-term funds between the money markets of the industrial countries, is examined first. An econometric analysis of factors influencing the Euro-dollar interest rate indicates that its movements are dominated by conditions in the United States and suggests a high degree of integration between the U.S. and Euro-dollar capital markets. However, the results also lend some support to the view that the Euro-dollar rate is influenced by conditions in Europe, particularly by bursts of speculation.

In order to explore the relationship between financial integration and the harmonization of interest rates, the article examines the effects of financial integration on the volatility in reserves over the country’s business cycle. Initially, at the low level of integration, increases in financial integration tend to be stabilizing to reserves, since capital movements tend to offset trends in the current account. Beyond a point, however, integration will destabilize reserves over a cycle. In a fixed exchange rate regime, countries will then be faced with difficult policy options. They may continue to pursue independent interest rate policies by sterilizing the liquidity effects of reserve movements or by increasing their intervention in capital movements. Since neither sterilization policies nor intervention in capital movements is likely to be entirely successful, highly integrated economies might also tend to determine their interest rates partly in the light of interest rate developments overseas.

Among alternative ways to evaluate the degree of financial integration and possible changes over time, the article examines the responsiveness of the premium and discount in the forward market to changes in interest differentials. With integrated money markets, a rise in the interest differential in a country’s favor will induce the movement of arbitrage funds, which will have the effect of increasing (reducing) the forward discount (premium) on the domestic currency. The greater this sensitivity, the more integrated may be the economy; also, changes in the degree of sensitivity may reflect changes in integration over time. Statistical tests of six industrial countries suggested a high degree of financial integration during the 1960s. There was also some evidence of possible increase in integration in Belgium and France between 1960 and 1965, but increased resort to controls of capital movements in the latter part of the 1960s may have arrested this process in several countries and may have even reversed it in others.

Main Features of the Employment Problem in Developing Countries1

Avinash Bhagwat

The worsening of the employment problem in the developing countries despite their good growth performance has become a cause for concern. The problem arises from both the absolute size and the growth rate of their labor force. In 1970 about two thirds of the world’s labor force lived in the developing countries, and the projected increase in the 1970s was about half of the total labor force in the developed countries in 1970. The burden of productively absorbing such a fast-growing labor force into relatively weak economies can be an unbearable one.

The concept of unemployment used in the developed countries can describe only a part of the employment problem in the developing countries. The main reason for this is the predominance of household enterprises and nonwage employment in the developing countries, particularly in their rural/traditional sectors. Moreover, the lack of productive employment opportunities even in urban areas is often reflected in long hours at unproductive work to eke out a living. Hence, the employment problem in the developing countries has two aspects: open unemployment (mainly an urban problem) and disguised unemployment.

Even on open unemployment, the available quantitative information in the developing countries is rather limited and only some tentative observations can be ventured. First, evidence is insufficient to establish that the rates of open unemployment in the developing countries have been rising. Second, these rates are high enough to be a matter for serious concern. Third, the urban unemployment rates generally range considerably higher than the country-wide rates. Fourth, the gap in the rate of real earnings between the urban/modern and the rural/traditional sectors plays an important role in determining the rate of open urban unemployment.

It is necessary to define disguised unemployment. From the different meanings given this term in the past, the so-called low-income, or productivity, approach appears to be most useful operationally. It would define the disguised unemployed as those whose annual real incomes are below a certain bench mark, which would be set in each country in accordance with its economic and social conditions. The acceptance of this income approach to the definition of disguised unemployment shifts the emphasis from the creation of jobs as such to the improvement in productivity, and thus in real earnings, on the lower rungs of income distribution, a shift that is both appropriate and desirable.

This survey of employment problems in the developing countries suggests the need for a three-pronged approach to ameliorative policies. Effective population policies are needed to limit the growth of the labor force; urban open unemployment must be tackled not only by job creation in the modern sector but also by measures to regulate the “urban drift” of the population; and high priority must be given to measures for increasing the productivity of labor in agriculture and other ancillary activities in the traditional sector.

Analysis of Proposals for Using Objective Indicators as a Guide to Exchange Rate Changes1

Trevor G. Underwood

This article surveys various proposals that have been made for introducing objective or quantitative criteria for guiding exchange rate adjustment as a supplement to the judgmental concept of fundamental disequilibrium incorporated in the Articles of Agreement of the International Monetary Fund (IMF). Two main categories of proposal are identified.

First, there are those proposals in which the indicator is employed to trigger international consultations as to whether a situation of fundamental disequilibrium does in fact exist. Sizable parity adjustments are still envisaged, and they are to continue to be restricted to situations of identifiable fundamental disequilibrium. The main modification to the existing Articles of Agreement is to cease to leave the initiative for proposing a change in par value to the country with an emerging payments imbalance. The foremost proposal of this type was made by Lord Keynes when the IMF was established.

The second category of proposal reflects a greater concern with the problem of making the criteria for still sizable parity changes more objective and, consequently, the change more predictable in a situation of considerable short-term international mobility of capital. These are the “crawling peg” proposals, in which the rate of crawl is guided by an objective indicator, such as the level of reserves or the exchange rate, rather than being predetermined or totally discretionary. Here, the main purpose is to introduce a greater measure of exchange rate flexibility, while constraining the maximum rate of change by this means within a safety margin of a few percentage points per annum. Objective indicators are used partly to stimulate parity changes where these are appropriate and partly as a safeguard against competitive manipulation of exchange rates. Although the essential feature of declared par values of the Bretton Woods system is retained and the need for occasional parity changes of substantial size may remain, parity changes no longer would be restricted solely to situations of fundamental disequilibrium and sometimes would be in the wrong direction, thus needing to be reversed later.

The analysis concentrates on this latter, more radical, group of proposals and in particular on the pros and cons of using exchange market indicators, such as the level of reserves or the exchange rate, in guiding the gradual adjustment of par values.

The Brazilian State Value-Added Tax1

Michèle Guerard

In 1967 the Brazilian states abolished the heterogeneous turnover taxes that they had levied for 30 years and replaced them with a unified sales tax of the value-added type. The reform was designed to overcome the defects of turnover taxation and to secure a greater degree of tax coordination among the states of the Federation. The Brazilian experience is of interest both as a case study of value-added taxation in a developing country and as an illustration of the problems posed by interstate tax coordination in the special setting of a federation characterized by huge regional disparities.

After a brief survey of the overall Brazilian fiscal structure, and a review of the historical background of the state tax reform, the study concentrates on an analysis of the technical features of the Brazilian state value-added tax (VAT), including scope, taxable base, rates, and exemptions; the treatment of agriculture, investment, and exports; and the operation of the tax-credit mechanism. This is followed by a quantitative estimate of the base of the tax and its sectoral allocation. Finally, the differential geographical impact of the tax and the interstate distribution of revenue are examined, in an effort to highlight the problem of regional fiscal imbalance, which constitutes one of the major issues of Brazilian federalism.

