A. The GFSM 2001 Analytical Framework

23. GFSM 2001 harmonizes the system used to report fiscal statistics with other macroeconomic statistical systems— most notably with the national accounts (the 1993 SNA) and, therefore, also the European System of Accounts 1995 (1995 ESA). In this way, GFSM 2001 formalizes and standardizes many adjustments made by Fund staff to the fiscal statistics provided by country authorities for inclusion in Fund staff reports. Moreover, fiscal accounting and reporting are brought closer into line with private sector practice (that is, a profit and loss statement, a balance sheet, and a cash statement). The linkages in the GFSM 2001 framework are summarized in Table 1.

Table 1.

The GFSM 2001 Analytical Framework

Table 1.

The GFSM 2001 Analytical Framework

Opening Balance Sheet		Government Operations		Other Economic Flows		Closing Balance Sheet
		Revenue
		-
		Expense
		=
		Net operating balance
		=
Net worth	+	Change in net worth due to revenue and expense transactions	+	Change in net worth due to other economic flows	=	Net worth
=		=		=		=
Non financial assets	+	Net acquisition of non financial assets	+	Change in non financial assets due to other economic flows	=	Non financial Assets
+		+		+		+
Net financial worth	+	Net lending/borrowing	+	Change in net financial worth due to other economic flows	=	Net financial Worth
=		=		=		=
Financial assets	+	Net acquisition of financial assets	+	Change in financial assets due to other economic flows	=	Financial assets
-		-		-		-
Liabilities	+	Net incurrence of liabilities	+	Change in liabilities due to other economic flows	=	Liabilities

View Table

Table 1.

The GFSM 2001 Analytical Framework

Opening Balance Sheet		Government Operations		Other Economic Flows		Closing Balance Sheet
		Revenue
		-
		Expense
		=
		Net operating balance
		=
Net worth	+	Change in net worth due to revenue and expense transactions	+	Change in net worth due to other economic flows	=	Net worth
=		=		=		=
Non financial assets	+	Net acquisition of non financial assets	+	Change in non financial assets due to other economic flows	=	Non financial Assets
+		+		+		+
Net financial worth	+	Net lending/borrowing	+	Change in net financial worth due to other economic flows	=	Net financial Worth
=		=		=		=
Financial assets	+	Net acquisition of financial assets	+	Change in financial assets due to other economic flows	=	Financial assets
-		-		-		-
Liabilities	+	Net incurrence of liabilities	+	Change in liabilities due to other economic flows	=	Liabilities

24. A key feature of GFSM 2001 is its distinction between transactions and other economic flows. Transactions cover all exchanges or transfers that take place by mutual agreement and the consumption of fixed capital (the economic equivalent of “depreciation”). Mutual agreement does not mean that transactions have to be entered into voluntarily (the payment of taxes is treated as a transaction despite being compulsory). Additionally, transactions cover monetary exchanges and in-kind activity (such as the receipt of commodity grants and noncash remuneration). Other economic flows are the result of events that affect the value of nonfinancial assets, financial assets, and liabilities but that are not exchanges or transfers. These flows can reflect either price changes (including exchange rate movements) or volume changes due to one-off events (such as mineral discoveries or natural disasters).

25. A set of well-defined relationships between flows and stocks—presented in integrated balance sheets—underpins GFSM 2001. Specifically, the framework reconciles the government’s opening and closing balance sheets with the flows derived from government operations and the other economic flows.

26. While countries can apply the GFSM 2001 framework to cash data, ideally, transactions and other economic flows should be recorded on an accrual basis. This means countries record these flows when the economic consequences associated with an event occur or when future consequences can be measured reliably. Thus they record an expense when the government uses resources. This recording will, in practice, usually be when government incurs an obligation to pay for those resources (e.g., when contracted goods are delivered as stipulated in the contract), rather than when it actually pays for the resources. An obligation to pay is distinct from a commitment, which occurs when contracts are signed, orders are placed, etc.

27. Similarly, in principle, revenue should be recorded when a liability to government is exchanged and not when payment is made. However, difficulties exist in identifying revenue on an accrual basis, and especially in determining precisely when an activity gives rise to a tax liability. In practice, governments would normally record a tax liability at the time of assessment and base it on the amount they realistically expect to collect. Governments also record transactions in nonfinancial assets, financial assets, and liabilities at the time assets change ownership and liabilities are incurred. Other economic flows accrue when an economic asset (or liability) is created, extinguished, or transformed. GFSM 2001 increases the comprehensiveness of fiscal data through accrual reporting, allows data inconsistencies to be detected, and thereby strengthens fiscal transparency—all of which will aid fiscal analysis.

