Book
Chapter

# VII Estimation of Level of Structural Balance

Ahsan Mansur, Richard Haas, and Peter Heller
Published Date:
May 1986
Show Summary Details

The analysis of fiscal developments undertaken in various issues of the Fund’s World Economic Outlook focuses on the fiscal impulse. Recently attention has been drawn to the level of the structural balance, which may be defined as the fiscal balance which would prevail if, ceteris paribus, the economy was neither in a recession nor in a boom but was instead moving along its “normal trend.”46 This may differ from the “cyclically neutral balance” (CNB), as defined in the fiscal impulse calculations, in two ways. First, the CNB is based on the tax and expenditure structure prevailing in the base year. As such, changes in the budget balance since that period owing to changes in structural phenomena and/or discretionary policy changes are residually attributed to the measure of fiscal impulse. Second, the economy may not be at its normal trend output level. In estimating a measure of the structural balance, one ideally should adjust the CNB prevailing in the base year (taken as a share of GDP) for the effects of structural and discretionary changes that have occurred since the base year.

An approximate estimate of the share of the structural balance Bs in current GDP, bs may be obtained by subtracting the fiscal stance measure, FIS, taken as a share of current GDP, from the base-year share of the structural balance

where

• T0 = total revenues in the base year
• G0 = total expenditure in the base year (inclusive of unemployment-insurance-benefit payments)
• fis = (Bn — B)/Y = the share of the fiscal stance measure, FIS, in GDP

In effect, if a structural deficit has prevailed in the base year (e.g., if ${b}_{0}^{s}<0$) and if fiscal policy (net of cyclical factors) has been expansionary since the base period (fist > 0), one obtains a larger structural deficit (${b}_{t}^{s}<{b}_{0}^{s}$).

There are at least three problems with this approach. First, it provides an estimate of the structural balance as a share of actual output, as opposed to the output level that would prevail if the economy were on its normal trend. This latter concept may be readily calculated as

where Bst = (t0g0) YNFIS, and YN denotes the trend output (neither peak nor trough). This will differ from the ${b}_{t}^{s}$ calculated in equation (19) by the sign and magnitude of the difference between Y and YN.47 The specification of the trend output level YN is obviously a key factor in measuring the structural deficit. In this paper’s empirical analysis, as well as that of the OECD, YN is defined as potential output. An alternative approach (also suggested by the OECD) would measure the structural balance at the “trend mid-cycle point” rather than at the cyclical peak.48

Second, the measure of the structural balance derived from equation (20) requires a judgment on the appropriateness or normality of the structural budget balance (as a percentage of GDP), since this study is adding successive impulses (expansionary or contractionary) to the base-year balance.49 It is obviously critical to ensure that the measure of the structural deficit in the base year is viewed as accurate.

Third, the fiscal stance measure used in equation (20) is a composite of the effects of fiscal drag, automatic stabilizers, and discretionary policy changes, where automatic stabilizers are defined as arising from structural revenue and expenditure elasticities to nominal GDP that are different from unity.

Past output gaps affect the size of fis and, thus, the measured bs, whereas conceptually bs should be measured at the normal trend level of output and should be independent of the past effect of those automatic stabilizers which are due to cyclical factors.50 To the extent that an expansionary fiscal impulse arises from the operation of such factors in a recessionary period, this component should, in principle, be excluded from the estimate of the structural balance and be seen as a temporary part of the actual budget balance. With recovery, this component of the balance would be eliminated and the true structural deficit (surplus) would be lower (higher). In principle, one can correct this problem by purging the fiscal stance measure of the effect of induced budget changes deriving from nominal revenue and tax elasticities different from unity. A problem arises, however, in obtaining valid measures of the structural revenue and expenditure-elasticity parameters.

The OECD implicitly derives tax-elasticity estimates from its Interlink model, which includes separate structural tax equations covering company and personal income taxes, indirect taxes, and social security transfers. It is difficult to judge the degree to which these estimates are true elasticity measures, accurately separating the effects of discretionary changes in the tax law. An alternative approach to estimating true tax elasticities has been suggested by Prest (1962), but the data requirements of this approach are substantial. Estimates of structural expenditure elasticities are even more questionable. Sufficient doubts exist as to the appropriate methodology for estimating such elasticities to perhaps make it desirable to accept the limitations of a unitary elasticity approach.

In Table 11, estimates of the structural deficit for general government for 1978–82 have been calculated based on equation (20), using the data available at the end of February 1984 and assuming a unitary tax elasticity. The potential output measure used in the Fund’s World Economic Outlook exercise is used for this study’s estimates of YN. The plausibility of these estimates may be tested by comparing them with the estimates provided in mid-1984 by the OECD.51 On balance, the estimates implied by equation (20) suggest higher (lower) structural deficits (surpluses) than would be suggested by the OECD estimates for Canada, the United States, Japan, and the United Kingdom, with the opposite bias observed in recent years for France, the Federal Republic of Germany, and Italy.

