Information about Europe Europa

2. Fiscal Accounts

International Monetary Fund
Published Date:
November 2005
  • ShareShare
Information about Europe Europa
Show Summary Details

a. Some basic concepts

According to economic theory, one important economic role of government is to produce certain goods and services that would not be supplied efficiently if production were determined by the market mechanism. This topic is a major theme of economic writings in the area of public finance, and it is discussed briefly below (Section c(6)). Stabilization policy is a second economic role. It focuses on the implications of government activity for the economy’s internal and external balance.

From the point of view of stabilization policy, there are two paramout aspects of government activity. Government spending on goods and services increases aggregate demand in the economy, and tax revenues reduce aggregate demand (not necessarily to an equal extent). The relative magnitudes of government spending and taxing determine whether the fiscal budget contains a positive or negative influence on overall output and the price level.

The second important aspect of government activity, from the viewpoint of stabilization policy, is the consequences of government borrowing. If a deficit is financed by borrowing from the central bank, there will be a link between monetary and fiscal policy; a fiscal deficit will occur at the same time as an increase in the money supply. Effects on the price level, and possibly on output, may be significantly increased. If, instead, the government deficit is financed by borrowing from commercial banks or the non-bank public, there will tend to be crowding out of private borrowing, and private-sector investment and growth may be reduced. If foreign financing is used, there will be consequences for the balance of payments in future years. While a quantity of foreign resources will be available to supplement domestic resources in the short run, offsetting inflationary pressure, the expenditure of foreign exchange will have to be reduced in the medium term, relative to receipts, to service and repay earlier borrowings. Whatever the source of finance, as government debt grows there are issues of sustainability, solvency, lender confidence, vulnerability to sudden speculative shifts, a restricted scope for use of fiscal policy for macroeconomic goals, and the threat of higher future taxes or reduced services.

b. Fiscal accounting

(1) Defining the government sector

The GFS Manual1 identifies several layers of government: (i) the central government; (ii) state or regional governments; (iii) local governments; and (iv) any supranational authority. The fiscal operations of the government sector include, at a minimum, the central government, which formulates national budgets. But in many countries fiscal operations are also performed by local and regional governments. The term general government includes all levels of government—central, provincial, and local. In order to emphasize the separation between the functions of the government and the financial sector, government finance statistics exclude any banking or monetary transactions that the government performs. In particular, all functions of the monetary authorities, irrespective of the institutions that carry them out (usually, but in some cases not exclusively, the central bank), are treated as activities of the monetary rather than the government sector. In order to distinguish fiscal policy from monetary policy, analysis must separate the activities of these two sectors.2

The broadest concept of government is the public sector, which includes the general government and nonfinancial public enterprises such as publicly owned railways and airlines or public utilities. Some of the central government’ revenues and expenditures typically reflect transfers from or to local governments and public enterprises. Such transfers have to be netted out so that “double counting” does not distort the sums when aggregates for the different levels of government operations are compiled. The accounting process of netting out and combining accounts at different levels of government is referred to as consolidation.

At each level of government, there are potentially three groups of operations: budgetary, extrabudgetary, and social insurance programs. Budgetary operations are, by definition, covered in the budget. Extrabudgetary operations, which are outside the budget, also raise resources through compulsory levies and provide nonmarket goods and services. For example, an extrabudgetary fund may be established to receive revenue from taxes on fuel and to spend these resources to maintain roads. Social insurance programs are a special category of extrabudgetary operations; these programs typically include a pension fund, an unemployment insurance fund, and a social insurance fund that benefits persons with physical disabilities or who are destitute.

In some countries, governments set up off-budget or extrabudgetary accounts to conduct what would otherwise be budgetary operations. The extensive use of extrabudgetary funds distorts the observable fiscal position, reduces flexibility in fiscal management by earmarking revenues for specific purposes, and, most importantly, leads to a loss of control over funds channeled through these accounts.3 To remedy this situation, countries create spending and revenue sub-categories within the budgetary framework to replace extrabudgetary accounts. In this way, a government may regain control, accountability, and scrutiny.

(2) Measuring government operations

Government finance statistics compiled in accordance with GFS standards record transactions on a cash basis rather than on an accrual basis as used in the national income and product accounts, the monetary accounts, and the balance of payments accounts. Receipts and payments recorded on a cash basis are documented as of the time of monetary settlement, while transactions recorded using the accrual method reflect the point in time at which a claim or liability arises or becomes due.

An example of a difference between transactions recorded on a cash basis and on an accrual basis is payments arrears. Arrears reflect a government’s inability to meet its expenditure commitments on time and result in inconsistencies between the accrual and cash deficits. An accrual deficit includes the government’s full expenditure obligations, while a cash deficit reflects expenditures that were actually made. A government with a cash deficit that is smaller than the accrual deficit is typically behind on its payment obligations. (See Box 2.2, below.)

(3) The GFS framework

The GFS framework covers primarily government receipts, payments, and unpaid obligations. The main principles underlying this framework are discussed below.

  • Revenues, receipts, payments, and expenditures. In order to assess the government’s overall fiscal position and the resultant macroeconomic impact, the first step is to note the distinction between receipts and revenues and between payments and expenditures. Revenues consist of all nonrepayable receipts other than grants. Receipts from loans to the government are not revenues, because the loans must be repaid. Similarly, not all payments are expenditures. By convention, a loan repayment is not an expenditure; it arises out of an obligation incurred when the loan was received. Interest payments, however, are an expenditure item. (Payments of grants or transfers to other governments, if any, are included in expenditure.)
  • Gross and net. As a general principle, revenues and expenditures are shown on a gross basis so that the statistics reflect the full magnitude and impact of the government’s revenue-raising operations and the disposition of revenues. Thus, school fees collected by government are not counted as an offsetting item to the cost of providing education, nor are the costs of collecting taxes deducted from tax revenues as “negative revenues.” The main exception to this rule is special lending and borrowing for policy purposes; in this case, sizable flows both from and to government make net lending the only meaningful item.
  • Requited and unrequited transactions. Requited transactions are those in which money is paid or received in exchange for goods or services. Unrequited transactions are those in which payment is made or received but nothing is received or given in return. Grants and certain transfers are examples of unrequited transactions.
  • Tax and nontax revenue. Tax revenues are defined as compulsory and unrequited receipts collected by the government for public purposes. Tax revenues (shown net of refunds of earlier overpayment) include compulsory social insurance contributions and profits transferred to the government by its fiscal monopolies. Nontax revenues include receipts from property income, fees and charges, fines, and operating surpluses of public enterprises (typically including the central bank).4
  • Grants. Grants are unrequited, nonrepayable, noncompulsory receipts usually from other governments or international institutions. The GFS Manual groups grants together with revenues “above the line” (as transactions that reduce the deficit) rather than with deficit financing items.5 Nevertheless, for purposes of fiscal analysis and budgetary planning, it is necessary to recognize that grants differ from revenues in a fundamental way. Because grants are not predictable or sustainable, expenditures planned and incurred on the assumption that grants will continue at current rates can seriously distort fiscal planning and force a severe adjustment in the future if the assumption turns out to be optimistic. In analytical presentations of the fiscal accounts, it may be advisable to show the fiscal deficit with grants both above the line (making the deficit smaller) and below the line, as financing (making the deficit larger).
  • Government net lending. Net lending (loans minus repayments) has a specific meaning in fiscal accounting. It refers to government lending undertaken to achieve public policy objectives.6 Examples of such lending include subsidized loans to farmers, students, or small businesses. Net lending is grouped with expenditures rather than with financing under GFS guidelines.7 This treatment reflects the differences between the reasons for government lending and those that motivate its borrowing.
  • Financing. Gross borrowing by government, and repayment of loans that are due, are the main components of transactions “below the line.” Borrowing and amortization flows are typically shown disaggregated by the source of finance (foreign, domestic banks, and domestic non-banks). Changes in the government’s holdings of currency and deposits are also part of financing.
  • The recapitalization of banks. In some countries, banks that are fully or partly owned by the government have been facing large losses and are now undercapitalized.8 In such cases, governments may inject capital into the banks or take over their debts. As noted earlier, the GFS framework is cash-based and therefore does not treat the government’s noncash assumption of debt as an expenditure at the time the debt is assumed. Cash interest payments are included as expenditure, but cash amortization payments are treated as negative financing. In other words, according to GFS conventions, only the interest payments on the assumed debt affect the size of the fiscal balance.
    • Privatization receipts. The GFS manual recommends treating the proceeds of the sales of equity in public sector assets as negative net lending. This is to be distinguished from sales of government physical assets which are recorded in nontax capital revenue. The rationale behind this thinking is that privatization proceeds are sales of government equity in enterprises acquired earlier, either through transfers or capitalization. This accounting treatment results in a onetime reduction in the fiscal deficit.While privatization receipts are formally treated as an above-the-line item, it is often useful to record them below the line in analytical presentations of the fiscal accounts, along with other transactions that affect the government’s net financial position. The one-time reduction in the fiscal deficit is offset by a smaller flow of nontax revenue, and thus larger deficits, in future years, reflecting the loss of government income from the enterprise’s remitted profits. These future deficits can be offset if the government uses the proceeds from the sale of enterprises to purchase other assets or to retire a portion of its own debt. In this sense, privatization proceeds resemble financing—the exchange of one asset for another, or use of an asset to redeem a liability.Under such a scenario, the government and the private sector have simply exchanged assets without affecting the demand for real resources. However, if the government uses the privatization proceeds to raise current expenditures, to cut taxes, or both, the deficit in the year of the sale will be unchanged while future deficits will be larger than otherwise, and the fiscal policy stance will be affected by privatization. The United Kingdom, which has had significant privatization receipts in recent years, has shown the fiscal deficit in the macroeconomic accounts both inclusive and exclusive of these receipts. This practice is useful because it clarifies the analysis by recognizing that privatization receipts represent an exchange of assets.9
    • Central bank profits. The profits of the central bank that are actually transferred to the government are treated as nontax revenue. However, these profits should not include unrealized profits stemming from the revaluation of holdings of foreign exchange or gold reserves. Moreover, profits remitted to government should cover the entire operations of the central bank, and not just selected operations (for instance, sales of appreciated foreign exchange or gold).

(4) Classifying revenues and expenditures

Government transactions are classified into broad categories under “revenues” and “expenditures.” The primary classifications are economic, including total, current, and capital revenues and expenditures. Within these categories, taxes are broken down according to the type of activity on which they are imposed, and expenditures according to their purpose—defense and education, for example. Expenditure may also be disaggregated by economic classification, which is more useful for fiscal analysis. The main categories of revenue and expenditure are given in Tables 2.1 and 2.2, respectively.

