Chapter 6. Deficit and debt of general government and public sector

Robert Dippelsman
Gabriele Semeraro
(Banca d’Italia)
João Cadete de Matos
(Banco de Portugal)
Julia Catz
Gabriel Quirós
Filipa Lima
(Banco de Portugal)

Governments play important roles in the economy. In the financial accounts and balance sheets, these include being major borrowers, issuers of securities, and funders of other entities. The chapter first dwells upon the delineation of the general government sector and the public sector, which includes general government and public corporations, both of which have separate economic roles. Governments often have complex structures, which give rise to specific classifications (for example, budgetary and extra-budgetary) and levels (for example, central, state, and local). This chapter also provides more detail on government deficit, financial assets owned by government and the various definitions of government debt, including the pitfalls when comparing such data internationally. It also deals with consolidation issues and some other specifics concerning the valuation of financial assets and liabilities, and the assessment of sustainability of government finance.


1. Delineating the government sector

Because of its role in providing non-market services (such as education, health care, defence and policing) and redistributing income (via subsidies and social benefits), the general government, or more broadly the public sector, is an important actor in the economy. Governments collect taxes and social contributions to pay for the provision of services and the redistribution of income. As the related amounts are usually substantial and vary over time, general governments also play an important role in the financial markets. Governments may run (temporary) surpluses or deficits, so can function as net lenders or net borrowers in the domestic economy.

General government versus public sector

There is a range of government-related groupings that may meet different needs for analysis. Such groupings include, amongst others, general government, public sector, social security funds, extra-budgetary agencies, the non-financial public sector, and the central government public sector. The most frequently used delineations relate to the general government, the central government, and the public sector, which are discussed in more detail below.

Government entities are grouped in a separate sector, because they differ in economic behaviour as compared to corporations and households. Government entities have policy motivations, such as influencing overall demand or income distribution, rather than maximising income. For example, both private financial corporations and general government may provide loans, but a commercial bank will provide a loan because of factors such as expected return from interest receipts and repayment of principal, while the general government may provide a loan with the objective to encourage a particular industry or to support a particular region. Government entities are significant in any macroeconomic analysis, by virtue of their impact on the whole economy. In addition, fiscal analysis is a field that looks at government in its own right, dealing with issues such as assessment of the effectiveness of government spending and design of tax systems.

The general government consists of institutional units that are non-market producers and that are controlled by the government. A non-market producer charges prices that are not “economically significant” (this term is described in further detail below). Government units are established by political processes and have legislative, judicial, or executive authority over other institutional units within a given area. The principal economic functions of government units are to:

  • assume responsibility for the provision of goods and services to the community or individual households primarily on a non-market basis;

  • redistribute income and wealth by means of transfers;

  • finance their activities primarily out of taxation or other compulsory transfers.

Fundamental to the distinction between general government and corporations is whether or not an entity produces goods or services for the market. Corporations that produce goods and services for the market are included in either the financial corporations’ sector or the non-financial corporations’ sector, depending on the type of goods and services they produce. Corporations that do not produce goods and services for the market and are controlled by government, are recorded as part of the government. This distinction between market and non-market producers critically depends on whether goods or services are sold at “economically significant prices”. The 2008 System of National Accounts (2008 SNA) further defines this criterion as “… prices that have a significant impact on the amounts that producers are willing to supply and on the amounts purchasers wish to buy” (paragraph 22.28). However, this still leaves much room for interpretation, a reason why in the European context a more quantifiable criterion has been agreed. In Europe, the sales of a market producer must cover more than 50% of the production costs (wages and salaries, intermediate consumption and depreciation) and consumers must be free to choose whether to buy and how much to buy on the basis of the prices charged. This criterion is being adopted by countries at an increasing rate.

Government’s activities affect the production, income, capital, and financial accounts. For instance, the financial accounts record how a government’s deficit (net borrowing) is funded. They also show funding that the government may provide for its public corporations or for the acquisition of financial assets. Governments’ funding requirements can affect the availability of funds to other domestic borrowers, or mean that governments must rely on foreign funding. Government debt includes securities, which play a key role as an investment vehicle on financial markets. The level and composition of debt has implications for future fiscal requirements related to interest payments and debt repayments.

The general government is usually broken down into different layers. The 2008 SNA identifies three possible tiers: central government, state government, and local government; as well as an optional fourth tier for having a separate subsector for social security funds. Government functions may be assigned differently across countries. Many countries may not have all levels. The state government, for example, is usually only relevant for large countries with federal constitutions like Germany and the United States, where considerable powers and responsibilities are assigned to state governments. On the other hand, some countries may have defined additional intermediate levels, depending on their national legislative structure, the importance of the various layers of government, and the policy interest for monitoring different layers of government. More generally, it can be noted that the degree of central control over the other levels differs from complete independence to heavy involvement. Finally, social security funds may be treated as a separate subsector or included among the other three subsectors. A majority of countries apply a separate recording of these funds.

The public sector is the most comprehensive government-related category, which cuts across three of the SNA’s main institutional sectors, encompassing all government-controlled units, as shown in Table 6.1. In addition to the general government sector, the public sector covers public corporations, which are combined with private sector non-financial and financial corporations in the primary institutional sector classification of the SNA. Public corporations are enterprises controlled by government that produce goods and services for the market. They also include central banks, which are public financial corporations that largely serve public policy functions (discussed in more detail in Chapter 3). Public corporations may vary, from being operated on an almost entirely commercial basis to being significantly involved in quasi-fiscal activities (that is, they carry out government operations at the behest of the government units that control them). The roles of public corporations vary from country to country, but may include serving as an instrument of public (or fiscal) policy for the government, generating profits for general government, protecting key resources, providing competition where barriers to entry may be large, and providing basic services where costs are prohibitive.

Table 6.1. The public sector and its relation to institutional sectors

Non-financial corporations

Financial corporations

General governmant










Examples of public non-financial corporations may include companies involved in public transport, various public utilities, or the exploration of minerals and energy. Whether or not they are actually controlled, or majority owned, by the government very much depends on the ways the economy is organised in a particular country. With the privatisation of many government-owned companies since the early 1980s, many of these corporations have become privately owned and controlled, and have ceased to be public corporations. Public financial corporations typically include the central bank, but may also include other banks, insurance companies, and pension funds. After the 2007-09 economic and financial crisis for example, governments had to come to the rescue of some very large banks and other financial corporations, some of which were nationalised in the process, becoming completely under the control of government. Public corporations can be major players in financial markets. This is obvious in the case of public financial corporations, but it is also true for public non-financial corporations which may issue substantial amounts of debt securities to cover their investments.

If public corporations cover more than 50%, but less than 100%, of their production costs by sales, such corporations would require additional sources of funding, often from the government, even though they are excluded from the general government sector. When these corporations run ongoing losses or have unsustainable debt liabilities, there can be contagion of funding problems from public corporations to the general government. For that reason, specific information on public corporations is relevant to understand possible implications for general government. See the Government Finance Statistics Manual (GFSM) 2014, paragraph 2.21-2.58, for more details on groupings within the public sector and different analytical uses.

