A. General Issues

6. The United States was selected for a pilot study in consultation with the U.S. Executive Director’s Office and the Western Hemisphere Department (WHD). While U.S. fiscal data are adequate for Fund’s surveillance, the adoption of the GFSM 2001 framework, which relies on the System of National Accounts 1993 (1993 SNA), would improve fiscal policy analysis by fostering fiscal transparency and consistency in the presentation of country data. This is achieved through three channels: (i) the use of standardized statistical concepts and classifications; (ii) the full integration of cash-based and accrual-based fiscal data; and (iii) the reconciliation of stock and flow data in an integrated balance sheet.

7. The reclassification of existing source data into consistent GFSM 2001 accrual and cash-based formats contributes to fiscal transparency. This is particularly important since substantial debate and uncertainty have recently surrounded the official fiscal data on the financial position of the U.S. federal government. The authorities publish various sets of fiscal data, on a cash and an accrual basis, that differ in coverage and methodologies used (see Box 2). The discrepancies between the core balances shown in these various sets of official data are a potential source of confusion for the user of the United States’ fiscal statistics.

8. The primary source data used for the study were statistics compiled by the Bureau of Economic Analysis (BEA) and the Federal Reserve Board (FRB). The Budget of the United States Government report and the Financial Report of the United States Government were used as complementary sources. Because none of these sources fully complies with the GFSM 2001 statistical concepts and definitions, the official figures were adjusted based on data provided by the authorities and staff estimates.¹

9. The institutional coverage in this pilot study is the consolidated central government, which in the United States is referred to as the federal government. It comprises the budgetary central government—including the government employee pension funds²—and the social security funds.³ Article IV staff reports use an institutional coverage for government that is similar to the pilot, with a few insignificant exceptions.⁴

10. The main discrepancy between the pilot study results and the data in Article IV staff reports relates to partially funded government employee pension funds.⁵ Article IV staff reports, which are based on fiscal data published in the Budget of the United States Government report, and the GFSM 2001 treat social security funds similarly (except for timing differences between the accrual and cash accounting). According to the international statistical standards, contributions to and payments from these funds are recorded as revenue and expense transactions, thus affecting the cash surplus/deficit of the budget. No liabilities for future social security benefits are recognized (although FASAB has proposed changing this).⁶ However, there are significant discrepancies between Article IV staff reports and the GFSM 2001 regarding the treatment of government employee pension funds.

11. Whereas Article IV staff reports treat government employee pension funds similarly to social security, GFSM 2001 considers government employee pension benefits to be deferred compensation.⁷ Following the GFSM 2001 methodology, to the extent that the employee retirement fund is unfunded, it is necessary to impute an expense transaction equal to the actuarial estimate of pension commitments made during the period minus reserves built up to meet the future payments. The imputed expense results in an increase in net lending/borrowing of the government, the counterpart of which is an increase in liabilities to government employees. According to the GFSM 2001 methodology, the net present value of future benefits to employees should be shown in the government’s balance sheet as a liability (see Appendix II for more detail).

12. Given the available data, migration to the GFSM 2001 framework is feasible in the short term. Although a complete migration may take time, the government accounting system and reporting formats are broadly adequate for the implementation of the GFSM 2001 methodology. Consequently, the advantages of migrating to the GFSM 2001 methodology in terms of enhancing fiscal transparency and improving fiscal analysis can be readily achieved. However, if the authorities decide to fully implement the GFSM 2001 methodology, the migration strategy would require coordinated efforts on the part of all the agencies reporting fiscal data, particularly the BEA, the FRB, and the Treasury to come to a set of consistent source data.

B. Preliminary Results

13. This section presents an overview of the advantages of using GFSM 2001 for the United States. The numerical results presented in this section should be considered illustrative due to unresolved data issues.⁸ Summary tables compiled in accordance with the GFSM 2001 (Tables 1, 2, and 3) are shown in Appendix I. Tables 1 and 2—the Statement of Government Operations and the Integrated Balance Sheet, respectively—are based on data published by the BEA, FRB, and the OMB. Table 3—the Statement of Sources and Uses of Cash—is based on data published in the Budget of the United States Government. The staff adjusted the official data, primarily in terms of institutional and transactional coverage, to conform to the GFSM 2001 statistical principles. Box 3 provides a description of the main aggregates and balances of analytical significance used in these tables.

14. The U.S. pilot study draws attention to:

• the larger federal government deficit if the obligations of unfunded government employee pension funds are recognized;

• the existence of data gaps that preclude deriving a measure of federal government net worth, as well as a full integration of stock and flow data; and

• the need to reconcile cash-based and accrual-based fiscal data in order to foster fiscal transparency.

