As of May 1, 1991, while this book was in press, the IMF’s Bureau of Statistics became the Statistics Department of the Fund. Consequently, all references to the Bureau of Statistics herein should be understood to refer to the Statistics Department.

Frontmatter Page

As of May 1, 1991, while this book was in press, the IMF’s Bureau of Statistics became the Statistics Department of the Fund. Consequently, all references to the Bureau of Statistics herein should be understood to refer to the Statistics Department.


© 1991 International Monetary Fund

Reprinted May 1995

Library of Congress Cataloging-in-Publication Data

The IMF's statistical systems in context of revision of the United Nations'A system of national accounts / Statistics Department; edited by Vicente Galbis

  • p. cm.

“A selection of papers presented at three expert group meetings sponsored by the Fund in 1987–88” –Foreword.

Includes bibliographical references.

ISBN 1-55775-159-5

1. National income—Accounting. I. Galbis, Vicente, 1942–II. International Monetary Fund. Statistics Dept. III. International Monetary Fund.

HC79.I5I44 1991

339.3'2'015195—dc20 91-2489


Both this book's cover and its interior were designed by the IMF Graphics Section.

The following symbols have been used throughout this book:

… to indicate that data are not available;

— to indicate that the figure is zero or less than half the final digit shown, or that the item does not exist;

– Between years or months (e.g., 1991–92 or January-June) to indicate the years or months covered, including the beginning and ending years or months;

/ Between years (e.g., 1991/92) to indicate a crop or fiscal (financial) year.

“Billion” means a thousand million.

Details may not add to totals shown because of rounding.

The term “country,” as used in this book, does not in all cases refer to a territorial entity that is a state as understood by international law and practice; the term also covers some territorial entities that are not states but for which statistical data are maintained and provided internationally on a separate and independent basis.

Price: US$35.00

Address orders to:

External Relations Department, Publication Services

International Monetary Fund, Washington D.C. 20431, U.S.A.

Telephone: (202) 623-7430 Telefax: (202) 623-7201

Cable: Interfund

internet: publications@imf.org

recycled paper


The International Monetary Fund has always attached considerable importance to the development of appropriate methodologies in its work in balance of payments, government finance, and monetary statistics. The Fund's contribution to the development of international statistics also takes other forms, including a system of statistical communications with member countries, technical assistance to member countries on statistical matters, cooperation in these matters with other international organizations, and compilation of its own statistical publications.

The role of the IMF within the international statistical community has allowed the Fund to participate actively in the ongoing revision of the United Nations' A System of National Accounts (SNA), last revised in 1968. The work on revision of the SNA may be seen as part of the Fund's commitment to achieve a greater degree of coordination in the statistical field with other international agencies. The Fund has also begun to revise the fourth edition of its Balance of Payments Manual, issued in 1977, and to refine further its methodology for monetary statistics, based on the draft of A Guide to Money and Banking Statistics. A Manual on Government Finance Statistics was published by the Fund in 1986. The Fund hopes that these efforts will not only lead to methodological improvements and a greater consistency in national statistics but will also promote harmonization of these statistical systems among themselves and with the SNA.

This volume contains a selection of papers presented at three expert group meetings sponsored by the Fund in 1987–88 for the revision of the SNA. These meetings dealt with issues in the balance of payments, government finance, and monetary statistics and were part of a sequence of expert group meetings organized by the interested international organizations to oversee the revision process. The papers place the Fund's statistical methodologies in the context of revision of the SNA and explain the contribution that the Fund has made, and can make, to the process of harmonization.

Although the papers represent the views of individual contributors and not necessarily those of their respective institutions, I trust that they will be of interest to those whose task it is to compile macroeconomic statistics and to develop new ideas to cope with the challenges of the interrelation, simplification, and evolution of these statistics.

Michel Camdessus

Managing Director

International Monetary Fund


At Its Twenty-First session in 1981 the Statistical Commission of the United Nations discussed proposals for improvement of the United Nations' A System of National Accounts (SNA), last published in 1968. Following this, an Expert Group Meeting on the Review and Development of the SNA, held in New York in March 1982, recommended that a long-term review of the SNA be undertaken to produce a revised SNA. To assist in this process, the Intersecretariat Working Group on National Accounts—a group comprising representatives of the UN, the Organization for Economic Cooperation and Development (OECD), the European Community (EC), the World Bank, the IMF and the five UN Regional Economic Commissions—was established. The main goals set for the revision of the SNA were to simplify the 1968 SNA so that it will be more directly relevant to the needs of member countries; to resolve methodological issues and internal inconsistencies noted since publication of the 1968 SNA; and to harmonize the SNA as much as possible with related statistical systems, especially the Fund's systems and methodologies relating to balance of payments, government finance, and monetary statistics.