One source of potential confusion in statistical terminology is that “public corporations” in the 2008 SNA are classified as such because of their economic function, i.e. being (non-)market producers, rather than because of their legal status. So, unincorporated entities having sufficient separate status and charging economically significant prices may be considered as public corporations, while heavily loss-making publicly owned corporations with a separate legal status may have prices that are not economically significant, and be included in general government. For example, in some cases, public corporations providing socially beneficial services, such as public transport companies, set prices so low that they are considered as not “economically significant”.

A final point in respect of the delineation between public corporations and private corporations concerns the definition of “control”. Governments can exercise control through ownership or in other ways, such as through funding. For further details, see paragraphs 4.77-4.80 and 4.92 of the 2008 SNA and Boxes 2.1and 2.2 of the GFSM 2014. The definition of control has been designed to include government-controlled entities in the public sector, when there is control without ownership. Instead of having a majority ownership, a government may decide to exercise control over some of its entities by, for example, appointing board members who have operational independence.

From the previous discussion, it is clear that the delineation of general government and the public sector is not always straightforward. Issues of identification and sector classification of units are often particularly difficult for governments, because of their complexity and administrative arrangements, control, and pricing strategies. There are often borderline cases which may lead to extensive debate. Therefore, specific manuals have been dedicated to clearly define and delineate the government and the public sector. The GFSM 2014 is a specialised statistical standard for general government and the broader public sector. The GFSM 2014 is harmonised with the 2008 SNA but goes beyond the 2008 SNA in that it provides more breakdowns and additional presentations designed to support fiscal analysis (see Appendix 7 of the GFSM for more details on differences). In the European Union, data for the general government are of very high political relevance, because of the Maastricht criteria for government debt and deficit. According to these criteria, the general government deficit should not exceed 3% of GDP, while gross debt levels should be below 60% of GDP or, if the debt-to-GDP ratio exceeds the 60% limit, the ratio shall, at a minimum, be found to have “sufficiently diminished and must be approaching the reference value at a satisfactory pace”. As these numbers depend on the delineation of general government, massive additional jurisprudence has been developed to arrive at internationally comparable standards for defining what belongs to the government and what does not. All of this has been laid down in the Eurostat Manual on Government Deficit and Debt, the latest version of which was published in 2016.

A summary of the criteria used for the delineation of general government and public corporations is presented in Table 6.2.

Table 6.2. Criteria for classification of units

Not economically significant prices (non-market producers)

Economically significant prices (market producers)

Government control

General government

Public corporation, financial or non-financial

Not controlled by government

Non-profit institutions serving households

Private corporation, financial or non-financial

Related to the complex structure of government is the issue of “consolidation.” Consolidation is the process of eliminating flows and stocks between units that belong to the same sector or subsector. Data for a consolidated grouping are shown as if that grouping was a single unit. On this issue, the 2008 SNA generally advises that data should be published in unconsolidated form to allow users to make their own choice over whether or not to consolidate, but also recognises the usefulness of consolidation for government data. The unit structure of government can involve a range of administrative and legal arrangements that have varying degrees of “separate” status and that can be somewhat fluid over time and between countries. To remove the effect of flows and stocks within a sector, it is often preferred for government statistics to be prepared on a consolidated basis; this format is also recommended in the GFSM, and in the definition of the Maastricht criteria. Without consolidation, there may be a ballooning of data, driven by administrative arrangements rather than the underlying economic reality. For instance, funds flowing between different parts of the general government do not change the financial position of the general government as a whole, but will appear to give larger values of revenue and expenditure, and assets and liabilities, in a country where there are a larger number of separate units. It can also be useful to consolidate the data for different levels of government, to avoid double counting the transactions and stocks between these different levels. However, unconsolidated data may also be suitable in some cases; for example, consolidated public sector data would exclude the non-financial public sector’s borrowings from public financial corporations, which is useful information. Consolidation is explained in more detail in Indicator 3 below, and in paragraphs 3.152-3.168 of theGFSM 2014.

Difference between budgetary and statistical presentations

National budgets usually present data in formats that differ from international statistical standards in terms of coverage of units and classifications. These formats often have public prominence because they are used in national political debates and are based on national laws and practices. As a result, almost all countries have two sets of fiscal data – the national format used in the budget and the internationally agreed standard formats of the 2008 SNA and GFSM. The variations in national formats across countries make international comparisons of national budgetary data difficult or invalid. Differences in classifications may mean that terms such as deficit or revenue have different meanings. Similarly, institutional coverage may differ between countries, particularly due to extra-budgetary agencies, which vary significantly in functions, number, and size from country to country. The metadata tables in the IMF’s Government Finance Statistics Yearbook identify budgetary and extra-budgetary agencies for each country. Since the scope of the budgetary and extra-budgetary subsectors is based on varying national practices, they do not play a role in the 2008 SNA, but they are both shown in GFSM presentations to allow links to be made with national budgetary data and to monitor the extent to which operations occur outside the budget.

The 2008 SNA and GFSM presentations provide standard classifications that allow consistent treatments for international data comparisons. Nevertheless, even if international standards provide harmonised data, different governmental structures and policies mean that international comparisons should be well-informed. For example, in some countries, a function may be recorded as being undertaken by public corporations, while the same function may be recorded as being undertaken by general government in others, because the strategy for pricing the products differs. Similarly, some functions may be allocated differently between central government and other levels of government.

2. Key financial accounts indicators for general government/public sector

Governments play an important role in financial markets, for example via the issuance of debt securities. These securities often provide a safe haven for investors. Government deficit and debt are therefore extensively monitored and high levels of deficit or debt can raise concerns about the future sustainability of government finance. Can the government raise enough income in the future to finance the debt service (the interest payments and the down payments)? This section discusses the analytical value and policy relevance of a number of indicators than can be derived from the government financial accounts and balance sheets. The following indicators are discussed: net lending/net borrowing, financial assets, gross and net debt, and net financial worth. These indicators allow for detailed analysis of government fiscal performance, understanding of financial (and non-financial) operations undertaken, tracking the evolution of components over time, and understanding how a government finances its deficit.

Net lending/net borrowing

Net lending/net borrowing is the balance of all incomes and expenditures of a government. In addition to current income and expenditure, such as taxes, social contributions and the payments of government employees, net lending/net borrowing also includes capital income and expenditure, mainly consisting of investments in non-financial assets (infrastructure, government offices, computers, etc.). The balance of current income and expenditure is referred to as saving. The GFS includes an additional measure operating balance, which is equal to the 2008 SNA balancing item “changes in net worth due to saving and capital transfers” (or equivalently, net lending/net borrowing plus net acquisition of non-financial assets).

Net lending/net borrowing, i.e. including all capital income and expenditure, is commonly known as the government deficit (or surplus), or the fiscal balance. In general, governments have tended to be net borrowers, rather than net lenders. If the government runs a deficit, it needs to borrow funds from other sectors. If total income is larger than total expenditures, government has a surplus, and it has funds available to invest in financial assets or to reduce liabilities. These transactions in financial assets and liabilities are recorded in the financial accounts. The financial accounts show the change in the government’s net financial worth as a result of the net acquisition of financial assets minus the net incurrence of liabilities. In other words, the financial accounts show the extent to which government is either putting financial funds at the disposal of other sectors or utilising financial funds generated by other sectors, either in the domestic economy or abroad.