The Brazilian state VAT is found to be less universal in coverage than the VAT that has been adopted in Western European countries. Services are excluded from its scope, along with a number of special sectors, such as construction and electric power. The tax is nevertheless a uniform, broad-based levy that succeeds in reaching, directly or indirectly, a considerable part of the value added by the Brazilian economy, and apparently involves only a small amount of double taxation. The rate of tax is the same for all commodities, and exemptions have been kept to a minimum. The administration of the tax does not appear to be unusually burdensome, a simplified estimate system being used to assess the tax on small retailers. A large proportion of the total VAT revenue is collected at the manufacturing stage.

Serious difficulties have been associated with the use of the tax at the state level within a federal framework. The tax has a common structure in all the states. The Federal Government prescribes most of its basic features, leaving relatively little scope for interstate variations in tax rates and exemptions that tend to disturb competitive equilibrium and lead to the dislocation of production and trade patterns. Yet, it has proved difficult to enforce a uniform system in all the states because of the marked regional economic differences. The tax has also tended to accentuate rather than to alleviate existing fiscal disequilibria. The use of origin rules in interstate transactions has been responsible for the heavy weighting in distributing the tax base in favor of the more developed states. The magnitude of the dilemma of interstate coordination is intensified by the lack of alternative sources of state finance that could be manipulated more flexibly on a local or regional basis.

Incomes Policy in Austria1

Erich Spitäller

Most industrial countries have recently found inflation to be more intractable than had generally been anticipated, and conviction has been growing that the authorities will have to supplement financial policies with more direct restraint on rises in wages and prices. Hence, the current interest in countries that have been pursuing incomes policies. This paper examines the experience that Austria has had with such policies. First, the paper considers some of the fundamental aspects of Austrian incomes policies and summarizes wage and price policies in the period between the end of World War II and the establishment of the current system of controls. Subsequently, it describes the mechanism of current incomes policies and the mixture of voluntary restraint and compulsion that underlies its operation.

The paper evaluates the effectiveness of Austrian incomes policies and finds that they may have been instrumental in the maintenance of good industrial relations and may have helped to avoid the disruption of growth and stability that is caused by strikes. In addition, the wage and price settlements effected under the system of controls may have restrained inflation, judging by the favorable developments in unit labor costs and the trade-off between unemployment and inflation in Austria compared with other industrial countries. At least in part, restraint appears to have been accomplished through the timing of settlements, involving a delay in wage and price increases and thus a reduction in inflationary pressures. The Austrian experience illustrates some basic interrelated conditions that may be conducive to the success of incomes policies: (1) comprehensive organizations of employers and employees; (2) highly centralized policy formation and execution in these organizations; (3) close association between the organizations and the administrative and political establishment; and (4) the willingness of the organizations to compromise.

The Full Employment Budget Surplus Concept as a Tool of Fiscal Analysis in the United States1

Daryl A. Dixon

The paper describes the concept of the full employment surplus and its method of estimation in the United States and analyzes (1) its general use as a summary measure of budget impact for purposes of policy evaluation and (2) its use in rules that govern the formulation of budgetary policy. The principal emphasis is on the first use of the concept, but the stabilization policy implications of its use to formulate budgetary policy are also described. In the analysis, wherever relevant, comparisons are made with the use of the actual budget surplus in the same context.

The full employment budget surplus is an estimate of what the budget result would be with a given expenditure and taxation program if the economy were at the full employment level of output. It is clearly a hypothetical construct designed to remove the effects of fluctuations in the actual level of economic activity from the budget result and to permit the identification of changes in government discretionary policy. One conclusion of the paper is that, subject to certain limitations, the full employment surplus is superior to the actual surplus for purposes of fiscal analysis in a static context. But in a dynamic context the superiority of the full employment surplus is less apparent except in certain situations described in the paper. In general, the defects associated with using the actual budget surplus to evaluate discretionary fiscal action, such as the possible variation of the size of the surplus as the result of discretionary action, are also relevant to the use of the full employment surplus in a dynamic setting.

The full employment surplus concept is presently used in the United States to determine the size of government expenditure to be included in a given budget. The stabilization policy implications of this use of the concept are favorable for achieving full employment compared with the use of the actual budget surplus in the same context. However, the use of the full employment surplus in such a way does not guarantee the achievement of the full employment objective.

RESUMES

Intégration financière et relations entre les taux d’intérêt des pays industriels, 1958-711

Victor Argy et Zoran Hodjera

Le présent article porte principalement sur les questions théoriques et de politiques posées par l’intégration financière entre pays industriels depuis le milieu des années cinquante; il examine en même temps les difficultés rencontrées pour mesurer l’intégration financière internationale et estimer ses tendances possibles au cours des années.

Les auteurs étudient d’abord le rôle du marché de l’Euro-dollar, intermédiaire financier d’une importance majeure qui alimente en fonds à court terme les marchés monétaires des pays industriels. Il ressort de l’analyse économétrique des facteurs qui ont une incidence sur le taux d’intérêt de l’Euro-dollar que les mouvements de ce taux sont influencés par la situation aux Etats-Unis—ce qui témoigne d’un degré élevé d’intégration entre les marchés financiers des Etats-Unis et de l’Euro-dollar. Cependant, les conclusions semblent montrer aussi que le taux de l’Euro-dollar est influencé par la situation économique en Europe, en particulier par les poussées de spéculation.

Afin d’étudier le rapport qui existe entre l’intégration financière et l’harmonisation des taux d’intérêt, les auteurs examinent les effets de l’intégration financière sur l’instabilité des réserves au cours du cycle conjoncturel. Au début, lorsque le niveau d’intégration est peu élevé, tout accroissement de l’intégration financière tend à avoir un effet stabilisateur sur les réserves, étant donné que les mouvements de capitaux tendent à compenser les tendances du compte courant. Mais au-delà d’un certain point, l’intégration aura un effet déséquilibrant sur les réserves au cours du cycle. Dans un régime de taux de change fixes, les pays se trouveront alors face à de difficiles options. Ils pourront continuer à appliquer des politiques indépendantes en matière de taux d’intérêt en stérilisant les effets de liquidité des mouvements des réserves ou en augmentant leur intervention dans les mouvements de capitaux. Etant donné que, fort probablement, ni les politiques de stérilisation ni l’intervention dans les mouvements de capitaux ne donneront entièrement satisfaction, des économies fortement intégrées pourraient aussi avoir tendance à déterminer leurs taux d’intérêt en fonction de l’évolution des taux d’intérêt à l’étranger.