The revision process has centered on a series of meetings of expert groups composed of national experts and the members of the Intersecretariat Working Group. In 1987 and 1988 the Fund sponsored three of these expert group meetings. They dealt with external sector transactions (March 23-April 2, 1987), public sector accounts (January 25–29, 1988), and financial flows and balances (September 6–15, 1988).

This volume contains a selection of the papers prepared for the expert group meetings sponsored by the Fund. The structure of the volume has three parts, corresponding to the issues discussed at each of the three meetings. In all, 32 papers are presented, most of them short. A brief introduction has been provided by the editor. Some of the papers have been modified for publication, with certain material, especially annexes and appendices, deleted to simplify and shorten the presentation or to avoid unnecessary repetition. The papers have not been revised, however, to reflect discussion during the expert group meetings. Thus, the contents of this volume should in no way pre-empt or prejudice the views expressed by participants at those meetings, the conclusions and recommendations of the meetings, or the solutions that may eventually prevail in achieving harmony between the revised SNA and the Fund's statistical methodologies. Rather, I would hope that these papers are viewed as research material that is interesting for its analytical approaches to some longstanding issues, and to new methodological issues in the field of economic statistics. The papers therefore represent solely the views of the authors and not the official views of the Fund or of other organizations.

Finally, I would like to thank the contributors to this volume not only for preparing the papers but also for their cooperation with the editor in the thankless task of streamlining the presentation by deleting some material, as noted above; this made it possible to publish the papers in a single volume without reducing their substance. I would also like to thank James McEuen of the Fund's External Relations Department for painstakingly reviewing the draft papers and for making many useful editorial suggestions at various stages of production of this volume.

John B. McLenaghan


Statistics Department

International Monetary Fund



Michel Camdessus


John B. McLenaghan


Vicente Galbis


1. Residents of an Economy

Arie C. Bouter

2. Treatment of International Organizations

Gérard G. Raymond

3. Change of Ownership and Time of Recording in the National Accounts

Robert McColl

4. Conversion of Balance of Payments Transactions as a Source of Valuation Changes: Problems, Principles, and Practical Solutions

Pierre Luigi Parcu

5. Currency Conversion in a Multiple Exchange Rate System

Marianne Schulze-Ghattas

6. Treatment of Exchange Rate Differentials in the National Accounts

Jan van Tongeren

7. Harmonization of the Classification of External Current Account Transactions

Arie C. Bouter and Jan van Tongeren

8. Proposed Treatment of Reinvested Earnings on Direct Investment

Geoffrey J. Robertson

9. Classification of International Transactions in Services, Income, and Unrequited Current Transfers

Arie C. Bouter

10. Measurement of a Nation's Terms of Trade Effect and Real National Disposable Income Within a National Accounting Framework

Mick Silver and Khashayar Mahdavy

11. Classification of Capital Account Transactions

Mahinder S. Gill

12. Classification of Corporate Enterprises

D. Keith McAlister

13. New Financial Instruments and the Balance of Payments

George H. Hoezoo

14. Classification of Transactions in Zero-Coupon Bonds, Junk Bonds, and Indexed Bonds in the Balance of Payments

Robert McColl

15. Some Issues in the Balance of Payments Presentation of Arrears and Debt Reorganization

John E. Thornton


16. A Discussion of Public Sector Accounts

Jonathan Levin, Jan van Tongeren, Brian Newson, and Derek Blades

17. An Example of Progress in Delineating Relationships Between Government Finance Statistics and National Accounts Concepts

Jonathan Levin and Jan van Tongeren

18. Overall Relationships Between the IMF's Government Finance Statistics and A System of National Accounts

Jonathan Levin

19. Derivation of System of National Accounts Value from Government Finance Statistics Data

Jan van Tongeren and Irene Tsao

20. Sectorization of Social Security Funds

Teresa Villacrés


21. Flow-of-Funds Accounts, A System of National Accounts, and Developing Countries

John C. Dawson

22. The Financial Sector

Edward W. Saunders

23. Monetary Concepts and Definitions

Emmanuel O. Kumah

24. Repurchase Agreements and Financial Analysis

Marta Castello-Branco

25. Recording of Financial and Operational Leases and of Income on Zero- and Low-Coupon Bonds

Balance of Payments Division, IMF Bureau of Statistics

26. Financial Leasing

Brian Newson

27. Treatment of Deep-Discounted and Index-Linked Bonds in the National Accounts

Brian Newson and Soren Brodersen

28. Principles of Valuation and Reconciliation Items in the IMF's Money and Banking Statistics and A System of National Accounts