Figure 6.1 shows net lending/net borrowing for the G7 countries for the period 2000-17. As noted before, governments in all countries almost consistently run deficits. Exceptions are Canada before the 2007-09 economic and financial crisis and Germany in more recent years. It is also clear that governments were heavily affected by the crisis, with a substantial increase in deficits for all G7 countries.

Figure 6.1. General government net lending/net borrowing for G7 countries, 2000-17
Percentage of GDP

Source: IMF (2017), World Economic Outlook Database, April 2017 (database),

Conceptually, the balance of non-financial (current and capital) transactions is equal to the balance of financial transactions as measured in the financial accounts, since a surplus leads to additional funds for investments or reducing debt, whereas a deficit has to be financed. In national accounts, both the balance of non-financial transactions and the balance of financial transactions are therefore referred to as net lending/net borrowing. Although in theory the balancing item from both financial and non-financial accounts should be equal, in practice they often differ. It is common for the non-financial and financial accounts to be compiled independently by different compilers, using different data sources and applying somewhat diverging statistical practices. Sometimes a difference is simply related to the timing of information used to compile the accounts.

The difference between the financial and non-financial accounts is known as the “statistical discrepancy”. While discrepancies also arise for other sectors, usually to a much more significant degree, government statistical discrepancies raise special issues, because the more limited number of units makes verification more feasible, and because government financing often has important impacts on the rest of the economy, as a consequence of which much attention is paid to the evolution of the government deficit. The size and bias of the discrepancy may be seen as an indicator of the quality of accounts. Its size in the final estimates of the accounts, after all source information has become available, must be reasonably small, in particular over a longer period of time, if the data are to be considered analytically useful and reliable. In some countries, variables such as other accounts receivable/payable on the financial accounts may be automatically adjusted in such a way to eliminate the statistical discrepancy between the non-financial and financial accounts. This practice ensures that only one figure for government deficit is released by the statistical authorities. In any case, compilers of the non-financial and financial government accounts should discuss and address the possible causes of divergences between the financial and non-financial accounts, and provide an indication to users about the extent of and the reasons for data inconsistencies.

GFSM terminology often refers to “above-the-line” and “below-the-line” transactions. Above-the-line transactions are revenue and expenditure, and correspond to the current and capital accounts of the SNA, while below-the-line transactions cover the financial transactions. Effectively, net lending/net borrowing is the “line” being referred to in the accrual version of the accounts, while it may also refer to the cash surplus/deficit in the “Statement of Sources and Uses of Cash” as defined in the GFSM.

Financial assets

Governments have several reasons to hold financial assets. There may be timing differences between cash receipts (payments) and government revenue (expenditure), as a consequence of which liquidity needs to be managed carefully. Governments may also set aside financial assets, to ensure the payments of future social benefits, such as those related to pensions for government employees or the society at large. They may provide loans to support particular projects for policy purposes (e.g. student loans or loans for social housing). Governments may also provide loans and equity to public corporations. Furthermore, taxes that have been assessed as due to be paid may already be part of government revenue, even though the equivalent cash payments have not yet been received. Such a timing difference is recorded under the financial asset “other accounts receivable”. Similarly, unpaid expenses or prepaid revenues are recorded as “other accounts payable”, as part of government liabilities. During the 2007-09 economic and financial crisis, governments tended to hold more financial assets. To support financial corporations facing financial problems, they acquired shares in those corporations, they appropriated (bad) debts, and/or they provided loans to those corporations.

Lending for public policy purposes has a different motivation from commercially motivated lending. As such loans have different likelihoods of repayment, it may be worth identifying them separately. GFSM 2014 does indeed recommend separating out these loans. It also provides an “overall fiscal balance” as a supplementary fiscal indicator, where additional policy lending is treated as an expense rather than as a transaction on the financial accounts. See Box 4.1 and paragraph 4.56 of the GFSM 2014 for further information.

Of the eight major 2008 SNA financial instrument categories, there are four main types of assets that are usually held by the general government. The first is equity reflecting the participation of the government in corporations. The second is debt securities, often held by social security funds. The third category relates to currency and deposits. Part of these deposits is held with the central bank. The final category concerns loans made to other countries or domestic corporations. The other types of financial assets – monetary gold and SDRs; insurance, pensions and standardised guarantee schemes; financial derivatives and employee stock options; and other accounts receivable – are usually of lesser significance, although they may be important in particular countries.

Figure 6.2 shows the total of general government financial assets, as a percentage of GDP, for a number of economies. It is clear that there is quite some divergence across countries, with countries within the Euro Area showing relatively high levels of financial assets held by the government, while the United Kingdom and the United States governments own relatively smaller amounts. The increase in the holdings of financial assets by Euro Area governments since 2007 has partly been driven by the support provided to the financial sector (through, for example, purchases of financial corporations’ equity), but also through the (re)classification of financial defeasance structures in the government sector. In the United Kingdom, the increase between 2007 and 2009 was largely driven by support measures to the financial sector.

Figure 6.2. Total general government financial assets, 2000-16
Percentage of GDP

Source: OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),; and ECB (2017), Statistical Data Warehouse,

Figure 6.3 presents the composition of the financial assets held by the United States, the Euro Area and the United Kingdom governments. In the Euro Area and the United Kingdom, the largest share of the financial assets is made up of equity and investment fund shares/units, whereas in the United States the government owns relatively little equities but many more loans and debt securities.

Figure 6.3. Composition of general government financial assets, by instrument, 2000-16
Percentage of total financial assets

Source: OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),; and ECB (2017), Statistical Data Warehouse,


One of the main indicators of the fiscal sustainability of a country is the country’s general government debt, or, more precisely, the general government’s debt-to-GDP ratio.1 There are various ways to define government debt. It is thus important to have a clear picture of how the debt has been defined. First, it is important to know whether the debt data refer to gross government debt, thus taking only liabilities into account, or whether it relates to a measure of net debt, i.e. whether assets owned by the government are also taken into account. Most publications of government debt data concern gross debt. In the SNA, the terms “net financial worth” and “net worth” are wider measures taking into account a wider range of assets and liabilities beyond debt. In this respect, net worth can be considered as the most comprehensive definition, as it includes all assets, non-financial as well as financial, minus all liabilities.

Looking at gross government debt, it is important to know which liabilities are included or excluded. An all-encompassing debt figure would include all liabilities excluding financial derivatives and equity. Equity is not very important for the general government in most countries. However, when analysing the public sector, the equity outstanding can be substantial, because of the inclusion of public corporations in this sector delineation. Financial derivatives are usually also not that significant, partly because only a small number of governments have engaged substantially in financial derivatives.