The following paragraphs discuss the findings of the pilot study on the basis of three tables, namely, the Statement of Government Operations and the Integrated Balance Sheet on an accrual basis, and the Statement of Sources and Uses of Cash.

15. The Statement of Government Operations (Appendix I, Table 1) shows the transactions of the federal government on an accrual basis. Two important balances can be derived from this table: (i) the net operating balance (NOB), which measures changes in net worth resulting from transactions and is a key indicator of the ongoing sustainability of government operations; and (ii) net lending/borrowing (NLB), which captures the financial impact of government operations on the rest of the economy.

16. Net worth has decreased by 14.4 percent of GDP over the four-year period under analysis, although at a diminishing rate. While the deterioration of net worth has been significant, as shown by the accumulated NOB over the period, the authorities achieved equilibrium in the primary net operating balance in 2006.⁹ This indicates an improvement in the sustainability of government operations. As the net acquisition of nonfinancial assets was maintained at low levels, the NLB closely mirrors the evolution of the NOB in terms of magnitude and trend.¹⁰ This net borrowing requirement has been financed by an increase in liabilities, primarily in the form of securities.

17. The net borrowing position of the federal government shown in Table 1 is higher than that reflected in the accrual-based NIPAs by 0.7 percent of GDP, on average. The discrepancy is primarily related to the treatment of obligations of government employee pension funds. In Table 1, imputed social contributions were included as compensation of employees affecting the NOB and the NLB accordingly, with the counterpart entry being the increase in liabilities as insurance technical reserves.

Box 3.

The GFSM 2001 Statements and Core Balances

The Statement of Government Operations (Appendix I, Table 1) records transactions on an accrual basis. The statement distinguishes among the following transactions:

Revenue	Transactions that increase net worth.
Expense	Transactions that reduce net worth.
Net acquisitions of nonfinancial assets	Transactions that affect the stock of nonfinancial assets, without changing net worth (acquisitions minus disposals).
Financing	Transactions that affect the stock of financial assets and liabilities, without changing net worth (net acquisition of financial assets minus net incurrence of liabilities)
The analysis of government operations is supported by two key fiscal indicators:
Operating balance	Summary measure of the effects of revenue and expense transactions on net worth. Net operating balance (NOB) equals revenue minus expense. The gross operating balance (GOB) equals revenue minus expense other than consumption of fixed capital.1
Net lending/borrowing	Financial resources that the government absorbs from, or releases to, other sectors of the economy on a net basis. It is calculated as the NOB minus the net acquisition of nonfinancial assets. Net lending/borrowing is also equal to the net acquisition of financial assets minus net incurrence of liabilities.
The Integrated Balance Sheet (Appendix I, Table 2) focuses on an assessment of the sustainability of government operations from a fiscal perspective. It includes the government’s net worth at the beginning and end of each fiscal year, as well as the related transactions and other economic flows. The sustainability of fiscal policy depends in part on how the government’s net worth changes over time. Changes in net worth can be explained not only by the government’s transactions but also by other economic flows attributable to gains or losses resulting from changes in the prices of assets and liabilities, as well as other changes in their volume.
Net worth	The total stock of assets minus liabilities. The net worth in period (t) can also be calculated as the net worth of the previous period (t-1), plus changes in net worth in period (t) due to transactions (the NOB), plus changes in net worth in period (t) due to other economic flows.
Net financial worth	The stock of financial assets minus liabilities.
The Statement of Sources and Uses of Cash (Appendix I, Table 3) shows purely cash flows associated with revenue and expense transactions and transactions in nonfinancial assets, which yield the cash surplus/deficit. The assessment of the government’s level of cash holdings and its determinants is a key element in analyzing fiscal policy, its interrelationships with the monetary policy and liquidity analysis.
Cash surplus/deficit	Net cash inflow from operating activities minus the net cash outflow from net acquisitions of nonfinancial assets.