Keith G. Dublin

29. Treatment of Output in the Banking Industry

Abul Siddique

30. A Further Look at the Treatment of Insurance in A System of National Accounts

André Vanoli

31. How to Treat Nonproduced Assets and Exceptional Events in the National Accounts? Considerations on the Variations in Wealth Accounting

Jean-Paul Milot, Pierre Teillet, and Andre Vanoli

32. Fixed Capital Formation by Owner and User

Heinrich Lützel




Vicente Galbis

Three broad goals have been set for the revision of the United Nations' A System of National Accounts (SNA).1 First, it is generally recognized that there is scope for a simplification of the system of accounts to make it more directly relevant for the purposes of compilation and the needs of users. Many countries have encountered difficulties in the implementation of some aspects of the 1968 (current) version of the SNA and in preparing current data for the national accounts. In addition, the continued expansion of the framework of the SNA through “Guidelines” and “Handbooks,” which complement the basic SNA document (known as the “Blue Book”), makes it advisable to take a fresh and comprehensive look at the entire system. In this context, a particular emphasis is planned to be given to the special concerns of the developing countries. Second, a need has been felt to take into account several relatively important methodological developments that have occurred since the publication of the 1968 SNA and that are not reflected in the current basic framework. These developments suggest that improvements in the methodological and analytical bases of the SNA are possible. Third, there is a need for harmonization of the SNA with related systems of macroeconomic accounts, among them the systems developed by the IMF relating to compilation of its balance of payments statistics (BOPS), government finance statistics (GFS), and money and banking statistics (MBS).

The three goals are closely interrelated. Simplification of the SNA is necessarily linked with methodological improvements, especially those that encompass concepts that direct attention to the broader aspects of the measurement of economic activity. In turn, methodological improvements, as well as practical advances in the measurement process, are often the basis for the harmonization of the various systems of macroeconomic accounts. For these reasons, all of the expert group meetings for the revision of the SNA had these three purposes in mind, although with different degrees of emphasis depending on the subject under discussion.2

The three expert group meetings for the revision of the SNA sponsored by the IMF in 1987 and 1988 placed particular emphasis on methodological improvements bearing on the harmonization issue. This was particularly appropriate because the Fund's main statistical systems—the BOPS, the GFS, and the MBS—cover in detail three of the broad macroeconomic sectors that are encompassed in the SNA. Data corresponding to these systems form part of the principal statistical publication of the Fund, International Financial Statistics (IFS), and of its other statistical publications.3 The principal focus of the papers included in this volume—originally presented as background papers for the discussions of the expert group meetings—is on harmonization. The objective of harmonization is broad consistency between these specialized statistical systems and the SNA on major methodological issues, including the appropriate sectorization, classification of transactions, and type of data. The development of the Fund's statistical methodologies and systems, while taking into account the overall structure and framework of the SNA, initially arose out of pragmatic responses to various challenges faced in the respective statistical areas, in different IMF departments, and over different periods.4 For this reason, as well as for complex issues of substance, there are a number of outstanding issues, of both form and method, for the process of harmonization. The papers included in this volume have served, as much as anything, to direct the efforts of the groups of experts in their deliberations on outstanding methodological issues with a view to bringing about the desired harmonization.5 They have also provided a basis for a discussion of methodological adaptations and innovations for possible incorporation in the SNA as well as in the revised Fund methodologies.

The IMF's Balance of Payments Manual (BPM) incorporates a conceptual framework relating the balance of payments to the external sector accounts of the SNA (BPM, Chapter 2, Section 6, paragraphs 44–51, and Appendix C). Although in principle the relationship between the two systems of accounts is clear-cut, the BPM (Appendix C) notes that there are many practical problems that are substantial enough to discourage the task of harmonization. Fortunately, however, this pessimistic view has been superseded by recent developments; for instance, a more recent examination of this issue suggests that “these two systems are consistent and interrelated.”6 A harmonized presentation is planned to be provided in the revised SNA, as well as in the fifth edition of the BPM; harmonization issues will also be taken into account in the Balance of Payments Compilation Guide, which is being prepared by the Fund's Bureau of Statistics (now the Statistics Department).7 The essays in this volume should serve to illustrate many of the outstanding issues on which substantial progress in this area has been made.

The Fund's A Manual on Government Finance Statistics (GFSM) discusses in detail the relationships between the GFS and the government sector of the SNA.8 To facilitate the compilation of national accounts for the government sector of the SNA from the GFS, the GFSM presents detailed “bridge tables” between the two data systems.9 This is possible because the underlying building blocks included in both systems should in principle agree. The main differences between the two systems are that the SNA records transactions on an accrual basis, has three separate but interrelated accounts, and classifies taxes and subsidies according to whether or not they are incurred in the course of production. The GFSM, on the other hand, provides for the presentation of accounts on a cash basis, organized into a single balanced account; taxes are classified by their base. Many of these differences and suggested ways of harmonization are discussed in the essays included in the present volume.