Government debt according to the “Maastricht”, or Excessive Debt Procedure (EDP), definition, as applied within the European Union, only includes the following debt instruments: currency and deposits, debt securities and loans. As shown in Figure 6.4, the difference between an all-encompassing government debt figure and a Maastricht-compliant government debt figure can be quite substantial. The share of debt not included in the European definition adds up to more than 20% of total debt liabilities in the United States, whereas for the Euro Area as a whole, the share ranges between 5 and 10% of total debt. A large part of the other debt liabilities of the US general government (between 17 and 18% in more recent years) relates to the underfunding of pension funds of civil servants for which the government has responsibility. In contrast to others, the United States has no loan liabilities, while loan liabilities had become by far the largest component for Greece by 2014, reflecting bailout funding and reduced access to securities markets.

Figure 6.4. Composition of general government (EDP) debt, by instrument, 2000-16
Percentage of total debt

1. Japanese data on government debt refer to unconsolidated data.

Source: OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),; and ECB (2017), Statistical Data Warehouse,

A special case that is important for monitoring and analysing government debt concerns the (non-)recognition and the (non-)recording of pension liabilities and contingent liabilities.2 In the financial accounts and balance sheets, contingent liabilities related to the payments of future pensions, which are provided by government to large parts of the population, also known as social security pension schemes, are generally not recognised and recorded in the central framework. Furthermore, the 2008 SNA gives some flexibility with regard to the recording of unfunded (or underfunded) government-sponsored employment-related schemes, but – as in the case of social security pension schemes – they are usually only measured and included in supplementary tables.3 The amounts involved can be enormous, representing much more than the traditional government debt. A further discussion on the recording of pensions can be found in Chapter 9.

It is also clear that the coverage of the government population, or the sector delineation, can have a significant impact on the debt measures, especially when looking at (gross) debt data. Is the data referring to central government, to general government, or the public sector? Or is the data referring to some national classification, such as budgetary government? Therefore, one always has to be aware of the population included, and the definition of debt used.

As an example of the magnitude of the differences that can result depending on which definition and population is used, Table 6.3 shows the impact of using different debt definitions and different populations for the gross government debt of Canada. It shows that, depending on the instrument and institutional coverage adopted, Canada could report “government debt” ranging from 35% to 129% of GDP. This range of values shows how important it is for statisticians and data users to be precise and clear about the definitions and populations used. Moreover, there is the potential for “hidden” liabilities when narrower measures are used, with the consequent inadequate recognition of possible risks and burdens.

Table 6.3. Consolidated gross government debt for Canada at nominal value, 31 December 2016
Percentage of GDP

Central government

General government

Public sector





Wide excl. IPSGS1








1. The narrow definition covers only currency and deposits, loans, and debt securities. The wide definition includes all debt instruments recognised by the 2008 SNA, excluding financial derivatives. The wide measure excluding IPSGS refers to the wide measure excluding insurance, pensions, and standardised guarantee schemes.

Source: OECD (2017), “Public Sector Debt: Public sector debt – consolidated”, OECD National Accounts Statistics (database).

In the international context, general government debt is the one that is most frequently used. However, some countries prefer to target a debt variable that not only pertains to the general government sector but also includes the debt of non-financial or all public corporations, as the latter may give rise to potential vulnerabilities for the government. If the public sector also consists of (large) financial corporations, it seems to make more sense to use a net debt concept (see below) for policy analysis, since a large part of the liabilities of these corporations will be offset by the financial assets acquired in the process of financial intermediation. However, gross debt can generally be seen as a more prudent measure of debt, since not all financial assets are liquid.

Another issue that is very relevant for data on gross debt, is whether the results concern consolidated data or non-consolidated data, that is, whether or not liabilities which are owned by other units within the sector under consideration are removed. While the standard presentation in the financial accounts and balance sheets is on a non-consolidated basis, government gross debt indicators usually provide data on a consolidated basis.4 The introduction already touched upon this issue, and more details are provided below.

A final point on the definition of government debt relates to the valuation of the liabilities. Generally, one can distinguish three types of valuation: market value, nominal value, or face value. Nominal value and face value are quite close to each other. They both refer to the contractually agreed amount that the debtor will have to refund to the creditor at maturity. If the debt is valued at nominal value, it will also include the accrued interest which is not yet paid; this is not included in the case of face value. As such, the nominal value is consistent with the accrual concept for interest income. The market value is the price that someone is prepared to pay for taking over the debt. In the central framework of the SNA, loans which are usually non-tradeable are to be valued at nominal value, while tradeable debt securities are valued at market prices. In times of financial crisis, the market value of government debt securities may fall due to perceived declines in creditworthiness of the government, even while the nominal and face values rise. Having said that, the most frequently used debt indicators for government usually apply a nominal valuation for all instruments. As a consequence, even if the population and the debt definition used are exactly the same, the results from the financial accounts and balance sheets, which typically use a valuation at market prices for tradeable instruments, may diverge from published government debt results.

Disregarding risks of defeasance, bankruptcy of the debtor, and currency developments, differences between the nominal value and the market value are caused by interest rate movements. Assume that one has a debt security of USD 1 000 with a fixed coupon payment of 5% per year. If, following the issuance of this security, the market interest rate changes to 2.5%, it is clear that a potential investor is willing to pay more than USD 1 000 for the existing security which gives him a return of 5% per year. One may thus observe more substantial differences between nominal and market value in the case of quickly changing interest conditions.

Figure 6.5 presents, for a number of economies, gross debt for the general government for the period 2000-17 according to the broadest coverage of instruments (excluding insurance, pensions and standardised guarantee schemes, IPSGS), a key indicator of the financial sustainability of government finance. All tradeable instruments are valued at market prices, while loans, deposits, and other accounts payable, where market prices are not readily observable, are valued at nominal prices. The size of the general government’s debt has changed considerably over time. It also differs, even more substantially, across countries. As the figure shows, all G7 countries experienced substantially increasing levels of general government debt during 2008-10, with debt growth levelling off since then in most countries. Only Germany showed a decrease of government debt relative to GDP after the 2007-09 economic and financial crisis. Comparing 2017 government debt as % of GDP to pre-crisis levels, increases were most significant for Japan and the United Kingdom, but increases were also substantial in France, Italy and the United States. In respect of the above, it should be noted that Japanese debt is overstated relative to other countries, because they refer to non-consolidated data, and may also include, for example, government bonds held by social security funds.

