IMF Staff Discussion Note
While key macroeconomic indicators such as Gross Domestic Product (GDP) or Consumer Price Index (CPI) are based on internationally accepted methodologies, indicators related to the debt of the public sector often do not follow international standards and can have several different definitions. As this paper shows, the absence of the standard nomenclature can lead to major misunderstandings in the fiscal policy debate. The authors present examples that show that debt-to-GDP ratios for a country at any given time can range from 40 to over 100 percent depending on the definition used. Debt statistics, for example, may include or exclude state and local governments and may cover all debt instruments or just a subset. The authors suggest that gross debt of the general government (―gross debt‖) should be globally adopted as the headline indicator supplemented by other measures of government debt for risk-based assessments of the fiscal position. Broader measures, including net debt and detailed information on contingent liabilities and derivatives, could be considered. The standard nomenclature of government and of debt instruments helps users understand the concepts in line with the Public Sector Debt Statistics Guide. Use of more standard definitions of government debt would improve data comparability, would benefit IMF surveillance, programs, and debt sustainability analysis, and would help country authorities specify and monitor fiscal rules. Data disaggregated by government subsector and debt instrument for 61 countries from the IMF‘s Government Finance Statistics Yearbook (GFSY) database are presented to illustrate the importance and viability of adopting this approach.
Most key macroeconomic indicators such as GDP, the consumer price index (CPI), data on monetary aggregates or balance of payments follow internationally accepted definitions. In contrast, countries often do not follow international guidelines for public debt data. As this paper shows, failure to apply global standards can lead to important misunderstandings because of the potentially large magnitudes involved. International guidelines on the compilation of public sector debt are well established and are summarized in the recently published Public Sector Debt Statistics Guide (Debt Guide). The Debt Guide also describes applications of these guidelines for the analysis of debt sustainability, fiscal risk, and vulnerability.
The authors seek in this paper to provide a more intuitive application of the various concepts and definitions found in the Debt Guide, and propose that global standard definitions of ―gross debt‖ referring to the ―general government‖ be adopted as a headline measure. As with other headline indicators, a variety of narrower and wider indicators remain valuable and useful for different purposes. The notion of gross debt will be familiar to macroeconomic statisticians, but, as a practical matter, the adoption of global standard statistical definitions of debt will require some development efforts in terms of source data availability and training for compilers of debt statistics. A particular challenge is complete coverage of all relevant institutions and financial instruments. Detailed information on contingent liabilities and derivatives should also be considered. Coordination across agencies that work with debt related data is also critical, as with other complex datasets such as GDP.
Many users are not aware of the extent to which differences in concepts and methods matter. Box 1 below highlights the four key dimensions of public sector debt. Countries publish data, for example, either including or excluding state and local governments, pension funds, and public corporations. Also, while much of the policy debate centers on government liabilities, some countries have begun to publish and focus policy analysis on net debt (financial assets minus liabilities). Debt data frequently only include two (of the six) debt instruments available: debt securities and loans. Debt instruments such as other accounts payable or insurance technical reserves are often not taken into account. In many cases the method of valuation is not explicitly mentioned even though market versus nominal valuation can be significantly different. Consolidation, which refers to the process of netting out intra-governmental obligations, is another important factor rarely specified in published data. And finally, debt data may be compiled using cash data and excluding non-cash items such as arrears or using accrual (or partial accrual) methods to reflect important non-cash obligations.
Box 1. Key Dimensions to Measure Government Gross Debt
Institutional Coverage of Government
Instrument Coverage of Debt
Valuation of Debt Instruments (market and nominal)
Consolidation of Intra-Government Holdings
Source: Public Sector Debt Statistics Guide.
The headline indicator for government debt should be defined as ―gross debt of the general government‖ or GL3/D4 in this paper‘s nomenclature. The authors suggest that countries should aspire to publish timely data on the broader concept of gross debt.
Data on the institutional level of the general government (GL3) would be consistent with a broad range of data uses and with the data requirements of other macroeconomic datasets, notably the national accounts. Including the full range of debt instruments is desirable particularly because some of these may expand in times of financial distress and could thus serve as valuable indicators of distress. Clarity of what the debt data cover would help build understanding of the data and their comparability across countries.
A global standard would facilitate communication on the main concepts in public sector debt statistics and it would bring greater precision to research on fiscal issues, and lead to improved cross-country comparison. This framework uses a nomenclature inspired by the approach in monetary data where M1 through M4 (monetary aggregates) reflect institutional and instrument coverage as well.
The methodological framework of government debt presented here is widely accepted among statisticians. The relevant definitions, concepts, classification, and guidance of compilation are summarized in GFSM 2001 and the Debt Guide. These standards are fully consistent with the overarching statistical methodology of the 2008 SNA and other international macroeconomic methodologies such as the Sixth Edition of Balance of Payments and International Investment Position Manual (BPM6) and broadly consistent with the European System of Accounts (ESA) manual and the more specialized manuals of deficit and debt that govern the Excessive Deficit Procedure.
However, the methodology is not always well defined in the policy debate. An international convention to view GL3/D4 as the desirable headline indicator of government debt, consistent with the international standards, would go a long way to create more transparency and better comparability of international data.
Our contribution is to provide a presentational framework and nomenclature that highlights the importance of different instruments, institutional coverage, and valuation and consolidation as key indicators of debt. Indeed, we have noted that other, more narrowly defined concepts can meaningfully supplement the comprehensive measure of debt. These narrower measures may be important for a risk-based assessment of the fiscal position, but they are not substitutes for a global indicator.
Further extensions of this work are the development of the statistical reporting of broader measures, for example net debt of the general government and the presentation of information on derivatives, and contingent liabilities.
The new debt database launched by the IMF and World Bank in 2010 is structured along government levels, debt instruments, consolidation and valuation as discussed in this paper. However, some countries report data only on the GL2 level and cover mostly D1. Developing data on the broader statistics will take some time, although Australia, Canada, and some other countries already publish or plan to publish GL3/D4 data or publish components that would allow the calculation of GL3/D4.
Debt statistics for various levels of government and instruments were shown for 61 countries and these data highlight some interesting patterns that merit further analysis such as the degree of fiscal autonomy of state and local government to issue debt, the degree of development of markets for government debt securities. The authors conclude that further research would be worthwhile on the advantages of a global standard of government debt for such topics as data comparability, IMF surveillance, programs, debt sustainability analysis, and the analysis of fiscal rules.