28. At the same time, it should be emphasized that the GFSM 2001 system continues to recognize the importance of monitoring cash flows and the crucial role these play in fiscal analysis. Therefore, a separate statement on the sources and uses of cash is an integral part of the GFSM 2001 statistical framework. This statement shows fiscal flows in terms of pure cash, while it follows the same basic structure as the statement on operations (Section III.B).

B. GFSM 2001 Fiscal Tables

29. The GFSM 2001 analytical framework can be summarized as a set of three fiscal tables.¹¹

• The Statement of Government Operations distinguishes between revenue and expense (or operating) transactions and among transactions in nonfinancial assets, financial assets, and liabilities. Revenue covers all transactions that increase net worth, and expense covers all transactions that decrease net worth (including importantly the consumption of fixed capital). In essence, this balance is akin to the government’s profit and loss statement for a given year. Transactions in nonfinancial assets, financial assets, and liabilities are not included in revenue or expense.¹² Under GFSM 1986, transactions in nonfinancial assets—and in financial assets for public policy (as distinct from liquidity management) purposes—are treated as expenditure. But under GFSM 2001, they are not treated as an expense because they are exchanges that do not affect net worth. This removes the asymmetry that exists in the GFSM 1986 system “above-” and “below-the-line.”¹³

• The difference between revenue and expense is the net operating balance. Subtracting the net acquisition of nonfinancial assets from the net operating balance yields net lending/borrowing, which in turn is equal to the net acquisition of financial assets less the net incurrence of liabilities (that is, government’s financing).

• The Balance Sheet shows the government’s net worth at the end of a fiscal year, which is equal to the stock of nonfinancial assets plus net financial worth (i.e., the difference between the stocks of financial assets and liabilities). The change in net worth during a year is the sum of changes due to revenue and expense transactions and to other economic flows. An integrated balance sheet shows the opening balance of assets and liabilities, as well as transactions and other economic flows in these assets and liabilities that explain the closing balance of these assets and liabilities.

• The Statement of Sources and Uses of Cash 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 the net change in the stock of cash. This cash flow statement provides useful information to link the fiscal impacts with relevant monetary variables.

C. Valuation

30. In general, flows and stocks should be valued at market prices in the GFSM 2001 system. This is the amount for which the goods, services, assets, labor, or the provision of capital are in fact exchanged (including the cash value of in-kind transactions). Flows should be valued at the prices current on the dates when the transaction accrues. Stocks should be valued at the prices current on the balance sheet date.¹⁴ Market prices are used to ensure consistency within the GFSM 2001 system and across macroeconomic statistics.

31. Although the balance sheet is to be valued at market prices, GFSM 2001 provides for reporting the nominal value of the debt as a memorandum item. The nominal value of the debt reflects the original (contractual) value of the debt and the impact of subsequent economic flows, such as transactions (e.g., accrual of interest, repayment of principal), revaluations, and other economic flows.¹⁵ As such, it measures the amount that debtors owe to creditors—relevant from the point of view of assessing the fiscal policy implications of debt. In particular, it is relevant to the extent to which a high debt level calls for fiscal adjustment to avoid solvency or liquidity problems or, more generally, to provide room for fiscal policy maneuver.

32. As a practical consideration, valuation of nonfinancial assets can present some difficulties. However, national accountants and engineers in line ministries apply techniques that often provide reasonable estimates for such assets and the associated consumption of fixed capital.¹⁶ Underlying accounting information may also be useful. Nonetheless, this usually needs to be adjusted to consider the economic value of resource flows rather than the accounting principles that may be sensitive to the tax environment.

D. Fiscal Indicators

33. The core GFSM 2001 fiscal indicators are the net operating balance, net lending/borrowing, net worth, and the cash surplus/deficit. With more than a single core balance, GFSM 2001 does not refer to “above-the-line” or “below-the-line” transactions.

• The net operating balance indicates the impact of fiscal policy from operating transactions that affect net worth (including the consumption of fixed capital). However, frequently, data on the consumption of fixed capital are not available and the gross operating balance is used. The consumption of fixed capital is the economic equivalent of depreciation. It measures the decline in the current market value of the stock of fixed assets during the accounting period as a result of physical deterioration, normal obsolescence, and accidental damage.¹⁷ The techniques to measure the consumption of fixed capital vary across countries; nevertheless, the gross operating balance always remains internationally comparable.

• Net lending/borrowing is perhaps the single most important fiscal indicator, since it reflects the government’s financing operations. It summarizes the way in which fiscal policy in the operating transactions and the nonfinancial asset transactions affects the rest of the economy and the rest of the world—indicating the impact on the government’s use of resources and on aggregate demand. It is the closest comparable measure to the overall balance used in GFSM 1986.