Table 11.Seven Major Industrial Countries: Comparison of IMF and OECD Measures of the General Government Structural Deficit, 1978–831(As percentages of potential GDP)
Country197819791980198119821983
IMF2.92.12.21.21.72.2
OECD2.91.61.7−0.51.21.9
United States
IMF−0.5−0.10.91.5
OECD−0.9−1.2−0.7−1.6−0.30.2
Japan
IMF5.55.04.44.13.53.1
OECD4.94.34.13.52.82.2
France
IMF1.90.8−0.8−0.2−0.1−0.5
OECD1.70.8−0.80.20.60.7
Germany, Federal Republic of
IMF2.53.33.32.40.1−0.8
OECD1.72.32.52.40.9−0.5
Italy
IMF9.710.18.810.79.27.3
OECD9.19.78.612.012.09.7
United Kingdom
IMF4.33.51.9−0.5−1.20.6
OECD3.83.21.1−1.8−3.3−1.6
Sources: Fund staff estimates as of February 29, 1984. The OECD estimates are provided in Muller and Price (1984).

Positive figures are deficits, and negative figures are surpluses.

Sources: Fund staff estimates as of February 29, 1984. The OECD estimates are provided in Muller and Price (1984).

Positive figures are deficits, and negative figures are surpluses.

Tables 12 and 13 provide estimates of the actual and structural deficits for the central and general governments of the major industrial countries. The structural deficits are provided for the conventional budget balance and for the two alternative inflation-adjusted measures of the budget balance, data permitting. The OECD’s recent inflation-adjusted estimates of the general government structural balance are also provided in Table 13. None of the methods adjust the balance for any biases arising from the effects of automatic stabilizers. All estimates are, of course, conditioned on the assumptions made as to the output gap prevailing in any given year.

Table 12.Seven Major Industrial Countries: Central Government Structural Deficits With and Without Adjustment for Inflation, 1975–851(As percentages of GDP)
19841985
Country197519761977197819791980198119821983Estimated
Actual deficit (as percentage of actual GDP)2.31.83.54.63.43.32.15.86.75.94.3
Structural deficit (as percentage of potential GDP)
Conventional method2.22.23.54.53.83.52.43.84.54.63.4
Percentage share of conventional structural deficit2 in actual deficit96.6127.7100.097.6110.5102.9113.264.365.973.872.1
United States
Actual deficit (as percentage of actual GDP)4.93.32.72.01.22.42.54.36.15.34.9
Structural deficit (as percentage of potential GDP)
Conventional method2.71.92.02.01.21.61.82.24.14.14.0
Percentage share of conventional structural deficit2 in actual deficit56.257.276.1100.0105.366.370.752.870.380.183.8
Japan
Actual deficit (as percentage of actual GDP)4.35.05.15.36.26.15.95.55.14.94.2
Structural deficit (as percentage of potential GDP)
Conventional method4.34.95.15.36.36.36.05.44.94.74.0
Percentage share of conventional structural deficit2 in actual deficit98.397.599.5100.0102.5103.4102.097.796.195.994.7
France
Actual deficit (as percentage of actual GDP)2.61.21.02.61.51.12.62.82.93.02.9
Structural deficit (as percentage of potential GDP)
Conventional method2.11.10.72.61.71.02.02.01.71.41.2
Percentage share of conventional structural deficit2 in actual deficit80.486.972.0100.0111.594.176.672.458.645.538.9
Germany, Federal Republic of
Actual deficit (as percentage of actual GDP)3.62.82.22.11.81.72.21.92.01.51.0
Structural deficit (as percentage of potential GDP)
Conventional method2.12.42.02.12.11.71.1−0.6−0.6−0.9−1.0
Percentage share of conventional structural deficit2 in actual deficit357.184.990.7100.0115.0101.347.3−36.1−40.0−69.2−105.1
Italy
Actual deficit (as percentage of actual GDP)10.79.19.014.611.110.912.915.116.817.3
Structural deficit (as percentage of potential GDP)
Conventional method10.710.09.314.611.611.512.212.912.912.9
Percentage share of conventional structural deficit2 in actual deficit100.5110.5103.6100.0104.9106.394.784.875.873.1
United Kingdom
Actual deficit (as percentage of actual GDP)7.95.53.15.05.34.94.12.94.93.42.9
Structural deficit (as percentage of potential GDP)
Conventional method7.14.72.45.05.63.31.1−0.12.21.00.5
Percentage share of conventional structural deficit2 in actual deficit390.284.675.5100.0104.465.718.4−17.139.320.96.2

Positive figures are deficits, and negative figures are surpluses.

Measured with respect to actual GDP.

A negative value implies that there is a structural budget surplus and an actual budget deficit.

Positive figures are deficits, and negative figures are surpluses.

Measured with respect to actual GDP.

A negative value implies that there is a structural budget surplus and an actual budget deficit.

