Economists and statisticians have long found it useful to separate the activities of government from those of the rest of the economy because the powers, motivation, and functions of government differ from those of other sectors. Governments have powers to raise taxes and other compulsory levies and to pass laws affecting the behavior of other economic units. They focus on public policy considerations rather than on profit maximization. The principal economic functions of general government are

Economists and statisticians have long found it useful to separate the activities of government from those of the rest of the economy because the powers, motivation, and functions of government differ from those of other sectors. Governments have powers to raise taxes and other compulsory levies and to pass laws affecting the behavior of other economic units. They focus on public policy considerations rather than on profit maximization. The principal economic functions of general government are

  • 1. To provide goods and services to the community on a nonmarket basis, either for collective consumption (such as public administration, defense, and law enforcement) or individual consumption (education, health, housing, and cultural services); and

  • 2. To redistribute income and wealth by means of transfer payments (taxes and social benefits).

When considering general government statistics, economists also look at the broader public sector, because governments often fulfill their public policy objectives through operating public enterprises (for example, railways, airlines, public utilities, and public financial corporations). A government may do so by requiring a corporation to service areas of the economy that would not be otherwise covered and by charging subsidized prices, including lending at low interest rates. As a result, the public corporation operates with a reduced profit or at a loss. Such public policy operations are known as quasi-fiscal activity.

For compiling and analyzing general government and public sector statistics, the Government Finance Statistics Manual 2001 (GFSM 2001; IMF, 2001) provides a comprehensive framework. To a very large extent,34 it has been harmonized with the 1993 SNA (Box 16).

The following sections discuss the (1) coverage of general government, (2) basis of recording, (3) analytic framework, and (4) major GFSM 2001 classifications.

Coverage of General Government

As with the other macroeconomic statistics, the coverage of the general government sector is based on classifying resident institutional units. The first criterion for national data compilers to use for sectorizing government units is “control”—if the government controls the unit, compilers classify it as within the public sector; otherwise it is a private sector unit.

Next, compilers classify the public sector units into the general government sector or the public corporations sector, on the basis of how the units’ output is financed. If the unit primarily sells goods and services at economically significant prices, it is a market producer, and it is classified as a (financial or nonfinancial) public corporation. If the unit does not sell most of its output at economically significant prices, it is a nonmarket producer of goods and services, and it should be classified as part of the general government sector.

Fundamentally, the general government sector covers all institutional units at the central government, state government, and local government levels. It also covers administrative units (budgetary and extrabudgetary), social security funds, and nonmarket nonprofit institutions controlled and mainly financed by government. Therefore, included in the general government sector are such institutional units as extrabudgetary funds, nonprofit institutions controlled by government, and public entities that, although legally having corporation status, do not sell their output at economically significant prices.

Basis of Recording

While governments have used cash transactions as the basis of recording government finance statistics, they have become increasingly concerned about the wider resource implications of fiscal actions. Traditionally, governments have kept their accounts on a cash basis. This tradition is reflected in the analytical framework of the 1986 edition of A Manual on Government Finance Statistics (GFSM 1986; IMF, 1986). The inclusion of only cash revenues and expenditures has the advantage of focusing governments’ attention on their liquidity constraint, which they viewed traditionally as the most important policy issue. However, as the sole basis for assessing fiscal policy, the use of cash transaction recording has brought increasing dissatisfaction (Box 17).35 As governments have become less constrained by liquidity in carrying out fiscal policy, they have become increasingly concerned about the wider resource implications of fiscal actions in addition to their financing. Cash transactions do not adequately record either the timing of the policy action or the impact of these decisions on the economy.

Relationship Between the GFSM 2001 and the 1993 SNA General


The Government Finance Statistics Manual 2001 (GFSM 2001; IMF, 2001) and the 1993 SNA are harmonized to a very large extent. The overall structure and the definition of units are the same in both (the overall structure signifying the recording of full balance sheets and flows; the definition of units distinguishing between transactions, holding gains, and other changes in volume). The classification of sectors and functions are identical, and the classification of transactions is very close. The accounting conventions are identical in both systems.

