Chapter

VI Alternative Accounting Systems and the Fiscal Impulse

Author(s):
Ahsan Mansur, Richard Haas, and Peter Heller
Published Date:
May 1986
Share
  • ShareShare
Show Summary Details

The Fund’s World Economic Outlook exercise provides an analysis of the fiscal policy of both the central and general governments of the principal industrial countries. For most countries, the budget balance of the central government is measured on a cash (CA) basis, while a national accounts (NA) basis is used in analyzing the general government. Should the central government’s fiscal position also be measured on a national accounts basis? This section describes the conceptual differences between the two data bases, evaluates the feasibility and the implications of choosing one or the other data base, and provides a set of NA-based fiscal impulses to compare with the current CA-based series for the central government.

The CA-based system is the accounting framework generally used by governments for financial control and is on a checks-paid or revenue-received basis. For many non-industrial countries, this system is the only way public sector accounts are kept; and for most countries (including some of the principal industrial countries), it offers the accounts, both current and prospective, in a more timely fashion than is possible using the NA-based system. Statistics on a CA basis can be reconciled, if desired, with the monetary accounts, which are also kept on a cash basis. The arguments for using a CA-based system for measuring a country’s fiscal posture have been outlined in the Fund’s Draft Manual on Government Finance Statistics and may be briefly noted

  • From the financial point of view actual payment between the government and the rest of the economy holds greatest significance. It is actual payments to suppliers, employees and others that increase the money supply and in this way activate or validate the community’s demand for goods and services. It is actual payments by taxpayers, similarly, that decrease their liquidity and demand for goods and services.41

While in principle the central government’s fiscal accounts could be constructed on a CA basis according to conceptually consistent Government Finance Statistics (GFS) standards across countries, it should be noted that, in practice, this has not been possible for the Fund’s World Economic Outlook exercise. There is not complete consistency across the Group of Seven countries, either in terms of what is included in expenditure and revenue or in the institutions included in the definition of central government. However, an attempt has been made to at least ensure that the measure of the fiscal balance, which is the basis for the fiscal impulse analysis, is roughly consistent.42

The NA-based system focuses on measuring and analyzing real economic activity, not financial flows. Under the NA-based system, expenditures occur when deliveries are made to the government; revenues occur when payment is due without penalty. Expenditures and revenues on an NA basis are frequently, but not always, mirrored by a financial counterpart. Use of the NA-based system offers the obvious advantage that measures of fiscal activity are consistent with both the aggregate measure of economic activity as well as measures of other sectors of the economy, as calculated on an NA basis (including the current account of the balance of payments). This avoids the double counting that can arise if investment derived from government net lending shows up as a private investment. It also allows comparisons with domestic and foreign savings flows.

Since most industrial countries ultimately prepare their government accounts on an NA basis, the NA-based approach offers the prospect of an internationally consistent measure of a country’s fiscal accounts, though there are often considerable lags in the preparation of data on this basis and not all countries have yet adopted the United Nations’ System of National Accounts (SNA) procedure in preparing their estimates. Most industrial countries do provide estimates of the current and prospective fiscal positions of their general governments on an NA basis. In the absence of any other data on the financial position of the general government, there are obvious reasons why the NA basis has been used for the analysis of the general government’s fiscal posture by the Organization for Economic Cooperation and Development (OECD), the Bank for International Settlements (BIS), and the Fund. However, only a few countries (including the United States and Canada) provide current and prospective estimates of the central government’s fiscal position43 on an NA basis.

For several reasons, analyses of budgets based on the two alternative systems will frequently yield quantitatively different indicators about the stance of fiscal policy and, in fact, may even differ qualitatively. That is, fiscal policy may be described as expansionary according to one scheme and contractionary according to the other.

To explain how this can happen, it is necessary to set out the key differences between the CA-based and NA-based budget systems. First, the CA-based system includes the sale of assets as an item “above the line.” Thus the sale of a nationalized industry will result in a CA-based surplus or reduced deficit because it will reduce the public sector financing requirement.44 The sale of a public asset is not treated as a revenue item in NA-based budgets because it is not directly related to economic activity. The government has simply changed the liquidity composition of its assets.

A second and similar issue concerns public sector lending. Governments normally make loans to carry out resource allocation, as opposed to liquidity management, and on a CA basis such loans are considered in many countries to be an expenditure item (comparable to a capital transfer).45 Inasmuch as it simply involves changing the indebtedness of one sector for that of another without directly affecting aggregate demand, the NA-based system excludes such lending activities from its budget. In the NA-based system it is the resultant activity that is counted. For example, public loans made to finance housing would be counted as private sector investment when, and if, the proceeds of the loan were reflected in new construction.

Advocates of the NA-based approach argue that in its net lending, the government is operating principally as a financial intermediary, providing loans to the private sector at rates generally below those which the latter could obtain in financial markets. They would argue that, at most, only the implicit transfer arising from any interest rate subsidies should be included as a government expenditure.