Although the current draft of A Guide to Money and Banking Statistics in International Financial Statistics (MBS Guide) does not contain a separate section on the relationships between the MBS and the SNA, many references are made in the text, as appropriate, to these relationships. The essays in this volume further elucidate these relationships and suggest that both the revised SNA and the draft MBS Guide, which is also planned for revision, will develop explicit principles for their harmonization.

The basic interrelationships between the Fund's main statistical systems and the SNA are clearly perceived from the perspective of the flow-of-funds (FOF) accounts, which may be viewed as a component of the broader SNA. The FOF accounts contain a closed system of financial flows among transactors by type of financial transaction and also connect financial flows to the overall surpluses (net savings) and deficits (net borrowing) resulting from the sectoral (nonfinancial) production, income, and outlay accounts of the SNA.10 Furthermore, by a coordinated approach to the sectorization of the SNA and the FOF accounts, it can be shown that the Fund's main statistical systems correspond to three macrosectors into which the economy may be disaggregate at the first level of disaggregation of the SNA, including as a component the FOF accounts (that is, the government, financial, and rest-of-world sectors; the remaining macrosectors comprise, in essence, households and corporations). This reinforces the interest in the harmonization of these systems (see Chapter 21 in this volume). The benefits of relating the Fund's main systems to the SNA should be seen in eliminating unnecessary discrepancies, as well as duplicative efforts in compiling statistics. In fact, the Fund's statistical systems should be able to provide much of the data required for the construction of the SNA and related FOF accounts; data for the SNA, BOPS, and GFS are largely in the form of flows and, therefore, comparable in principle. Although the MBS data are mainly stock data, financial flow data can be derived from stock data by adjusting changes in stock data to eliminate elements of revaluation changes and other reconciliation factors.

A main challenge, therefore—and the one addressed by the papers in this volume—is to harmonize as much as possible the sectoral information from the SNA (and FOF) with the corresponding sectors covered by the IMF's main statistical systems. Successful efforts in the harmonization process will be of great benefit to compilers and users of macroeconomic statistics.

I. External Sector Transactions

The first six papers in Part I of this volume deal with some general methodological principles in compilation of statistics on the external sector of the SNA and the balance of payments.

The crucial concept of the residents of an economy—a concept needed to distinguish external (rest-of-world) from domestic transactions—is the subject of the papers by Bouter and Raymond. Bouter (Chapter 1) presents a comprehensive review of the residency criteria given in the SNA and the BPM; he shows that, although the main concepts regarding the territory of an economy and the “center of economic interest” of a given entity remain valid, some supplementary rules of thumb are required to pin down the criteria in special cases—for example, when the entity changes territory, especially when it does so frequently, and when it remains permanently outside the territory of any national economy. He goes on to specify in some detail the entities that are to be considered residents of an economy. Broad economic groups are examined for this purpose: the general government, private nonprofit bodies serving individuals, enterprises, and international organizations. The specific issues raised by the treatment of the last of these are discussed in greater detail by Raymond (Chapter 2). After reviewing the criteria in the SNA and the BPM, Raymond advocates a more refined definition of an international organization as one whose authority derives directly from the authority of its member countries and that has a sovereign status (that is, the laws and regulations of the country or countries in which it is located do not apply to it). However, the decision to treat an entity as an international organization for statistical purposes can only be made on a case-by-case basis. International organizations are not considered residents of any national economy; each is treated as a separate economy. The only resident entity of an economy comprising an international organization is that organization itself.

The change-of-ownership principle, which is used to determine whether a transaction has in fact taken place, and the time for recording a transaction are discussed by McColl (Chapter 3). He first points out that ownership is presumed to be identical to legal ownership except in three cases: when legal ownership cannot be granted, as in the case of illegal transactions; when the transactions are between two transactors that are part of the same legal entity; and when the resource is provided under a financing arrangement (including financial leasing) that transfers economic control and most of the risks of ownership of the asset to another party without transferring legal ownership. Other proper accounting requirements are that the two sides to a transaction must be recorded simultaneously by the two transactors and that the time of recording must have economic significance (that is, it must refer to the change of control or ownership over economic resources). McColl concludes that the only accounting framework to meet these criteria is accrual accounting, whereas cash accounting recognizes only one transaction—payment—that may not be concomitant with the transfer of economic control and ownership. He goes on to apply accrual accounting methods to the recording of transactions between affiliate companies, financial leases, interest income (particularly in the case of deep-discounted bonds), and overdue obligations.