Parmi les diverses manières qui s’offrent d’évaluer le degré d’intégration financière et les changements susceptibles de se produire au cours du temps, les auteurs examinent la mesure dans laquelle le report et le déport sur le marché à terme réagissent aux changements intervenant dans les écarts entre les taux d’intérêt. Dans le cas de marchés monétaires intégrés, une augmentation de l’écart entre les taux d’intérêt en faveur d’un pays amorcera le mouvement de fonds d’arbitrage, ce qui aura pour effet d’accroître (de réduire) le déport (report) à terme sur la monnaie nationale. Plus cette sensibilité est grande, plus l’économie sera intégrée; de même, tous changements dans le degré de sensibilité pourront traduire des changements se produisant dans l’intégration au cours du temps. Il ressort de tests statistiques effectués sur pays industrialisés que l’intégration financière a atteint un degré élevé pendant les années 60. Ces tests indiquent aussi un accroissement possible de l’intégration en Belgique et en France entre 1960 et 1965, mais il se peut que le recours accru à des contrôles sur les mouvements de capitaux pendant la dernière partie des années soixante ait mis fin à ce processus dans plusieurs pays et l’ait renversé dans d’autres.

Le problème de l’emploi dans les pays en voie de développement: ses principales caractéristiques1

Avinash Bhagwat

Malgré un taux de croissance relativement élevé, les pays en voie de développement voient avec inquiétude s’aggraver leur problème de l’emploi. L’existence de ce problème provient du chiffre élevé de la population active et de son taux d’accroissement. En 1970, les deuxtiers environ de la population active du monde vivaient dans les pays en voie de développement et on estimait que son augmentation pendant les années 1970 représenterait à peu près la moitié du total de la main-d’œuvre des pays développés en 1970. Aussi, la tâche qui consiste pour des économies relativement faibles à faire absorber par les secteurs productifs une main-d’œuvre dont le nombre croît rapidement, peut devenir intolérable.

La notion du chômage telle qu’elle est utilisée dans les pays développés n’explique qu’en partie le problème de l’emploi auquel se heurtent les pays en voie de développement, et ceci principalement en raison du grand nombre d’entreprises familiales et de non-salariés, en particulier dans les secteurs ruraux et traditionnels. En outre, le manque de possibilités d’emploi productif, même dans les régions urbaines, se traduit souvent par de longues heures de travail peu productif passées à trouver une maigre subsistance. C’est pourquoi, le problème de l’emploi dans les pays en voie de développement se présente sous deux aspects: le chômage déclaré (essentiellement un problème urbain) et le chômage déguisé.

Même sur le chômage déclaré, les données quantitatives dont disposent les pays en voie de développement sont relativement limitées, c’est pourquoi les remarques que l’on peut faire sont d’un fondement quelque peu incertain. Premièrement, on n’a pas suffisamment de preuves pour affirmer que les taux du chômage déclaré dans les pays en voie de développement ont augmenté. Deuxièmement, ces taux sont suffisamment élevés pour soulever de sérieuses inquiétudes. Troisièmement, les taux de chômage dans les régions urbaines sont en général beaucoup plus élevés que sur le plan national. Quatrièmement, l’écart entre le taux de revenu réel dans les secteurs urbains et modernes et dans les secteurs ruraux et traditionnels influe de façon décisive sur le taux de chômage urbain déclaré.

Une définition du chômage déguisé s’impose. Des diverses acceptions données à cette expression dans le passé, c’est la méthode appelée méthode des faibles revenus ou de la productivité qui semble sur le plan opérationnel la plus utile. Elle englobe dans le chômage déguisé ceux dont le revenu réel par an est inférieur à un certain montant déterminé dans chaque pays en fonction de ses conditions économiques et sociales. Si l’on accepte cette définition du chômage déguisé, ce n’est plus à la création d’emplois en soi que l’on accorde de l’importance, mais bien à l’accroissement de la productivité et donc du salaire réel des groupes à revenus les plus faibles; cette modification des priorités semble à la fois appropriée et souhaitable.

De cette étude des problèmes de l’emploi dans les pays en voie de développement ressort la nécessité d’attaquer le problème de l’emploi sur trois fronts: il faut d’abord appliquer une politique démographique efficace pour limiter l’accroissement de la main-d’œuvre; il faut ensuite aborder le problème du chômage urbain déclaré, non seulement en créant des emplois dans le secteur moderne, mais également en prenant des mesures visant à réglementer l’exode rural de la population; il faut enfin donner une priorité élevée aux mesures destinées à accroître la productivité de la main-d’œuvre dans l’agriculture et autres activités connexes du secteur traditionnel.

Examen des propositions concernant l’emploi d’indicateurs objectifs comme critères pour les modifications des taux de change1

Trevor G. Underwood

Le présent article examine les diverses propositions faites pour adopter des critères objectifs ou quantitatifs permettant de modifier les taux de change, critères qui viendraient s’ajouter au concept de déséquilibre fondamental énoncé dans les Statuts du Fonds Monétaire International (FMI). Deux grandes catégories de propositions sont analysées ici.

Il y a, d’une part, celles qui proposent d’utiliser un indicateur pour déclencher un processus de consultations internationales visant à établir s’il existe réellement un déséquilibre fondamental. Ces propositions continuent d’envisager des ajustements de parité considérables qui ne devront toujours s’appliquer qu’aux situations où l’on peut établir l’existence d’un déséquilibre fondamental. La principale modification aux Statuts actuels du Fonds consiste à ne plus laisser au pays où commence à se manifester un déséquilibre de paiements l’initiative de proposer la modification de son pair. C’est Keynes qui le premier fitune proposition de ce genre lors des discussions qui portèrent création du FMI.

La seconde catégorie de propositions témoigne du désir d’augmenter l’objectivité des critères permettant d’effectuer encore de considérables modifications de parité et, par conséquent, de rendre ces modifications plus prévisibles lorsque se produisent d’importants mouvements de capitaux internationaux à court terme. Ces propositions visent à établir une “parité mobile” dans laquelle le taux de mobilité est fonction d’un indicateur objectif tel que le niveau des réserves ou du taux de change et non pas déterminé à l’avance ou laissé complètement à la discrétion des intéressés. Le principal objet de cette proposition est de rendre les taux de change plus souples tout en limitant ainsi le taux de change maximum à l’intérieur d’une marge de sécurité de quelques points de pourcentage par an. Les indicateurs objectifs sont utilisés en partie pour stimuler des modifications de parité là où elles s’imposent et en partie comme garantie contre la course à la dévaluation. Bien que cette proposition—et c’est là une caractéristique essentielle du système de Bretton Woods—fasse encore obligation aux membres de déclarer un pair et leur permette toujours de le modifier dans une mesure considérable, elle ne limite plus les modifications de parité aux situations où existe un déséquilibre fondamental et conçoit même l’adoption de modifications contraires, exigeant une opération inverse à une date ultérieure.

L’analyse porte sur ce dernier groupe de propositions, qui sont d’un caractère plus radical, et en particulier sur les avantages et inconvénients d’utiliser des indicateurs relatifs au marché des changes comme le niveau des réserves ou du taux de change, pour guider l’ajustement progressif des pairs.