The main specifics of government finance statistics (GFS) differences are the following:

  • The focus is on areas of prime importance to fiscal analysis. This implies, among other things, that the GFS do not include a production account.

  • The GFS have fewer core balances than the 1993 SNA, defining only those essential for fiscal analysis. However, the GFS identify several supplementary analytical balances.

  • The GFS break down several transaction categories in greater detail than in the 1993 SNA, and the classification criteria may also differ in some cases.

  • In the GFS, tables specify the sources and uses of cash.

  • The GFS emphasize the public sector more than the 1993 SNA does.

  • The GFS emphasize consolidated statements more than the 1993 SNA does.


The GFS and the 1993 SNA differ in quite a number of other respects. They have different backgrounds, as listed below.

Differences that are simplifications or the consequences of simplifications

  • In the GFS, inputs for own-account capital formation are not distinguished by type (such as use of goods and services, compensation of employees, and consumption of fixed capital).

  • In the GFS, the value of social benefits excludes goods and services produced by government units themselves. This is also the case for transfers of individual nonmarket goods and services to households.

  • The GFS treat some in-kind transactions as if they were made in cash, followed by a sale. This applies only to goods or services produced by government itself. Such imputations are made for (1) in-kind wages to government employees; (2) in-kind social benefits provided by government employer schemes; (3) in-kind grants to other government units and international organizations; and (4) in-kind compensation for damage to property or personal injury, or settlement of an insurance claim.

  • The GFS do not adjust interest flows for financial intermediation services indirectly measured (FISIM).

  • The GFS do not break down gross non-life-insurance premiums into insurance services and net premiums.

  • For employer social insurance schemes that provide retirement benefits, the GFS do not record social contributions (whether actual, imputed, or supplementary) in government revenue. Similarly, the GFS do not record pensions paid as government expense. In the case of unfunded pension schemes, the GFS record these receivables and payables in the balance sheet of the government under insurance technical reserves.

  • The GFS do not make imputations for reinvested earnings on direct foreign investment.

Differences that are theoretical improvements

  • The GFS treat defined-benefit retirement schemes that are unfunded in a theoretically more adequate way than the 1993 SNA does. This has various consequences. Where the 1993 SNA balance sheets record only actual available reserves under the item insurance technical reserves, the GFS record the present value of the schemes’ pension obligations. Contributions to the schemes reflect payments made in the 1993 SNA, but in the GFS they also reflect the increase in promised future benefits in the GFS balance sheets. As for property income attributed to insurance policyholders, the 1993 SNA records the income earned on the reserves held by the schemes. The GFS record instead the increase in pension liabilities that results from the passage of time.


  • Whereas the 1993 SNA records certain flows as taxes paid by general government units, the GFS classify them as miscellaneous other expense in the accounts of individual units and eliminates them in consolidation for the general government sector.

  • Whereas the 1993 SNA records certain flows as subsidies paid to general government units, the GFS classify them as grants.

The GFSM 2001 extends the cash-based analytical framework by emphasizing the recording of data on an accrual basis. That is, it recommends that government accountants record an economic event when economic value is created, transformed, exchanged, transferred, or extinguished—not just when the cash flow takes place. This gives a comprehensive picture of government activity, by allowing for the recording of noncash transactions (such as barter or in-kind transfers), internal transactions, other economic flows, the accrual of interest on discounted securities, and any payment arrears.

