I. Public Expenditure and Sustainable Fiscal Policy
- Ke-young Chu, and Richard Hemming
- Published Date:
- September 1991
What factors determine whether public expenditure is too high?
What are the macroeconomic consequences of taxation and alternative forms of deficit financing?
How are judgments about the sustainability of fiscal policy affected by public debt accumulation?
That public expenditure should not exceed 60 percent of national income may seem a surprising conclusion to those who have had to argue the case for reductions in expenditure from already much lower levels. Yet it is a conclusion for which Milton Friedman is well known, and other commentators on the limits to public expenditure in a number of countries have been guided by it. However, the most surprising aspect of this conclusion is not so much the number quoted or the arguments used to defend it. Rather, it is the fact that such a rule of thumb has been suggested at all, since it must use as its benchmark the level of public expenditure that can be sustained in the context of broad growth, inflation, and balance of payments objectives. Given that these objectives are country specific, so is the sustainable level of public expenditure.
The following discussion abstracts from a number of issues. First, public expenditure can be measured in a variety of ways, and in the presence of a wide range of measures the notion of a sustainable level of public expenditure is necessarily an uneasy one. The discussion proceeds on the basis of an agreed and clearly understood measure of public expenditure. The note on Public Expenditure Measurement discusses some of the issues that arise in arriving at such a measure. Second, microeconomic concerns related to economic efficiency are not fully taken up—these are covered in detail in the note on Public Expenditure and Resource Allocation. Clearly, if the rate of return on public expenditure is sufficiently high, the question of sustainability is virtually defined away. It is, however, assumed that the macroeconomic constraint is binding. It nevertheless remains the case that the microeconomic impact of public expenditure has a significant bearing on its macroeconomic sustainability. The link between the microeconomic and macroeconomic implications of public expenditure is discussed in the note on Public Expenditure, Stabilization, and Structural Adjustment.
Paying for Public Expenditure
The issue of the sustainability of public expenditure cannot be separated from questions related to how it is paid for. These are questions that are usually considered in two alternative but closely related ways. The first alternative focuses on the government’s (or public sector’s) financial balance, which can be written
where T = tax revenue, Cg = government consumption, Ig = government investment, Bgp = government borrowing from the private sector, ΔH = the change in the stock of high-powered money, and government borrowing from foreigners. The left-hand side of equation (1) is the fiscal deficit. If the government wishes to increase expenditure, this can be paid for with higher taxation without affecting the fiscal deficit. If a higher deficit results, this has to be financed by a combination of borrowing from the private sector, money creation and borrowing from foreigners. There are also other sources of financing, some of which do have a bearing on the analysis of sustainability. For example, if the government were to run down foreign reserves, this may be indicative of an impending foreign exchange crisis. In most respects, however, the economic impact of depleting reserves is the same as that of increased foreign borrowing, and the former possibility is not discussed in its own right below. Two other sources of financing with less obvious implications are discussed in subsequent notes: these are financing through the sale of public sector assets (see the note on Privatization) and financing through the accumulation of arrears (see the note on Expenditure Arrears).
While equation (1) lists the options available to pay for public expenditure, some aspects of the economic impact of these options are more readily analyzed in the context of the second alternative, the economy’s savings-investment balance. This can be written
where Sp = private saving, Ip = private investment, M = imports, and X = exports. (M - X) corresponds to the external current account deficit. Equation (2) then indicates that the fiscal deficit is the sum of the private sector’s saving-investment gap and the external current account deficit. The link between (1) and (2) is seen by noting that
where Bpf = private sector borrowing from foreigners. Equation (3) says that the private sector’s excess saving is measured by the extent to which it lends to the government and holds additional high-powered money less its foreign borrowing. Equation (4) says that the external current account deficit is financed by government and private sector foreign borrowing, or foreign savings. Substituting (4) and (3) into (2) yields (1).
An increase in expenditure matched by an equivalent increase in taxation leaves the fiscal deficit unchanged. According to the balanced budget multiplier this will have a positive impact on aggregate demand and output. In a closed economy, aggregate income rises by an amount exactly equal to the expenditure increase; in an open economy, the increase is lower as part of the increase in expenditure leaks out of the economy through imports. The expansionary consequences of balanced budget expenditure increases on output are, however, limited. In particular, aggregate supply may fail to keep pace with aggregate demand. For example, this may be the case if there are capacity constraints. It may also arise if increasing taxation is associated with disincentives to work and save. At a theoretical level, the behavioral response of labor supply and saving to higher taxes is ambiguous, being the net outcome of the disincentive effect of higher marginal tax rates and the usually offsetting income or wealth effect of higher average tax rates. At the empirical level, labor supply responses have in general been found to be small. However, important dimensions of labor supply, such as work effort and employment in the underground economy, are not reflected in these estimates, which may underestimate damaging supply responses. Similarly, tax-induced changes in interest rates generally have a small impact on the level of savings but significantly influence the composition of savings, with a possible bias against activities with the highest rate of return. An emerging mismatch between aggregate demand and aggregate supply, with its adverse implications for inflation and balance of payments, would point to a level of public expenditure that is unsustainable despite being paid for entirely by higher taxation.