La taxe sur la valeur ajoutée dans les Etats du Brésil1

Michèle Guerard

En 1967, les Etats du Brésil abolissaient les impôts hétérogènes sur le chiffre d’affaires qu’ils prélevaient depuis 30 ans pour les remplacer par un impôt commun sur les ventes, du type taxe sur la valeur adjoutée. Cette mesure visait à éliminer les inconvénients de l’imposition sur le chiffre d’affaires et à assurer une meilleure coordination fiscale entre les Etats de la Fédération. Le cas du Brésil a le double intérêt de fournir un exemple d’utilisation de la TVA dans un pays en voie de développement et d’illustrer les difficultés de la coordination fiscale entre Etats dans le cadre particulier d’une Fédération accusant d’énormes disparités régionales.

Après avoir fait l’esquisse de la structure fiscale du Brésil et présenté le contexte historique de la réforme, l’auteur s’attache à analyser les caractéristiques techniques de la TVA adoptée au Brésil et notamment sa portée, son assiette, ses taux et ses exonérations; le régime accordé à l’agriculture, aux investissements et aux exportations ainsi que le régime des déductions d’impôt. Il procède ensuite à une estimation quantitative de l’assiette de l’impôt et de sa répartition entre les divers secteurs. Pour terminer, l’auteur examine la répartition géographique de l’assiette et la distribution entre les Etats des recettes de l’impôt, afin de mettre en lumière le déséquilibre fiscal régional qui constitue un des principaux problèmes du fédéralisme brésilien.

L’auteur constate que le champ d’application de la TVA brésilienne n’est pas aussi vaste que celui de la TVA adoptée par les pays de l’Europe occidentale puisque les services en sont exemptés ainsi qu’un certain nombre de secteurs particuliers, comme la construction et l’électricité. La TVA brésilienne est néanmoins un impôt uniforme et général qui parvient à atteindre, directement ou indirectement, une grande partie de la valeur ajoutée par l’économie brésilienne; en outre, elle ne semble entraîner que dans une faible mesure une double imposition. Le taux de la taxe est le même pour tous les produits et le nombre des exonérations limité au minimum. L’administration de la taxe ne semble pas trop difficile, un système d’estimation simplifié étant utilisé pour déterminer le montant dû par les petits détaillants. Une grande partie des recettes de la TVA est perçue au stade de la production.

Dans une fédération, l’utilisation de la TVA au niveau des Etats membres soulève de sérieuses difficultés. La taxe a une structure commune dans tous les Etats. Le Gouvernement fédéral fixe la plupart de ses caractéristiques de base, limitant ainsi nettement les possibilités de variations entre Etats des taux et des exonérations—variations qui risqueraient de perturber l’équilibre concurrentiel et de provoquer un bouleversement des structures de la production et des échanges. Il s’est toutefois avéré difficile d’appliquer un système uniforme dans tous les Etats car ceux-ci présentent de fortes disparités économiques. En outre, la taxe a plutôt tendance à aggraver qu’à atténuer les déséquilibres fiscaux existants. Le fait que les transactions entre Etats sont imposées en fonction de la provenance des produits entraîne une répartition de l’assiette imposable nettement favorable aux Etats les plus développés du pays. Le dilemme que pose la coordination entre Etats prend une ampleur d’autant plus grande du fait que ceux-ci ne disposent pas d’autres ressources fiscales susceptibles d’être manipulées de façon plus souple à l’échelle locale ou régionale.

La politique des revenus en Autriche1

Erich Spitäller

La plupart des pays industriels ont récemment trouvé l’inflation plus difficile à maîtriser qu’il n’avait été prévu en général. On s’est convaincu de plus en plus que les pouvoirs publics devront doubler leur politique financière d’une action plus directe destinée à entraver la hausse des prix et des salaires. C’est ce qui explique l’intérêt qui se manifeste à l’heure actuelle dans les pays qui ont adopté une politique des revenus. La présente étude examine l’expérience qu’a faite l’Autriche. Elle traite d’abord des aspects fondamentaux de la politique des revenus en Autriche et résume les politiques qui ont été suivies en matière de salaires et de prix entre la période postérieure à la Deuxième Guerre mondiale et l’établissement du système de contrôle actuellement en place. L’étude décrit ensuite le mécanisme des politiques des revenus présentement appliquées et le dosage de restriction volontaire et de contrainte que comporte son fonctionnement.

L’étude évalue l’efficacité de la politique des revenus en Autriche et elle constate que la politique des revenus a sans doute contribué au maintien de bonnes relations entre employeurs et salariés dans l’industrie et permis d’éviter l’interruption de la croissance et l’instabilité qu’entraînent les grèves. Au surplus, il se peut que les arrangements en matière de salaires et de prix conclus dans le système de contrôle aient restreint l’inflation, à en juger par l’évolution favorable des coûts de main-d’œuvre unitaires et le compromis entre le chômage et l’inflation en Autriche par comparaison avec d’autres pays industriels. Il semble que le freinage ait été obtenu au moins en partie par le calendrier des règlements, c’est-à-dire en différant les hausses des salaires et des prix et en réduisant ainsi les poussées inflationnistes. L’expérience autrichienne montre bien que certaines conditions fondamentales sont favorables au succès d’une politique des revenus: il est souhaitable 1) qu’il existe des groupements d’employeurs et de salariés ayant une large base; 2) que l’élaboration de la politique et son exécution au sein de ces groupements soient fortement centralisées; 3) que les groupements professionnels et les institutions administratives et politiques travaillent en étroite coopération; et, enfin, 4) que ces groupements aient la volonté de se faire mutuellement des concessions.

L’excédent budgétaire de plein emploi, instrument de l’analyse budgétaire aux Etats-Unis1

Daryl A. Dixon

Dans ce document l’auteur examine le concept d’excédent de plein emploi et la méthode utilisée aux Etats-Unis pour l’estimer, puis il analyse 1) son utilisation générale en tant que mesure sommaire de l’incidence budgétaire aux fins de l’évaluation de la politique économique, et 2) son utilisation dans les directives qui régissent la formulation de la politique budgétaire. L’accent est mis principalement sur la première utilisation de ce concept mais l’auteur décrit également les implications pour la politique de stabilisation qui résultent de son emploi dans l’établissement de la politique budgétaire. Au cours de l’analyse, des comparaisons sont effectuées lorsqu’il y a lieu avec l’utilisation dans le même contexte de l’excédent budgétaire réel.