Moving from GFSM 1986 to GFSM 2001

Many countries currently compile and present their government finance statistics on a cash basis, using either the framework of A Manual on Government Finance Statistics 1986 (GFSM 1986; IMF, 1986) or their own national frameworks. The definitions used in the GFSM 1986 to distinguish revenue, expenditure, and financing result in changes in the overall balance that do not properly convey the economic impact of certain fiscal operations. Focusing attention on the government’s liquidity position, the GFSM 1986 framework does not adequately capture the economic activity of government, either in terms of the timing of policy actions or its impact on the economy. Moreover, from an operational standpoint, the GFSM 1986 lacks an explicit linkage between flows and stocks, which is important for measuring fiscal sustainability. It is important, therefore, that steps be taken to implement the comprehensive framework of the Government Finance Statistics Manual 2001 (GFSM 2001; IMF, 2001).

Implementation of the GFSM 2001 involves three distinct sets of actions that can be initiated simultaneously but, depending on the relative complexities involved, will be accomplished with different time horizons. These actions are

  • 1. Changing the presentation of the existing data (near term),

  • 2. Improving data reporting (medium term), and

  • 3. Fully implementing accrual reporting and the associated underlying systems (long term).

Of these actions, adopting the GFSM 2001 presentational format is the easiest to conduct. It essentially involves reclassifying existing fiscal data in the appropriate analytical framework. This activity would require few resources and can be accomplished relatively quickly.

The second action—reporting according to the GFSM 2001 framework—poses more of a challenge and could be constrained by a lack of statistical capacity. Close collaboration between the national authorities and IMF staff could be required to develop this capacity through training. This could take considerable time.

The third action—full implementation of the GFSM 2001—is only achieved with introducing an accounting system based on accrual principles and a modern public expenditure management framework. This would ensure classification systems that are capable of supporting statistical reporting compliant with the GFSM 2001 framework. Accomplishing this action is a major task for most countries, requiring careful planning and management to avoid disrupting the flow of fiscal statistics. To effectively manage the constraints to implementation at every stage, countries should develop a migration strategy for implementing GFSM 2001 tailored to institutional capacity.

Warranting special mention is the recording of tax revenue on an accrual basis. The GFSM 2001 requires accountants to record as revenue only those taxes that countries can reasonably expect to collect. Over time, a close relationship should develop between the accrual recording of tax revenue and the cash received.

Analytical Framework

For its analytical framework, the GFSM 2001 supports the balance sheet approach (BSA) to analyzing economic policy—bringing together stocks and flows in a transparent and consistent framework.Table 15 illustrates the features of this approach. By presenting fiscal data in an integrated framework, it enables analysts to assess fiscal sustainability—that is, to evaluate how net worth evolves through a series of balance sheets. Moreover, the GFSM 2001 framework yields of government saving, investment, and consumption are largely harmonized with the national accounts framework.

The GFSM 2001 framework is composed of summary tables—similar to business financial accounting reports—comprising an operating statement, a balance sheet, and a cash statement. As illustrated in Table 15, the framework consists of four analytic statements: the statement of government operations, the statement of other economic flows, the balance sheet, and the statement of sources and uses of cash (not shown here). Within the tables are the four main balances of the system—four core fiscal indicators—the net operating balance, net lending/borrowing, and net worth, plus the cash surplus/deficit (not shown here).

In addition to identifying the main balances of the system, the GFSM 2001 identifies other analytic measures of potential interest to analysts. These include the overall primary balance—adding back the interest expense to the overall fiscal balance to give a measure of the outcome from government’s discretionary activities—and the gross debt position— measuring the stock of all debt liabilities, as defined in the GFSM 2001 (that is, all liabilities except shares and other equity, and financial derivatives;Box 18).36

Table 15.

GFSM 2001 Analytical Framework

article image

GFSM 2001 Analytic Statements

Of the four analytic statements of the GFSM 2001 framework, three are based on accrual data for transactions, other economic flows, and balance sheets, while the fourth is cash-based, as follows:

  • 1. The statement of government operations (Table 16) summarizes all transactions by general government units and derives the important analytic balances of net operating balance and net lending/borrowing (NL/B) from the following information.

Table 16.

