Abstract

The IMF, in cooperation with other concerned organizations, set up a working party to investigate and improve the statistical procedures being used, and to recommend compilation procedures that would make nations' balance of payments statistics more consistent with one another. In addition to detailed explanations of its findings and recommendations, the Report contains extensive statistical appendices and 109 tables.

Statistical Appendices

I The Fund’s Balance of Payments Statistics

The balance of payments statistics that are assembled and published by the Fund’s Bureau of Statistics are the basic data for this appraisal of world payments imbalances and provide the measures of the discrepancies that are the focus of the Working Party. The principal published form of the Fund’s statistics is the Bureau’s Balance of Payments Statistics Yearbook, which becomes available in the second half of each year and contains data for most countries through the preceding calendar year.

The Yearbook has two parts, which are published in separate volumes. Part 1 is a collection of balance of payments statements for individual countries, each stated in the standard categories described in the Fund’s Balance of Payments Manual and expressed in SDRs. Countries included in the 1985 Part 1 consisted of about 125 Fund members and Switzerland. Some Fund members were missing—notably Iran, Iraq, Qatar, and the United Arab Emirates—and there were no statements for Eastern European nonmembers or for Hong Kong, Bermuda, and other economies that do not produce national balance of payments statements. Part 1 also omits international organizations such as the Fund.

Part 2 of the Yearbook is a separate publication that summarizes the information in Part 1 across countries to produce regional and world balance of payments statements. Part 2 contains the world balances that concern the Working Party. Several technical aspects of the construction of this publication need to be mentioned in order to permit readers to understand the nature of the numbers it contains, the relationships among its tables, and the adjustments needed to reach consistency across countries and across types of transaction.

Part 2 begins with a set of “A” tables that organizes the data into a number of subtotals, or “analytic” balances, including the current account, that show country, regional, and world amounts for each type of balance. These tables are followed by a set of “B” tables that give regional balance of payments statements, structured to summarize major types of transactions rather than analytic balances. Finally, there is a set of 46 “C” tables giving country and regional data for each of the main individual categories of current or capital transactions in the Fund’s standard balance of payments presentation.

Part 1 of the Balance of Payments Statistics Yearbook reproduces country data essentially as reported by the countries, though the Fund staff may rearrange the national submissions to fit as closely as possible the Fund’s standard classifications. In constructing the tables for Part 2, however, the Fund staff (including both the area departments and the Bureau of Statistics) provides estimates for reports that have been delayed and for countries that do not regularly report to the Fund. These estimated quantities are not shown separately in the Part 2 tables, but are carried into the regional and global aggregates. One consequence is that the country data shown do not add up to the regional or global totals.

For its assignment the Working Party required both an authentic published measure of the total world current account (whether correct or not) and a consistent set of accounts for individual transactions that added up to that aggregate. The current account as shown in the A tables of Part 2 was the logical candidate, but it was necessary to make an adjustment to eliminate from those current account totals the balances estimated by the Fund staff for certain countries (mainly in Eastern Europe) for which the Fund did not prepare estimates of the transactions components of the current account.

The C tables contain the country data for the individual types of transactions on which the Working Party concentrated (investment income, shipment, other transportation, and unrequited transfers) and were the main focus of the analysis. This set of tables was the basis for Tables 1 and 2 of Chapter I. To facilitate its research, the Working Party had available the full country breakdown of these tables, including the staff estimates for missing items. Moreover, the Working Party was able to obtain these data over extended time periods and in U.S. dollar terms. For its purposes the Working Party prepared estimates, where necessary, for the income and shipping transactions of countries not included in the C tables.

There are other published measures of the world current account balance besides those in the Balance of Payments Statistics Yearbook, all of which show approximately the same levels and changes. Table 80 compares the 1985 Yearbook balance with those published in the Fund’s World Economic Outlook and the OECD’s Economic Outlook. The World Economic Outlook figures, although closely related to Yearbook statistics, are put together in the Fund’s Research Department; they are prepared at a different date and are somewhat independent of Bureau of Statistics figures. The OECD states that its balances are based heavily on information supplied by OECD members, and that same information is used in constructing data for nonmembers. OECD results thus differ more than World Economic Outlook numbers do from the Yearbook numbers, which are derived from national sources and Fund estimates. For 1978 to 1982, the several series are closely parallel. For 1983-85, the differences reflect to a large extent successive revisions, ranging from the inflow of new information to the use of different estimates when current data were incomplete.

Table 80

WORLD BALANCES ON CURRENT ACCOUNT: COMPARISON OF ALTERNATIVE VERSIONS,1 1978–85

(In billions of U.S. dollars)

article image

Each series omits Council for Mutual Economic Assistance (CMEA) countries that were not Fund members, and the OECD total also omits Hungary and Romania The first column is the series adopted by the Working Party.

Adjusted to include official transfers.