• Net worth and the changes in net worth are relevant to analyzing fiscal sustainability in that, as opposed to focusing on debt alone, they take into account both the government’s assets and its liabilities.¹⁸ However, debt and debt sustainability remain important, since governments can run into solvency and liquidity problems independently of their net worth, in part because some assets may not be readily marketable. For that reason, fiscal analysts will focus mainly on net financial worth (the difference between financial assets and liabilities), which abstracts from changes in nonfinancial assets.

• The cash surplus/deficit measures the change in the government’s liquidity position owing to revenue and expense transactions and transactions in nonfinancial assets, based purely on cash flows. It is thus the cash equivalent of net lending/borrowing. The net change in the stock of cash reflects the cash flow associated with the net acquisition of financial assets (other than cash) and the net incurrence of liabilities. It also measures the change in the government’s overall liquidity position. As such, it better indicates the cash- flow implications of government operations.

34. GFSM 2001 recognizes that a wider range of supplementary fiscal indicators may continue to be useful in particular circumstances. For example, analysts can adjust net lending/borrowing to yield a primary balance and other indicators judged analytically useful in the previous framework.¹⁹ Another possibility would be to derive an overall balance from net lending/borrowing. That is, analysts could group transactions in financial assets, undertaken for public policy purposes, with transactions in nonfinancial assets, treating sales of nonfinancial assets as transactions in financial assets.²⁰ The principal advantage of deriving such an overall balance would be that it signals the impact of fiscal policy in a familiar way. Thus, as a transitional arrangement, its use should continue only until some experience has been gained and a certain comfort level achieved using GFSM 2001 fiscal indicators. In the traditional fiscal analysis, problems have arisen when recording economic events, such as investment in nonfinancial assets with dubious value and loans that are unlikely to be repaid. In the past, staff have had to make adjustments to the data to account for these events. However, by using market prices (Section III.C) and recognizing holding gains or losses as other economic flows, GFSM 2001 contains mechanisms to record such events consistently and transparently.

E. Advantages of GFSM 2001

35. The GFSM 2001 framework can enhance the quality of fiscal analysis:

(1) Foremost, it provides a basis for strengthening fiscal sustainability analysis through evaluation of the changes in net worth. To this end, the Fund’s debt sustainability template would need to fully integrate stocks and flows comparably to the GFSM 2001 framework.

(2) GFSM 2001 yields measures of government saving, investment, and consumption that are better suited for inclusion in a quantitative macroeconomic framework than the analogous GFSM 1986 concepts were. This is because GFSM 2001 has been harmonized with the national accounts framework in the 1993 SNA. In the past, staff have had to adjust (often quite arbitrarily) the traditional fiscal variables for this purpose.²¹

(3) Since the net operating balance and net lending/borrowing in GFSM 2001 are counterparts to the current balance and the overall balance respectively, they are well suited to drawing more attention to public investment. The framework also provides a basis for recognizing that public investment creates assets and that debt accumulated to finance public investment need not reduce net worth. More generally, GFSM 2001 supports a shift to a balance sheet approach to analyzing economic policy.

(4) GFSM 2001 standardizes many adjustments already made in staff reports. The framework eliminates the asymmetries in the previous set of fiscal data, enhancing clarity in the analysis. More comprehensive analytically, the framework captures noncash transactions coherently and consistently.

(5) Finally and in a similar vein, GFSM 2001 provides a “common language” that fiscal analysts can use to develop a consistent approach to handling new, and often complex, government operations—operations that create challenges in fiscal reporting and analysis (e.g., bank recapitalization, securitization, and public-private partnerships).

A. Fiscal Tables in Fund Reports

37. The ultimate objective is that fiscal tables in Fund reports will be presented in GFSM 2001 format and in accordance with GFSM 2001 concepts. Fund staff reports typically present the government accounts on a basis consistent with the national presentation (often similar to GFSM 1986 as well), albeit with some modifications, such as for privatization proceeds. Countries that have well-developed accrual accounting, recording, and reporting systems may be able to produce a set of GFSM 2001 tables. However, the vast majority of countries will need time to achieve this result. In the interim, all countries can quickly make a simple mapping of their existing data into the GFSM 2001 format. Thus, staff propose to pilot this gradual adoption of GFSM 2001 presentation in staff reports, beginning initially with countries that volunteer for this exercise (see Section V).

38. The easiest first step will be to merely reclassify the existing data. This involves moving most of the capital revenue and expenditure items to net acquisition of nonfinancial assets and moving lending minus repayments to net acquisition of financial assets, as illustrated in Table 2. If possible, data compilers could make distinctions between financial assets and liabilities. This would immediately yield a GFSM 2001 statement of government operations. This and two other fiscal tables are outlined in Section V, which explores using GFSM 2001 for fiscal analysis. Whatever fiscal tables are presented, it is necessary at all times for compilers to indicate the table coverage, the accounting basis of the information reported (i.e., pure cash, full accrual, or something in between), and any significant departures from the coverage and basis indicated.