Issues of conversion and exchange rate differentials in balance of payments statistics and the SNA are the focus of the papers by Parcu, Schulze-Ghattas, and van Tongeren. The conversion of transactions denominated in different currencies into a common valuation currency is the specific subject discussed by Parcu (Chapter 4). This is a subject of considerable interest—and controversy—in recording external transactions in the SNA and the balance of payments because the increased variability of exchange rates since the introduction of the floating exchange rate system has made the problem of conversion more difficult. The analogy between valuation changes resulting from price changes and those stemming from exchange rate changes clarifies the nature of the analytical problem for statistical recording. The general recommendation has been to choose for conversion the exchange rate prevailing at the date of the contract. Unfortunately, the practical applicability of this principle is severely limited by lack of adequate statistical information. For this reason, in practice, transaction values are usually converted into the unit of account at an average rate for the period in which the transaction occurs. Parcu suggests that this practice inevitably distorts the statistics and makes them less reliable, the more so the wider the fluctuations in exchange rates. The difficulties for statistical recording are even more severe in countries with a multiple exchange rate system—as discussed by Schulze-Ghattas (Chapter 5). In principle, two possible methods of conversion may be applied in the presence of multiple rates: conversion of each individual transaction at the exchange rate at which it was effected, or conversion of all transactions at a unitary conversion rate. The latter method has the merit that it ensures the recording of the taxes or subsidies that are implicit in a multiple exchange rate system. The main practical problem, however, is the determination of the appropriate unitary conversion rate. A reasonable proxy may be a weighted average of the different exchange rates observed in a multiple exchange rate system, the weights being the shares of the transactions effected at each particular rate.

A different set of issues on the treatment of exchange rate differentials is raised by van Tongeren (Chapter 6). Focusing not only on the external accounts but also on other sectors of the national accounts and on the effects on gross domestic product (GDP) and its components, he proposes a distinction between three components of exchange rate differentials: bank service charges, capital gains or losses, and taxes or subsidies. His conclusions are in line with those of Schulze-Ghattas in that a unitary exchange rate is proposed in circumstances of multiple exchange rate regimes that generate taxes or subsidies. However, van Tongeren suggests that no adjustments are indicated for changes in exchange rates over time, with the valuation of purchases and sales of foreign currency being recorded at their actual exchange rates. Revenues derived from those changes should be treated as capital gains or losses and included in the revaluation account.

The nine remaining papers in Part I deal with the classification of transactions in both the external sector of the SNA and the balance of payments, and taken together they cover both current and capital account transactions as well as new financial instruments. In their joint paper, Bouter and van Tongeren (Chapter 7) focus on the harmonization of current account transactions between the SNA and the balance of payments. They first identify outstanding issues for harmonization by reference to the characteristics of the standard components of the current account in the SNA and the BPM and the relationships between them. Next, they discuss the results of a survey of the practices of selected countries, issue by issue, pointing out the differences between the SNA and the BPM components and, where necessary, suggesting modifications of the components that would result in a closer harmonization of the two systems. The main outstanding issues identified are that the SNA deviates in some respects from the change-of-ownership principle, whereas the BPM is consistent in its application; that reinvested earnings of direct-investment enterprises that have foreign branches or subsidiaries are not included in the SNA, but are included in the BPM; that imports are valued on a c.i.f. (cost, insurance, freight) basis in the SNA and on an f.o.b. (free on board) basis in the BPM; that monetary and nonmonetary gold are treated differently in the two systems; that the SNA has a separate identification of labor and property income received by factors of production, but the BPM does not; and that the SNA requires various imputations for insurance transactions that are not made in the BPM. In Chapter 8, Robertson discusses the need to include reinvested earnings of direct-investment enterprises in the SNA—as is already done in the BPM—exemplifying very well one of the harmonization issues raised by Bouter and van Tongeren.

In a separate paper on the treatment of services, income, and unrequited transfers (Chapter 9), Bouter proposes a revised classification of the standard components for these current account items (42 in the BPM), taking into account the need to make improvements without substantially increasing the number of components and the desirable harmonization of the underlying transactions between the BPM and the SNA. One of his main proposals is to separate investment income into interest income and dividends, a breakdown indispensable for compiling debt service ratios.

In contrast to the other papers, Silver and Mahdavy (Chapter 10) focus on a specific statistical issue related to the current account—the measurement of the terms of trade in real terms—and its implication for the measurement of real national disposable income. They show that the measurement of real values is dependent on the price indices on which the terms of trade are based and that the results are therefore subjective. The diverse implications of different price deflators for the terms of trade effect are shown.