Figure 6.5. General government gross debt, 2000-17
Percentage of GDP

1. Japanese data on government debt refer to unconsolidated data.

Source: OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),

The difference between the change in debt and the deficit/surplus during a certain period is known as the “deficit-debt adjustment” (DDA) or, more generally, as the “stock-flow adjustment”, or, in 2008 SNA terminology, the “other flows” relating to changes in debt. It is widely known that deficits contribute to an increase in (gross) debt levels, while surpluses reduce them. However, the change of government gross debt is also affected by other elements such as non-debt liabilities, the purchase of financial assets, revaluations and other changes in the volume of assets. In formulas, this can be expressed as follows:

(1a) Government surplus = Balance of financial transactions =

= Net purchase of financial assets minus Net incurrence of total liabilities


(1b) Government deficit = Net incurrence of total liabilities minus Net purchase of financial assets


(1c) Net incurrence of total liabilities = Government deficit plus Net purchase of financial assets

The total change in stocks of debt liabilities can then be (re-)written as follows:

(2a) Total change in stocks of debt liabilities =

= Net incurrence of debt liabilities plus Revaluations of debt liabilities plus Other changes in the volume of debt liabilities =

= Net incurrence of total liabilities minus Net incurrence of non-debt liabilities plus Revaluations of liabilities plus Other changes in the volume of liabilities

= Government Deficit plus Net purchase of financial assets minus Net incurrence of non-debt liabilities plus Revaluations of liabilities plus Other changes in the volume of liabilities

= Government Deficit plus Deficit-Debt Adjustments (DDA)


(2b) Deficit-Debt Adjustments (DDA) = Net purchase of financial assets minus Net incurrence of non-debt liabilities plus Revaluations of liabilities plus Other changes in the volume of liabilities

As long as the components of the DDA are sound, the difference between the change in debt and deficit is explained and does not raise concerns regarding data quality. In normal circumstances, with debt at nominal value, revaluations and other changes in the volume will be relatively small, and differences in deficit and changes in the stocks of debt liabilities can be explained by net purchases of financial assets. More generally, in statistical terms, DDA is the standard statistical analytical tool to compare flows and stocks which is common to all monetary and financial statistics.

Net (financial) worth/debt

The concepts of net financial debt and net financial worth are closely linked. Net financial debt only pertains to the net positions in debt instruments, whereas net financial worth is defined as the difference between all liabilities and all financial assets. For the general government, which in most countries does not have any equity liabilities, the main difference between these two indicators will concern equity holdings of government. Another difference, which may be more substantial in a limited number of countries, relates to financial derivatives and other accounts receivable/payable. A government’s net financial worth changes over time due to transactions (the net impact of which is captured by net lending/net borrowing), revaluations, and other changes in volume (e.g. debt write-offs) of the assets and liabilities included in the definition of net financial worth.

It can be very useful to analyse net financial debt/worth, in addition to gross debt. From a fiscal sustainability point of view, a government owning substantial amounts of financial assets is better off than a government with minor holdings, given that they have similar levels of gross debt. Figure 6.6 shows total liabilities, total financial assets and net financial worth for general government of the United States, the Euro Area, Japan and the United Kingdom. Japanese gross government debt is unconsolidated, as a consequence of which government’s net financial worth is better comparable with other countries than their gross debt. For Japan, the United States and the United Kingdom, net financial worth was close to minus 120%, minus 100% and minus 90% of GDP, respectively, whereas for the Euro Area it was around minus 70% of GDP by the end of 2016. However, net financial worth is not negative for all general governments. For Norway, the general government’s net financial worth is quite large and positive: almost 295% of GDP by the end of the first quarter of 2017.5

Comparison of net financial worth, as can be done using Figure 6.6, is still incomplete in that it does not take into account holdings of non-financial assets. The balance of total assets, financial as well as non-financial, and total liabilities is represented by net worth. For instance, a government owning small values of financial assets but large values of natural resources, like oil or gas, may appear to have low net financial worth and might be perceived by the public as relatively “poor”. However, when one takes into account non-financial assets, which include the natural resources mentioned before, this may alter the picture completely. The main reason to exclude non-financial assets from the government’s wealth is the unavailability of data on these assets. While the value of “known” natural resources is relatively straightforward to measure if there is a market price for the commodity exploited, the value of non-financial assets in general is quite difficult to estimate due to the absence of markets where the relevant assets are traded. The valuation of these assets therefore usually depends on the accumulation of past investments, appropriately depreciated with assumptions about, for example, asset lives and depreciation patterns that are subject to uncertainty. Furthermore, although it may be possible to readily value some government assets like buildings, they may concern historical buildings that are not likely to be sold, and are unique and irreplaceable, so are difficult to value. Finally, public infrastructure like railway tracks, sewerage systems, infrastructure for utilities, etc. have a high value from the perspective of generating economic benefits for the country, but may not have a commercial value, because the government prohibits the commercial exploitations of these assets. Forthat reason, its value is also derived on the basis of the depreciation method mentioned before.

Figure 6.6. Total general government financial assets, liabilities and net financial worth, 2000-16
Percentage of GDP

1. Japanese data for financial assets and liabilities refer to unconsolidated data.

Source: OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),; and ECB (2017), Statistical Data Warehouse (database),

Box 6.1. The Excessive Deficit Procedure in the European Union

The Stability and Growth Pact (SGP) is a rule based fiscal framework intended to ensure fiscal discipline in the European Union (EU). The core principle of this Pact is that EU countries should avoid excessive deficits. Hence, the government deficit-to-GDP ratio is allowed to exceed a reference value of 3 per cent of GDP under exceptional circumstances only. The government debt-to-GDP ratio is not allowed to be higher than a reference value of 60 percent of GDP, unless the ratio is sufficiently diminishing and approaching the reference value at a satisfactory pace. These two fiscal indicators, and their reference values of 3 per cent and 60 per cent of GDP, are also the subject of the fiscal convergence criteria for entry to the European Union and to Stage Three of the Economic and Monetary Union (EMU, or Euro Area).

Government debt, also known as “Maastricht debt” or “EDP debt”, is defined in the Protocol on the Excessive Deficit Procedure (EDP) annexed to the Maastricht Treaty and in Article 1 (5) of Council Regulation (EC) No. 479/2009 as the total general government gross debt at nominal value outstanding at the end of the year with the following characteristics:

  • Sector delineation: EDP debt comprises the (consolidated) liabilities of the general government sector, therefore including all levels of government: central government, local government, social security funds, and when applicable state government. This means that the debt of public corporations (government-controlled units classified in the financial and non-financial corporations’ sectors) is excluded from the measurement of government debt in the EU.

  • Gross debt: EDP debt is “gross” debt which means that financial assets of general government units are not subtracted in the calculation of government debt.

  • Coverage of instruments: EDP debt consists of the following liabilities of general government: currency and deposits, debt securities, and loans, as defined in the 2010 European System of Accounts (ESA 2010). This means that it excludes the remaining financial instruments, namely monetary gold and SDRs; insurance, pensions and standardised guarantee schemes; financial derivatives and employee stock options; and other accounts payable.

  • Valuation rules: EDP debt is measured at what it calls “nominal value” and equals the contractually agreed amount that the government will have to refund to creditors at maturity. In the 2008 SNA, GFSM 2014 and Public Sector Debt Statistics: Guide for Compilers and Users 2011, the term “nominal value” is used differently, while the EDP’s method of valuation is known as “face value”. The latter valuation method means, in particular, that the government debt is not affected by changes in market yields, and excludes unpaid accrued interest. EDP debt is thus measured differently than most government liabilities according to the 2008 SNA, which are recorded at market value.

  • Consolidation: EDP debt is consolidated across the general government sector, which implies that government debt instruments held as assets by other general government units are not included in the calculation of government debt.

All EU countries are legally required to report the breakdown of government debt by instrument and initial maturity to Eurostat, the statistical office of the European Union. Furthermore, the ECB Guideline on Government Finance Statistics (ECB/2014/21) requires all Euro Area national central banks to report additional breakdowns of government debt to the ECB: by residual maturity, by holding sector, and by currency.