If higher expenditure results in larger fiscal deficits, then the issue of sustainable expenditure coincides with the issue of a sustainable fiscal deficit. The commonly held view is that this latter issue rests on the impact of the associated financing, be it borrowing from the private sector, money creation, or foreign borrowing. There is, however, an alternative view that financing does not matter.
According to the debt neutrality (or Ricardian equivalence) hypothesis, borrowing is no more than deferred taxation, and insofar as the private sector recognizes this, it will adjust its consumption/savings behavior accordingly and the financial impact of borrowing will be reduced to that of the equivalent amount of taxation. While this argument is easy to understand in the context of borrowing from the private sector or foreigners, it less obviously applies in the case of borrowing from the central bank. However, the resulting money creation will lead to a higher price level which will require holders of money balances to increase their nominal money holdings to preserve their real balances. This phenomenon is often referred to as the inflation tax. If money creation is equivalent to taxation, the debt neutrality view is readily seen to extend not only to the prospect of borrowing financed by future taxation but also to the prospect of a future inflation tax. Notwithstanding the theoretical attraction of this view, it clearly assumes a degree of rationality on the part of private agents that is most unlikely to exist in practice. Moreover, the contention that taxpayer behavior is affected in the same way by different taxes—including one that is levied through higher inflation—is difficult to accept as an empirical proposition. Not surprisingly, the debt neutrality hypothesis gets little support from the available evidence in either industrial or developing countries. Despite its powerful theoretical implications, for all practical purposes it has to be assumed that the hypothesis is of limited relevance and that financing indeed matters.
Equation (1) indicates that an increase in the fiscal deficit can be financed by increased borrowing from the private sector, while equation (2) points to the fact that this could reflect an increase in private saving of equivalent amount. However, as already indicated, the available evidence suggests that the elasticity of savings with respect to interest rates is small, making it difficult to raise voluntary savings substantially in the short run. In consequence, any adjustment in the private sector’s savings-investment gap will tend to come from the crowding out of private investment. The precise mechanism by which this crowding out takes place can vary. If public and private investment are close substitutes then the latter may be directly displaced by the former. Alternatively, there may be financial crowding out as the availability of credit to the private sector to finance investment is rationed or the higher deficit forces up interest rates which, in turn, lowers investment. Crowding out is discussed in more detail in the note on Public Expenditure, Stabilization, and Structural Adjustment.
While there is limited scope for increased voluntary saving by the private sector, there may be potential for increased forced saving. Money creation is the most direct way of achieving such an increase in saving—note from equation (3) that increased money holdings are part of private saving. Provided that the private sector’s demand for real balances is inelastic with respect to inflation and the economy is sufficiently closed, the government can increase its spending power as the private sector accumulates nominal balances. There are two elements to this. First, the government collects seignorage revenue as real balances rise in response to changes in real income, interest rates, and financial structure. Second, it collects the inflation tax. Seignorage revenue is of limited potential. The standard quantity theory identity implies that
where v is the velocity of circulation of high-powered money, which is assumed to be stable, Δ
indicating that seignorage revenue (which is defined at π = 0) is increasing in the growth rate but decreasing in income velocity. In developing countries v can be as high as 20 or more; therefore at a growth rate of 5 percent, seignorage revenue is unlikely to exceed 0.25 percent of GDP. With an inflation rate of 5 percent, revenue would double, and increase further with higher inflation rates. There are, however, limits on the feasibility of continued recourse to financing the fiscal deficit through the inflation tax.
In industrial countries, a generally low tolerance for high rates of inflation and the narrowness of the noninterest-bearing component of the monetary liabilities of the banking system in relation to GDP mean that inflationary financing could amount only to a small share of GDP. In developing countries, the share of monetary liabilities of the banking system that are noninterest bearing will be larger. However, increases in inflation may depress other tax revenues owing to collection lags. More importantly, the scope for inflationary financing will be limited by the elasticity of real balances with respect to inflation; the demand for real balances will tend to decrease over time as the private sector seeks alternatives to monetary assets. While there may be scope for collecting significant revenue in the short run through the inflation tax, in the longer term the impact of reduced money holdings will dominate. Moreover, in an open economy, foreign exchange can increasingly be used for domestic transactions.