Bridge Between Typical Staff Report andGFSM 2001 Presentation of Operations

Citation: Policy Papers 2005, 082; 10.5089/9781498330916.007.A001

a/ Consists of sales of fixed capital assets, strategic stocks, land, and intangibles in this example, thereby assuming capital transfers received from nongovernment sources are zero. Capital transfers from nongovernment sources are classified to GFSM 2001 Revenue.b/ Consists of purchases of fixed capital assets, strategic stocks, land, and intangibles.c/ GFSM 2001 Revenue = Total revenue and grant minus Capital revenue (sales of fixed assets, strategic stocks, land, and intangibles). [62,300=62,600-300]d/ GFSM 2001 Expense = Total expenditure and net lending minus Purchases of fixed assets minus Net lending. [72,700=78,500-5,100-700]; or GFSM 2001 Expense = Current expenditure plus capital transfers. [72,700=68,800+3,900]e/ Gross operating balance = Revenue minus Expense (excluding Consumption of fixed capital). [-10,400=62,300-72,700] Net operating balance = Revenue minus Expense (including Consumption of fixed capital). In this example, Consumption of fixed capital is not available and thus the Net operating balance is not calculated.f/ GFSM 2001 Net acquisition of nonfinancial assets = Purchases of fixed capital assets, strategic stocks, land, and intangibles minus Capital revenue (Sales of fixed capital assets, strategic stocks, land, and intangibles). [4,800=5,100-300]g/ Net lending (+) / borrowing (-) = Gross (Net) operating balance minus Net acquisition of nonfinancial assets. [-15,200=-10,400-4,800] Note that in the staff report, the overall balance is equal, but with an opposite sign, to total financing; whereas the GFSM 2001 Net lending/borrowing is equal and with the same sign as Net acquisition of financial assets minus Net incurrence of liabilities.h/ GFSM 2001 Net acquisition of financial assets comprises domestic and foreign Net lending, Sales of equity (such as privatization proceeds), and Changes in cash and deposits. In this example, for simplicity, it is assumed that there are no transactions in financial assets held abroad. Accordingly, in this example, domestic Net lending [700] and Sales of equity [3,500] are classified, along with Changes in cash and deposits held domestically [4,400] to equal Net acquisition of financial assets (domestic): [700-4,400-3,500=-7,200]. In the staff report, a decrease in financial assets classified as financing is shown with a positive sign and an increase with a negative sign. In the GFSM 2001 presentation, a decrease in financial assets is shown with a negative sign and an increase with a positive sign.i/ GFSM 2001 Net incurrence of liabilities = Net domestic borrowing and Net foreign borrowing. [8,000=2,400+5,600].

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Bridge Between Typical Staff Report andGFSM 2001 Presentation of Operations

Citation: Policy Papers 2005, 082; 10.5089/9781498330916.007.A001

a/ Consists of sales of fixed capital assets, strategic stocks, land, and intangibles in this example, thereby assuming capital transfers received from nongovernment sources are zero. Capital transfers from nongovernment sources are classified to GFSM 2001 Revenue.b/ Consists of purchases of fixed capital assets, strategic stocks, land, and intangibles.c/ GFSM 2001 Revenue = Total revenue and grant minus Capital revenue (sales of fixed assets, strategic stocks, land, and intangibles). [62,300=62,600-300]d/ GFSM 2001 Expense = Total expenditure and net lending minus Purchases of fixed assets minus Net lending. [72,700=78,500-5,100-700]; or GFSM 2001 Expense = Current expenditure plus capital transfers. [72,700=68,800+3,900]e/ Gross operating balance = Revenue minus Expense (excluding Consumption of fixed capital). [-10,400=62,300-72,700] Net operating balance = Revenue minus Expense (including Consumption of fixed capital). In this example, Consumption of fixed capital is not available and thus the Net operating balance is not calculated.f/ GFSM 2001 Net acquisition of nonfinancial assets = Purchases of fixed capital assets, strategic stocks, land, and intangibles minus Capital revenue (Sales of fixed capital assets, strategic stocks, land, and intangibles). [4,800=5,100-300]g/ Net lending (+) / borrowing (-) = Gross (Net) operating balance minus Net acquisition of nonfinancial assets. [-15,200=-10,400-4,800] Note that in the staff report, the overall balance is equal, but with an opposite sign, to total financing; whereas the GFSM 2001 Net lending/borrowing is equal and with the same sign as Net acquisition of financial assets minus Net incurrence of liabilities.h/ GFSM 2001 Net acquisition of financial assets comprises domestic and foreign Net lending, Sales of equity (such as privatization proceeds), and Changes in cash and deposits. In this example, for simplicity, it is assumed that there are no transactions in financial assets held abroad. Accordingly, in this example, domestic Net lending [700] and Sales of equity [3,500] are classified, along with Changes in cash and deposits held domestically [4,400] to equal Net acquisition of financial assets (domestic): [700-4,400-3,500=-7,200]. In the staff report, a decrease in financial assets classified as financing is shown with a positive sign and an increase with a negative sign. In the GFSM 2001 presentation, a decrease in financial assets is shown with a negative sign and an increase with a positive sign.i/ GFSM 2001 Net incurrence of liabilities = Net domestic borrowing and Net foreign borrowing. [8,000=2,400+5,600].