The paper by Gill (Chapter 11) turns to the general classification of capital account transactions and reviews the criteria for compilation of this complex part of the balance of payments. The author concludes that a substantial modification of the structure of the capital account would be required to serve the IMF's analytical and operational needs, to harmonize the balance of payments with the rest-of-world sector of the SNA, and to promote harmonization between the Fund's systems for balance of payments, government finance, and money and banking statistics, among other purposes. One of the amendments proposed by Gill is to distinguish between current and capital transfers in the balance of payments, as is currently done in the SNA. Another amendment would consist in merging portfolio investment, currently a separate item, into other (nonreserve and nondirect investment) capital movements and to distinguish the relevant capital flows in terms of the domestic sector of the creditor or the debtor instead of following an instrument-by-instrument approach. This revision would facilitate reconciliation with related bodies of economic statistics, provide better material for a variety of economic analyses, and facilitate the link between stock and flow data. In the new order, the current classifications of capital movements—by long term and short term and by financial instrument—would be relegated to secondary distinctions. This shift would recognize the lesser roles now played by these concepts, as secondary markets make the long-short distinction less relevant and the differences between instruments blur as a result of financial innovations. It is, however, proposed that the sectorization follow the functional approach of the GFSM and the draft MBS Guide, instead of the institutional approach espoused in the SNA.

The paper by McAlister on the classification of corporate enterprises (Chapter 12) deals with criteria for distinguishing between resident-owned and foreign-owned enterprises, particularly in relation to the definition of direct investment, a separate component of the capital account. Specifically, the paper examines the notion of foreign ownership with reference to the concept of direct investment as currently formulated in the BPM.

The papers by Hoezoo, McColl, and Thornton elaborate on the proposed reclassification of the capital account of the balance of payments suggested by Gill. Hoezoo (Chapter 13) reviews the treatment of new financial instruments in the balance of payments—including note issuance facilities, swaps, options and futures, and Eurobonds—and shows that their classification can be handled within the methodological guidelines provided by the current BPM. The issues arising from the securitization of debt and off-balance-sheet transactions and debt capitalization are also discussed and suggestions made for their treatment within the parameters of the BPM. A more specific discussion of the treatment of zero-coupon bonds, junk bonds, and indexed bonds in the balance of payments is provided by McColl (Chapter 14). His main conclusion in this respect is the same as that of Hoezoo: that all of these bonds, although very different from each other and different from standard bonds, do not exhibit any features for which the BPM does not already contain appropriate recommendations for proper classification. Finally, Thornton (Chapter 15) provides a framework for the presentation of arrears and debt reorganization in the balance of payments. Building on the general methodology of the BPM to this effect, he shows that a proper treatment requires the use of accrual rather than cash accounting to allow for the identification of arrears and debt reorganization transactions. This new treatment requires an increase in the detail given in the accounts beyond the current list of standard components in order to identify these transactions, which constitute part of the “exceptional financing” of the balance of payments recorded “below the line.”

II. Public Sector Accounts

Part II of the volume contains five papers that are quite different in size and content but cover most of the issues of harmonization between the government sector of the SNA and the GFSM. The first paper (Chapter 16)—the result of efforts by staff members of four international agencies—reviews most of the outstanding harmonization issues in this field; it served as a basic discussion paper and an annotated agenda for the Expert Group Meeting on Public Sector Accounts held in January 1988. The other papers address specific issues included in the group's agenda.

The discussion paper begins with an examination of the different objectives served by the government sector of the SNA and the GFSM and concludes that, although further harmonization would be useful, it is unlikely that either system can be entirely replaced by the other. The GFS is a system directly derived from government accounting records needed for operational purposes, and is largely a cash-based system. The SNA, on the other hand, requires the government sector to be treated symmetrically with other sectors of the economy and uses accrual accounting. In addition, there are many minor conceptual differences between the two systems, as well as data discrepancies arising from different compilation procedures.

The discussion paper deals next with the coverage of transactors, with special emphasis on borderline issues between the government sector and other institutional sectors and within subsectors of the government sector. Among the most difficult external borderline issues are those concerning the treatment of government departmental enterprises and ancillary enterprises as belonging to either the enterprise sector or the general government; the attribution of community production of capital goods to the sector responsible for their upkeep; the rerouting of central and commercial banking activities performed by government to the financial sector, a practice supported by both the GFSM and the draft MBS Guide but not indicated by the SNA; the attribution of government employees' pension funds and welfare schemes to either the financial sector or the government; the assignment of nonprofit institutions to various institutional sectors; and the appropriate classification of supranational authorities and other international organizations. The main intrasectoral issues concern the classification of central, state, local, and other government levels; and the treatment of social security funds, which are consolidated with the corresponding level of government in the GFSM but constitute a separate subsector in the SNA.