View Table

Revenue	Transactions that increase net worth.
Expense	Transactions that reduce net worth.
Net acquisitions of nonfinancial assets	Transactions that affect the stock of nonfinancial assets, without changing net worth (acquisitions minus disposals).
Financing	Transactions that affect the stock of financial assets and liabilities, without changing net worth (net acquisition of financial assets minus net incurrence of liabilities)
The analysis of government operations is supported by two key fiscal indicators:
Operating balance	Summary measure of the effects of revenue and expense transactions on net worth. Net operating balance (NOB) equals revenue minus expense. The gross operating balance (GOB) equals revenue minus expense other than consumption of fixed capital.1
Net lending/borrowing	Financial resources that the government absorbs from, or releases to, other sectors of the economy on a net basis. It is calculated as the NOB minus the net acquisition of nonfinancial assets. Net lending/borrowing is also equal to the net acquisition of financial assets minus net incurrence of liabilities.
The Integrated Balance Sheet (Appendix I, Table 2) focuses on an assessment of the sustainability of government operations from a fiscal perspective. It includes the government’s net worth at the beginning and end of each fiscal year, as well as the related transactions and other economic flows. The sustainability of fiscal policy depends in part on how the government’s net worth changes over time. Changes in net worth can be explained not only by the government’s transactions but also by other economic flows attributable to gains or losses resulting from changes in the prices of assets and liabilities, as well as other changes in their volume.
Net worth	The total stock of assets minus liabilities. The net worth in period (t) can also be calculated as the net worth of the previous period (t-1), plus changes in net worth in period (t) due to transactions (the NOB), plus changes in net worth in period (t) due to other economic flows.
Net financial worth	The stock of financial assets minus liabilities.
The Statement of Sources and Uses of Cash (Appendix I, Table 3) shows purely cash flows associated with revenue and expense transactions and transactions in nonfinancial assets, which yield the cash surplus/deficit. The assessment of the government’s level of cash holdings and its determinants is a key element in analyzing fiscal policy, its interrelationships with the monetary policy and liquidity analysis.
Cash surplus/deficit	Net cash inflow from operating activities minus the net cash outflow from net acquisitions of nonfinancial assets.

^1/ The NOB/GOB exclude net acquisition of nonfinancial assets. The latter does not affect net worth because it represents only an accumulation of assets in exchange for an accumulation of liabilities or use of existing assets.

18. Table 1 indicates a high degree of vertical integration—that is, consistency between the data on operating transactions and financing. In general, statistical discrepancies may arise between the NLB and its financing (net acquisition of financial assets minus net incurrence of liabilities) due to data deficiencies. The BEA and the FRB periodically reconcile both datasets and, in the case of the federal government, the statistical discrepancy accounts for only 0.2 percent of GDP, on average.

19. However, the Integrated Balance Sheet (Appendix 1, Table 2), shows horizontal integration to be less robust on account of specific data gaps revealed by the pilot study. When compiling the Integrated Balance Sheet, data gaps are particularly apparent in the lack of data on nonproduced assets, while the valuation of the stock of assets and liabilities precludes a complete explanation of differences between opening and closing balance sheet. In this regard GFSM 2001, by presenting stock and flow data in an integrated framework, as shown in Table 2, fosters data consistency and facilitates the analysis of changes in net worth.

20. The lack of data on the stock of nonproduced assets precludes deriving a measure of net worth for the federal government. This is particularly important in light of the significant stock of natural resources owned by the U.S. federal government, which would increase its net worth.¹¹ However, an estimate of the change in net worth due to transactions can be obtained from NOB as presented in Table 1 (Statement of Government Operations). In the absence of a comprehensive stock measure, the NOB constitutes an indicative measure of the sustainability of fiscal policy. In addition, a more restrictive concept, the net financial worth, which is equal to the difference between financial assets and liabilities, shows a significant decline throughout the period, although at a diminishing rate (see Figure 1).

View Full Size

Changes in Net Financial Worth and its Determinants

(percent of GDP)

Citation: IMF Staff Country Reports 2007, 266; 10.5089/9781451839708.002.A001

• Download Figure
• Download figure as PowerPoint slide

21. Data constraints imply that a full explanation of changes in net financial worth cannot be provided. Table 2 explains changes in net financial worth by transactions (NLB) and holding gains and losses and other volume changes. However, holding gains and losses cannot be calculated because of the valuation criteria used to report assets and liabilities. GFSM 2001 prescribes the valuation of assets and liabilities at market prices. However, the NIPAs and the Flow of Funds Accounts value financial assets and liabilities at historic cost.¹² Consequently, the deterioration of the net financial worth is primarily explained by changes in transactions (NLB), as it is shown in Figure 1. It should be noted that in preparing the U.S. pilot study, the staff partially estimated the so-called “other economic flows.” In particular, based on data published in the Financial Report of the U.S. Government, any changes in the net present value of future benefits to government employees—insurance technical reserves—that were not explained by imputed social contributions were recorded as other economic flows.¹³ These include primarily changes in the discount rates used in the calculation of the net present value, as well as changes in the eligibility conditions of retirement programs.

22. Despite existing data gaps, the Integrated Balance Sheet (Table 2) provides an important insight into the composition of government liabilities. Following GFSM 2001 concepts and definitions, the net present value of future benefits to government employees (insurance technical reserves) is included as a liability of the federal government. In the case of the federal government, this liability represents 18 percent of GDP, on average (see Figure 2). These obligations are not recorded at all in a cash-based system, and they are only partially recognized in the accrual-based data reported by the BEA and the FRB (8 percent of GDP, on average).