• Download Figure
• Download figure as PowerPoint slide

Bridge Between Typical Staff Report andGFSM 2001 Presentation of Operations

Citation: Policy Papers 2005, 082; 10.5089/9781498330916.007.A001

a/ Consists of sales of fixed capital assets, strategic stocks, land, and intangibles in this example, thereby assuming capital transfers received from nongovernment sources are zero. Capital transfers from nongovernment sources are classified to GFSM 2001 Revenue.b/ Consists of purchases of fixed capital assets, strategic stocks, land, and intangibles.c/ GFSM 2001 Revenue = Total revenue and grant minus Capital revenue (sales of fixed assets, strategic stocks, land, and intangibles). [62,300=62,600-300]d/ GFSM 2001 Expense = Total expenditure and net lending minus Purchases of fixed assets minus Net lending. [72,700=78,500-5,100-700]; or GFSM 2001 Expense = Current expenditure plus capital transfers. [72,700=68,800+3,900]e/ Gross operating balance = Revenue minus Expense (excluding Consumption of fixed capital). [-10,400=62,300-72,700] Net operating balance = Revenue minus Expense (including Consumption of fixed capital). In this example, Consumption of fixed capital is not available and thus the Net operating balance is not calculated.f/ GFSM 2001 Net acquisition of nonfinancial assets = Purchases of fixed capital assets, strategic stocks, land, and intangibles minus Capital revenue (Sales of fixed capital assets, strategic stocks, land, and intangibles). [4,800=5,100-300]g/ Net lending (+) / borrowing (-) = Gross (Net) operating balance minus Net acquisition of nonfinancial assets. [-15,200=-10,400-4,800] Note that in the staff report, the overall balance is equal, but with an opposite sign, to total financing; whereas the GFSM 2001 Net lending/borrowing is equal and with the same sign as Net acquisition of financial assets minus Net incurrence of liabilities.h/ GFSM 2001 Net acquisition of financial assets comprises domestic and foreign Net lending, Sales of equity (such as privatization proceeds), and Changes in cash and deposits. In this example, for simplicity, it is assumed that there are no transactions in financial assets held abroad. Accordingly, in this example, domestic Net lending [700] and Sales of equity [3,500] are classified, along with Changes in cash and deposits held domestically [4,400] to equal Net acquisition of financial assets (domestic): [700-4,400-3,500=-7,200]. In the staff report, a decrease in financial assets classified as financing is shown with a positive sign and an increase with a negative sign. In the GFSM 2001 presentation, a decrease in financial assets is shown with a negative sign and an increase with a positive sign.i/ GFSM 2001 Net incurrence of liabilities = Net domestic borrowing and Net foreign borrowing. [8,000=2,400+5,600].

• Download Figure
• Download figure as PowerPoint slide

39. The transition from the current presentation to the GFSM 2001 presentation should not compromise the quality of fiscal policy analysis in Fund staff reports. In view of country familiarity with the overall balance as the principal fiscal policy indicator in Fund work, analysts should make available a bridge table explaining the differences between GFSM 2001 “net lending/borrowing” and the traditional overall balance used in a particular country. Analysts should also document this bridge table and the ways in which source data are used. The bridge table and its documentation will be needed until the information requirements of GFSM 2001 are well understood and can routinely be met.

40. The changeover to the GFSM 2001 presentation could nevertheless produce a break in the time series of fiscal data. If past fiscal statistics are available in sufficient detail, data compilers should find it straightforward to bring the past presentation into line with the GFSM 2001 presentation. However, in practice, the required data may not be available, and the prospects for constructing historical series could range from difficult to impossible. Consequently, a break may have to exist in the time series of fiscal data. For the period (or periods) in which this occurs, some overlapping fiscal data should be presented and the differences should be fully documented. A break in the data would limit the scope to analyze longer-term fiscal trends. However, as countries continue along the path of implementing GFSM 2001, this break will fade to the past, as better data become available.