Third, the systems differ in their timing. Tax receipts in the NA-based system are counted as they accrue, while in the CA-based budget they are counted as they are collected. The same is true for expenditures. Over short periods, this factor alone can lead to substantial differences. Over longer periods, this may not be as much of a problem as long as the lags are fairly constant.

These three items can lead to both different fiscal balances and different impulses in the CA- and NA-based systems. Other differences exist that do not affect the balances. These would only become important if one were to adjust various budgetary items in addition to the balance. For example, a capital-consumption allowance is both an expenditure and a revenue item in the NA-based system but is omitted from the CA-based system.

Based on the points made above, and subject to the availability of data, it would seem that the preferred budgetary system will be determined by the objectives envisioned for the fiscal impulse measure. If the important policy questions center around the short-run financial pressures brought on by the financing requirements of the government, then a strong case exists for a budget balance measure which accurately reflects these pressures. The CA-based budget data dominate in this regard. This is particularly relevant to developing countries, where public lending is an important element of government expenditure and net lending.

However, if the major concern is to analyze the effects of government expenditure and revenue policy on aggregate macroeconomic variables, such as consumption and investment, then a strong case can be made for using NA-based data in constructing the fiscal impulse. In this instance the budgetary data is more systematically related to aggregate demand. However, by using the NA-based approach, one may lose the ability to measure the influence of the government’s net lending policy in causing shifts in the observed behavior of other sectors of the economy as reflected in the national accounts.

Table 10 presents the results of central government fiscal impulses for 1976–81, calculated on both CA and NA bases. Note that the results of the Fund’s World Economic Outlook exercise shown in Table 10 correspond to those obtained in the July 1983 exercise and thus may differ slightly from the results of other tables in this paper (reflecting data as of February 1984). In both the United States and the United Kingdom, the CA- and NA-based fiscal impulse measures differ in sign one third of the time, with no systematic pattern discernible. In the United Kingdom, where it is possible to isolate the net lending effect, this factor appears to operate in the anticipated direction; nevertheless, it does not explain all of the differences between the CA and NA fiscal impulses for each year, even though the values for 1977 and 1981 are very close. In France, Japan, and the Federal Republic of Germany, the measured policy impulse differs in sign for the two series in at least one instance for each country over 1976–81. In two of the four cases where this occurs, the NA measure is expansionary and the CA measure contractionary over 1976–81. It is also possible to calculate the fiscal impulse with CA data, exclusive of net lending, for the Federal Republic of Germany, the United Kingdom, Italy, and France. However, no clear pattern emerges when this is done. In Canada there is generally little difference between the two measures. This is as it should be. Both measures are NA based. The difference must be attributed to differences between the Canadian national accounts, as provided by the Fund staff, and the NA coverage and corrections of the OECD.

Table 10.Seven Major Industrial Countries: Alternative Measures of the Budget Balance and the Fiscal Impulse of the Central Government, 1976–811(As percentages of GDP)
Country197619771978197919801981
Canada2
SNA3-WEO1−0.021.041.43−0.45−0.37−1.02
National accounts basis0.051.101.50−0.37−0.35−0.75
United States
Cash basis-WEO1−0.870.16−0.790.350.16
National accounts basis0.65−0.03−0.14−0.720.50−0.40
Japan
Cash basis-WEO10.630.310.381.090.11−0.19
National accounts basis−0.400.580.320.85−0.15−0.21
France4
Cash basis-WEO1−1.09−0.340.69−0.09−0.780.94
Cash basis (without net lending)−1.12−0.040.90−0.31−0.470.62
National accounts basis−1.430.490.64−0.49−0.970.27
Germany, Federal Republic of
Cash basis-WEO10.29−0.400.17−0.04−0.36−0.69
Cash basis (without net lending)0.56−0.270.11−0.15−0.35−0.72
National accounts basis0.17−0.110.150.11−0.37−0.90
Italy4
Cash basis-WEO1−0.65−0.775.30−2.98−0.070.67
Cash basis (without net lending)0.88−2.009.93−1.72−0.350.79
National accounts basis−1.97−0.235.78−1.38−1.571.48
United Kingdom
Cash basis-WEO1−2.20−2.642.720.39−2.17−2.90
Cash basis (without net lending)−0.34−1.532.59−0.16−2.16−1.66
National accounts basis1.18−1.451.78−1.09−1.64−1.61

Based on the data available for the Fund’s July 1983 World Economic Outlook (WEO) exercise.

In Canada, the central government fiscal impulse is calculated using national accounts data.

SNA denotes the United Nations’ System of National Accounts.

The French and Italian national accounts data exclude the social security budget. This is consistent with the Fund’s current World Economic Outlook practice.

Based on the data available for the Fund’s July 1983 World Economic Outlook (WEO) exercise.

In Canada, the central government fiscal impulse is calculated using national accounts data.

SNA denotes the United Nations’ System of National Accounts.

The French and Italian national accounts data exclude the social security budget. This is consistent with the Fund’s current World Economic Outlook practice.

    Other Resources Citing This Publication