View Full Size

Composition of Federal Government Liabilities

(percent of GDP)

Citation: IMF Staff Country Reports 2007, 266; 10.5089/9781451839708.002.A001

• Download Figure
• Download figure as PowerPoint slide

23. Differing treatment of government employee pension funds explains a 10 percent of GDP discrepancy between total liabilities in the pilot study and in the official figures. The BEA and the FRB, following the NIPAs methodology, recognize a federal government liability to employees as long as dedicated financial assets exist.¹⁴ This results in a partial recognition of total liabilities owed to government employees of 8 percent of GDP on average. However, the GFSM 2001 recognizes the full liability with government employees calculated as the net present value of future benefits (18 percent of GDP). Consequently, the adjustment made in the pilot study to account for the unfunded part of the scheme represents 10 percent of GDP, on average.

24. Unlike the government employee pension funds, social security schemes do not give rise to liabilities in the GFSM 2001 framework. Nevertheless, in order to provide a more informed basis for conducting long-term fiscal policy and fiscal sustainability analysis, the GFSM 2001 recommends including them as a prominent memorandum item to the balance sheet.¹⁵ Even if they are not strict obligations in a legal sense, there is a general expectation that contributions in current periods will lead to benefits being paid to those making the contributions in the future. Because of timing differences, contributions and benefit payments will initially result in net revenues, but over time this flow will be reversed and the government will accumulate substantial future obligations as expense exceed revenue. Following the international standards, these obligations, which in the case of the U.S. represented 320 percent of GDP in 2006, are not included in the balance sheet of the federal government.¹⁶

25. The Statement of Sources and Uses of Cash (Appendix I, Table 3) shows cash flows associated with revenue, expense, and nonfinancial assets transactions. The main source for this table is the Budget of the United States Government. These data were adjusted to ensure internal consistency with the data presented in Tables 1 and 2.¹⁷

26. Although the cash deficit in Table 3 is similar to that in the Budget, revenue and expense categories have been adjusted to account for discrepancies in coverage. On the one hand, adjustments have been made to the institutional coverage to exclude: (i) transactions of the U.S. territories and Puerto Rico, and (ii) current surplus of nonfinancial public corporations. On the other hand, adjustments made to the transactional coverage exclude: (i) employee contributions to government pension funds; (ii) subsidy expenses related to direct loans; (iii) deposit insurance expense for resolving failed banks and thrift institutions; (iv) proceeds from sales of nonproduced assets; and (v) other coverage adjustment.

27. Cash deficits were financed primarily by new issuance of government securities, repayments of direct loans, and, to a lesser extent, reductions in the stock of cash. The latter has been reduced from 0.9 percent of GDP in 2003 to 0.5 percent in 2006. The stock of cash is included in the Integrated Balance Sheet (Table 2) as part of total financial assets. In this way, both cash-based and accrual-based measures of fiscal performance are clearly linked using the GFSM 2001 framework.

28. Figure 3 suggests a difference between cash-based fiscal operations and the comparable accrual aggregate (NLB) shown in Table 1. Fiscal analysis based exclusively on cash or accrual data may produce inaccurate signals with regard to the overall fiscal position. In the case of the U.S. federal government, the accrual balancing item, the NLB, is higher than the cash deficit in every year. The difference between the two measures is a result of: (i) the accrual data including expense items which do not have a cash equivalent, and (ii) timing differences in the recording of revenue and expense. The former item includes the imputed accruing liabilities for the government employee pension obligations during the period (see Table D in Appendix II for details).

View Full Size

Net Lending/Borrowing and Cash Surplus/Deficit

(percent of GDP)

Citation: IMF Staff Country Reports 2007, 266; 10.5089/9781451839708.002.A001

• Download Figure
• Download figure as PowerPoint slide

C. Next Steps

29. The BEA broadly agreed that they would work toward improving the dissemination of fiscal data in accordance with GFSM 2001. In this regard, the authorities plan to continue research to improve the consistency of U.S. fiscal data with the GFSM 2001 standards for the purposes of reporting to the Fund.

30. To migrate the U.S. fiscal data into the GFSM 2001 framework, staff will continue to work with the authorities to compile more comprehensive and detailed data in the following areas:

• Other economic flows (Table 2): staff will continue to work with the BEA to expand the range of data available on other economic flows, particularly that disseminated in the context of the International Investment Position (IIP).

• Valuation of stocks at market value (Table 2): staff will work with the BEA to improve the valuation of stocks of assets and liabilities to conform to the GFSM 2001 concepts and definitions.

• Cash payments for operating activities by economic classification (Table 3): readily available source data on a cash basis did not allow for the classification of cash payments for operating activities by economic type. Staff will continue to work to obtain the full detailed set of categories set out in Table 3.

• Coverage: staff will continue to work with the BEA to extend the results of this pilot to the state and local governments.