Table 13.Seven Major Industrial Countries: General Government Structural Deficits With and Without Adjustment for Inflation, 1975–851(As percentages of GDP)
19841985
Country197519761977197819791980198119821983Estimated
Actual deficit (as percentage of actual GDP)2.51.72.43.11.82.51.15.35.95.03.4
Structural deficit (as percentage of potential GDP)
Conventional method2.32.52.32.92.12.21.21.72.12.21.2
United States
Actual deficit (as percentage of actual GDP)4.12.10.90.61.20.93.84.23.83.5
Structural deficit (as percentage of potential GDP)
Conventional method1.20.30.1−0.5−0.10.91.52.22.2
Japan
Actual deficit (as percentage of actual GDP)2.73.73.85.54.84.24.03.63.33.12.5
Structural deficit (as percentage of potential GDP)
Conventional method2.63.63.85.55.04.44.13.53.12.92.3
France
Actual deficit (as percentage of actual GDP)2.20.50.81.90.7−0.31.82.63.12.92.9
Structural deficit (as percentage of potential GDP)
Conventional method1.60.40.51.90.8−0.8−0.2−0.1−0.5−1.7−2.0
Germany, Federal Republic of
Actual deficit (as percentage of actual GDP)5.73.52.42.52.83.13.93.52.91.91.3
Structural deficit (as percentage of potential GDP)
Conventional method3.52.92.12.53.33.32.40.1−0.8−1.3−1.4
Italy
Actual deficit (as percentage of actual GDP)11.69.07.99.79.68.111.711.911.912.5
Structural deficit (as percentage of potential GDP)
Conventional method12.210.68.49.710.18.810.79.27.37.9
United Kingdom
Actual deficit (as percentage of actual GDP)4.75.03.34.33.23.62.82.13.62.82.3
Structural deficit (as percentage of potential GDP)
Conventional method3.84.12.54.33.51.9−0.5−1.20.60.1−0.4

Positive figures are deficits, and negative figures are surpluses.

Positive figures are deficits, and negative figures are surpluses.

As a share of potential GDP, the share of the structural deficit bs*, adjusted for the effects of inflation, equals

where ∆V0 reflects the particular inflation-adjustment method used and FIS** denotes the fiscal stance measure corresponding to the particular inflation-adjustment measure.

Table 12 shows that the central government of the United States had a secular increase in its structural budget deficit since 1979, as one would have expected. After declining through 1981, Canada’s structural deficit rose to its 1978 level in 1983 and was projected to decline by 1985. The Italian structural deficit fell to 11.5 percent of potential GDP in 1980, but it is projected to have risen by more than a percentage point by 1983. After reaching approximate structural balance in 1982, the United Kingdom’s structural deficit reached more than 2 percent of potential GDP in 1983, but was projected to decline to 0.5 percent by 1985. The Federal Republic of Germany’s structural balance improved from a deficit of 2.1 percent of potential GDP in 1978 to a surplus of 0.6 percent to 1.0 percent of potential GDP during 1982–85. Japan’s structural deficit has steadily declined from 6.3 percent of potential GDP in 1979 to less than 5 percent in 1983, and 4.0 percent was projected for 1985. In four of the countries studied (the United States, Canada, France, and Italy), the structural deficit in 1983 is two thirds to three quarters of the actual deficit, while in the United Kingdom, it is only 40 percent. In the Federal Republic of Germany, there is a structural surplus coinciding with an actual deficit. In Japan, however, the structural deficit accounts for almost all of the actual deficit.

At the general government level, the United States ran a structural balance or surplus up to 1981, shifting in 1982 to an increasing structural deficit, which is projected to continue through 1985. In the European countries and Japan, there has been a clear pattern of reduction in the structural deficit, which in some cases has been reflected in the emergence of structural surpluses (in France, the Federal Republic of Germany, and the United Kingdom). The structural deficit is clearly highest in Italy, though its structural deficit has been reduced slightly since 1978, from 9.7 percent of potential GDP in that year to about 7.5 percent in 1983–84. In Canada, the structural deficit has fluctuated between 1.2 percent and 2.2 percent of potential GDP over the same period.

The effect of the two inflation-adjustment procedures is to reduce uniformly the observed structural deficit, with particularly large adjustments in specific years. For the United States, for example, the 1982 central government structural deficit is 2.2 percent of GDP without the inflation correction, 0.6 percent using the full inflation-adjustment method, and -0.2 percent using the real-interest-adjustment method.

The conventionally measured budget balance, when adjusted for inflation, generally shows a smaller deficit or a larger surplus, and since successive impulses are added to the adjusted base-year balance, the estimated structural balance with adjustments for inflation also shows similar trends. Periods of higher inflation are generally characterized by lower structural deficits or higher surpluses compared with the baseline World Economic Outlook estimates (e.g., the United States, France, the United Kingdom, Canada, and Italy). As between the two methods of adjustment for inflation, structural deficits based on full adjustment for inflation appear to be more sensitive to changes in the rate of inflation, as estimates of inflation-induced changes in the real value of outstanding government debt decline sharply with increased inflation.