Beveridge, W.A., “Fiscal Adjustment in Financial Programs Supported by Stand-By Arrangements in the Upper Credit Tranches, 1978-79” (unpublished, International Monetary Fund, July 1, 1981).
Beveridge, W.A., and Margaret R. Kelly, “Fiscal Content of Financial Programs Supported by Stand-By Arrangements in the Upper Credit Tranches, 1969-78,” Staff Papers, Vol. 27 (June 1980), pp. 205–49.
Donovan, Donal J., “Macroeconomic Performance and Adjustment Under Fund-Supported Programs: The Experience of the Seventies,” Staff Papers, Vol. 29 (June 1982), pp. 171–203.
Heller, Peter S., “The Problem of Recurrent Costs in the Budgeting and Planning Process” (unpublished, International Monetary Fund, June 1, 1982).
Khan, Mohsin S., and Malcolm D. Knight, “Some Theoretical and Empirical Issues Relating to Economic Stabilization in Developing Countries” (unpublished, International Monetary Fund, June 21, 1982).
Milne, Elizabeth, “The Fiscal Approach to the Balance of Payments,” Economic Notes, Monte dei Paschi di Siena, Vol. 6 (No. 1, 1977), pp. 89–106.
Reichmann, Thomas M., and Richard T. Stillson, “Experience with Programs of Balance of Payments Adjustment: Stand-By Arrangements in the Higher Tranches, 1963-72,” Staff Papers, Vol. 25 (June 1978), pp. 293–309.
Tanzi, Vito (1976), “Fiscal Policy, Keynesian Economics and the Mobilization of Savings in Developing Countries,” World Development, Vol. 4 (No. 10 and 11, 1976), pp. 907–17.
Ms. Kelly, Assistant Division Chief in the Stand-By Operations Division of the Exchange and Trade Relations Department, was Assistant Division Chief in the Fiscal Review Division of the Fiscal Affairs Department when this paper was written. She is a graduate of the University of New England (Australia) and the Australian National University and was an economist in the Research Department of the Reserve Bank of Australia before joining the Fund staff.
The concept of the government sector is discussed in Section II.
Donovan (1982) compares macroeconomic variables in the program year with the preprogram year (one-year comparisons) and also the average of three years beginning with the program year with the average of three years prior to the program period (three-year comparisons). Results in this paper are based on one-year comparisons.
This approach would also enable actual outcomes to be compared with hypothetical outcomes associated with alternative policy actions. In this context, however, it should be stressed that the determination of what alternative policies would have been followed in the absence of a Fund program involves major elements of subjective judgment.
This methodology is outlined briefly in Section I.
This does not imply that for programs in which an increase in the current account deficit is planned in the program year fiscal targets and subceilings are not useful.
Including net transfers (except public transfers) and workers’ remittances, as well as net factor incomes.
The overall government deficit as defined in this paper, unlike the definition in the Fund’s Draft Manual on Government Finance Statistics (Washington, June 1974), includes foreign grants “below the line” as a financing item. This is equivalent to treating foreign grants as a use of foreign savings (that is, the same as foreign borrowing) in the balance of payments. If foreign grants are included as a current transaction in the balance of payments, for consistency, they should be included together with revenue in the overall government deficit.
For ease of exposition in the rest of the paper, it is assumed that the author is dealing only with countries with balance of payments current account deficits and overall government deficits.
If external payments arrears exist, the overall balance of payments equals the change in reserves plus the change in outstanding arrears.
The treatment of balance of payments financing (for example, Eurocurrency loans) in this context is discussed in the subsection policy aspects in Section I.
Equations (2) and (3) focus on the relationships between the domestic savings and investment (income and expenditure) gap and the balance of payments. Alternatively, the current account balance and overall balance of payments outcomes can be expressed in terms of financial variables (assuming that changes in “other items net” in the monetary survey are zero).
domestic credit NDC = net domestic credit
Lpg = borrowing by government from the nonbank domestic sector
K = net foreign assets
M = money supply
Δ= a one-period change in the variable
Equation (4) shows that financing for the balance of payments current account deficit equals net domestic credit plus foreign capital inflows (including government borrowing and foreign grants). Equation (5), which is the familiar monetary survey identity, shows that the overall balance of payments outturn is equal to domestic credit creation offset by any increase in the money supply.