(5) The conventional fiscal deficit10

In terms of the cash flow of government operations, total receipts are always equal to total payments, and therefore the government’s fiscal accounts are always, in one sense, in balance. However, for analytical and policy purposes, it is more useful to focus on the difference between government revenues and grants, and government expenditures including net lending. The conventional concept of the fiscal balance (also called the “overall balance”) does precisely this, defining balance as total revenue plus grants minus total expenditure:

A government deficit thus represents the portion of expenditure and net lending that exceeds receipts from revenue and grants. The government covers the deficit by borrowing and/or by running down its liquid asset holdings. This conventional concept of the fiscal deficit is of central importance to fiscal analysis because it offers a comprehensive picture of the government’s overall financial position. Nevertheless, it suffers from a number of shortcomings when used as a measure of the impact of the budget on aggregate demand, of the allocation of resources in the economy, and of the distribution of income. First, the same deficit level may have substantially different macroeconomic effects depending on the associated structure of taxation and expenditure. It has long been established that changes in taxes affect macroeconomic aggregates differently than equal changes in expenditures. Second, different ways of financing a given fiscal deficit clearly have different macroeconomic effects. As explained below, central bank financing of the fiscal deficit has substantially different macroeconomic effects than, say, nonbank financing or foreign financing.

The conventional fiscal balance must be equal to the figure for financing but with the opposite sign (a fiscal deficit will be matched by positive financing). It may be presumed that, if the balance is a deficit, financing is made up mostly of borrowing. However, in any period a typical indebted government will make some repayment of past debt. Therefore, it is gross new borrowing net of amortization that is related to the fiscal deficit. Several other details are worth noting. In the identity,

“cash and deposits” refers to the government’s own short-term liquid assets (usually held at the central bank). In practical terms, seigniorage is invisible and may well be included as part of gross borrowing, although the two can be separated analytically (see Appendix I to this chapter for a full discussion). The “other” item could contain arrears if expenditures are recorded on an accrual basis (typically they would not be); “other” could also contain proceeds of the sale of gold or minerals from state-owned deposits (though usually profits from government enterprises are part of nontax revenues); finally, this item could contain privatization receipts (not according to GFS conventions, but reflecting analytical judgments in an individual reporting country).

c. Fiscal analysis

(1) The government saving-investment gap

As with other sectors, two key relationships define the fiscal sector’s resource balance. One of these is in terms of income and absorption, and the other is in terms of savings and investment. Whichever way the resource gap is viewed, it can be shown to be the same in principle as the fiscal balance (the conventional surplus or deficit, defined above). The two identities are derived in this section.

Let Tr represent government transfer payments (interest, subsidies, and social insurance benefits), and let net lending be grouped with capital expenditure. The government budget may then be written as:

where CB is the conventional balance and T is tax and non-tax revenue and grants.

By definition, T - Tr is disposable income of government, Ydg;Cg + Ig is absorption; and income minus consumption is saving. Substituting these definitions into the above expression yields,

Total income in the economy (Y = GDP) can be separated into government and nongovernment parts by recognizing that private disposable income includes transfers and excludes taxes:

Since consumption and investment disaggregate in a straightforward way into government and nongovernment parts, then from

where X and M refer to goods and services, one may write, substituting from above,

If Y is defined as GNDI instead of GDP, then (2.5) and (2.6) will have the current account balance on the right-hand side instead of X - M (as shown in Chapter 5).

(2) Measures of fiscal imbalance11

Although the conventional GFS concept of deficit is widely used, there is no single or best measure of the fiscal deficit. Alternative concepts can also be useful depending on the analytical purpose at hand. The conventional, or overall, deficit measured as a proportion of GDP is often used as a summary measure of fiscal performance and is perhaps the most widely accepted indicator of a country’s fiscal situation. But even this conventional ratio should be supplemented by some or all of the other measures discussed in this section.

  • The public sector borrowing requirement (PSBR) is the conventional fiscal deficit defined for the entire public sector. This comprehensive deficit concept can be applied at each level of government. The central government borrowing requirement and the general government borrowing requirement are contained within the public sector borrowing requirement. In practice, the PSBR can be derived by combining the general government and the public enterprise sector borrowing requirements with appropriate netting out of transfers between the two.
  • The current fiscal deficit (current account deficit) is defined as current revenue minus current expenditure. It is often used as a measure of government saving and therefore as a measure of the government’s contribution to the economy’s total saving. This measure can be written as
    There are, however, limits to the usefulness of the concept of the current deficit in practical fiscal analysis. It depends on the distinction between capital and current expenditures and revenues. Capital expenditures include purchases of assets that are expected to be used in production for a period of more than one year, plus capital transfers; all other expenditures are classified as current. Thus, spending on education (teachers’salaries) is counted as consumption by society, however essential it is for long-term growth, and only tangible physical assets like buildings and roads are included as investment. The conventions for classifying expenditures as current or capital are inherently somewhat arbitrary.12 More importantly, the underlying assumption that all government investment spending contributes to growth, and is therefore desirable, is questionable in view of the many instances of highly wasteful public investment projects. Moreover, spending on investment to the detriment of expenditure on maintenance of the existing capital stock may result in a net decrease in the latter, and ultimately reduce (rather than raise) the rate of economic growth.Thus, while attractive at first glance, the current fiscal deficit is not necessarily a very useful indicator. It also is not helpful in analyzing the impact of fiscal developments on either the external or the domestic macroeconomic balance.
  • The primary or noninterest deficit accurately measures the effects of current discretionary budgetary policy by excluding interest payments from the conventional measure of the deficit.13 This type of deficit indicates how recent fiscal actions of the government affect the allocation of resources in the economy and government debt, in abstraction from past borrowing decisions. This measure can be written as
  • The operational deficit is the conventional deficit less the inflation-induced portion of interest payments or, equivalently, the primary deficit plus the real component of interest payments. Inflation lowers the real value of the stock of public debt. Creditors are compensated for this loss through high nominal interest rates, which restore the buying power of the original loan principal. But a part of government interest payments is therefore actually amortization; in real terms the value of the principal of the debt is decreasing, as though it were being repaid. If the implicit amortization is not separated from interest payments above the line, the deficit will be overstated. The concept of the operational deficit overcomes this problem. This concept is particularly important in countries with high inflation and large public debt.

The relationship can be expressed as,


(3) Financing the deficit

The macroeconomic impact of a government deficit depends on the way the deficit is financed. There are essentially three ways of financing a deficit: (i) borrowing from the central bank, or “monetizing” the deficit; (ii) borrowing from the rest of the banking system or from the domestic nonbank sector; and (iii) borrowing abroad.

In a broad sense, each form of financing is associated with a major macroeconomic imbalance: excessive money creation with inflation; excessive foreign borrowing with an external debt problem; and excessive domestic borrowing with high real interest rates. In addition, explosive growth in public debt may occur because of the dynamic interactions between higher interest payments, larger deficits, and greater debt. Complicating the analysis, government and banking-sector data on government debt may not be consistent.14

  • Borrowing from the central bank (monetizing the deficit). Government borrowing from the central bank is equivalent to the creation of high-powered money.15 Creating money at a rate that exceeds demand at the current price level creates excess cash balances and eventually drives up the overall price level. In economies in which government deficits are monetized, this process is likely to be the principal source of inflation.Not all money creation is inflationary, but whether inflationary or not, it provides government with a claim on the resources of the economy in addition to revenue. Figuratively speaking, the government can “spend the new money first,” or part of it, and thus finance part of its expenditure other than by taxation. The claim on resources that the government realizes as a consequence of increasing the money supply is called “seigniorage.” It is discussed in Appendix I of this chapter.
  • Borrowing from the rest of the banking system. Unlike borrowing from the central bank, borrowing from deposit money banks (commercial banks) does not automatically lead to the creation of high-powered money. If the central bank accommodates extra demand for credit from the deposit money banks by supplying them with additional reserves, then this type of borrowing is similar to borrowing from the central bank. But if the central bank does not accommodate the extra demand for credit, the deposit money banks will be forced to reduce credit to the private sector in order to meet the higher demand for government credit. This phenomenon, which is referred to as crowding out of private spending, takes place principally through interest rate increases.
  • Nonbank borrowing. The government may sell securities to residents other than banks. Nonbank borrowing allows the government to finance a deficit in the short run without increasing the monetary base or raising the level of foreign debt. Nonbank domestic borrowing is, thus, considered an effective way to avoid both the risk of inflation and of external crisis that normally accompany a large, continuing fiscal deficit. However, like bank borrowing, borrowing from the public directly crowds out borrowers from the private sector to some extent by putting upward pressure on domestic interest rates (Box 2.1). Not only do high real interest rates hurt economic growth by inhibiting investment, but issuing public debt at high rates adds to the cost of future debt service and thus to future fiscal deficits. If the average real interest rate exceeds the rate of real economic growth, debt service will grow as a share of GDP, so that public debt will threaten to outrun the economy’s ability to service it. The debt burden can be described as “unsustainable” in this case.In countries experiencing high inflation, the value of government bonds erodes rapidly if nominal yields are less than the rate of inflation; voluntary demand for such bonds is limited. Governments are often tempted to coerce banks (directly or indirectly), and even the public, to hold these bonds, but such actions can severely damage the government’s credibility.
  • External borrowing. Governments can finance deficits abroad by issuing bonds to nonresidents or through direct borrowing from foreign banks, governments, or international organizations. For many developing (and some transition) economies, overborrowing in the past and a lack of creditworthiness severely limit this source of financing for the present. Even when available, foreign credit from commercial sources carries interest rates that tend to be high.

Box 2.1.Debt Neutrality1

According to the debt neutrality or “Ricardian equivalence” hypothesis, borrowing is no more than deferred taxation. Specifically, for given government expenditure, a reduction in current taxes will clearly raise thebudget deficit and hence borrowing. Insofar as the private sector recognizes that increased government borrowing today means higher taxes in the future, it increases private saving in order to provide for increases in taxes in the future. The Ricardian equivalence hypothesis in its pure form stipulates that a shift from tax financing to debt financing does not change total national saving, because the initial reduction in government saving will be fully offset by an increase in private sector saving. Thus, a tax cut financed by government borrowing does not reduce the tax burden; it only postpones it. Nor does it stimulate spending on a net basis.