II Investment Income Questionnaire

Several chapters of the Report use the results of a special questionnaire designed by the Working Party covering investment income flows and the stocks of cross-border assets and liabilities that generate income flows. The questionnaire was sent by the Working Party to about 60 countries in an effort to learn more about the figures that the countries have sent to the Fund’s Bureau of Statistics: in particular, the sources of the data, recent revisions, and unpublished detail on the types of investment income that were being reported. Replies were received from 46 countries. These replies brought out many questions of consistency and completeness, and served as a basis for discussions with the respondent countries that involved not only the questionnaire replies but also turned up related problems. The principal uses of the questionnaire results appear in Chapters II, IV, and V.

The questionnaire itself is reproduced on pages 140-46. It asks for income flow data for the two years 1979 and 1983, which were, respectively, before and after the sudden growth of income flow discrepancies, and on lines 40 to 69 it asks for stocks of international assets and liabilities for the one year 1983. It asks in particular for cross-border banking information—interest flows and positions of banks vis-à-vis banks and nonbanks. The banking distinction, when available, allows a comparison to be made between balance of payments data reported by countries and the body of International Banking Statistics (IBS) that is compiled and published by the Fund.38 That comparison can reveal inconsistencies across countries in their data or estimates of stocks and related income for this major sector of international claims and liabilities. Lines 30 to 39 asked about data sources in a schematic form, and the results of that survey appear in Chapter II.

Table 81 is a broad summary statement of the results of the questionnaire. The table compares questionnaire income flows with those published in the 1985 Balance of Payments Statistics Yearbook both for the respondent countries and for all countries. Lines 25-27 show that the 46 questionnaire replies came from countries with about 90 percent of the Yearbook totals of investment income. Industrial countries had a heavier weight in the replies than in the Yearbook, and as a result virtually all credits for direct investment income were covered while debits were less well represented. Because of this weighting in direct investment income, the balance in questionnaire countries is a sizable net credit, dominated by the reinvested earnings credit balance, both in the questionnaire (line 2 of Table 81, rightmost column) and in the Yearbook (line 9), whereas for all countries direct investment income had a small net debit (line 23) in 1983. For other income the respondent countries had a larger net debit than the Yearbook totals, mainly because a $25 billion net credit included in the Yearbook for Middle Eastern oil exporters was only thinly represented in questionnaire responses.

Table 81

COMPARISON OF QUESTIONNAIRE REPLIES WITH BOPS1 YEARBOOK, 1979 and 1983

(In billions of U.S. dollars)

article image

Balance of Payments Statistics.

Ratio of 1985 BOPS Yearbook amounts for questionnaire respondents to totals for all countries in the Yearbook.

Differences from the Balance of Payments Statistics Yearbook for the respondent countries, on lines 15 to 21 of Table 81, reflect a combination of new information, reclassifications, and statistical revisions. For reinvested earnings the larger amount in the questionnaire comes principally from data for the Netherlands and the United Kingdom, both debits and credits, that are in the questionnaire but not in the Yearbook. “Other” direct investment income is also larger because of a U.K. change that was in the questionnaire but not in the 1985 Yearbook. For the United Kingdom, these questionnaire revisions resulted mainly from direct investment income of the petroleum industry, which was entirely excluded from direct investment income totals in the 1985 Yearbook and was instead combined with other investment income in both credits and debits. At the world level, this was an asymmetry with other-country recording of these flows. The United Kingdom changed the procedure in 1985 to shift U.K. petroleum industry earnings into direct investment income. The U.K. data in the revised version that appears in the 1986 Yearbook include data for the petroleum industry, including reinvested earnings, in direct investment income beginning with 1984, so that hereafter the U.K. data will be comparable with the data of partner countries. For the questionnaire, U.K. statisticians provided estimates for the petroleum industry on a global basis for 1983 that were comparable with data for 1984 and later years as published in the 1986 Yearbook. However, country detail for the U.K. petroleum industry in 1983 is not available, so that the bilateral comparisons in Chapter IV are based on U.K. data excluding the petroleum industry.

The U.K. shift of petroleum direct investment, included in the questionnaire but not in the 1985 Yearbook, produces an equal and opposite difference in other investment income in Table 81, lines 19 and 21. That shift is obscured, however, by another U.K. change, this one entailing inclusion of income from Eurocurrency positions on a gross basis rather than tile net basis used in Yearbooks up through 1985. These grossing changes had much larger effects on credits and debits separately than the petroleum income shift and are the dominant component of line 21 of Table 81.

The U.K. grossing of Eurocredit income has no effect on net balances in the right-hand columns of Table 81, and the shift of direct investment income has only offsetting effects on the two components on lines 16 and 19. While the net effect of the questionnaire for all countries together was small for 1983 on line 15, questionnaire data raised the 1979 balance by $5 billion, resulting in an even steeper slide into net debit than the Yearbook had shown. Six countries were the source of most of the slide, including the Netherlands (reinvested income) and the United Kingdom (data revisions beyond procedural changes). All of the major statistical revisions in the questionnaire acted to deepen the 1983 net debit on total investment income relative to 1979.