Source: Eurostat Manual on Government Deficit and Debt; IMF et al., Public Sector Debt Statistics: Guide for Compilers and Users 2011.

3. Going into more detail

This section discusses some more detailed issues related to government and public sector financial accounts and balance sheets. Subsequently, the following topics are discussed: i) consolidation; ii) some difficult cases potentially affecting the measurement of government deficit and related financial transactions; and iii) valuation and use of financial accounts and stocks. Attention is also paid, in Box 6.2, to the Government Finance Statistics Manual (GFSM) 2014 and how this international standard for government statistics relates to the 2008 SNA.


Similar to business accounting, national accounts and government finance statistics can be presented in a consolidated version, where the accounts for a group of units are reported as if they constitute one single entity. The main concept has been introduced earlier on in this chapter, along with a discussion of consolidation’s advantages and disadvantages, in contrast to the 2008 SNA general policy to publish unconsolidated data for all sectors.

Consolidation refers to the elimination, from both assets and liabilities, of financial transactions and positions between units that are grouped together. Consolidation should be applied to financial as well as non-financial accounts: if a bond position between government units is removed from the financial accounts and balance sheets, then the corresponding coupon payments should be deleted from the revenues and expenses of the same units as well.

Consolidation commonly occurs when the accounts of subsectors of general government are combined. Consolidation of the general government sector can be performed by removing financial transactions and positions between units within the same subsector of general government: this is referred to as intra-sectoral (or “within”) consolidation. In this case, operations of local government with central government, for example, are not deleted; only the operations between local government entities are. Other methods, often used in national-based publications and primary sources, also require removing of assets and liabilities between different subsectors of government. This process is referred to as inter-sectoral (or “cross”) consolidation, and is necessarily more thorough than “within” consolidation; that is, it eliminates more transactions than “within” consolidation. Table 6.4 summarises these two forms of consolidation.

Table 6.4. Methods of consolidation

Central government (CG)

Local government (LG)

General government (GG)







CG Bonds issued to LG

CG Bond held by LG

LG Bonds issued to non-LG

CG Bonds issued to non-government

LG Bonds held by LG

Loans to non-government

Bonds issued to non-government

Loans to non-government

LG Bonds issued to LG

Bank loans

Loans to LG

Loans from CG

Bank loans

To be deleted under “within” consolidation


To be deleted under “cross” consolidation

It should be noted that cross-consolidation might affect balances at a subsector level, whereas consolidation within subsectors is always neutral: it preserves the balancing items of the unconsolidated accounts, such as net lending/net borrowing and net (financial) worth, while only impacting the magnitude of the aggregate totals. This is shown in Table 6.4 as well, where the consolidating amounts marked in grey on the asset (bonds) and liability (loans) sides of local government do not need to be the same (and neither do the corresponding amounts on the central government balance sheet). Consolidated figures for the balancing items of general government are the same as the unconsolidated numbers, regardless of the method used for its subsectors.6

The reasons in favour of consolidation for government statistics are discussed in detail in O’Connor et al. (2004). The main arguments relate to international comparability, i.e. the need to eliminate the distorting effects on aggregates of differing administrative arrangements across countries.

Box 6.2. Differences and similarities between the SNA 2008 and GFSM 2014, with specific reference to financial accounts and balance sheets

An alternative version to the 2008 SNA sequence of accounts is the government finance statistics (GFS) presentation. While using definitions and rules that are the same as, or compatible with, the 2008 SNA, the GFS provides a different, but still integrated, picture of the accounts, specifically designed to capture government (or public sector) features. The GFS is less focused on the production of goods and services and the linkages with other sectors, and more focused on both the impact of economic events on the finances of government and – in the opposite direction – the impact of government activities on the economy through taxing, spending, borrowing and lending.

In the 2008 SNA presentation, financial accounts and balance sheets, combined with non-financial accounts, are part of a seven-account sequence that starts with production, and then shows income generation and distribution, use of income for final consumption, and accumulation, focusing on the linkages between sectors of the economy. The GFS presentation starts with a statement of revenue and expense (thus not imputing government output, government consumption, and other items less familiar to policy makers and budgetary accountants), and then presents overviews of transactions in non-financial and financial assets and liabilities. As in the SNA, these transactions and other economic flows are integrated with balance sheets.

Table 6.5 shows three of the four statements recommended by the GFS, showing that all changes in the balance sheets result either from the Statement of Operations, or from the Statement of Other Economic Flows (i.e. revaluations or other changes in volume, not resulting from transactions). In addition, the GFS includes a Statement of Sources and Uses of Cash, to provide key information on liquidity.

Definitions and concepts in the GFS framework are fully compatible with the ones in the 2008 SNA. Technical differences between the 2008 SNA and the GFS Manual include the practices for counterpart entries: similar to business accounting, the GFS is based on double-entry recording, whereas the 2008 SNA is based on quadruple-entry recording, with all transactions recorded from the point of view of all units involved. This kind of difference reflects the fact that the GFS is aimed at providing an exhaustive presentation of a single sector, rather than a picture of the overall economic processes and interactions in the economy. For example, the 2008 SNA partitions interest into a service element and a “pure” interest element, and also partitions the insurance premiums into a service element and the part of the premiums meant to cover the insurance claims. All of this is necessary to measure the output of financial services, but is not relevant to general government revenue or expense and is not observable from government accounts. More information on differences is available in the GFSM 2014, Appendix 7, “GFS and Other Macroeconomic Statistics”.

The overviews provided by the GFS add value, because they present information in a way more oriented to the questions of fiscal policy and analysis. In the sequence of tables, the full reconciliation with the cash-flow statement, an element not relevant to the SNA, marks the policy-oriented nature of the GFS for liquidity management and for identifying payment timing issues.

Table 6.5. Structure of the GFS presentation


Source: GFSM 2014, Figure 4.9.

Some difficult cases potentially affecting the measurement of government deficit and related financial transactions

In the 2008 SNA and the GFSM 2014, a key role of providing summary information on the overall performance of general government (or the public sector) is played by the main balancing items: the net lending/net borrowing figure and the change in net (financial) worth. The former indicator, commonly known as the “government deficit”, is also adopted as a key target for administrative use in the fiscal surveillance of the European Union (see Box 6.1).

Financial accounts and balance sheets for the government sector are based on the same general principles as financial accounts and balance sheets for other sectors in the national accounts framework. In order to properly identify financial instruments, specific provisions may be needed, reflecting that the powers, motivation, and functions of government are sometimes different from those of other sectors. Such special provisions should not be seen as exceptions: they are simply additional rules, able to provide guidance for borderline cases where different 2008 SNA criteria would be appropriate. An important example is provided by the specific provisions for “capital injections”, which are payments by the government to public corporations on a large and irregular basis, where formal elements of financial transactions coexist with elements of capital transfers. Of course, both types of transactions are covered by 2008 SNA principles, but a different classification, depending on whether “capital injections” are deemed to be more similar to financial transactions or capital transfers, would lead to different figures and economic analysis. For these reasons, additional criteria, as developed in the GFSM or the Eurostat Manual on Government Deficit and Debt,7 are required in order to select the appropriate 2008 SNA treatment. Capital injections are considered to be expenses – rather than financial investments – when they do not result in an effective financial claim on the debtor. This is the case when the purpose of the capitalisation is to cover accumulated past losses, rather than increase the value of government equity (as a normal investor would require).