See Khan and Knight (1982) for a discussion of these interrelationships. In this context, however, the paper by Milne (1977) is of interest. Milne investigates the validity of the Cambridge New School fiscal approach to the balance of payments by testing the empirical relationship between the government deficit and the current account deficit. This approach assumes that the private sector deficit/surplus is essentially constant and that changes in the balance of payments reflect primarily changes in the government deficit. For this purpose, Milne estimated the equation (Z - X) = α + β (G - T) for 38 countries using annual data for the period 1960-75, where (Z - X) = trade balance, (G - T) = government deficit, α = the private sector surplus and is assumed to be constant for each country. The equation was also estimated allowing for a lagged response of the balance of trade to changes in the deficit for 32 countries. The results lent support to the fiscal approach to the balance of payments in slightly more than half the countries. That is, β was found to be large and statistically significant. (For the lagged version of the equation, this included approximately half the less developed countries included in the sample.) As noted in Milne, the existence of an empirical relationship between the government deficit and balance of trade does not imply causality. Nevertheless, the policy implications of these results are of interest in the present context. For countries where the p coefficient (or sum of the β coefficients for equations using β polynomial distributed lag) is greater than unity (which is true for all countries analyzed by Milne where β is statistically significant), an increase of 1 per cent in the government deficit eventually results in an increase that is greater than 1 per cent in the trade deficit. While this does not imply that policies that affect the private sector deficit have no effect on the balance of trade, it shows that such effects tend to be swamped by the effects of changes in the government deficit. Thus, for such countries, fiscal targets would take on an overriding importance in the context of balance of payments adjustment.
Similarly, the ex post measure of the change in the private sector deficit does not measure solely the impact of nonfiscal policies on the balance of payments.
Looking at the balance of payments from the point of view of financial variables, equation (4) indicates that a reduction (increase) in the fiscal deficit accompanied by a reduction (increase) in domestic bank and foreign borrowing by the government will lead to a reduction (increase) in the current account of the balance of payments, ceteris paribus. An increase in the government deficit financed by nonbank borrowing will have an expansionary impact if the monetary authorities accommodate the resulting increased needs of the private sector. If the monetary authorities do not accommodate these needs, private sector activity is possibly “crowded out.” Of course, the effects of changes in the fiscal deficit on the balance of payments depend not only on accompanying monetary policies but also on other policies, in particular, exchange rate and trade policies, and income and price policies. For example, if a flexible exchange rate policy is employed or the balance of payments is protected by quantitative restrictions, an increase in the government deficit financed by bank borrowing or foreign borrowing will have its major effect on inflation unless the monetary authorities reduce private sector credit.
In fact, in a number of Fund programs foreign commercial borrowing is treated in this manner in specifying ceilings on total credit and subceilings on credit to government.
For a fuller discussion of the “permanency issue” as it relates to the concept of fiscal equilibrium, see Tanzi (1982). The implication of Tanzi’s argument is that a reduction in the government deficit achieved through a reduction in current expenditure is likely to be less permanent than a reduction achieved through cuts in capital expenditure. It might be argued that tax increases (excluding those introduced on a temporary basis) might also have more permanent effects on the budget balance, and hence on the balance of payments.
These issues are discussed more fully in IMF (1976).
See Tanzi (1976) for a different view and a discussion of these issues.
Where the tax level is already high, the inflationary effects of further tax increases also become crucial. Tanzi (1981) in a review of this issue concludes that when the tax level is already high, the cost-increasing effect of tax increases can neutralize the effect on prices of the deflationary impact on demand.
The absence of a market test for the government sector often allows waste on prestige projects and ill-considered projects. To a large extent, the same is true for the private sector (and the nonfinancial public enterprise sector) in many less developed countries where price distortions (owing to interest rate, exchange rate, subsidy, and minimum wage rate policies) may result in a large divergence between private and social rates of return.
See International Bank for Reconstruction and Development, World Development Report, 1980 (Oxford University Press, 1980), p. 8 and pp. 48-64 for estimates of rates of return on education, health, and research and development.
See Heller (1982) for a discussion of recurrent costs in the budgeting and planning process.
The definition of the program period, which is the same as in Donovan (1982), is explained in the Appendix.
As discussed earlier, the term “fiscal contribution” used in this context is not synonymous with “fiscal policy contribution.”
Data on average changes in the current account are identical to those used in Donovan (1982).
Balance of payments financing is included as foreign financing (when utilized for budgetary purposes) in fiscal data.