If there are limits to the extent to which higher public expenditure can be financed by domestic saving, equation (2) indicates that the only alternative is to rely more on foreign saving, or an increase in the external current account deficit. The link between the public sector deficit and the external current account deficit has tended to be a strong one, especially in heavily indebted countries. While there are examples of countries running large external deficits and fiscal surpluses—reflecting the strength of private sector imports, for example—these are exceptions that serve to illustrate that the link between the twin deficits is an imperfect one. Moreover, this link can be influenced by policy choices; in particular, the monetary policy that accompanies a fiscal expansion can, through its effect on the interest rate and the exchange rate, affect the resulting impact on the external current account deficit. But when there is a close link between the fiscal and external deficits, sustainability of the fiscal deficit requires an assessment of the external debt and debt service implications of running the implied external deficit. Indeed, if the government accumulated all the economy’s external debt and no domestic debt or liabilities to the central bank, this would be all that is required. In reality, external debt and public debt do not coincide. Even if the accumulation of external debt is not threatening balance of payments viability, and deficit financing is not inconsistent with immediate growth and inflation objectives, rising public debt may create its own problems.
Public debt dynamics
The dynamics of public debt accumulation can be described by the following expression for the change in the debt-to-GDP ratio
where r is the real interest rate, and
Equation (7) has a straightforward interpretation. The first term indicates that if the real interest rate exceeds the growth rate, rising interest payments will cause the debt ratio to increase; similarly, a primary deficit will also cause the debt ratio to increase. New debt need not be accumulated to the extent that the overall fiscal deficit (i.e., the primary deficit plus interest payments) can be financed by the creation of high-powered money. Otherwise, if the inflationary implications of money creation limit its acceptability, the debt ratio can be contained while running a primary deficit only if the growth rate of the economy exceeds the real interest rate; conversely, if the real interest rate exceeds the growth rate, containing the debt ratio without resort to money creation requires that the government run a primary surplus.
The dynamics of public debt are clearly sensitive to the relative magnitudes of the growth rate and the real interest rate. Prolonged periods with a growth rate higher than the real interest rate can accommodate an increasing level of debt. Although economic theory provides little guidance as to the general outcome, it is unlikely that a growth rate in excess of the real interest rate can persist indefinitely. In all likelihood, interest rates will rise and growth will be depressed as debt increases. Moreover, attempts to avoid such a result by maintaining interest rates at an artificially low level are likely to be counterproductive; this will either inhibit growth by fostering inefficient resource allocation, lead the government to borrow abroad at higher interest rates to maintain the growth momentum, or force increased reliance on money creation.
Public debt and inflation
A principal objection to rising public debt is that if there is an upper limit to the debt burden, deficits and debt will ultimately have to be financed through the inflation tax and that the required inflation rate will be unacceptably high. To demonstrate this, let the upper limit to the debt relative to GDP be denoted D. At this upper limit, money creation necessary to finance the fiscal deficit is given by setting
which on substitution into equation (5) implies that the inflation rate can be written
With v = 20, r = 5 percent, g = 3 percent, and a primary deficit equal to 1 percent of GDP, a permanent inflation rate of 37 percent is required to hold
The issues that arise in assessing the sustainability of public debt are taken up in more detail in the note on Interest Payments. In summary, it is shown that such an assessment should ideally be based upon whether the government (or public sector) is solvent. If the government’s net worth, taking into account the present value of future receipts and payments, is negative then public debt—and by implication fiscal deficits and the stance of fiscal policy—can be regarded as unsustainable. This judgment is closely linked to the earlier conclusion that the sustainability of expenditure—and therefore fiscal deficits and fiscal stance—should be assessed by reference to its consistency with the stable evolution of growth, inflation, and the balance of payments. The continuous pursuit of fiscal policy that is unsustainable from a macroeconomic standpoint will likely result in insolvency, while an insolvent government is generally expanding aggregate demand at a rate that exceeds the economy’s productive capacity.
AnandR. and S. van Wijnbergen“Inflation and the Financing of Government Expenditure: An Introductory Analysis with an Application to Turkey,” The World Bank Economic Review 3 1989.
Buiter W.H. and C.R. Bean The Plain Man’s Guide to Fiscal and Financial Policy (London: The Employment Institute1987).
FriedmanM. “The Line We Dare Not Cross: The Fragility of Freedom at 60 Percent,” Encounter1976.
Tanzi Vito “Fiscal Management and External Debt Problems,” in External Debt Managemented. by H. Mehran (Washington: International Monetary Fund1985).