Statement of Government Operations

article image

The net operating balance summarizes the change in net worth owing to government transactions and is ultimately a measure of the sustainability of government policies affecting revenue and expense. It is conceptually equivalent to the national accounting concept of savings plus capital transfers. If the balance is positive, it indicates the government has generated surplus revenue from its current operations, resulting in an increase in its net worth. The government may use this surplus revenue to acquire assets and/or decrease liabilities. If the balance is negative, it indicates that government current operations have fallen short, necessitating the incurrence of liabilities or liquidation of assets to finance the shortfall.

Public Sector Debt

The public sector debt concepts, definitions, and classifications in the Government Finance Statistics Manual 2001 (GFSM 2001; IMF, 2001) are broadly consistent with those prescribed in External Debt Statistics: Guide for Compilers and Users (IMF, 2003; the Guide) and the 1993 SNA. Debt is a subset of liabilities, and according to the GFSM 2001 (page 129), public sector debt “consists of all liabilities that require payments of interest and/or principal by the debtor to the creditor at a date or dates in the future.” Thus, all liabilities of public sector units are debt instruments—except for shares and other equity and financial derivatives, because they do not require payments of interest and/ or principal. The GFSM 2001 recommends valuing these debt instruments at current market values, except for loans, which are normally valued at nominal prices. (The Guide recommends valuing tradable debt investments at nominal prices as well as market prices.) Contingent claims on the public sector are not recognized as liabilities in their balance sheets but may be reported as memorandum items.

Total debt instruments issued by public sector units comprise “gross public debt.” At the same time, the public sector can hold, as financial assets, the debt instruments issued by institutional units outside the public sector. Then the public sector’s net position on debt instruments (that is, assets minus liabilities) constitutes the “net public debt.” The public sector’s net position (net worth), over time, is an indicator of the sustainability of government policies. In compiling the balance sheet for the public sector, national compilers should consolidate counterpart claims and liabilities among public sector units with respect to debt instruments (that is, intrapublic sector holdings of debt instruments)—thus excluding them from the gross and/or net public sector debt.

Governments seek to ensure that public debt is sustainable, in the context of broader policy objectives, and that the debt can be serviced under a wide range of macroeconomic circumstances. In this regard, not only the level but the composition of public debt becomes relevant. Vulnerability and sustainability analyses require accurate and timely data on the composition of the public debt in terms of financial instruments, currency composition, residency, and maturity structure. Although the GFSM 2001 framework prescribes the breakdown of assets and liabilities by residency and instrument, compilers can usefully supplement it with information on currency composition and maturity.

The international statistical community is making efforts to further improve the analytical usefulness of public debt data. A draft public sector debt template, as well as general guidelines for reporting data, is available on the IMF’s website at http://www.imf.org/external/pubs/ft/gfs/manual/comp.htm. The template’s framework is designed to cover all institutional units within the public sector, covering data for domestic and external debt.

Subtracting the net acquisition of nonfinancial assets from the net operating balance gives a second balance, NZ/B. NL(+)/B(–) is a summary measure indicating, in essence, the extent to which government is either putting financial resources at the disposal of other sectors in the economy or using the financial resources generated by other sectors. Readers may therefore view the balance as an indicator of the financial impact of government activity on the rest of the economy. It is conceptually equivalent to the national accounting concept of NL/B.

In other words, it may be said that the net operating balance focuses on the “activity” side (by considering transactions in revenue and expense), whereas NL/B focuses on the “financing” side (by considering transactions in financial assets and liabilities).

The statement of other economic flows presents information on changes in net worth arising from economic flows other than transactions. Economists classify these flows as either changes in prices (revaluations) or changes in the volume of assets and liabilities.

  • 2. The balance sheet represents the stocks of assets, liabilities, and net worth at the beginning and the end of the accounting period.