The debt neutrality hypothesis gets limited support from available empirical evidence. In industrial countries, there seems to be a tendency for a partial offset, but not a full offset, of a decrease in government savings. In developing countries the hypotheses finds no support.


This is also sometimes referred to as the Barro-Ricardo debt neutrality theorem. See Robert J. Barro, “The Ricardian Approach to Budget Deficits,” Journal of Economic Perspectives, Vol. 3 (Spring, 1989).

If fiscal accounts are presented on an accrual basis, another form of financing enters the picture: payments arrears. In some economies, governments have incurred large arrears, resulting in cash deficits that are significantly lower than deficits measured on an accrual basis. If the fiscal accounts are on cash basis, arrears usually do not appear explicitly but can still constitute a form of finance. If arrears become sizable and protracted, they directly undermine faith in a government’s unwillingness or ability to keep its commitments. (Box 2.2).

(4) The sustainability of fiscal policy

In recent years the issue of the sustainability of fiscal policy has attracted considerable attention. It was formerly argued that debts of government were riskless as long as the creditor would accept payment in domestic currency, since the government could always print enough currency to meet its obligations. The experience of countries with protracted, large fiscal deficits in an environment of international capital mobility has demonstrated that the earlier view is no longer sufficient.

If a government uses note issue or other money creation to finance a large, recurring deficit, inflation may result. Because money loses value rapidly during high inflation, money holders will prefer to hold as little as possible, and the resulting increase in velocity will add to the inflation rate. In an environment of accelerating inflation, the change in inflationary expectations will have an effect on the government’s ability to borrow from the nonbank public. Potential bond holders will insist on a positive expected rate of return, and experience suggests they will require a large safety margin in the nominal yield in case the erratic inflation rate should jump higher. Bond holders appear to insist on such a cushion even when the maturity of government debt is quite short—for example, three months. At any rate, this interpretation is consistent with yields that have reached the range of 15—40 percent in real terms in Turkey and in Latin America in the 1980s.

Box 2.2.Expenditure Arrears1

The phrase “in arrears” means to be overdue in the discharge of a financial obligation to suppliers or creditors. Since fiscal accounts typically are kept on a cash basis, government expenditure arrears cause an underestimate of spending and of the size of the fiscal problem facing a country. Since arrears are a form of forced lending, the government’s borrowing requirement is also understated, leading to a distorted picture of the sources of credit expansion in the economy. While deficit financing can allow the government to absorb more of the economy’s resources than would otherwise be possible, this initial effect may be offset as the rest of the economy responds by raising suppliers’ prices or holding back payments for taxes and fees. Ultimately, expenditure arrears raise the cost of providing government services.

The accumulation of government arrears may also have an adverse impact on the private sector’s confidence in the soundness of government finances. Private consumers and investors may anticipate an increase in the tax rate, higher inflation, or a general worsening of the financial situation in the medium term. Arrears may spread throughout the economy as a result of government arrears, with severe consequences for the stability of the financial and production systems and prospects for economic growth.

Arrears are often caused by an overly optimistic revenue forecast or a lack of proper mechanisms for monitoring or controlling government spending. A well-functioning treasury can therefore play an important role in preventing the emergence of payment arrears.


Adapted from Chu and Hemming, Public Expenditure Handbook.

Thus, the nominal rates of interest required to make government securities attractive on financial markets may be very high, especially in times of marked changes in the observed inflation rate, and especially in an environment of capital mobility Domestic securities have to compete with foreign bonds, on which the real rate of return is more certain and capital gains, if they occur, are more likely to be positive than negative. (The currency of the high inflation country is more likely to be devalued, in real terms, than revalued.)

The high nominal rates of interest required to make domestic government debt attractive will tend to add significantly to government interest payments and to the size of the deficit to be financed. This adds to the quantity of bonds that must be sold as an alternative to more inflationary central bank finance. The ratio of government debt to GDP will rise, perhaps rapidly. At some point financial markets form the view that the government may become unable to borrow sufficient funds to pay the interest that is due without unleashing enough extra inflation to wipe out the risk cushion incorporated in current nominal interest rates. Securities will no longer be marketable. As soon as this situation is foreseen (perhaps long before it occurs), a full-blown financial crisis will take place.

Economic theory does not provide a straightforward answer to the question, what size of fiscal imbalance is sustainable? Instead of a concrete number, there is broad agreement that fiscal policy is not sustainable if the present and prospective fiscal stance results in a persistent and rapid increase in the ratio of government debt to GDP. A constant debt ratio is therefore one concept of fiscal sustainability. (Obviously, whether the ratio is large or small must also matter, although it is not clear how to incorporate this information analytically).16

To see how this notion has been useful in the development of an operational indicator of sustainability, consider the government’s budget constraint. It can be shown17 that the budget constraint can be written as



dt : debt-to-GDP ratio in period t,

pd : primary fiscal deficit as a fraction of GDP,

r : real interest rate on government debt expressed as a fraction,

g : rate of growth of real GDP expressed as a fraction, and

s : seigniorage as a fraction of GDP.

The expression for the change in the debt-GDP ratio (equation (2.7)) implies the following points:

  • The change in the debt-GDP ratio is determined by the primary deficit, seigniorage, and the built-in positive or negative momentum of the debt-GDP ratio. When interest rates exceed the growth rate, the debt ratio will tend to rise by feeding on itself since interest payments add more to public debt than growth adds to GDP, unless the primary deficit is kept below what can be financed by seigniorage. Put differently, when interest rates exceed the growth rate, it becomes unsustainable for the government to run a permanent primary deficit in excess of revenues that can be raised through seigniorage. The more the policy adjustment is delayed, moreover, the higher the debt-GDP ratio will be and the lower the room to maneuver for the government.
  • If, on the other hand, the real interest rate is lower than the growth rate, then the country can grow out of its debt. Alternatively, the country can afford to run a primary deficit permanently in excess of the desirable level of seigniorage without making the fiscal imbalance unsustainable. It does not mean that there is no constraint on the debt-GDP ratio. If the country increased its borrowing substantially, it would risk raising interest rates to a point where they exceed growth rates. This would convert sustainable policies into unsustainable ones.
  • Whether fiscal policy is sustainable depends not only on factors that the fiscal authorities control, such as revenue and spending, but also on other factors such as the interest rate on government obligations, the long-run growth rate of the economy, and demographic trends.
  • The proper measure of debt for the government budget identity is net rather than gross debt. Net debt comes close to measuring the net worth of the government (assets minus liabilities), although it records financial assets at book value and may not adjust for unfunded liabilities and nonfinancial assets.
  • The government budget identity can be used to calculate budget targets that would achieve specific debt objectives, such as stabilizing or making specific reductions in the debt-GDP ratio. It can be shown that stabilizing the debt-GDP ratio requires that the ratio of the deficit to GDP, inclusive of interest payments on the debt, not exceed the initial debt-GDP ratio multiplied by the growth rate of nominal GDP (see Appendix II).

(5) Analysis of revenues

Tax elasticity and buoyancy

The elasticity of a tax is defined as the relative change in revenues from that tax compared with the relative change in the tax base, holding the structure of the tax (tax rates, brackets, coverage, exemptions, and deductions) unchanged. It can be written as

If GDP is taken as a proxy for the actual tax base, then elasticity with respect to GDP is


AT : is the tax receipts from an unchanged tax system,18 and

Δ the change during a period.

Elasticity provides a tax system with built-in flexibility. A tax system is elastic when it has an elasticity value greater than one, suggesting that tax revenues are increasing at a higher rate than GDP without new taxes or increases in tax rates—that is, with no discretionary change in tax policy. Elasticity greater than one may be desirable in a tax system and should be encouraged in countries in which government expenditures tend to increase more rapidly than GDP. The tax system is likely to be elastic with respect to GDP when taxes are levied on growing economic sectors and when tax rates are progressive. The tax system will be ineslastic if taxes are specific rather than ad valorem,19 and/or when taxes are not collected promptly. This last point is especially important in the case of high inflation, when an unduly long lag between the assessment and collection of taxes will erode the real value of tax revenues.

An elastic tax system is useful from the point of view of economic growth, which generally calls for a sustained rise in spending on social and economic infrastructure and maintenance. If these increases in expenditures are not accompanied by rising revenues, they can lead to undue reliance on deficit financing, either external or domestic. With an elastic tax system, there is usually no need for the frequent and unanticipated tax changes that can adversely affect anticipated real wage and property income.

The buoyancy of a tax is defined simply as the increase in the revenue collected compared with the relative increase in GDP. The change in revenue includes any effects of changes in the tax system, including discretionary changes in the tax structure. The algebraic expression of the formula for tax buoyancy is

In the case of buoyancy, the numerator, ΔT, measures the change in actual tax revenues over the period; in the elasticity formula, ΔAT measures the change in tax revenues adjusted for the estimated impact of changes in the tax system over the period (that is, excluding the impact of all discretionary changes). The elasticity of a particular tax or tax system will usually tend to be approximately constant over time, but there is no presumption that buoyancy will remain constant. (See also Chapter 7.) If the changes in the tax system are revenue enhancing, then buoyancy will exceed elasticity, because the actual tax revenue will exceed the amount that would have been generated in the absence of changes in the tax system.

Assessing a tax system

Analysts engaged in reviewing a country’s tax system will want to consider the appropriateness of that system and explore possible directions for improving it. While the primary objective of taxation is to pay for government services and transfers, it has also been used to correct market failures, to achieve external and internal balance, and to help redistribute incomes. A number of criteria have been developed to assess how well a tax system works in generating revenue. The Tanzi Diagnostic Test (summarized in Box 2.3), although largely subjective, can be a useful guide to evaluating a country’s tax system.20

Besides revenue productivity, the analyst may assess a tax system from the point of view of the value of revenues collected relative to GDP. Called “tax effort analysis,”21 this approach relies on intercountry comparisons. However, using the actual tax ratio to judge these efforts can be misleading in cases in which GDP is not an accurate indicator of relative taxable capacity. Taxable capacity is defined as the level of income that would produce revenue comparable to an inter-country average using average tax rates. The amount of tax revenue actually collected as a proportion of the taxable capacity therefore indicates tax effort. Three major factors that can determine capacity have been identified: (i) the degree of openness of an economy; (ii) the level of development and income; and (iii) the composition and distribution of income.