In order to test whether the government acts similarly to a private investor/shareholder, so that the capital injection can be treated as an increase in equity, the expected return on investment is a crucial factor. A realistic rate of return is indicated by a rate of return that is sufficient to generate dividends or holding gains at a later date and encompasses a claim on the residual value of the corporation. This simple requirement takes into account neither the relationship between risks and rewards, nor the opportunity cost with respect to alternative investments. Nevertheless, it has allowed a number of relevant cases to be ruled out in the past.

More recent experience, notably in the field of government intervention in favour of the banking sector, shows that transfers may often take the form of viable investment (i.e. the government is able, in principle, to generate dividends on its investment), but at out-of-market rates after taking into account risk exposure. Future evolution of the rules, in order to properly partition financial investments and identify transfer components, might imply a major use of instruments taken from financial theory. For example, a standard Capital Asset Pricing Model – CAPM (Bodie et al., 2005)8 might be appropriate in deriving a risk-adjusted rate of return, to be compared with average market returns. It should be noted that the complexity of such an approach is exacerbated by the use of the general government as the main target population for indicators on fiscal sustainability. The broader population of the public sector would show balancing items less subject to the impact of borderline transactions, as in this case the relevant transactions would be eliminated upon consolidation if the corporation is or becomes a public corporation, and therefore part of the public sector.

An opposite example, in which the government receives payments, concerns the transfer of pension obligations, which has been observed in several countries. A detailed discussion on pension entitlements is included in Chapter 9. Here, the example is limited to the issue of lump-sum payments related to transfers of pension obligations. A number of governments, in order to facilitate a public corporation’s privatisation campaign or new stock market issuance, have “cleaned up” that corporation’s balance sheet by assuming that corporation’s pension commitments to its employees, receiving in return a lump sum payment corresponding to the financial burden of the future pension payments. The economic substance of this transaction is an equal exchange of cash for the incurrence of an obligation that is a liability. Therefore, the transaction should be treated as a financial transaction and not improve government net lending/net borrowing or net worth. However, depending on the asset boundary, the obligation to pay for future pensions may not appear as a liability on the balance sheet of either of the units transferring or assuming the obligations.

These transactions, owing to their size and relevance for government accounts, had a great influence on the debate on the reform of statistical standards that led to the 2008 SNA, ESA 2010 and GFSM 2014. However, some lack of harmonisation between the SNA, the ESA and GFSM survived in the classification of liabilities to be recorded in the central framework and implicit liabilities which are not recognised in the central framework. In the GFS, pension entitlements of (unfunded) government sponsored, employment-related defined benefit schemes are fully recognised, whereas these schemes are excluded from the central framework of the ESA, and the 2008 SNA provides some flexibility on the treatment of these entitlements. Notwithstanding these differences, the balancing items of government in national accounts can no longer improve as a consequence of pension transfers. Under ESA, the lump sum payment, if not classified as an explicit pension liability, should now be viewed as a prepayment of social contributions for future pension payments and be classified as “other accounts payable” (to the pensioners).9

Apart from pension entitlements, the government is particularly likely to be involved in contingent liabilities and in the provision of one-off or standardised guarantees, traditionally not recognised in the national accounts. The international standards for compiling government statistics recognise that some arrangements that are not included as liabilities, such as one-off guarantees and other contingent liabilities, may be analytically important and provide specific guidance for them. For example, the Guide on Public Sector Debt Statistics provides a categorisation and supplementary table for explicit contingent liabilities.

Valuation and use of financial accounts and stocks

Assets and liabilities are valued at market prices in the financial accounts and balance sheets, except for three specific instruments: currency and deposits, loans and other accounts receivable/payable, which are valued at the amount that the debtors are contractually obliged to repay to the creditors, regardless of changes in market prices and rates. Unlike non-financial assets, which for governments may include historic monuments and infrastructure assets that may be difficult to value, financial assets and liabilities are less problematic to value. However, a few exceptions can be observed, including equity stakes in unlisted public corporations, which are without a private equivalent.

An issue of greater interest may instead relate to proper use of financial balance sheets, and to their relationship with fiscal sustainability analysis. The concept of fiscal sustainability concerns a situation in which a government is expected to be able to continue servicing its debts in the long run and to sustain its current spending, tax and other policies, without defaulting on some of its liabilities or promised expenditures, or relying on unrealistically large future corrections to the balance of income and expenditure. One way to express sustainability is by means of a solvency condition in terms of discounted present value of future income and expenditure flows, compared to the current debt, as follows:

picture (1)


Dt0 : Debt at time t0


where all variables are expressed in GDP terms (D, PB in % of GDP; r is defined as 1 + r = (1 + i)/(1 + y), where y is the rate of growth of nominal GDP, and i the nominal interest rate. Primary expenditure excludes interest, which is a useful concept as interest is an outcome of expenditure decisions in previous years.

The above equation illustrates that the discounted value of the total future primary balances should be (at least) equal to the government debt at the starting point t0. One key issue concerns the stock component (the term “D” in equation (1)), and which debt measure is more adequate to analyse fiscal sustainability: nominal gross debt, debt at market prices in line with the national accounts balance sheets, Maastricht Debt, net (financial) worth, or some other measure.

An alternative characterisation of sustainability was proposed by Buiter (1985), who argued that sustainable fiscal policy should maintain the ratio of net worth to GDP at its current level. Other proposals pointed out the inadequacy of traditional government indicators and put more emphasis on the role of balance sheets and flow of funds, suggesting an extension of coverage for the items included in the accounts (Buiter [1983]; Kotlikoff [1984] and Cadete de Matos et al. [2015]). These proposals recommended including assets as well as liabilities in measures of sustainability (in order to show the total “net worth”). In addition, they proposed that the liability side should comprise implicit entitlements, even though these have a contingent nature.

On the other hand, no clear preference has emerged in relation to which specific rules for the valuation approach should be used in fiscal sustainability analysis. In the specific context of the above equation and its alternative formulations, market valuation might not be helpful. For example, for a government approaching a default situation, debt valued at market prices will automatically decrease, as investors are willing to pay less and less for the relevant debt instruments because of the risk of non-payment. But this phenomenon of decreasing debt at market value does not necessarily signal that the government has a better ability to sustain its current spending in the long run. For this kind of analysis, measures closer to the concept of gross debt at nominal value might provide better information for analysing sustainability.

With this caveat, market valuations – fundamental in the national accounts context – still play an important role in GFS and the 2008 SNA. Data at market value enables a full reconciliation with data from counterparties (which typically value their holdings at market value) and is thus an instrument for cross-checking. It also expresses the cost for terminating units or buying back liabilities, with possible policy relevant implications. Finally, it can be useful in studying the counterparties of government, for example in the analysis of portfolio shifts of investors.