B. Reporting Fiscal Data to the Fund

41. Member countries are developing the capacity to report fiscal statistics to STA using the GFSM 2001 framework. STA has disseminated the GFSM 2001 methodology among data compilers through training courses and technical assistance. Dissemination also takes place during the annual interaction between staff and member country compilers when preparing data submissions for publication in the Fund’s Government Finance Statistics Yearbook (GFSY). Evidence of the progress is reflected in the growing number of countries that have reported annual data for the GFSY using the GFSM 2001 framework. In 2002, STA assisted 22 countries (mainly in Europe) to report data in the GFSM 2001 framework, by taking available fiscal data (e.g., reported to Eurostat) and bridging them to the GFSM 2001 framework. Of the 99 member countries that reported data for publication in the GFSY for the 2003–04 yearbooks, about 79 percent (78 countries) submitted data using the GFSM 2001 framework (reporting both flows and stocks). Among those reporters, about 70 percent were in developing and emerging market countries.

42. In addition, STA has initiated work on developing a higher frequency fiscal database (monthly and quarterly), using the GFSM 2001 framework. STA is focusing initially on collaborating with Eurostat and developing an integrated data transmission framework to be used for monitoring fiscal performance. While this work is intended to affect fiscal data presented in the International Financial Statistics, work has not yet begun to modify the presentation of fiscal data in the World Economic Outlook database.

43. Reporting to area departments in the GFSM 2001 format has just begun. The main task will be to determine how this can be done without compromising surveillance and program activities in the short term, yet making surveillance and programs more effective over the longer term. Country compilers could use various approaches to report fiscal data to area departments for surveillance and program purposes, depending on country circumstances. Most countries would reclassify their currently available fiscal data into GFSM 2001 format. Others will rely mainly on accrual-based national accounts data that will need to be supplemented with additional information gathered by ministries of finance and others. A few countries that have developed accrual-based data collection systems should be able to move quickly to reporting fully on a GFSM 2001 basis. While these options are clearly appropriate for surveillance countries, they may not be appropriate in a program context. Introducing the GFSM 2001 framework in the context of ongoing Fund programs could pose risks if two sets of fiscal data were to be presented for analysis simultaneously.

C. Fiscal Targets Under Fund-Supported Programs

44. Shifting to GFSM 2001 fiscal tables and indicators can benefit program design and monitoring. Clearly, using fiscal statistics that are fully GFSM 2001 compliant for program purposes (provided they are timely) would bring all the advantages that GFSM 2001 offers for fiscal transparency and fiscal analysis more generally. It would reduce the country-specific adjustments needed in staff reports, enhance cross-country comparability by eliminating asymmetries in fiscal statistics, and promote intersectoral consistencies. By integrating stocks and flows, GFSM 2001 allows earlier identification and preparation of measures to address fiscal data quality problems, as well as possible circumvention of fiscal performance criteria and misreporting, than at present. This advantage increases with the amount of accrual information that is reflected in fiscal statistics. Also increased is the usefulness of GFSM 2001 fiscal indicators—compared to the overall balance—in conveying information about the impact of fiscal policy.²²

45. Nevertheless, data quality is a key consideration in designing fiscal performance criteria under Fund-supported programs. Of the various dimensions of data quality, timeliness is usually the binding constraint, given the needs of program design and monitoring. This has meant that program targets are mainly specified in terms of cash-based or commitment/adjusted cash-based fiscal indicators. Moreover, no reason exists to expect any significant difference in this regard while countries are implementing GFSM 2001. The only likely difference in the short term is that countries will make a start toward defining fiscal targets and performance criteria in terms of GFSM 2001 fiscal indicators. However, in this connection, a considerable premium has to be placed on building confidence in working with GFSM 2001 fiscal tables and indicators and on developing an awareness of operational risks they might entail. The risks arise primarily from additional opportunities for creative accounting that result from the shift to fiscal indicators; such accounting includes accrual adjustments subject to some discretion (e.g., in calculating consumption of fixed capital).²³ This being the case, using the GFSM 2001 statistical framework in new programs would only start after country teams gain more experience. Moreover, once a decision is taken to shift to the GFSM 2001 presentation, changes in the fiscal tables and indicators used for program purposes should only be considered for new programs

D. Coverage of Fiscal Statistics

46. The core focus of GFSM 2001 is the general government. All countries should achieve comprehensive coverage of central government on a cash basis—as a minimum reasonable objective—for fiscal reporting purposes. However, when subnational governments are of clear fiscal policy relevance, countries would be well advised to expand coverage to include at least the most significant parts of subnational governments. Many countries already do so. Countries should routinely expand coverage from the central to the general government. For many, the capacity to implement GFSM 2001 will be greater at the central government level than for subnational governments; however, capacity limitations in subnational governments will ultimately determine the overall pace of GFSM 2001 implementation. Countries should eventually seek to report GFSM 2001-compliant fiscal statistics for the general government.