Tax effort analysis, while useful in assessing tax performance, has obvious limitations. It is not a normative measure;a country with below-average “tax effort” may nevertheless find its tax ratio appropriate in light of national preferences. Tax effort analysis is also purely static in nature because it does not take into account changes in tax ratios and tax efforts over time. In the dynamic sense, the elasticity of tax revenue with respect to GDP is often regarded as a more useful indicator.

(6) Analysis of expenditure22

Government-produced goods and services

Government expenditure involves the provision of goods and services and income transfers. The goods and services provided by government are sometimes categorized according to whether they are consumer goods (national parks) or producer goods (support for research and development). However, most government-produced output combines both characteristics—consider education, health, defense, and infrastructure (highways). Some public expenditures—pensions, unemployment compensation, and production subsidies, for example—are not related to the supply of goods and services but represent direct transfers to households or business enterprises.

Four aspects of government spending commonly are considered in an assessment of expenditure:

  • an aggregate spending level and fiscal balance that are consistent with macroeconomic objectives;
  • judicious use of the private sector in the delivery of certain goods and services;
  • analysis of capital and current spending within a program or a sector to ensure a balanced allocation for each; and
  • an analysis of budgetary institutions to ensure that incentives and rules help to control aggregate spending and facilitate efficiency and equity in the composition of spending.23

A macroeconomic adjustment program may require a degree of fiscal retrenchment. The adverse effects of higher rates for existing taxes and a lack of alternative revenue sources may dictate that a substantial share of this retrenchment be achieved through expenditure cuts. Thus, adjustment programs typically have to focus on spending priorities. Governments seeking to reform spending must identify areas in which markets can effectively substitute for public sector involvement, and consider how scarce government-sector resources can best be utilized in those areas where public sector involvement is considered appropriate.

Types of public expenditure

The main economic categories (as distinct from functional categories) of public expenditure are wages and salaries, goods and services, subsidies, interest payments, other transfers, and capital. It may be noted that both interest payments and subsidies are types of transfer payment if the latter is broadly defined. The purchase of goods and services that appears in the fiscal accounts covers inputs to be used in the production of government services: uniforms for policemen, typewriter ribbons and printer cartridges for clerks, weapons for soldiers. See Box 2.3, “Quasi-Fiscal Operations,” for a class of expenditure that tends to be omitted from the accounts.

  • Wages and salaries. Government policies regarding civil service employment and pay have a major impact on the efficiency of government expenditures. Low pay and inadequate salary differentials for skilled and technical staff may contribute to poor recruitment and low productivity in the government sector. At the same time, the widespread practice of using the public sector as employer of last resort may substantially increase wage costs. Recent reforms in several countries have sought to reduce the government wage bill through a variety of measures, including a civil service census and the elimination of “phantom” workers; the elimination of vacancies and temporary positions; hiring freezes; the suspension of employment guarantees; voluntary retirement programs; wage cuts, caps, and freezes; and the most difficult measure, layoffs. Attempts have also been made in some countries to increase salary differentials in favor of senior staff.
  • Goods and services. Although substantial economies are often possible under this heading, it is important to ensure that cuts in goods and services do not compromise the efficient delivery of government services. An important part of current spending on goods and services goes for the operation and maintenance of capital stock. Inadequate spending on supplies can lead to low levels of effectiveness in areas such as education and health. Similarly, inadequate spending on maintenance (for example, of highways) can lead to the rapid deterioration of physical capital. Pursuing a policy that focuses on creating new capacity while allowing existing infrastructure to deteriorate can be costly and counter productive. Similarly, across-the-board cuts in materials, supplies, and services in the context of macroeconomic adjustment programs should be avoided in order not to undermine the effectiveness of government functions.
  • Capital expenditures. Growth-oriented adjustment requires productive government investment combined with policies to correct distortions in relative factor and commodity prices. It is important to ensure that the investment program is well designed and that projects are economically sound because the cost of poorly designed or inefficiently implemented projects can be high. In general, emphasis should be given to government investment that complements and supports rather than competes with market-determined activities. Education, health, urban services, and rural infrastructure are priority areas for government involvement.
  • Subsidies.24Subsidies are defined as government assistance to producers or consumers for which the government receives no compensation in return. Subsidies can take many different forms, including (i) direct payments to producers or consumers (cash grants); (ii) loans at interest rates below the government borrowing rate and with government guarantees (credit subsidies); (iii) reductions in specific tax liabilities (tax subsidies); (iv) the provision of goods and services at below market values (in-kind subsidies); (v) government purchases of goods and services at above-market prices (procurement subsidies, price supports); (vi) implicit payments through government regulatory actions that alter market prices or access (regulatory subsidies); and (vii) maintenance of overvalued or undervalued currencies (exchange rate subsidies to consumer or exporters).

Box 2.3.Quasi-Fiscal Operations1

In some economies, central banks and other public financial institutions are involved in activities—including significant loss-making activities—that give rise to financial transactions with government that have an effect on the true fiscal deficit. Some of the main types of quasi-fiscal operation are:

  • exchange rate subsidies administered through the foreign exchange system;
  • subsidized lending to government, public enterprises, or private entities;
  • unfunded or contingent liabilities such as loan or exchange rate guarantees;
  • support for the exchange rate through central bank interventions in the presence of strong market expectations of a devaluation; and
  • borrowing by the central bank at high interest rates from the public in an effort to sterilize excessive foreign exchange receipts and prevent rapid money growth.

Where these quasi-fiscal operations are significant, their costs should be included in any comprehensive measure of the public sector deficit, for several reasons. In many countries, central and public sector bank losses are so large that they contribute to financial instability. Their existence also means that the conventional measures of government’s fiscal balance are misleading indicators of the role of fiscal operations in the economy. The taxes and subsidies associated with quasi-fiscal operations may have distortionary effects on resource allocation.

Although they are not always easy to quantify precisely, quasi-fiscal operations need to be extracted from the accounts of the central bank and public financial institutions. To the extent possible, quasi-fiscal activities should be transformed into normal budgetary operations—that is, quasi-fiscal taxes and subsidies should be replaced with explicit taxes and subsidies. The long-term objective should be to address the root causes of quasi-fiscal operations, such as lending activities, that are essentially substitutes for explicit subsidies that otherwise would be in the budget, and the lack of a unified exchange system. The legal authority of the central bank may need to be revised to limit the extent to which it can carry out quasi-fiscal operations. Such structural reforms are essential to minimize and eventually eliminate the need for quasi-fiscal operations.


See International Monetary Fund, “Quasi-Fiscal Operations of Public Financial Institutions,” SM/95/65 (Washington: IMF, Fiscal Affairs Department, 1995). Also see International Monetary Fund, “Guidelines for Fiscal Adjustment,” IMF Pamphlet Series, No. 49 (Washington: IMF, Fiscal Affairs Department, 1995).

Whether explicit (direct) or implicit (indirect), subsidies are a major drain on government budgets in some economies. Subsidies are explicit if they are fully reflected in the budget as expenditures, and implicit if they are not. Implicit subsidies may arise as a result of administered prices that are set at levels below or above free market values—for example, for energy—or may support unrealistic interest rates and overvalued exchange rates. Since a significant portion of government subsidies are implicit, budgets do not fully reflect their range and value. Subsidies also affect resource allocation by reducing the flexibility of the economy, and are often an obstacle to structural adjustment. One example of the distortions subsidies can cause is the pricing of energy products below market levels, leading to wasteful energy consumption.

Any assessment of subsidies should take account of the following:

  • Effectiveness. Not all subsidies are bad, but only those that meet given policy goals by transferring a minimum of resources with a minimum of distortions to the incentive system can be called effective. Effective subsidies are also well targeted—that is, they do not spill over onto groups and activities for which they are not intended. If bread is subsidized in general, then government funds are being used to support the rich as well as the poor.
  • Duration. How long subsidy programs last is a serious concern because people change their behavior in order to be included in these programs and may resist being excluded when their circumstances change. It is this behavior that makes many subsidies ineffective over time. While some subsidies should be limited from the outset, effectively implementing a subsidy program requires periodic reassessments of the rationale behind the subsidy, and, if needed, revision, retargeting, or termination of the program.
  • Transparency. The size of a subsidy program and the implied financing requirement should be made explicit in public budgets. Transparency is desirable from both public and private perspectives in order to identify clearly the benefits and costs. As a general rule, governments should aim to identify subsidy expenditures explicitly in public budgets and should to the extent possible provide them as cash grants rather than as procurement, tax, interest, or regulatory supports. Only cash grants provide both government and beneficiary with a clear and explicit picture of the amounts involved. In turn, this transparency provides a basis for judging the affordability and desirability of subsidies.
  • Financing. To foster transparency and accountability, subsidies should be financed through the budget. Attempting to solve financing problems with extrabudgetary instruments such as government marketing boards or parastatal agencies or through the central bank is dangerous. Such methods hide the costs of subsidy from voters and protect recipients from public scrutiny of the special benefits granted by the legislature. This lack of transparency, in turn, encourages unofficial private payments from beneficiaries to administrators or legislators.
  • Selecting a pragmatic approach. Subsidy programs must be consistent with a government’s institutional and administrative capabilities. Implicit (indirect) subsidies may be more difficult to control effectively than explicit (direct) subsidies. To simplify administrative burdens, subsidy programs should be made as explicit as possible.

Public expenditure and the role of the state

Governments often engage in production activities because markets have failed to fill a need. Markets will not provide public goods, such as national defense, law enforcement, or national parks, because these goods are consumed collectively and do not enable producers to charge individual users (or to exclude from service those who do not pay). The theory of public finance provides other examples of market failure leading to a role for government: natural monopolies (supply of water, gas, electricity), externalities (pollution, job training), and investment in projects with extraordinary business risk (the search for a cure for cancer).

The government’s economic role may also include macroeconomic stabilization, redistribution of income, initiative-taking in the case of certain very risky investments, and influencing consumption patterns in useful or fashionable ways (campaigns to urge citizens to stop smoking).25

The growth implications of public investment and its financing26

Public expenditure affects both aggregate supply (production) and aggregate demand (spending). Productive public investment in physical and human capital directly increases the capacity of the economy to produce output. In addition, in many cases public investment is complementary to private investment, so that strategic government investment can result in an increase in the return to private investment. For both reasons, government investment tends to increase the rate of real GDP growth.

The effects on supply may be fairly immediate, such as when a few key infrastructural bottlenecks are removed. More often, however, the returns to public investment are fully realized only in the longer term, especially in such areas as education. In the short run, the public sector competes with the private sector for resources, and public expenditure (whether financed through taxation or borrowing) tends to crowd out private expenditure, including private investment. Many economies are limited in their ability to finance expenditures through either method. In the short run, they have neither the administrative and political capacity to raise taxes nor the well-developed domestic capital markets necessary to support public borrowing, and borrowing from the central bank is disadvantageous because it is inflationary.