L’excédent budgétaire de plein emploi est une estimation des résultats budgétaires qui seraient obtenus en appliquant un programme de dépenses et d’imposition donné, si l’économie se trouvait au niveau de production de plein emploi. C’est là bien entendu une conception théorique destinée à faire disparaître des résultats budgétaires les effets des fluctuations du niveau effectif de l’activité économique, et à permettre d’identifier l’incidence des modifications de la politique autonome du gouvernement. L’auteur conclut notamment que l’excédent de plein emploi donne dans certaines limites de meilleurs résultats que l’excédent réel lorsqu’on effectue une analyse budgétaire statique. Mais dans une analyse dynamique, la supériorité de l’excédent de plein emploi est moins évidente, sauf dans certaines situations décrites par l’auteur. En général, les difficultés liées à l’utilisation de l’excédent budgétaire réel pour évaluer les mesures budgétaires autonomes, par exemple les variations éventuelles du montant de l’excédent occasionnées par ces mesures, surgissent également lorsque l’on a recours, aux fins de l’analyse dynamique et dans le même contexte, à l’excédent de plein emploi.

Le concept d’excédent de plein emploi est actuellement utilisé aux Etats-Unis pour déterminer le montant des dépenses publiques à inclure dans un budget donné. Les implications pour la politique de stabilisation qui résultent de l’utilisation de ce concept permettent de se rapprocher davantage du plein emploi que si l’on utilise l’excédent budgétaire réel aux mêmes fins. Toutefois, une telle utilisation de l’excédent de plein emploi ne garantit pas que l’on atteindra l’objectif du plein emploi.

RESUMENES

Integración financiera y relaciones entre los tipos de interés en los países industriales, 1958-7l1

Victor Argy y Zoran Hodjera

Este artículo se concentra en las cuestiones de teoría y de política planteadas por la integración financiera entre los países industriales desde mediados del decenio de 1950. Se examinan también los problemas de medición de la integración financiera internacional y los de estimación de su tendencia a lo largo de los años.

En primer lugar se examina la función del mercado del eurodólar como uno de los intermediarios principales para canalizar fondos a corto plazo entre los mercados monetarios de los países industriales. Un anälisis econométrico de los factores que influyen en el tipo de interés del eurodólar indica que sus variaciones vienen dominadas por las condiciones en Estados Unidos y sugiere la existencia de un alto grado de integración entre el mercado de capitales de Estados Unidos y del euro-dólar. No obstante, los resultados apoyan también la idea de que el tipo de interés del eurodólar estä influido también por las condiciones en Europa, en particular los brotes de especulación.

Con el objeto de estudiar la relación entre la integración financiera y la armonización de los tipos de interés, se examinan en este trabajo las repercusiones de la integración financiera en cuanto a la volatilidad de las reservas a lo largo del ciclo económico de un país. Al principio, con un bajo grado de integración, los aumentos de integración financiera tienden a estabilizar las reservas, ya que los movimientos de capital tienden a contrapesar las tendencias de la cuenta corriente. Sin embargo, pasado cierto punto, la integración desestabilizarä las reservas a lo largo de un ciclo. En un régimen de tipos de cambio fijos, los países se enfrentarían con difíciles opciones de política. Pueden continuar la aplicación de políticas independientes de tipos de interés, esterilizando los efectos de liquidez de las variaciones de las reservas, o interviniendo en mayor grado en los movimientos de capital. Como no es probable que las medidas de esterilización ni la intervención en los movimientos de capital tengan completo éxito, las economías muy integradas podrían también tender a fijar sus tipos de interés, en parte a la luz de la evolución de los mismos en el exterior.

Entre distintos medios de evaluar el grado de integración financiera y las posibles variaciones con el tiempo, se examina en este trabajo la sensibilidad de las primas y los descuentos en el mercado a término ante las variaciones de las diferencias entre tipos de interés. De haber mercados monetarios integrados, un aumento en la diferencia entre los tipos de interés a favor de un país pondrä en movimiento fondos de arbitraje, cuyo efecto serä elevar (reducir) el descuento (prima) a término para la moneda nacional. Cuanto mayor sea esa sensibilidad, más integrada puede estar la economía; también las variaciones en el grado de sensibilidad puede que obedezcan a variaciones de integración en el tiempo. La verificación estadística en seis países industriales indicó un alto grado de integración financiera en la década de 1960. También se obtuvieron pruebas de un posible aumento de la integración en Bélgica y Francia de 1960 a 1965; sin embargo, el mayor uso de controles a los movimientos de capital en los últimos años de la década de 1960 puede que haya detenido ese proceso en varios países, y hasta puede haber causado un retroceso en otros países.

Principales características del problema del empleo en los países en desarrollo1

Avinash Bhagwat

El agravamiento del problema del empleo en los países en desarrollo, a pesar de la actuación satisfactoria de éstos en cuanto al crecimiento económico, ha pasado a ser motivo de preocupación. El problema proviene tanto del tamaño absoluto como de la tasa de crecimiento de la mano de obra. En 1970, unas dos terceras partes de la fuerza laboral mundial vivía en los países en desarrollo, y el aumento previsto en ese grupo en las proyecciones para la década de 1970 equivale a la mitad de la fuerza laboral total en 1970 de los países desarrollados. La carga de absorber productivamente esta fuerza laboral que crece tan rápidamente en economías relativamente débiles puede resultar insoportable.

El concepto de desempleo que se usa en los países desarrollados sirve para describir sólo una parte del problema del empleo en los países en desarrollo. La principal razón de ello es que en los países en desarrollo predominan las empresas de una unidad familiar y el empleo no asalariado, especialmente en los sectores rurales tradicionales. Además, la falta de oportunidades de empleo productivo, aun en las zonas urbanas, a menudo da por resultado largas horas de trabajo improductivo para ganarse la vida sea como sea. Por consiguiente, el problema del empleo en los países en desarrollo tiene dos aspectos: el desempleo abierto (principalmente un problema urbano) y el desempleo encubierto.

Aun en el caso del desempleo abierto, la información cuantitativa de que se dispone en los países en desarrollo es bastante limitada, y sólo pueden aventurarse algunas observaciones tentativas. Primero, la evidencia es insuficiente para afirmar que las tasas de desempleo abierto han ido aumentando en los países en desarrollo. Segundo, estas tasas son suficientemente altas como para constituir un motivo de seria preocupación. Tercero, las tasas de desempleo urbano generalmente se encuentran a niveles mucho más altos que las del desempleo de todo el país. Cuarto, la diferencia de nivel de ingreso real entre los sectores urbano/moderno y rural/tradicional desempeña una importante función en la determinación de la tasa de desempleo urbano abierto.

Hay que definir el desempleo encubierto. De los distintos significados que se le ha dado a esta expresión en el pasado, parece que el más útil desde el punto de vista práctico es el método llamado de ingreso bajo o productividad. Según este método se definiría el desempleo encubierto como el de las personas cuyo ingreso real anual sea inferior a cierto nivel establecido, que se fijaría en cada país de acuerdo con sus condiciones económicas y sociales. Al aceptar este método del ingreso para definir el desempleo encubierto, se pondrá menos énfasis en la creación de empleos como tal, y más en el aumento de la productividad, y por lo tanto de los ingresos reales, en los escalones inferiores de la distribución del ingreso, cambio de enfoque que resulta apropriado y conveniente.