  • 3. The statement of sources and uses of cash (not illustrated here) shows purely cash flows associated with revenue and expense transactions and transactions in nonfinancial assets, which yields the cash surplus/deficit. Adding cash flow transactions in financial assets (other than cash) and liabilities to the cash surplus/deficit gives net change in the stock of cash.

Tables 17 and18 summarize some worked examples for the GFSM 2001 framework, based on data examples in the 1993 SNA. They highlight the structure, key elements, and balances (except for the sources and uses of cash) needed for fiscal analysis and policy.

Table 17.

Statement of General Government Operations

article image

Assuming no nonautonomous government employee pension fund is included in general government.

Presented after consolidation and including capital grants.

Includes output adjusted for collective consumption.

Assuming no own-account capital formation.

Major GFSM 2001 Classifications

This section outlines the three major classifications of the GFSM 2001: (1) economic, identifying the types of outlays governments incur; (2) functional, identifying the purpose for which governments undertake the outlays; and (3) counterparty, identifying the counterparty sectors.

Table 18.

Integrated Balance Sheet for General Government

article image

Economic Classification

The statement of government operations presents government activity using what is known as an “economic” classification. The statement presents information based on the type of revenue, expense, nonfinancial asset, and financial instrument:

  • Revenue identifies separately tax revenue, forming the dominant share of revenue for most governments; social contributions, consisting of receipts from employers or employees that secure employees’ entitlements to future social benefits; grants, equaling noncompulsory transfers from other government units or international organizations; and other revenue, comprising property income, sales of goods and services, and miscellaneous revenue.

  • Expense identifies separately compensation of employees, consisting of the government’s labor cost; use of goods and services, equaling its cost of materials and services used in production; consumption of fixed capital, comprising its cost of using of fixed assets;37 interest, its cost of borrowing; transfers in the form of subsidies, grants, and social benefits; and finally, other expense, covering all expense transactions not covered elsewhere.

  • Transactions in nonfinancial assets identify separately the acquisition and disposal of fixed assets, inventories, valuables, and nonproduced assets.

  • Transactions in financial assets and liabilities identify the transactions in financing, classified by the instrument involved (for example, loan, deposit, currency, and so forth).

The balance sheets and statement of other economic flows also use the economic classification of nonfinancial assets and use the type of instrument to classify financial assets and liabilities.

Functional Classification

The economic classification informs users on the type of outlay incurred by the government, but many users are also interested in the purpose for which the outlay was undertaken. The GFSM 2001 framework satisfies this interest by applying a functional classification. The GFSM 2001 framework applies the Classification of the Functions of Government (COFOG) to both expenses and the net acquisition of nonfinancial assets.

COFOG classifies in detail the functions, or socioeconomic objectives, that general government units aim to achieve through kinds of outlays. COFOG permits analysts to examine over time the trends in government outlays on functions or purposes, regardless of the organizational structure of the government.

COFOG classifies objectives at three levels:

  • At the highest level, COFOG identifies 10 divisions:

    • 1. General public services

    • 2. Defense

    • 3. Public order and safety

    • 4. Economic affairs

    • 5. Environmental protection

    • 6. Housing and community amenities

    • 7. Health

    • 8. Recreation, culture, and religion

    • 9. Education

    • 10. Social protection.

  • It identifies within each division several groups.

  • It identifies within each group one or more classes of activities, by which the broad objectives are achieved.

Counterparty Classification

The GFSM 2001 also presents stocks of financial assets and liabilities and transactions in financial instruments by counterparty sector. This is to satisfy the interest of users in the sectors in which the government has outstanding claims/obligations and with which it is engaged in financial transactions. The economic classification classifies stock positions and transactions in financial assets and liabilities by types of financial instruments.

Resident counterpart sectors are the central bank, other depository corporations, financial corporations not elsewhere classified, nonfinancial corporations, households, and nonprofit institutions serving households (NPISH). Nonresident sectors are general government, international organizations, financial corporations, and other nonresidents.