In this situation, a government’s attempts to spend more may actually worsen its financing difficulties by triggering higher inflation, inducing taxpayers to delay tax payments and thus reducing the real value of tax revenue. Moreover, an increase in public sector borrowing can raise domestic interest rates, increase the cost of private investment along with government’s borrowing costs, reduce the country’s long-term growth potential, and create a debt burden. Unless the government uses borrowed resources productively, generating the output and income required to raise tax revenues sufficient to finance future debt repayments and interest, the public sector will have to finance the debt burden by cutting services in the future. With international capital mobility, an increase in domestic interest rates can induce large capital inflows, an exchange rate appreciation, and a deterioration in the country’s external competitive position.

d. The structure of the public sector in Turkey27

There are five main parts of the public sector in Turkey: the central government, the extra-budgetary funds, the social insurance system (plus a number of other, generally small, revolving funds), local governments, and public sector enterprises. The budget of the central government includes revenues and expenditures of a number of budgetary funds; these funds are contained in an annex to the budget. Administratively, the budgetary funds are separate from government, but analytically they are part of central government. The budget is referred to as the “consolidated budget” rather than the central government budget, meaning central government plus budgetary funds.28

(1) General government

“General government” refers to the consolidated accounts of the first four of the above-listed subsectors. Public sector enterprises are not included as part of general government and in discussions of Turkey are usually referred to as “state economic enterprises” (SEEs). This grouping is understood to exclude state-owned financial enterprises, which are part of the banking sector or the financial sector.

(2) Extrabudgetary funds

The proliferation of extrabudgetary funds started during 1984–86. The funds were part of a strategy of “fiscal decentralization.” (At the same time there were significant increases in the share of revenue collected by the central government and transferred to local governments, along with some devolution of responsibilities to local government.) The objective of the creation of extrabudgetary funds was to give selected government operations autonomous funding and expenditure control. Extrabudgetary funds received revenues automatically, either through the authority to charge user fees or from a predetermined share of (mainly indirect) tax revenue, such as import duties, and determined their own spending priorities. Originally it was argued that this type of organization offered strong benefits in the form on improved effectiveness and flexibility resulting from the avoidance of long parliamentary delays and the public spotlight of parliamentary budget debate.

The typical extrabudgetary fund has as its objective the implementation and administration of a specific government program—the incentive programs for exports and investment, social and housing programs, highway construction, subsidies to agriculture through above-market minimum output prices and other means, and support of the domestic defense industry. For example, the Public Participation Fund, which started operation in 1985, has as its purpose the financing of infrastructure investment and is also intended to carry out privatization of selected SEEs. Its main source of revenue is the operating income of several dams and the first Bosporus bridge. The main source of finance for the Public Participation Fund is the sale of revenue sharing certificates, which have a fixed maturity but no guaranteed divided and are available to small investors. In addition, the Fund undertakes conventional borrowing, mainly in the form of project credits from abroad.

Over time it seemed that the creation of extra-budgetary funds led to a loss of control over government revenue and expenditure by the executive and legislative branches of government. In 1988, a provision was enacted to retrieve a proportion of the revenue formerly transferred automatically to certain funds. This step provided a measure of control over the funds’ operations since it limited their incomes; they may not borrow except to cover capital expenditure. However, the measure was weakened by the fact that revenue accruing to the funds was not recorded systematically in the fiscal accounts nor reported to the central authorities. In 1993, 63 extrabudgetary funds were brought under central government control (although not on-budget), accounting for an estimated 85 percent of the aggregate revenue of extrabudgetary funds. By 1995, the number of funds had reached 126 including the 63 controlled by government.

(3) Local government

Local government in Turkey is made up of 80 provinces, 15 larger municipalities, and 2,827 smaller municipalities.29 Provincial governments are administered by the central government; the two are separate only statistically and administratively. Municipal governments are independent, on the other hand. They receive a large share of their revenues from the central government, based on predetermined tax shares, and about a third of their income results from local taxes and nontax revenue sources that they control.30 Municipal governments have autonomy in determining their expenditures. The aggregate deficit of local governments has been less than 0.5 percent of GDP in recent years. Local governments can borrow abroad if the loans are guaranteed by the central government. In practice this has occurred only for large infrastructural projects of the largest cities—the Ankara subway, for example.

(4) State economic enterprises

In addition to a major role in the supply of energy, transportation, and telecommunications, services that are regulated or operated by the state in many countries, SEEs in Turkey have a large presence in manufacturing as a result of the import substitution policy pursued until 1980 (see Part I). As part of the reforms of the early 1980s, SEEs were to be better managed, supported less by budget transfers and more competitive. Nevertheless, few changes appear to have been made in these areas. Concrete plans to privatize selected SEEs have largely stalled for 20 years because of technical, legal, and political problems, although some sales have occurred. Cumulative cash receipts from sales of former SEEs amounted to US$2.6 billion by 1995.31

The main weaknesses of the operation of SEEs in Turkey include pressure from government to postpone price increases (though it is difficult to establish the magnitude of actual effects), granting unusually large wage increases during the 1990s (especially following the threat of a general strike in the runup to a strongly contested national election), non-application of bankruptcy provisions in the country’s commercial legal code, and easy bank and central bank credit and transfers from the central government budget. The aggregate SEE deficits in Turkey also reflect the typical failures of public sector producers: overemployment, low productivity, and inadequate management autonomy. In Turkey, the largest loss-makers are firms involved in production of iron and steel, the state coal mines, the state railroad, and the electricity utility, plus SEEs involved in agricultural price supports.

By the 1990s, the value added by SEEs in Turkey amounted to 10 to 12 percent of GDP. By comparison, their borrowing requirement in the late 1980s was about 4 percent of GDP (excluding budgetary transfers), or half the average PSBR for the period. (The SEE deficit role to 8 percent of GDP in 1991 after three successive years of large real wage increases.) In some years, SEEs have resorted to exceptional financing methods—they have postponed dividends, tax payments, and social insurance contributions.

(5) Social insurance and the pension system

The national pension system in Turkey is threatened by demographic trends, as is also the case in other countries in Europe. The average age of the population is increasing while the age of normal retirement is not increasing commensurately, so that the number of persons contributing to the pension system is rising by less than the number of beneficiaries. The Turkish system relies on current contributions to pay for current benefit payments so the underlying demographic trend alone will result in growing deficits. In addition, the provisions of the Turkish system have been made substantially more generous through legislative changes in the 1980s and early 1990s, especially in 1992. The minimum retirement age has been effectively lowered—for some, to the early forties—while the system has been reorganized in a way that makes it vulnerable to abuse. While the level of pension support for a retired individual is not very high—typical retired persons are forced to find at least part-time work—nevertheless, the financial basis of the system has been so weakened that it has become unsustainable. The deficit of the pension portion of the social insurance system amounted to 1 percent of GDP in 1995 and is projected to rise to about 2 percent in 1996, to 5 percent in 2002, and to increase further thereafter.32

There are three parts to the Turkish social insurance system: one branch covers employees in the private sector and SEEs; a second covers civil servants; and a third covers the self-employed, municipal officials, agricultural workers, and housewives.33 In total, about half of all eligible workers are covered (see Section e of Chapter 1). Contribution rates are ostensibly one third of total wage or salary receipts, of which 20 percentage points is to be paid by the employer. However, payment is capped at 1.8 times the minimum wage, which in turn is equal to about one fifth of the average wage. As a result, the actual contribution rate amounts to about 12 percent of wages, not 33 percent, for a worker earning the average wage. Agricultural workers may choose the level of benefits they wish to receive, with contributions determined by benefits; however, they may opt for the minimum benefit initially (qualifying for health benefits immediately) and increase their pension option as they approach retirement age

In general, men work 25 years to qualify for a pension and women 20 years. However, private sector and SEE participants can receive pension benefits after only 14 years of contributions. There is an incentive not to work officially past the earliest qualifying retirement date since contributions are required whereas pension benefits increase very little, if at all. Wages tend to be under-reported until the last five years of employment, the reference period on which the benefit stream is based. Employers have a strong incentive to collude in the under-reporting since doing so reduces their contributions as well. The average cost of benefits has grown from four times the value of contributions to five times because benefits are adjusted for inflation more promptly than are contributions and because benefits are indexed to civil service wages rather than to the price level, so that there were very large real increases in benefit levels during the years 1989–91. While the level of benefits is not generous relative to the beneficiary’s salary prior to retirement, nevertheless, the benefits are generous relative to contributions. There is no coordination by the authorities of social insurance contributions and tax payments.