Este estudio de los problemas del empleo en los países en desarrollo indica la necesidad de que la política destinada a introducir mejoras actúe por tres frentes. Se necesitan medidas demográficas eficaces para limitar el crecimiento de la fuerza laboral; el desempleo urbano abierto debe atacarse no sólo mediante la creación de empleos en el sector moderno sino también con medidas tendientes a regular el desplazamiento de la población hacia las ciudades; y se debe dar alta prioridad a las medidas destinadas a elevar la productividad de la mano de obra en la agricultura y en las otras actividades subordinadas en el sector tradicional.

Análisis de las propuestas para utilizar indicadores objetivos como guía en las modificaciones de los tipos de cambio1

Trevor G. Underwood

En este artículo se examinan varias propuestas que se han hecho para adoptar criterios objetivos o cuantitativos que sirvan de guía a los ajustes de los tipos de cambio, como complemento del concepto de desequilibrio fundamental incorporado, como elemento de juicio, en el Convenio Constitutivo del Fondo Monetario Internacional. Se definen dos categorías principales de propuestas.

Primero, propuestas con arreglo a las cuales el indicador se emplea para iniciar consultas internacionales en las que se decida si existe de hecho una situación de desequilibrio fundamental. Se prevén todavía ajustes importantes de la paridad, que continuarán limitándose a situaciones de desequilibrio fundamental identificable. La principal modificación que se haría al Convenio Constitutivo actual es que la iniciativa de proponer una modificación de la paridad ya no le correspondería al país en el que surja un desequilibrio de pagos. La principal propuesta en este sentido fue la presentada por Lord Keynes cuando se creó el FMI.

La segunda categoría de propuestas refleja una mayor preocupación con el problema de hacer más objetivos los criterios para modificaciones aún mayores de la paridad y, en consecuencia, lograr que las modificaciones sean más previsibles en una situación de considerable movilidad internacional del capital a corto plazo. Estas son las propuestas de “paridad móvil”, en que la tasa de movilidad se guía por un indicador objetivo, como el nivel de las reservas o el tipo de cambio, en vez de ser predeterminada o totalmente discrecional. A este respecto la principal finalidad es la de introducir un grado mayor de flexibilidad cambiaría, restringiendo al mismo tiempo la tasa máxima de variación a un margen de seguiridad de unos pocos puntos porcentuales al año. Los indicadores objetivos se emplean, en parte, para estimular modificaciones de la paridad cuando éstas sean pertinentes y, en parte, como salvaguardia contra la manipulación competitiva de los tipos de cambio. Aunque se retiene la característica esencial de paridades declaradas, del sistema de Bretton Woods, y puede seguir habiendo necesidad de hacer modificaciones ocasionales de magnitud sustancial en la paridad, las modificaciones de la paridad no se limitarían ya solamente a situaciones de desequilibrio fundamental y a veces se harían en sentido erróneo, con lo cual habría que invertirlas de sentido más tarde.

Este análisis se centra en este último grupo de propuestas más radicales, y en particular en las ventajas y desventajas de utilizar los indicadores del mercado cambiario, tales como el nivel de reservas o el tipo de cambio, como guía en el ajuste gradual de las paridades.

El impuesto de los estados brasileños sobre el valor añadido1

Michèle Guerard

En 1967 los estados brasileños abolieron los diferentes impuestos sobre el volumen de negocios que habían regido por 30 años, y los sustituyeron por un impuesto común sobre las ventas del género valor añadido. El objetivo de la reforma fue superar los defectos de la tributación sobre el volumen de negocios y conseguir mayor grado de coordinación tributaria entre los estados de la Federación. La experiencia brasileña reviste interés como estudio ilustrativo de la tributación sobre el valor añadido en países en desarrollo y como ejemplo de los problemas que plantea la coordinación tributaria interestatal en el marco especial de una federación caracterizada por enormes disparidades regionales.

Luego de efectuar un breve examen de la estructura fiscal brasileña en general y de pasar revista a los antecedentes históricos de la reforma tributaria de los estados, el estudio se concentra en el análisis de las características técnicas del impuesto de los estados brasileños sobre el valor añadido (ICM), su alcance, base imponible, tipos y exenciones; el trato de la agricultura, la inversión y la exportación; y el funcionamiento del mecanismo de créditos tributarios. A ello le sigue una estimación cuantitativa de la base del impuesto y su distribución por sectores. Finalmente se examinan las diferencias del impacto del impuesto por zonas geográficas, y la distribución de la renta entre los estados, tratando de poner de relieve el desequilibrio fiscal regional, que constituye uno de los mayores problemas del régimen federal brasileño.

Se llega a la conclusión de que el ICM de los estados brasileños tiene una cobertura menos universal que el impuesto sobre el valor añadido adoptado en los países de Europa occidental. Excluye los servicios y diversos sectores especiales, tales como la construcción y la energía eléctrica. No obstante, el impuesto es un gravamen uniforme, de base amplia, que consigue incidir, directa o indirectamente, en una parte considerable del valor añadido por la economía brasileña y parece entrañar pocos casos de doble tributación. La tasa del impuesto es la misma para todos los productos, y las exenciones se han reducido a un mínimo. La administración del impuesto no parece ser excesivamente gravosa, y se emplea un sistema simplificado de estimaciones para determinar el impuesto sobre los pequeños minoristas. Una proporción considerable de la renta total del ICM se recauda a nivel del fabricante.

En un sistema federal, la aplicación del impuesto a nivel de los estados ha planteado graves problemas. El impuesto tiene una estructura común en todos los estados. El Gobierno Federal prescribe la mayor parte de sus caracteres básicos, dejando relativamente pocas posibilidades para las variaciones interestatales de los tipos del impuesto y la exenciones que tienden a alterar el equilibrio competitivo y conducen a la dislocación de la estructura de la producción y el comercio. Aun así, ha resultado difícil hacer cumplir un sistema uniforme en todos los estados debido a las marcadas diferencias económicas regionales. El impuesto también ha tendido a acentuar, y no a moderar, el desequilibrio fiscal existente. El criterio de la procedencia de la mercancía aplicado en las transacciones interestatales ha favorecido considerablemente a los estados más desarrollados en la distribución de la base imponible. La magnitud del problema de la coordinación interestatal se intensifica por la falta de otras fuentes de recursos financieros de los estados que puedan manipularse más flexiblemente a nivel local o regional.

Política de ingresos en Austria1

Erich Spitáller

La mayoría de los países industriales han comprobado en los últimos años que el problema de la inflación es más difícil de lo que habían previsto en general, y se ha ido llegando a la conclusión de que las autoridades tendrán que complementar la política financiera con una contención más directa de los precios y salarios. De ahí el interés actual en los países que han seguido una política de ingresos. En el presente estudio se examina la experiencia que ha tenido Austria con dichas medidas. Primero, se analizan algunos aspectos fundamentales de la política de ingresos de Austria y se resume la política de precios y salarios seguida desde el fin de la Segunda Guerra Mundial hasta que se estableció el actual sistema de controles. A continuación se describe el mecanismo de la política actual de ingresos, y la combinación de contención voluntaria y obligatoria en que se basa el funcionamiento de la misma.