Another section of Chapter 16 deals with various issues in the registration of transactions, including the choice of accrual versus cash accounting, procedures for consolidation of the accounts, gross versus net recording of entries, and imputations and rerouting. The analytical framework of the government accounts, focusing in particular on the distinction between current and capital accounts, and the definitions of government saving and overall deficit or surplus are discussed next. The final section of the paper deals with selected classification issues, including the treatment of taxes, social security contributions, fees, property income, indexation payments, and the functions of government.

A concise historical account of the main efforts to date on the harmonization between the GFSM and the SNA, with special attention to the “bridge tables” between the two systems and the further links sought through joint country questionnaires, is provided in the joint paper by Levin and van Tongeren (Chapter 17).

The overall relationships between the GFSM and the SNA are the subject of the next paper, Chapter 18, by Levin. Apart from the rerouting of some government monetary authority and deposit-taking functions to the financial sector in the GFSM, there are some other major differences between the two systems, namely (1) the classification of government lending together with expenditures “above the line” in the GFSM, while government lending is “below the line” (that is, in the finance account) in the public sector accounts of the SNA, mainly because of the need for symmetry with the treatment of other sectors; (2) the consolidation of social security funds in the GFSM with the level of government at which they operate, while the SNA groups all social security funds into a single subsector of the government; (3) the cash basis of GFS data, in contrast to the accrual basis of the SNA, which covers some transactions not reflected in cash payments as well as some imputations.

The paper by van Tongeren and Tsao (Chapter 19) on the derivation of SNA value from GFS data is important for the harmonization issue because it shows how GFS data can be adjusted to obtain estimates of the public sector accounts of the SNA.

The differences of treatment of the sectorization of social security funds between the GFSM and the SNA continue to be a matter of controversy. Villacres (Chapter 20) discusses this issue in light of comprehensive statistics on the operations of such funds in different countries and concludes that both classifications are useful. A separate compilation of statistics on social security funds, as in the SNA, may serve to examine, among other things, the financial health of these institutions. At the same time, because national social security funds have reached significant proportions in many countries, their consolidation with central government data, as in the GFSM, can provide a more comprehensive view of the influence of central government policies on the economy.

III. Financial Flows and Balances

The papers in Part III of the volume cover much ground on the relationships between the Fund's MBS and the SNA.11 Dawson sets the stage for the analysis of these relationships by means of presenting an overview (Chapter 21) of the financial accounts of the SNA and a useful analytical framework for the study of the relationship between the FOF, conceived as an element of the SNA, and other SNA elements. As noted earlier, Dawson shows, by an appropriate sectorization of the economy, that the three main statistical systems of the Fund provide most of the information needed to compile a simplified but useful FOF accounts and to relate these accounts to other elements of the SNA. This finding reinforces the interest in the harmonization among all these interrelated statistical systems, and therefore between the MBS and the SNA.

The appropriate sectorization of the financial sector is the specific subject of the paper by Saunders (Chapter 22), who discusses the precise coverage of the MBS and contrasts it with that of the financial sector of the SNA. A primary difference is the inclusion, in the central bank subsector of the MBS, of financial activities encompassing monetary authority functions carried out by government, whereas such functions are retained in the government sector of the SNA. This rerouting is done to identify certain key financial aggregates. Other differences relate to the subsectorization of the accounts within the financial sector. In both cases a harmonization is contemplated in the revision of the SNA and the draft MBS Guide.

Kumah (Chapter 23) describes the evolution of monetary concepts and definitions and shows the shift of emphasis, from narrow definitions of money to broader monetary aggregates, that has taken place in many countries over the years. This has reflected both the relatively faster growth of other financial intermediaries in relation to deposit money banks and the development of new financial instruments—by both bank and nonbank financial intermediaries—that are close substitutes for money.

The statistical treatment of selected new financial instruments is examined in four papers that deal respectively with repurchase agreements (by Castello-Branco), financial leases and zero- and low-coupon bonds (by the Balance of Payments Division of the IMF's Bureau of Statistics), deep-discounted and indexed bonds (by Newson and Brodersen), and financial leases (by Newson); the last three overlap to some extent. Repurchase agreements (“repos”) involve the sale of a financial asset or group of assets, with the agreement to reverse the transaction at some specified date in the future. Castello-Branco (Chapter 24) proposes that repos be treated in the SNA as a new financial instrument, similar to a collateralized loan, following the current practice in the MBS. However, this treatment would not be consistent with that in the BPM, which treats repos as a direct transfer of ownership of the underlying securities.