Key Points

  • Governments make revenue, expenditure and financing decisions that affect the whole economy. They differ from other units in that they have a public policy motivation. In some cases, the government makes its decisions on spending with the objective of influencing the rest of the economy by reducing or expanding aggregate demand.

  • Government deficit is a key variable for analysis because of its need for financing and its contribution to future debt. Government debt may raise particular concerns in relation to financial sustainability – fiscal insolvency has caused a number of financial crises and, unlike business units, expenditures which increase government debt are often related to non-commercial motives, and thus may generate burdens that are passed on to (future) tax paying units. Measures of government debt can vary according to instrument coverage, institutional coverage, valuation, consolidation, and the degree of netting of assets, so these aspects need to be considered in comparisons.

  • Governments have complex structures that differ from other entities – budgetary and extra-budgetary components; social security funds; central, state and local levels; and public corporations that have financial implications for government and are motivated to varying degrees by non-commercial motivations. Governments are also able to exercise control over units in more complex ways than in the private sector. The 2008 SNA includes various criteria to identify government control to reflect this.

  • Different components of the government may be considered separately for different purposes, such as budgetary central government being the focus of fiscal policy, and local governments being analysed separately to focus on their specialised functions.

  • The Government Finance Statistics (GFS) presentation is an alternative version to the 2008 SNA sequence of accounts, specifically designed to capture government (or public sector) features. For example, measures of total revenue and expenditure are key indicators in the GFS, but are not shown in the SNA, where they are spread over several accounts. Definitions and concepts in the GFS framework are almost fully compatible with those in the SNA. However, GFS is more closely related to business accounting, providing a bridge to published national budget accounting data. Furthermore, these issues mainly affect “above-the-line” figures, with financial accounts and balance sheets being the same in the 2008 SNA and the GFSM 2014. The one exception to this is the more inclusive recognition of pension liabilities in the framework for GFS. Pension liabilities have been a particular concern in fiscal risk analysis.

  • In addition to recording transactions on an accruals basis, which is in line with the 2008 SNA, the GFS provides an additional Statement of Sources and Uses of Cash. Such a statement, using a cash basis recording, is more in line with the standard practice in many countries’ government accounting systems. It also supports management of cash and government budget liquidity.

  • In the specific case of government, there are good reasons to deviate from 2008 SNA’s general policy to publish unconsolidated data. Consolidation allows analysts to relate government aggregates to the economy as a whole, eliminating the “internal churning” of funds and to compare results across countries, regardless of the degree of centralisation of government functions.

  • Another exception from 2008 SNA general policies relates to valuation: nominal value might provide better information with which to assess sustainability. However, market valuation still plays an important role in GFS and the 2008 SNA; for example, market valuation is used for the full reconciliation of balance sheets across the whole economy; to express the cost for terminating units or buying back liabilities, with possible policy relevant implications; and to study the behaviour of the counterparties of government.


Buiter, W.H. (1983), “Measurement of the public sector deficit and its implications for policy evaluation and design.” International Monetary Fund Staff Papers, Vol. 30, No. 2, (June), International Monetary Fund, Washington, DC, pp. 306-349.

Buiter, W.H. (1985), “A Guide to Public Sector Debt and Deficits”, Economic Policy, Vol. 1, Oxford University Press, Oxford, pp. 612-35.

Cadete de Matos J., S. Branco and F. Morais (2015), Conceptual Issues Related to the Definition of Government Debt, presented at the 60th World Statistics Conference, 26‐30 July, Rio de Janeiro.

Dippelsman, R., C. Dziobek and C. Gutierrez (2012), “What lies beneath: The statistical definition of public sector debt. An overview of the coverage of public sector debt for 61 countries”, IMF Staff Discussion Notes, 12/09, International Monetary Fund, Washington, DC,

ECB (2017), Statistical Data Warehouse, European Central Bank, Frankfurt, http://sdw.ecb.

Elton, E.J. et al. (2003), Modern Portfolio Analysis, J. Wiley and Sons, Hoboken.

Estrada, M., D. Igan and D. Knight (2014), “Fiscal Risks and Borrowing Costs in State and Local Governments”, in IMF Country Report No. 14/222, International Monetary Fund, Washington, DC.

European Commission (2014), “Assessing public debt sustainability in EU member states: A guide”, European Economy, Occasional Papers, No. 200, European Commission, Brussels,

Eurostat, Manual on Government Deficit and Debt 2016, Eurostat, Luxembourg,

IMF (2017), World Economic Outlook Database, April 2017, International Monetary Fund, Washington, DC,

IMF (2011), Modernizing the Framework for Fiscal Policy and Public Debt Sustainability Analysis, International Monetary Fund, Washington, DC,

IMF (2002), Assessing Sustainability, International Monetary Fund, Washington, DC,

IMF and other international organizations, Public Sector Debt Statistics: Guide for Compilers and Users 2011, International Monetary Fund, Washington, DC,

Kotlikoff, L.J. (1984), “Economic Impact of Deficit Financing”, IMF Staff Papers, Vol. 33, International Monetary Fund, Washington, D.C., pp. 549-581.

O’Connor, L., E. Weisman and T. Wickens (2004), GFSM 2001 Companion Material, Consolidation of the General Government Sector, International Monetary Fund, Washington, DC,

OECD (2017), “Financial Balance Sheets, SNA 2008 (or SNA 1993): Consolidated stocks, annual”, OECD National Accounts Statistics (database),; and“Public Sector Debt: Public sector debt – consolidated”, OECD National Accounts Statistics (database),


← 1. Debt sustainability analysis also takes into account maturity, currency composition, off-balance sheet risks such as guarantees, and public corporation debt. See IMF, Modernizing the Framework for Fiscal Policy and Public Debt Sustainability Analysis.

← 2. See Dippelsman et al. (2012) which shows that institutional and instrument coverage of data published as government debt varies widely. For example, some countries include pension liabilities, while in other countries these are recorded in supplementary tables or not at all. It adopts a summary terminology (D1, D2, etc.) to highlight the differences.

← 3. These supplementary tables only become available at the end of 2017, or in the course of 2018, for OECD-countries.

← 4. Exceptions of the latter treatment within the OECD are Japan and Korea.

← 5.

← 6. For neutrality of consolidation on the GFSM balancing items (including the net operating balance, net lending/net borrowing and net worth, or net debt) for the aggregate public sector, see IMF et al. Public Sector Debt Statistics: Guide for Compilers and Users 2011, Box 8.1.

← 7. The Eurostat Manual on Government Deficit and Debt (MGDD) provides very detailed guidance on the recording of specific cases. See

← 8. Bodie, Z., A. Kane and A.J. Marcus (2005), “Investments”. McGraw-Hill; see also Chapter 13 in Elton, E.J, M.J. Gruber, S.J. Brown and W.N. Goetzman (2003), “Modern Portfolio Analysis”, J. Wiley and Sons.

← 9. See ESA 2010, paragraph 20.275. This interpretation would allow to neutralise the effects of pension transfers even when the pension obligations are merged with a social security scheme, for which no core-liability is recognized in any statistical standard.