e. Exercises and issues for discussion

  • (1) The illustrative data given below show that the government’s cash flow is in balance—total receipts equal total payments. Reclassify the information in an analytical way to calculate the conventional deficit and the primary deficit. Show where each item should be classified (that is, as part of revenue, expenditure, net lending, or financing).Assume that the average debt stock in this hypothetical economy was 5,000, so that the average interest rate paid on government borrowing for the period shown was 10 percent. If inflation in this economy was 8 percent in the year in question, by how much would the operational deficit differ from the conventional deficit?
    Cash Flow of the Government
    Tax receipts1,000
    Bank loan receipts550
    Interest receipts on loans to public enterprises100
    Proceeds from sale of bonds to the public250
    Principal repayments by public enterprises on government loans300
    Loan receipts from abroad200
    Other nontax receipts400
    Grants from abroad200
    Wages and salaries1,200
    Purchases of goods and services250
    Loan repayments to the banking system150
    Lending to public enterprises400
    Redemption of maturing bonds50
    Repayment of foreign loans100
  • (2) On the basis of the data in Table 2.4, calculate Turkey’s primary deficit for the period 1991–95. Analyze and compare the developments in the conventional and primary deficits during the period.
  • (3) Using Table 2.4, prepare answers to the following questions:
    • (i) Calculate general government consumption and saving in 1993–95 in percent of GDP. What were the main changes in 1995 relative to the previous years?
    • (ii) How large was the public sector deficit as a percent of GDP in Turkey in these years? What were the main reasons for the deterioration or improvement in the fiscal position during 1993–95?
  • (4) How was the public sector deficit financed in 1995? What were the main changes relative to 1993 and 1994?
  • (5) With reference to the hypothetical data shown below, compute in both absolute terms and in relation to GDP
    • (i) the seigniorage in 1994 and 1995; and
    • (ii) the inflation tax in 1994 and 1995.What explains the change in the observed inflation tax? In your answer,
    • (iii) explain the observed decline in seigniorage; and
    • (iv) show that when the rate of growth of nominal money balances is in line with the rate of inflation, seigniorage will equal the inflation tax.
      (Monetary data in billions of U.S. dollars)
  • (6) Calculate the real growth rate of GDP that would have to occur in Turkey in order that the ratio of government debt to GDP not rise. Do this for 1992 and 1994–95 (the interest-rate datum for 1993 is not available) using equation (2.7) in the text.
    • For nominal interest use the end-year rates on three-month treasury bills given in the table below;
    • for inflation, use the through-the-year increase in the CPI, also shown below;
    • for seigniorage, use the nominal change in “currency issued,” plus “residents’ lira demand deposits” excluding central government, plus “residents’ lira time and savings deposits,” all from Table 4.5, divided by GDP;
    • compute the primary deficit as a percent of GDP from Tables 2.3 and 2.4; and compute debt from Table 7.5 (summing “central government” and “other government”, foreign—in liras—plus domestic).
      Nominal interest rate,
      end-year, three-month,
      treasury bills, in
      CPI, percent change,
      through the year71687112670
  • (7) Comment on the evolution of the Turkish tax-GDP ratio during 1991–95. Is it a good measure of the Turkish authorities’ tax effort? Comment on the structure of revenues and expenditures on the basis of Tables 2.32.4.
  • (8) Calculate the buoyancy of total tax revenues for 1992–95. Could you have deduced the buoyancy on the basis of the observed changes in the tax-GDP ratios over the period? Is the available information sufficient to calculate the elasticity of tax revenues with respect to GDP? If not, why not?
  • (9) It has been argued that the main government contribution to growth is through public investment. With reference to the discussion on the concept of the government current account balance, give examples in which public investment may not contribute to growth and examples of current expenditure that may have significant impact on growth.
  • (10) In forecasting tax revenues, the IMF has often been reluctant to take into account expected revenue gains from planned improvements in tax administration. Is this attitude justified?

Figure 2.1Turkey The General Government Deficit and the Public Sector Borrowing Requirement, 1985-1995

(In percent of GDP)

Source: IMF: Turkey--Recent Economic Developments, 1996.

Table 2.1Government Revenue and Grants
I.Total revenue and grants (II + VII)
II.Total revenue (III + VI)
III.Current revenue (IV + V)
IV.Tax revenue
Tax on income, profits, and capital gains
Social security contributions
Taxes on payroll and work force
Domestic taxes on goods and services
General sales tax or VAT
Profit of fiscal monopolies
Taxes on specific services
Taxes on use of goods
Other taxes on goods and services1
Taxes on international trade
Import duty
Export duty
Other taxes on international trade
Other taxes
V.Nontax revenue2
VI.Capital revenue3
Source: International Monetary Fund, A Manual on Government Finance Statistics (Washington, D.C: IMF 1986).

Including business and professional licenses and motor vehicle taxes.

Including property income, administrative fees and charges, fines, and operating surpluses of public enterprises and the central bank.

Including the sales of fixed assets, sales of stocks, land and other intangible assets and capital transfers.

Source: International Monetary Fund, A Manual on Government Finance Statistics (Washington, D.C: IMF 1986).

Including business and professional licenses and motor vehicle taxes.

Including property income, administrative fees and charges, fines, and operating surpluses of public enterprises and the central bank.

Including the sales of fixed assets, sales of stocks, land and other intangible assets and capital transfers.

Table 2.2.Economic Classification of Government Expenditure and Net Lending
I.Total expenditure and net lending (II + V)
II.Total expenditure (III + IV)
III.Current expenditure
Expenditure on goods and services
Wages and salaries
Employer contributions to social security and pension
Purchases of goods and services
Interest payments
Subsidies and other current transfers
To nonfinancial public enterprises
To financial institutions
To nonprofit institutions, household, and abroad
IV.Capital expenditure
Acquisition of fixed capital assets
Purchase of stocks and land
Capital transfers
v.Lending minus repayments
Source: International Monetary Fund, 1986, A Manual on Government Finance Statistics (Washington.: IMF).

Subsidies include all unrequited, nonrepayable transfers on current account to private industries and public enterprises and the cost of covering the cash operating deficits of departmental enterprise sales to the public.

Source: International Monetary Fund, 1986, A Manual on Government Finance Statistics (Washington.: IMF).

Subsidies include all unrequited, nonrepayable transfers on current account to private industries and public enterprises and the cost of covering the cash operating deficits of departmental enterprise sales to the public.

Table 2.3.Turkey: General Government Tax Revenue, 1991-95(In billions of Turkish liras)
19911992199319941995 1/
I. Central government tax revenue78,644141,602264,273587,7061,084,311
Direct taxes41,09471,393128,324283,733460,437
Taxes on income40,41870,134125,793278,074435,999
Personal income tax33,35560,056106,661181,884329,795
Corporate income tax7,06310,07819,13243,976103.241
Additional Taxes52,2142,963
Taxes on wealth6761,2592,5315,65924,438
Indirect taxes37,55070,209135,949303,973623,874
Taxes on domestic goods and services24,67947,34189,447214,353429,232
Domestic VAT14,54127,05350,892110,918212,119
Supplementary VAT (excises)6051873888,02916,937
Petroleum consumption tax2,3706,76912,79146,625103,180
Financial transactions tax2,1193,9227,12916,46725,340
Stamp duty2,4574,1537,97113,67729,197
Other indirect taxes2,5875,25710.27618,63742,459
Taxes on imports12,86422,84846,21389,596194,609
Customs duties1,0361,71513,17121,84248,433
VAT on imports8,29115,03430,98565,824142,861
Other duties and levies3,5376,0992,0571,9303,315
Abolished Taxes7202892433
II. Local Government and EBF Taxes27,74453,87895,062148,172297,555
Direct Taxes6,41113,30722,18048,84797,984
Indirect Taxes21,33340,57172,88299,325199,571
Total General Government Taxes106,388195,480359,335735,8781,381,866
Memorandum item:
Gross domestic product630,1171,093,3681,981,8673,868,4297,554,758
Source: IMF, Turkey-Recent Economic Developments, 1996 and 1997.
Source: IMF, Turkey-Recent Economic Developments, 1996 and 1997.
Table 2.4.Turkey: General Government and Public Sector Operations, 1991-95(In billions of Turkish liras)
1991 1/1992 1/199319941995
Total revenues119,764215,335518,2231,034,5881,899,903
Nontax revenues13,37619,855158,888298,710518,037
Total expenditures and net lending165,020332,348762,7641,339,6292,428,083
Current expenditures112,095253,022663,5731,193,1022,140,366
Current transfers41,13189,700204,778358,415603,987
Of which: social security14,000108,940200,398376,882
Interest Payments26,84045,307125,984326,387653,288
Quasi Fiscal Outlays and
extrabudgetary transfers 2/27,58664,10757,07981,453
Other current expenditure44,12490,429268,704451,221801,638
Wages and salaries217,192346,490592,796
Purchases of goods and services45,23670,181164,432
Subsidies, n.i.e.6,27634,55044,410
Capital expenditures51,21977,89198,514146,102290,660
Capital transfers18,48519,70211,21629,51375,380
Net lending 2/1,7061,435677425-2,943
Balance (revenue - expenditure)-45,256-117,013-244,541-305,041-528,180
Financing of general government balance45,256117,013214,541305,041528,180
Consolidated central government1,9214,03821,062-67,174-79,600
Local governments5303,6926,3943,72116,974
Domestic borrowing42,742104,424213,482350,883551,893
State economic enterprises (SEEs)
Total public sector borrowing
Financing of public sector borrowing requirement67,779151,745299,145364,587469,704
Domestic borrowing64,025136,743275,502403,856495,615
Source: IMF, Turkey—Recent Economic Developments, 1996 and 1997.

The consolidation methodology among parts of general government has been modified for the period 1993-95 in order to present all operations on a gross basis. The data for 1991 and 1992 are not strictly comparable with subsequent years.

Budgetary data have been modified to include certain quasi-fiscal outlays and off-budget transfers to entities outside the general government.

Source: IMF, Turkey—Recent Economic Developments, 1996 and 1997.

The consolidation methodology among parts of general government has been modified for the period 1993-95 in order to present all operations on a gross basis. The data for 1991 and 1992 are not strictly comparable with subsequent years.

Budgetary data have been modified to include certain quasi-fiscal outlays and off-budget transfers to entities outside the general government.


In feudal times, it was the right of the ruler of an estate (the “seigneur” or lord) to mint coins for general circulation. The ruler might keep the amount of precious metal in the coins high, so that money was literally “worth” the value stamped upon it; the coins could be melted down and the metal sold for equal value. Alternatively, the ruler might use a smaller proportion of precious metal, with base metals (like iron) substituted for gold or silver. Such coins were said to be “debased”. In this case, the cost of producing coins was less than the face value, and the difference was called “seigniorage”. Since the ruler normally spent the new money first, in exchange for goods and services or to pay debts, seigniorage constituted a source of implicit revenue to the estate, like tax receipts.

The larger part of “fiat money” in circulation in modern economies is paper currency and is devoid of any promise to be redeemable in precious metal (it has no “gold backing”). Fiat money is accepted in exchange for merchandise, for example, or labor, not because of its intrinsic value but because the state has declared that all residents are obliged to accept it in settlement of financial claims against one another, and also because people are accustomed to using it. In the case of paper money, the cost of producing the medium of exchange is negligible, and seigniorage per unit of currency is high. Modern coins contain little precious metal and therefore also yield an amount of seigniorage. Besides currency and coins, checking-account balances in commercial banks serve as money. Here, too, there is a large element of seigniorage since, in a fractional-reserve banking system, the marginal cost of producing additional deposit balances is approximately zero.1 (The average cost of money creation may be higher than zero; this is because the cost of central banking functions, such as bank supervision, can be attributed to creation of bank-deposit money.) Despite possible costs of money creation, the value of seigniorage in moderm economies can be substantial. In low inflation countries, the portion of seigniorage accruing to government is on the order of 0.5 percent of GDP per year; in high inflation countries and economies undergoing rapid structural change, it can amount to more than 5 percent.2

The seigniorage resulting from money growth in a fractional-reserve banking system does not all accrue to government. Increases in bank reserves, plus coins and currency issued,3 cause seigniorage gains directly for the central bank: the central bank can create banknotes and reserve deposits at little or no cost and exchange them for other assets. Ultimately the financial gains resulting from this process tend to be passed along to the central government. The financial benefit to the central bank can be viewed as a process whereby the central bank purchases assets (such as loans to government and foreign exchange) with mere paper or bookkeeping entries (the currency issue or commercial bank reserves), so that the net worth of the central bank is increased with the stroke of a pen. Alternatively, seigniorage benefits may be seen as the acquisition by the central bank of higher-interest-earning portfolio items in exchange for lower-interest (including zero-interest) items. In the end, these two views of seigniorage can be shown to amount to the same thing.