Se evalúa la eficacia de la política austríaca de ingresos, concluyéndose que puede haber sido uno de los factores en el mantenimiento de las buenas relaciones industriales y puede haber contribuido a evitar las perturbaciones del crecimiento y de la estabilidad originadas por huelgas. Asimismo, las decisiones sobre precios y salarios implantadas con el sistema de controles pueden haber contribuido a contener la inflación, a juzgar por la marcha favorable en los costos laborales unitarios y en la relación de correspondencia entre el desempleo y la inflación en Austria, en comparación con otros países industriales. La contención parece haberse logrado, al menos en parte, mediante el control de la entrada en vigor de las decisiones sobre precios y salarios aplazando las subidas, con lo que se reducen las presiones inflacionistas. La experiencia austríaca ilustra algunas condiciones básicas interrelacionadas, que pueden ser conducentes al éxito de una política de ingresos: 1) amplias organizaciones patronales y obreras, 2) un alto grado de centralización en la formulación y ejecución de la política de dichas organizaciones, 3) estrecha relación entre éstas y las instituciones administrativas y políticas, y 4) el que las mencionadas organizaciones hayan estado dispuestas al compromiso.

El concepto de superávit presupuestario en pleno empleo como instrumento de análisis fiscal en los Estados Unidos1

Daryl A. Dixon

En este trabajo se describen el concepto del superávit en pleno empleo y su método de estimación en los Estados Unidos, y se analizan 1) su uso general para medir de forma concisa el impacto del presupuesto, a fin de evaluar la política y 2) su uso en las directrices por las que se rige la formulación de la política presupuestaria. Se hace hincapié en el primer uso de este concepto, pero también se describen las repercusiones en la política de estabilización que tiene su empleo en la formulación de la política presupuestaria. En el análisis, y cuando es del caso, se hacen comparaciones con el uso del superávit real del presupuesto en el mismo contexto.

El superávit presupuestario en pleno empleo es una estimación de lo que sería el resultado presupuestal con un programa dado de gasto y tributación si la economía se hallara al nivel de producción de pleno empleo. Es, desde luego, una construcción hipotética destinada a eliminar del resultado presupuestal los efectos de las fluctuaciones del nivel real de actividad económica, permitiendo la identificación de variaciones en la política discrecional del gobierno. Una de las conclusiones del trabajo es que, sujeto a ciertas limitaciones, el superávit en pleno empleo es mejor que el superávit real para fines de análisis fiscal en un contexto estático. Pero en un contexto dinámico la superioridad del superávit en pleno empleo es menos evidente, salvo en ciertas situaciones que se describen en el estudio. En general, los inconvenientes que van unidos al uso del superávit real del presupuesto como forma de evaluar una medida fiscal discrecional, tales como la posible variación de la magnitud del superávit a consecuencia de las medidas discrecionales, existen también en el caso del uso del superávit en pleno empleo cuando hay una situación dinámica.

El concepto de superávit en pleno empleo se usa actualmente en los Estados Unidos para determinar la magnitud del gasto del gobierno que ha de incluirse en un presupuesto dado. Desde el punto de vista de la política de estabilización, las repercusiones del uso de este concepto son favorables al logro del pleno empleo, en comparación con el uso del superávit real del presupuesto en el mismo contexto. Sin embargo, tal uso del superávit en pleno empleo no garantiza que se cumpla el objetivo del pleno empleo.

In statistical matter (except in the résumés and resúmenes) throughout this issue,

  • Dots (…) indicate that data are not available;

  • A dash (—) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist;

  • A single dot (.) indicates decimals;

  • A comma (,) separates thousands and millions;

  • “Billion” means a thousand million;

  • A short dash (-) is used between years or months (e.g., 1955-58 or January-October) to indicate a total of the years or months inclusive of the beginning and ending years or months;

  • A stroke (/) is used between years (e.g., 1962/63) to indicate a fiscal year or a crop year;

  • Components of tables may not add to totals shown because of rounding.

International Monetary Fund, Washington, D.C. 20431 U.S.A.

Telephone number: 202 393 6362

Cable address: Interfund

Voting and Decisions in the International Monetary Fund

An Essay on the Law and Practice of the Fund

by Joseph Gold

Pp. xii + 368

$6.50

This book will be of interest not only as an account of the working of the International Monetary Fund but also as an examination of the process by which decisions are taken in one field of international activity. The process for arriving at international decisions is the subject of continuous debate in a world in which the number of independent states has increased rapidly, and in which important functions have been vested in a growing number of international organizations.

There are at present 125 member countries in the Fund, in all parts of the world, and in all stages of development. The voting power of each member is weighted on the basis of its quota in the Fund, and there are therefore great variations in voting power among members. The basic rule for decisions is that they are taken by a majority of the votes cast. The system of weighted voting power is buttressed by provisions that deal with adjustable voting power, reserved powers, special majorities, and other components of an elaborate framework for taking decisions. The origin, negotiation, and purpose of these provisions are explained. Within this framework, the Fund has been able, in the exercise of all of its regulatory, financial, and other functions, to take decisions by compromise and consensus, or at least by broad agreement, and without resort to voting. The book describes both the legal framework and the practice, and discusses, therefore, the elements of authority, power, and influence.

Mr. Gold describes the composition, powers, and procedures of the organs of the Fund, the Board of Governors, and the Executive Directors. He deals also with the role of the Managing Director and the staff. Finally, there is a discussion of the work of committees in the Fund, including the one most recently established, the Committee on Reform of the International Monetary System and Related Issues. Twelve appendices provide comprehensive information on various aspects of the voting power of members, the voting strength of individual governors and executive directors, the election of executive directors, and other material relevant to the adoption of decisions.

Address all orders and inquiries to

The Secretary

INTERNATIONAL MONETARY FUND

19th and H Streets, N.W.

Washington, D.C. 20431

U.S.A.

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.

This paper is the second in a series of studies of the techniques of fiscal analysis applied in selected industrial countries that has been undertaken by the Fund’s Fiscal Affairs Department. The first paper, “Techniques of Fiscal Analysis in the Netherlands,” also by Mr. Dixon, appeared in the November 1972 issue of Staff Papers (pp. 615–46).

Committee for Economic Development, Taxes and the Budget: A Program for Prosperity in a Free Economy (New York, November 1947). For a comprehensive description of the uses of the concept and its evolution, see Arthur M. Okun and Nancy H. Teeters, “The Full Employment Surplus Revisited,” Brookings Papers on Economic Activity: 1 (1970), pp. 77-110, and the references contained therein.