Even if data were available to adjust the government cash deficit to an accrual basis, it is not clear that this adjustment would necessarily ensure consistency of the fiscal and balance of payments data used in this report, in that balance of payments and national income data actually reported to IFS are a mixture of cash and accrual data.
However, as noted in Section III, the two outlying cases in Chart 1 (where the current account and the budget deficit moved in opposite directions by substantial amounts) were related to an inconsistent treatment of arrears in the budget and the balance of payments.
As noted in Donovan (1982), where a program came into effect within a calendar year, the change in a variable during a one-year arrangement was spread over two calendar-year periods.
Unpublished data for this comparison are obtained from questionnaires completed for Beveridge and Kelly (1980), Beveridge (1981), and other unpublished data.
Data for total savings and investment, changes in the current account of the balance of payments, changes in international reserves, and changes in inflation and the rate of growth are identical to those used in Donovan (1982).
Fiscal problems in this context may have resulted from nonimplementation of fiscal or other policies, incorrect specification of functional relationships, or incorrect forecasts for exogenous variables.
While such a comparison is not made in this paper, it is an area in which future research could usefully be conducted.
For the two outlying observations (Zaire, 1977 and Togo, 1979) in Chart 1 in which the budget deficit and current account moved in opposite directions, it is interesting to note that for Zaire, external payments arrears increased substantially, whereas they had been programed to decline. This was reflected (on an accrual basis) in a larger than programed current account deficit. The recorded improvement in the budget deficit partly reflects the fact that transactions in the fiscal accounts were recorded on a cash basis. For Togo, the government deficit on a commitment basis did not change in relation to GNP. The increase on a cash basis reflects the effect of arrears, which increased more in the preprogram year than in the program year.
For all programs, the average increase in the rate of inflation in the year of the program compared with the preprogram year was 1.8 percentage points; on the basis of a comparison of the three-year period prior to the program with the three-year period after the program, the average increase was 3.2 percentage points (Donovan (1982), p. 184, Table 5). The comparable average increases for programs in which the current account deficit declined and the budget deficit increased were 1.4 and 3.6 percentage points, respectively.
Milne used time-series data to estimate the equation (Z-X)t=α + β (G - T)t. In this paper, the change in the current account is regressed on the change in the government overall deficit using cross-sectional data.
The constant term in equations where it is included is not significantly different from zero at the 95 or 99 per cent levels.
The coefficient p is not significantly different from unity in any of the equations.
This result is not surprising in that exogenous shocks are likely to have increased both the current account deficit and the fiscal deficit.
The proportionate adjustment for each sector is calculated as the ratio of the change in its deficit during the program period to the size of its deficit in the preprogram year, both measured in relation to GNP.
Table 5 is derived from questionnaires used in Beveridge and Kelly (1980), Beveridge (1981), and unpublished data. For comparability with Tables 1 and 2, only 5 of the 17 stand-by arrangements actually approved in 1980 are included.
This is the weighted average of 3.7 (programs in which current account and fiscal balance improved) and 1.3 (programs in which current account deteriorated and fiscal balance improved) in Table 1.
This is the weighted average of -1.8 (programs in which current account and fiscal balance deteriorated) and -1.6 (programs in which current account improved and fiscal balance deteriorated) in Table 1.
This does not imply that data for the current account (and hence implicit data for the rest of the economy) are not affected by the definition of “program period.”
The change in overall balance of payments is measured relative to exports, and the change in the domestically financed budget deficit is measured relative to GNP.
Alternatively, if the overall balance of payments is in small surplus in the preprogram year, it might be expected to show a higher surplus in the program year.
For many of these programs, the overall budget deficit is also expected to decline.
For this purpose, government investment is defined as the sum of government capital expenditure and net lending, and government savings is obtained as the difference between current revenue and current expenditure. Data for total savings and investment are identical to those used in Donovan (1982).
Changes in rates of growth are the average of three years, beginning with the program year, compared with the average of the three years prior to that year.
Changes in capital/output ratios can be of importance for countries in which the structure of production is changing from one that is less capital intensive to one that is more capital intensive.
A plus sign (+) denotes that external payments arrears existed; an asterisk (*) indicates that fiscal data are obtained from IFS; and double asterisks (**) indicate that fiscal data are obtained from unpublished or national sources.
See Donovan (1982), p. 199, for further details.