Increases in broad money over and above the increase in currency and bank reserves represent increases in demand and time deposits of commercial banks and cause seigniorage gains for these banks. Like the central bank, commercial banks exchange low-interest liabilities (deposits) for higher interest assets (loans to their customers). In a competitive system, the seigniorage benefit to banks tends to be shared with the deposit-holding public and borrowers. Potentially, seigniorage results in subsidized banking services or higher deposit rates than would occur if banks did not accept deposits (if banks could raise funds, say, only by selling securities in financial markets).

The value of seigniorage to government can be seen as a way of formalizing the analysis of what happens when a government uses newly-printed money or central bank advances to finance a deficit in its accounts. To the extent that the government pays lower than market interest on its central bank loan, the seigniorage value is transferred from the central bank to the government in this form. The central bank created money and spent it on a non-remunerative asset; thus, it neither gains nor loses. The government, however, will spend the loan proceeds on goods and services, civil servant salaries, transfers, or debt repayment. Clearly, the government benefits compared, say, to borrowing from commercial banks.

If, on the other hand, the central bank charges the government a market-related rate of interest, the profits of the central bank will be higher for the period, and therefore the amount that it remits to the government in the form of nontax revenue (or corporate profits) will also be higher, other things being equal. The central bank may not be obliged to transfer all of its profits to government in the immediate run; it may save some funds in reserve for anticipated expenses or losses from other activities. However, seigniorage tends to be passed along to government eventually, irrespective of the interest rate charged by the central bank.

Once in circulation, the newly created money may meet an increasing demand for means of payment. If aggregate output is growing, and the increase in the money supply is in line with the growth in economic activity, the new money may be added willingly to the wallets and bank accounts of enterprises and households. This change in the demand for money holdings, in real terms, is called “pure seigniorage”. It means that the government can spend the newly created money without risk of causing inflation because the economy needed an increase in the money stock to accommodate an expanding value of transactions.

However, if the increase in the money stock is large (compared to the growth in output, or to other exogenous factors influencing money demand), there will be an additional effect. The new money will stimulate extra spending, resulting in increases in the quantity of output (GDP) and/or the average level of prices. To the extent that the money increase is inflationary, it reduces the buying power of money holdings throughout the economy. In this way, money holders lose part of their wealth to government in the competition to acquire goods and services from the limited amount that can be produced per unit of time. The seigniorage gains realized by the government are matched by “revenue” from an “inflation tax” on money holders. Total seigniorage is equal to the sum of pure seigniorage and the revenue from the inflation tax on money holders.

It is possible to formulate this analysis more precisely.4 In the following derivation, if M is interpreted as representing the stock of reserve money, then the resulting expression for seigniorage, S, refers to the amount potentially receivable by government; if M stands for broad money, then S reflects seigniorage resulting from activities of the banking system as a whole. In either case, the change in money that enters the expressions given below should be adjusted by subtracting any costs associated with increasing the money supply if they are substantial (notably, the present value of interest paid on deposit balances in commercial banks). To simplify the algebra, it is assumed that money creation is costless.

Define ΔMt, as the change in the stock of money from the end of period t-1 to the end of period t: ΔMt = Mt - Mt-1. Seigniorage is usually defined as the real change in the money stock, St = ΔMt / Pt. One may decompose the term on the right-hand side into two parts:

Add and subtract the ratio Mt-1/Pt-1 on the right-hand side, and rearrange:

Let ∏t, represent the rate of inflation expressed as a fraction, defined to equal ΔPt/ Pt-1. The second bracketed term in the above expression may then be written, in terms of ∏, as [1 - 1/(1 + ∏t)]. After rearranging, this expression becomes,

If m is used to denote M/P, the stock of real money balances, and it is recognized that ∏ ͌ ∏ (1 + ∏) for a relatively low rate of inflation, one may rewrite the above expression as,

which is how it is usually encountered.

The first term in this expression represents pure seigniorage, the desired change in real money balances by all holders of domestic money.5 If economic agents have a larger real value of transactions to finance, for example, they will need a larger real money stock to accommodate the increase; if they expect future inflation to reduce the purchasing power of their money balances, they will find ways to economize on these holdings and desired m will decrease. The second term in (2.A.2) represents the inflation tax on money holdings and is (approximately) equal to the rate of inflation times the level of money balances; to the extent that inflation reduces the buying power of households’ and enterprises’ liquid balances, these economic transactors must spend less in real terms. In an indirect way, resources to make consumer and goods and services are taken away from households and private-sector producers by the effect of inflation on their real money balances and diverted to produce government services or real transfers instead.6

There are two ways to view the impact of seigniorage on the fiscal accounts. If the government borrows from the central bank, one way or another these loans will be interest-free since central bank profits are remitted to government.7 In this view, the seigniorage benefit to government takes the form of a flow of reduced interest expenditures that stretches indefinitely far into the future.8

The other interpretation is to view seigniorage as a loan to government that will not be repaid (or as a grant, or a windfall gain to the institution that has been given the monopoly power to create domestic money). The government obtains currency or increases in its transactions account at the central bank, and it spends these funds; somewhere below the line in its accounts there will be a financing item in the same amount—a loan, possibly central bank advances, and/or simply a reference, “From notes and coin”. Seigniorage thus substitutes for the sale of government securities. It shows up in the fiscal accounts in the sense that a central-bank-financed deficit equal to some (small) percent of GDP is the economic equivalent of fiscal balance.

Thus, in the first view seigniorage is a flow, a stream of subsidies on interest payments on government borrowings or a stream of higher non-tax revenues because of larger central bank profits remitted to government. In the other view, seigniorage is a one-time financial grant or loan or benefit. There is a rate of discount that will equate the “present value” of the stream of reduced government net interest payments to the stock of resources obtained by government in the present period from the increase in reserve money. In this way the two views come together.

If inflation accelerates in the economy because of a high rate of money creation, the inflation-tax portion of seigniorage increases—the second term in equation (2.A.2) becomes larger. However, at higher inflation rates the desired level of real money holdings in theory will fall—the first term in (2.A.2) decreases and may turn negative. Since total seigniorage will be the sum of these two effects, one growing and one declining, it is reasonable to ask whether there is an optimal inflation tax from the point of view of generating fiscal “quasi-revenue.” The expression above does, indeed, give rise to a U-shaped curve with a maximum value, although the inflation rate corresponding to the highest point on the curve varies among countries depending on monetary institutions, expectations, and aversion to holding money in times of inflation. There is a general presumption that with high foreign currency substitution, the optimal rate of the inflation tax is forced down; money holders will switch to foreign currency holdings to avoid the tax.9


If the central bank pays interest on the reserve deposits of commercial banks, then the marginal cost of money creation is increased by the interest payments on bank reserves. Typically, however, interest on these deposits is zero.


For seigniorage calculations for industrial, transition, and developing countries over various time periods, see Paul R. Masson, Miguel A. Savastano and Sunil Sharma, “The Scope for Inflation Targeting in Developing Countries,” IMF Working Paper No. 97/130 (Washington: International Monetary Fund, October, 1997).


This sum is referred to as “reserve money”, “high-powered money”, or the “monetary base” (“base money”). See Chapter 4.


This analysis of seigniorage is due to Phillip Cagan, “The Monetary Dynamics of Hyperinflation,” in Milton Friedman, ed., Studies in the Quantity Theory of Money (Chicago, Illinois: The University of Chicago Press, 1956), p. 77 ff.


We may impose this interpretation on the algebra because ultimately money demand must adjust to equal supply. See alsoChapter 4.


The seigniorage definition given here is widely used for the simple case in which it is assumed that there is a zero cost of producing additions to the money supply and an instantaneous period of adaptation to changes in the rate of money creation.

Expression (2.A.2) is not a behavioral relationship, but a decomposition of the definition of total seigniorage into two parts; it holds for any specified change in the nominal money supply. To extend the analysis to incorporate economic behavior, one can substitute appropriately from a money-demand equation for the first term on the right-hand side of (2.A.2). This will indicate how a specific increase in the supply of money will be divided between the two parts of seigniorage based on the factors determining money demand-—such as the level of real activity and the inflation rate.


This is true even if the government does not borrow directly from the central bank. If the central bank buys government securities on the secondary market—from commercial banks, or from the non-bank public—it is implicitly lending to government. The interest income that it receives from holding treasury bills will ultimately be refunded to government itself Suppose the government did not borrow from anyone. To increase the money supply, the central bank could buy corporate securities. But the income from these assets would be remitted to government, and thus “net interest payments” of government would decrease even in this case.


Analogously, for commercial banks seigniorage takes the form of a low interest rate, and therefore low interest income, paid on deposits relative to the interest rate received on loans. Sometimes, in fact, seigniorage is characterized as the result of an “interest-free loan” on the part of money holders to money creators (the banking system). However, it is not obvious that money holders get nothing in return for their implicit loans to money creators; households and enterprises acquire a medium of exchange and a store of value, which doubtless have positive utility.


Extensions of the basic idea of seigniorage and the fiscally optimal inflation rate in the presence of foreign currency substitution and other complications are surveyed in Pierre-Richard Agénor and Peter J. Montiel, Development Macroeconomics (Princeton, N.J.: Princeton University Press, 1996), p. 111 ff.

The Accounting Approach to Fiscal Sustainability

In a technical survey for the World Bank, John Cuddington quotes Max Corden’s quip: “The growth of a child is not sustainable, but is desirable nevertheless.”1 The issue of fiscal sustainability—whether a deficit means trouble in the longer run, even if not in the short run—is one that economic theory can address only partially and obliquely. Probably the governments of most countries do not intend that large deficits will recur indefinitely, whereas the formal analysis of sustainability focuses on this case. One approach to sustainability studies whether the ratio of government debt to GDP is increasing. This approach is based on the assumption that a continuing increase in this ratio will eventually undermine the confidence of lenders in the government’s ability to service its obligations, while stability of this ratio seems more manageable, at least through the medium term. This is sometimes called the “accounting approach” to sustainability, and the two key constraints that follow from it are derived in this appendix.