47. GFSM 2001 also encourages countries to report fiscal statistics covering the public corporations sector and the public sector as a whole. Given that any public enterprise is a potential source of fiscal risk (if not now, then possibly in the future), countries should report and monitor the operations of all public enterprises in the fiscal statistics. Achieving this objective could meet significant resistance, especially if countries have not done this in the past. Resistance may reflect information and capacity problems, a need to strengthen fiscal management, or different priorities. (For example, it may be more important to cover extrabudgetary or subnational government fiscal activities.) The aim should be to focus in the first instance on enterprises (and subsectors of the general government) that pose the largest fiscal risk. Coverage can then be expanded further as priorities change. However, in principle, fiscal indicators and targets used for Fund surveillance and program purposes²⁴ should cover all public enterprises that pose significant fiscal risk.

E. Institutional Requirements

48. Implementing the full GFSM 2001 remains the responsibility of the national authorities. All countries will be encouraged by the Fund staff to develop a migration strategy that has, as its ultimate objective, producing fiscal tables and the corresponding detailed data consistent with GFSM 2001.

49. Fully implementing GFSM 2001 will be a major task for most countries. The task will require careful planning and management to avoid disrupting the flow of fiscal statistics. Further, the capacity of governments to train, recruit, and retain skilled staff to work for national statistical agencies, ministries of finance, other government agencies, and, where appropriate, the wider public sector will be important at every stage of migration to GFSM 2001.

50. The proposed approach to implementing GFSM 2001 should be a migration strategy tailored to institutional capacity. The basic elements of such a strategy are the following:

1. reorganizing and presenting existing statistical information using the GFSM 2001 framework;

2. developing a legal environment that provides for compiling cash- and accrual-based fiscal statistics and that assigns institutional responsibility for compiling statistics on the public sector and its subsectors; and

3. introducing improvements in the underlying accounting and classification systems, including introducing a chart of accounts, a general ledger, accrual-based accounting standards, and an automated financial management information system.

51. In developing a migration strategy, lessons are to be learned from the experience of those countries that have already shifted to accrual reporting. The box ahead summarizes a staff review of country experiences with incorporating accrual accounting, recording, and reporting into the statistical framework. Australia and Iceland were the first to introduce full accrual reporting in line with GFSM 2001. New Zealand has had long experience with accrual accounting but has only recently initiated a project to compile data using GFSM 2001. More recently, the United Kingdom introduced a comprehensive system of resource accounting on an accrual basis. Also, other countries in the OECD, Latin America, and Asia have initiated adoption of accrual accounting for reporting government operations. Incorporating accrual accounting in the methodology of compiling fiscal statistics is part of a broader initiative of public sector accountants and statisticians to achieve greater harmony between international accounting and statistical standards.²⁵ These experiences underscore the long-term nature of the process to implement accrual reporting.

Summary of Desk Studies

54. To test the strength of the GFSM 2001 framework relative to the traditional approach, the staff recast fiscal data shown in Article IV staff reports into three fiscal tables in GFSM 2001 format for Uruguay, Malaysia, and Mauritius.²⁶ A summary of the analysis of fiscal policies follows, and the appendix provides more detail.

55. The Uruguay desk study illustrates that the GFSM 2001 presentation captures the impact on the public sector of the banking crisis in 2002 more comprehensively than the staff’s best efforts to estimate an “augmented deficit.” The latter, which appeared as a memorandum item, did not include public sector capital transfers (other expense) equivalent to about 3 percent of GDP that affected the gross operating balance and net lending/borrowing. The GFSM 2001 presentation highlights the reduction in net worth (in the balance sheet) and the increase in the cash deficit (in the cash statement). It also shows clearly that government liquidity support is merely a financial transaction, altering the composition of the balance sheet but not the core balances. Finally, the GFSM 2001 presentation explicitly records write-offs as other economic flows that negatively affect net worth and net financial worth in the balance sheet. These losses in the value of assets were difficult to capture in the traditional analysis in staff reports.

56. The GFSM 2001 presentation of the fiscal data for Malaysia does not alter the major trends reflected in the staff reports but does suggest that the magnitudes of the key balances are slightly different. The authorities and the staff registered privatization proceeds as part of non-tax capital revenue. However, the transactions represent the sale of equity shares in public corporations; therefore, they should not affect the core operating balances. Reclassifying privatization proceeds out of revenue (and to the disposal of shares and other equity) results in an increase of the net borrowing position of ½ percent of GDP during 1997–2003. Privatization is neutral to net worth because it is an exchange of one asset (equity) for another (cash), as clearly shown in an integrated balance sheet. At the same time, the sale of domestic equity contributes to the increase in the stock of cash, reflected in the cash statement.