The Budget of the United States Government, Fiscal Year 1972 (Washington, 1971), p. 7.

See, in particular, Okun and Teeters, op. cit.; Nancy H. Teeters, “Estimates of the Full-Employment Surplus, 1955-1964,” The Review of Economics and Statistics, Vol. XLVII (1965), pp. 309-21; Arthur M. Okun, “Potential GNP: Its Measurement and Significance,” 1962 Proceedings of the Business and Economic Statistics Section, American Statistical Association (Washington, 1962), pp. 98-104; Michael E. Levy, Fiscal Policy, Cycles and Growth, National Industrial Conference Board, Studies in Business Economics, No. 81 (New York, 1963), Chapter 6, pp. 82-97; and Keith Carlson, “Estimates of the High-Employment Budget: 1947-1967,” Federal Reserve Bank of St. Louis, Review, Vol. 49 (June 1967), pp. 6-14.

See William H. Oakland, “Budgetary Measures of Fiscal Performance,” The Southern Economic Journal, Vol. XXXV (April 1969), pp. 347-58; and E. Gerald Corrigan, “The Measurement and Importance of Fiscal Policy Changes,” Federal Reserve Bank of New York, Monthly Review, Vol. 52 (June 1970), pp. 133-45.

For expositional convenience, it is assumed throughout this paper that a budget surplus may be either positive or negative. When it is negative, it is, of course, the conventionally defined deficit.

For a more detailed description, see Robert Solomon, “A Note on the Full Employment Budget Surplus,” The Review of Economics and Statistics, Vol. XLVI (1964), pp. 105-108, where the philosophical bases of this are explored succinctly.

Teeters, however, has recently identified changes in federal payments (corresponding to changes in the level of unemployment) for retirement, welfare, veterans’ benefits, Medicare, Medicaid, and aid to families with dependent children. Because these induced changes in the level of expenditure are not presently accounted for in measuring the full employment surplus, Teeters estimates that in the fiscal year 1972 the full employment surplus would have been approximately $8 billion larger than the official estimates. See Nancy H. Teeters, “Built-in Flexibility of Federal Expenditures,” Brookings Papers on Economic Activity: 3 (1971), pp. 615-48.

The U.S. budget for 1972 on a unified accounts basis was presented as one balanced at full employment. In this computation, no adjustments were made for fluctuations in the level of unemployment benefits. Ibid., p. 647. Such fluctuations have been allowed for, however, in the official calculations of the size of the 1973 full employment surplus.

Total expenditure will be lower at full employment than at levels of output at less than full employment because, for any given rates of unemployment compensation, the level of compensation will decrease as employment increases.

For a summary description of the method of computing the full employment surplus, see also the Economic Report of the President (Washington, February 1971), pp. 71-72.

See also Okun and Teeters, op. cit., pp. 82-84.

Ibid., p.93.

The U.S. Budget in Brief, Fiscal Year 1972 (Washington, 1971), p. 6. See also Richard E. Slitor, “The Full-Employment Budget Concept and Its Current Usage by the Federal Government” (paper presented to the West Virginia Tax Institute Conference at Pipestem, West Virginia, September 28, 1971).

The term “summary measure (or indicator) of fiscal action” is used in the following sections for expositional convenience. It refers to the use of the summary measure for either or both the ex post evaluation of policy and ex ante policy formulations.

The principal concern of the economists who were advising the Administration early in the 1960s was with the extent of the fiscal drag and the resulting need for a tax cut. See, for example, the views expressed in Walter W. Heller, New Dimensions of Political Economy (Harvard University Press, 1966).

For expositional convenience, it is here assumed that no expenditure (that is, including unemployment compensation) varies automatically with the level of economic activity. The relaxation of this assumption does not affect the conclusions of this section.

For a discussion of the program function that relates GNP to the budget surplus, see The Annual Report of the Council of Economic Advisers (Washington, 1962), pp. 77-81.

If the revenue system has a constant income elasticity over a considerable range of income, the schedule will be nonlinear over this range, because with expenditure assumed to be invariant, any increase in revenues will increase the budget surplus as a percentage of GNP.

Discretionary fiscal action will normally result in a change in the level of GNP and in a corresponding automatic change in government revenues. However, in separating the discretionary component of government action, it is generally assumed that the level of income at which it is desired to measure the discretionary change does not vary because of the budgetary action. This assumption is also adopted in this paper.

See Oakland, op. cit., pp. 353-54.

Okun and Teeters, op. cit., p. 83.

Robert Solomon, “The Full Employment Budget Surplus as an Analytical Concept,” in 1962 Proceedings of the Business and Economic Statistics Section (cited in footnote 4), p. 108. See also F. C. Miller, “The Full Employment Budget Surplus and Canadian Fiscal Policy, 1957-1962,” Canadian Tax Journal, Vol. XII (1964), pp. 445-46.

See pages 211-14.

For a further discussion of this point, see Miller, op. cit.

Okun and Teeters, op. cit., pp. 90-96.

Okun and Teeters have reviewed the case for weighting the full employment surplus in the United States at some length, ibid., pp. 84-88. See also Edward M. Gramlich, “The Behavior and Adequacy of the United States Federal Budget, 1952-1964,” Yale Economic Essays, Vol. 6 (Spring 1966), pp. 99-159.

See Slitor, op. cit., pp. 4-5.

The Netherlands has based its expenditure decisions on such a rule since early in the 1960s. See Daryl A. Dixon, “Techniques of Fiscal Analysis in the Netherlands,” Staff Papers, Vol. XIX (1972), pp. 615-46.

Naturally, any redefinition of the concept of full employment as part of a trade-off to achieve compatibility of full employment with other government objectives would necessitate a corresponding redefinition of potential output.

These conclusions are subject to modification if it is technically possible for the comparison of alternative policies at the full employment level of income to indicate different results from comparisons at other levels of income. See the section, Comparison with the actual surplus.

See Okun and Teeters, op. cit., pp. 90-96.

The complete text of this article begins on page 1.

The complete text of this article begins on page 78.

The complete text of this article begins on page 100.

The complete text of this article begins on page 118.

The complete text of this article begins on page 170.

The complete text of this article begins on page 203.

Le texte intégral de cet article en anglais commence à la page 1.

Le texte intégral de cet article en anglais commence à la page 78.

Le texte intégral de cet article en anglais commence à la page 100.

Le texte intégral de cet article en anglais commence à la page 118.

Le texte intégral de cet article en anglais commence à la page 170.

Le texte intégral de cet article en anglais commence à la page 203.

El texto completo en inglés de este artículo comienza en la página 1.

El texto completo en inglés de este artículo comienza en la página 78.

El texto completo en inglés de este artículo comienza en la página 100.

El texto completo en inglés de este artículo comienza en la página 118.

El texto completo en inglés de este artículo comienza en la página 170.

El texto completo en inglés de este artículo comienza en la página 203.

Other Resources Citing This Publication