The other approach, called the “present-value constraint,” is not presented here. It is based on the comparison between the size of the present debt and the governments’ ability to repay in the long run if expenditure and revenue remain at current proportions of GDP. This last assumption may be somewhat artificial—as illustrated by the studies in the early to mid-1990s that claimed to demonstrate that the U.S. deficit was unsustainable (see Cuddington).

The accounting identity for financing a fiscal imbalance was given in Section b(5) as

The change in government net indebtedness can be defined as gross new borrowing net of amortization payments and net of the change in government bank deposits and cash balances. The change in debt defined in this way will equal the conventional fiscal balance (with sign reversed) after allowing for seigniorage gains and miscellaneous financing sources such as gold sales or privatization receipts, if any. To simplify the algebra that follows, let seigniorage be combined with miscellaneous financing items. Also, let the algebraic signs of both sides of this equation be changed (on the left-hand side, a deficit is a positive number, to match positive borrowing on the right). Using the notation,

Dt: net government debt at the end of period t,

St: seigniorage accruing to government during period t (plus miscellaneous financing items, if any), and

CDt: the conventional fiscal deficit for period t as defined implicitly in Section b(5),

and where Δ Dt - Dt-1 , the financing identity above may be written

Adding ΔDt=Dt-1 to both sides, and rearranging, yields

which becomes, after dividing by GDP,

Let lowercase letters denote the ratio of a variable to GDP in the same period (for example, st = St/GDPt). If one also defines,

gt : the growth rate of real GDP, and

t : the rate of inflation measured by the GDP deflator

(both expressed as fractions, not in percent, to simplify the algebra), then GDPt/ GDPt-1 = (1 + gt)(1 + ∏t), and the above expression may be rewritten as

It follows that the change in d is given by

If the rate of nominal GDP growth is low, the value of the bracketed factor in (2.A.5) does not change much if the denominator is taken to equal unity approximately. This is the basis for the statement in Section c(4) that stabilizing d requires “the ratio of the deficit to GDP, inclusive of interest payments on the debt, not exceed the initial debt-GDP ratio multiplied by the growth rate of nominal GDP” (p.20). This statement assumes that the deficit is equal to CD-S (or that S equals 0). This result is an approximation and involves a significant error if the growth rate of nominal GDP is larger than a few percent per year, whereas equation (2.A.5), above, involves no approximation.2

It is possible to disaggregate the conventional deficit in (2.A.4) into the primary deficit (PD) and interest expenditure. Utilizing the assumption of a lag of around six months between loan disbursal and interest payment, one may rewrite cd in (2.A.5) as

where i is the nominal interest rate paid on average in period t on government debt. If rt is defined to be the average real rate of interest on government debt, so that

then (2. A.6) can be written as

and (2.A.5) as

by collecting terms on dt-1. After simplifying, the bracketed factor becomes (r-g)/(1+g), so that the precise result is,

The assumption that g is small makes the bracketed term in (2.A.7) roughly equal to (r - g). Then (2.A.7) reduces to the version presented as equation (2.7) in Section c(4) of the chapter.


John Cuddington, “Analyzing the Sustainability of Fiscal Deficits in Developing Countries,” Policy Research Working Paper 1784 (Washington: The World Bank, June, 1997) p.l.


The statement in the body of the chapter is based on calculus and applies to an indefinitly small change in d, in which case the error of approximation is negligible.


The government finance statistics accounting framework, which was developed to assist in the compilation of fiscal accounts, is set out in A Manual on Government Finance Statistics (Washington: International Monetary Fund, 1986).


For an elaboration of these criteria for distinguishing government sector activities from nongovernment, see the GFS Manual, Chapter 1.


There are two legitimate uses of off-budget accounts: as trust funds (money held on behalf of nongovernment entities) and advance or suspense accounts (funds held against future payments).


In the fiscal accounts of Turkey, social insurance contributions have been included only in recent years and have been counted as nontax revenues.


The GFS Manual places grants in a line by themselves, separate from revenues. This treatment recognizes that the government has little control over the amount or sustainability of grants, unlike other receipts.


Loans made by government to a foreign government for public policy purposes are also included in net lending and thus affect the size of the deficit. The borrowing country, however, treats the loan as foreign borrowing and includes it below the line.


It is important to distinguish between incurring debt and merely guaranteeing it. Guaranteed debt is not included as a government transaction even though it constitutes a contingent liability of government.


In finance, this means that a realistic value of a bank’s assets is so low, relative to the value of its liabilities, that the initial contribution of cash by the bank’s owners (“capital”) is not large enough to provide a prudent cushion against possible negative values of the difference, total assets minus total liabilities. The bank’s “net worth” is small or negative, and the bank is insolvent or nearly so.


See Ke-Young Chu and Richard Hemming, eds., Public Expenditure Handbook (Washington: International Monetary Fund, 1991).


For a discussion of the intertemporal shortcomings of the conventional deficit, see Mario I. Blejer and Adrienne Cheasty, “The Measurement of Fiscal Deficits: Analytical and Methodological Issues,” Journal of Economic Literature, Vol. 29 (December, 1991).


See Blejer and Cheasty, “Measurement of Fiscal Deficits.”


Such conventions also differ sharply among countries, making intercountry comparisons of public investment quite difficult, even for the major industrial countries.


Generally total interest payments are subtracted from total expenditures to calculate the primary balance. However, conceptually, only the net interest payments by the government (net of interest receipts) should be subtracted.


In practice, reconciling fiscal and monetary data on financing gives rise to a number of problems, including (i) timing; (ii) the definition of government; (iii) the treatment of noncash transactions (such as debt assumption by the government) that is not reflected in fiscal accounts but is covered in monetary statistics; (iv) discrepancies in coverage between bank and nonbank financing; and (v) errors in data.


“High-powered money” refers to liquid liabilities of the central bank, and includes commercial bank deposits. In a fractional reserve banking system, a change in high-powered money tends to result in a much larger change in the total money supply. See Chapter 4.


A country with a low but rapidly rising debt ratio might appear more likely to run into trouble than a country with a larger but slowly diminishing ratio.


AT is usually a data series that is not observed. It is derived from the actual tax revenue series T “by adjusting for the effects of changes in the tax system that occurred during periods under consideration. The resulting data series shows the revenues that would have been collected if all tax rates, exemptions, brackets, and other elements of the tax system had remained constant over the period. Any changes in revenues from year to year would have been caused by increases or decreases in the tax base. See Chapter 7.


Specific taxes are those expressed as a fixed currency value per unit—for example, TL10 per pack of cigarettes; ad valorem taxes are expressed in terms of percent of the nontax price—50 percent on cigarettes. If a country has zero inflation, revenues will be maintained under either form of tax, but with inflation the real value of revenue of specific taxes will decline.


For a more detailed analysis of tax policy, see Tax Policy Handbook, Parthasarathi Shome, ed. (Washington: International Monetary Fund, 1995).


For a detailed review and discussion of the “tax effort” approach, see “Tax Ratios and Tax Effort in Developing Countries, 1969-71,” by Chelliah Raja, Hessel Baas, and Margaret Kelly, 1975 Staff Papers, Vol. 22, (Washington: International Monetary Fund, March), pp. 187-205.


This section draws on Ke-Young Chu, “Public Expenditure Policy: An Overview of Macroeconomic and Structural Issues,” in Fiscal Adjustment in Eastern and Southern Africa, (Washington: International Monetary Fund, 1994).


For a detailed exposition of these principles of public expenditure analysis, see Sanjay Pradhan, “Evaluating Broad Allocations of Public Spending: A Methodological and Data Framework for Public Expenditure Reviews,” (Washington: World Bank, 1995).


This section is adapted from Gerd Schwartz and Benedict Clements, “Note on Subsidies: Evaluation and Impact,” IMF Seminar on Fiscal Adjustment in Eastern and Southern Africa, September-October, 1994. For a discussion of the issues relating to subsidies, see Benedict Clements, Rejane Hugounenq and Gerd Schwartz, “Government Subsidies: Concepts, International Trends, and Reform Options,” IMF Working Paper WP/95/91(Washington: International Monetary Fund, September, 1995). See also “Social Safety Nets for Economic Transition: Options and Recent Experiences,” IMF Paper on Policy Analysis and Assessment, PPAA/95/3 (Washington: International Monetary Fund. February. 1995).


For a general discussion of the appropriate economic role of government, see Joseph E. Stiglitz, Economics of the Public Sector (New York: W. W. Norton, 1986), especially Chapters 4 and 5.


This topic is explored in more detail in G.A. MacKenzie and others, “The Composition of Fiscal Adjustment and Growth,” IMF Occasional Paper 149 (Washington: International Monetary Fund, 1997).


This section draws on International Monetary Fund, “Turkey: Recent Economic Developments” (Washington: unpublished IMF staff reports, various years), and on recent published reports as noted below.


“Consolidated” in this case is used partly in the conventional accounting sense. That is, transactions between entities covered in the budget document are treated on a net basis. For example, if the central government makes a transfer payment to support one of the budgetary funds, the income of the fund is not added to total revenue of the central government in computing total revenue in the consolidated budget.


International Monetary Fund, Turkey-Recent Economic Developments, IMF Staff Country Report No. 98/104 (Washington: IMF, September, 1998), p. 112.


Municipal governments receive receipts from a property tax, but assessed valuation is controlled indirectly by the central government. Infrequent tax re-assessments of real estate in an environment of high inflation causes large fluctuations in the real value of the revenue flow. Municipalities also receive nontax revenues, especially a kind of monopoly premium on electric power sold to the public. IMF, RED, September 1998, pp. 112–114.


IMF, RED, September 1998, Table 9, p. 91.


The finding of unsustainability has been made by the International Labor Office and the World Bank as well as the IMF and the OECD. World Bank, Turkey: Challenges for Adjustment (Report #150760TU, April 1, 1996), Chapter 3; IMF, Turkey: Recent Economic Developments, IMF Staff Country Report No. 96/122, November 1996, Appendix II, p. 111; OECD, OECD Economic Surveys: Turkey, 1996, p. 58.


The three plans are known by their initials or by an abbreviation of their Turkish names: SSK, ES, and Bag-Kur. Benefits levels are set by government, which also provides some financial support, IMF, RED, September 1998, p. 62 ff.

    Other Resources Citing This Publication