The treatment of leases has long been a subject of controversy because of the wide variety of legal and financial arrangements covered under this instrument, but in general a distinction is made between financial and operational leases. As pointed out in the paper by the Balance of Payments Division (Chapter 25), the BPM defines a financial lease as an arrangement that provides for the recovery by the lessor of all (or substantially all) of the cost of the goods, together with carrying charges. The BPM also indicates that, in practice, a recovery of three fourths of the cost of the goods together with the carrying charges is taken as evidence that a change in ownership from lessor to lessee is intended. All other leases that do not meet these requirements are considered to be operational leases implying that the legal owner retains control over the assets. In financial leasing, by contrast, such control passes effectively to the lessee, who for statistical purposes becomes the owner. In this way the “ownership concept” is broadened for financial leases to encompass an element of control over the assets. However, Newson (Chapter 26) suggests new qualitative rather than quantitative criteria to differentiate between the two types of leases.

With respect to the recording of income on zero- and low-coupon bonds, the accrual method of accounting is proposed by the Balance of Payments Division; this method results in interest being imputed periodically over the life of the debt while the value of the principal reflects the increasing debt associated with accumulating interest. By contrast, cash accounting would attribute all interest (or capital gain) to the last period. However, there are other problems that affect classification—for instance, the difficulty of distinguishing interest from capital gain when bonds issued at deep discount, as in the case of low- or zero-coupon bonds, are traded in secondary markets or when indexing contracts are involved. A method to deal with these issues is presented in the paper by Newson and Brodersen (Chapter 27).

The principles of valuation and reconciliation in the MBS and the SNA are discussed by Dublin (Chapter 28). Changes in balance sheets over time include not only all transactions that have taken place during the period but also nontransaction items such as changes in the valuation of assets and liabilities (revaluation), changes in classification, and errors and omissions in the recording of transactions. All these nontransaction items are collectively assigned to the reconciliation accounts in the SNA. The valuation principles for recording transaction and nontransaction flows, as they apply to financial instruments, are now similar for both the SNA and the MBS, emphasizing the symmetrical treatment of assets and liabilities. Short-term instruments are recorded at book value, long-term domestic currency assets and liabilities at market prices, and foreign currency assets and liabilities at the prevailing exchange rate. The practices for the derivation of transaction flows from balance-sheet positions in the absence of direct transactions data are also illustrated.

A miscellaneous group of papers, dealing with some of the most difficult and controversial issues raised on money and banking statistics, completes this volume. Both Siddique and Vanoli deal with the treatment of financial intermediaries from the point of view of their contribution to output and income in the SNA. Siddique (Chapter 29) discusses the issue of the treatment of output in the banking industry, comparing current SNA practice with some proposals for change. Alternative methods of deriving banking output are reviewed, and an illustration from data for Luxembourg is provided. As opposed to the current practice in the SNA of allocating banking services to depositors only, the paper proposes allocating such services to both borrowers and depositors, highlighting thereby the intermediation function of the banking industry. Siddique also addresses an important issue concerning the allocation of imputed banking services between intermediate and final demand—in contrast to the present SNA, which allocates all imputed bank service charges to intermediate demand. Moreover, although Siddique's paper refers to “banks,” its analysis applies equally well to all other financial intermediaries except for insurance entities, for whom a specific treatment is necessary. The appropriate treatment of insurance in the SNA is the subject of Vanoli's paper (Chapter 30), which proposes that insurance services should be defined as gross premiums earned, plus net investment income, less technical charges, the last of these being computed as the claims due plus the changes in technical reserves.

General insights into broader concepts of the SNA that help clarify the relationships between flows and stocks are provided in the papers by Milot, Teillet, and Vanoli and by Lützel. Understanding of these relationships is needed to relate the real sector SNA flows to the balance sheets, which necessarily comprise both real and financial assets in the same account. Milot, Teillet, and Vanoli (Chapter 31) provide an analysis of various factors, besides real and financial transactions, that explain variations in wealth, including nonproduced assets, such as land, forests, and rivers, and exceptional events, such as earthquakes.

Lützel's paper on the treatment of fixed capital formation by owner and user (Chapter 32), deals with the issue of allocating fixed capital formation (or fixed assets) to investing sectors (producers). The conclusion that emerges is that whether the owner or the user concept should be applied for recording fixed capital formation by investors should depend on the intended purpose of the data. Thus, although preference should be given to the ownership concept in the SNA, the user concept should also be included as an additional recording devise in some cases. An application is suggested to the issue of recording financial leasing, which, as noted earlier, has gained considerable importance. Lützel’s principles suggest that the revised SNA should consider financial leasing as an acquisition of a fixed asset on long-term credit. The relevant commodity should therefore be shown as part of the fixed capital of the borrower, who would be treated for statistical purposes as an owner.

Author: Vicente Galbis