57. The Mauritius desk study shows that the GFSM 2001 presentation provides a more comprehensive framework for the analysis of fiscal sustainability, including the dynamics of public debt. The 2002 selected issues paper contained a GFSM 2001 presentation, showing the evolution of budgetary central government operations. It showed the effects of those operations on its balance sheets, including debt liabilities during 1994/95–2000/01 and (staff) projections through 2006/07. These data in the selected issues paper reveal the country’s steady incurrence of debt liabilities during 1995/96–1999/2000, primarily through the accumulation of domestic debt in loans and government paper. The projections reinforce the staff report’s main argument that the baseline fiscal path would not yield sustainable debt; however, the GFSM 2001 analytical framework may have presented the case more forcefully.

B. The Case for Pilot Studies

58. While the desk studies illustrate the potential gains for fiscal analysis from using the three GFSM 2001 tables, this framework has not been fully tested across the membership. To broaden experience with the use of this framework as the standard for Fund operational work, the staff, in collaboration with country authorities, could undertake pilot studies in countries that have migrated to accrual reporting (such as Australia, United Kingdom, Iceland, New Zealand) and countries that draw heavily on the national accounts for developing their fiscal statistics (such as EU countries and the United States). They could also undertake pilot studies in countries that continue to have cash-based statistics but either compile their data applying the GFSM 2001 framework (for instance, South Africa, Thailand) or map their data into the GFSM 2001 framework after their data have been compiled using another methodological framework. (This process was followed by many countries in the context of the GFSY.) Thus, the pilot studies would review the experiences of countries across the spectrum of accounting and statistical sophistication.

59. The specific pilot countries would be selected on a voluntary basis, taking into account the existence of significant policy issues in the fiscal area to which the staff and the authorities could usefully apply the GFSM 2001 framework. In particular, staff and the authorities could focus on using the analytical framework to draw policy conclusions about the fiscal implications of bank restructuring, debt restructuring, relations with public corporations (including equity injections/recapitalizations or privatization), and the portfolio management of public sector assets and liabilities to maximize economic growth and the public sector’s financial soundness and fiscal sustainability. Initially, the selection of volunteer countries for the pilot studies would include surveillance and program countries, the latter being restricted to countries entering new programs. The pilots also should cover a spectrum of countries across income levels. Staff and the authorities would need to collaborate closely to identify volunteers and conduct the studies.

60. In conducting these pilot studies, the staff considers it important that the three GFSM 2001 tables appear in parallel with the traditional fiscal tables when possible, allowing comparisons of the data and analytical approaches. While the basic skeleton of the tables would be similar because they follow GFSM 2001, the details listed for any individual country would depend critically on the fiscal policy issues under discussion. GFSM 2001 data could appear initially as an appendix to the staff report, as part of the accompanying selected issues papers, or directly in the main staff report. While a parallel presentation is feasible for surveillance countries, it would cause confusion in program countries. This being the case, GFSM 2001 fiscal tables would not be presented in program staff reports or other program documentation.

61. It is envisaged that three-and-a-half staff years could be required to complete these pilot studies over the course of two years and to undertake related training and support activities, which would be absorbed by STA and FAD within the existing resource envelope. Collaboration with area department staff and the authorities would be an integral component of a pilot study. The proposed pilot studies would explore the modalities for presenting and analyzing GFSM 2001 statistics in Fund reports. Staff would report to the Board on the outcome of these pilot studies and, on this basis, propose next steps on using the GFSM 2001 to strengthen fiscal analysis in the Fund.

C. Training and Support

62. While the responsibility for compiling and reporting fiscal statistics remains with the national authorities, IMF training and support for their efforts to implement GFSM 2001 will be required. STA will continue to provide technical support to the national authorities, through statistical products and preparation of papers on methodological issues as companion materials to GFSM 2001 that are posted on the Fund’s website.²⁷ Simultaneously, FAD would continue to assist countries seeking to strengthen their accounting and classification systems. Together, STA and FAD would work with the authorities to promote sound fiscal data when the data are reclassified into the GFSM 2001 framework. In addition, FAD and STA would advise country compilers and area departments on how to properly record economic events in the fiscal data and, collaborating with the INS, train them on these matters. The INS will review internal and external training courses (such as financial programming) that might need modifying to reflect the changeover to GFSM 2001, and efforts in this area will be increased.

63. In addition, a joint STA/FAD Advisory Group on GFSM 2001 would act as a one-stop resource for area departments. Area departments will continue to serve as a conduit for the technical dialogue between the Fund (including STA and FAD) and the national authorities regarding fiscal statistics. In addition, they will increasingly use the GFSM 2001 framework for fiscal analysis in Fund operational work. The Advisory Group will also seek to ensure that related work deriving from each department’s specific responsibilities is properly taken into account in implementing GFSM 2001. Responsibilities include the harmonization of statistical standards in the case of STA, Fund surveillance and program design in the case of area departments, PDR and FAD, and the implementation of fiscal transparency standards in the case of FAD.