Lines 20 and 21 show a sizable shift from private to official debits within total other income. As mentioned in Chapter V, the division between private and official income is statistically weak and difficult to use on a world basis, and this questionnaire result was therefore not used in Chapter V.

Tables 82 and 83 give for flows and stocks, respectively, the full detail of the questionnaire replies, with the regional division that is used in Chapter V. Components are less than the totals shown in these tables because some replies did not include all detail asked for, but the large gaps are mainly in official income and official assets and liabilities.

Table 82

REGIONAL TOTALS OF INCOME QUESTIONNAIRE: INCOME FLOWS, 1983

(In billions of U.S. dollars)

article image
Table 83

REGIONAL TOTALS OF INCOME QUESTIONNAIRE: STOCKS OF ASSETS AND LIABILITIES, 1983

(In billions of U.S. dollars)

article image

The stock data from the questionnaire, in Table 83, constitute a fairly complete inventory of information currently available from national sources on cross-border financial positions. Position data are not published in nearly as systematic a form as international payments statements, and for stocks there has been no single volume, such as the Balance of Payments Statistics Yearbook, that cataloged the data available. Although the Yearbook had increased its coverage of stock data, there were still in 1985 only a few countries showing position statements. Asset and liability positions are useful in calculating income credits and debits, as Chapter V illustrates, and even sketchy and incomplete estimates can be applied in estimating some part of income flows. However, the discussion in Chapter V illustrates that many of these national estimates of positions are inadequate when compared with data from international sources, and the questionnaire can be seen as suggesting an agenda for further development of data on international investment positions.

The countries included in these results are listed in Table 84 and 85, with total income flows and stocks reported for each country in direct investment and other investment, respectively. Most of the 46 replies to the questionnaire included some stock information, if only reserves and bank positions of the kind published in the Fund’s International Financial Statistics. Beyond the banking sector, 32 countries included private nonbank debt to foreigners, and 25 reported some amounts for nonbank assets abroad.

Table 84

DIRECT INVESTMENT INCOME, 1983: SUMMARY OF QUESTIONNAIRE REPLIES

(In billions of U.S. dollars)

article image
article image

Reported amounts are included in totals but not shown separately.

Table 85

OTHER INVESTMENT INCOME, 1983: SUMMARY OF QUESTIONNAIRE REPLIES

(In billions of U.S. dollars)

article image
article image

Regional ratios omit countries missing income or stocks.

Reported amounts are included in totals but not shown separately.

The direct investment results in Table 84 illustrate the limitations of statistics in this field. Many countries combined direct investment income with other income and cannot show it separately. Several industrial countries, however, give separate debits for direct investment income but no credits, even though they are likely to have direct investment assets. In a few cases asset stocks are given without income credits, suggesting that part of the missing credits in direct investment is connected with these countries. Stock data for direct investments are not systematically related to earnings for several reasons, including the variations in the valuations used (usually, but not necessarily, book values), and variations in earnings experience from year to year or across countries because of changes in economic conditions in different industries. Also, multinational enterprises have considerable freedom to make arbitrary allocations of earnings among their subsidiaries.

Table 85 shows a more complete set of country statistics for other investment income, although in many cases these numbers include direct investment income as well as other income. Most countries supplied stock figures to some extent, as well as income figures. Even when the relationship seems reasonable, however, there is the problem, discussed in Chapter V, that while income and stocks as measured in national statistics may be consistent with one another within a given country, they can be inconsistent with data available from other countries.

Tables 86 and 87 compare questionnaire replies on bank-related positions with corresponding International Banking Statistics (IBS) positions. For banks, in Table 86, the differences are principally in institutional coverage. Industrial country amounts are very close to those that appear in BIS banking statistics, which reflect a more narrow definition of bank than the Fund’s IBS. For offshore centers the questionnaires exclude most of the international banking activity that is also generally excluded from the balance of payments accounts of these areas, but these balances are included in BIS statistics, IBS, and the Chapter V adjustments. The nonbank figures in Table 87 are relatively much less complete and generally much lower than the independently derived estimates based on IBS that are explained in Appendix III. For countries where comparisons are possible, questionnaire reports covered only 60 percent of IBS figures for nonbank liabilities and just 40 percent of the IBS asset totals as reported by nonresident banks.

Table 86

COMPARISON OF QUESTIONNAIRE WITH IBS1 BANK POSITIONS, 1983

(In billions of U.S. dollars)

article image
article image

International Banking Statistics, which appear in the Fund’s monthly publication. International Financial Statistics.

Regional totals include IBS amounts only for countries reporting corresponding questionnaire amounts. IBS data are also shown for replies lacking the item.

Table 87

COMPARISON OF QUESTIONNAIRE WITH IBS1 BANK POSITIONS, 1983

(In billions of U.S. dollars)

article image
article image

International Banking Statistics, which appear in the Fund’s monthly publication. International Financial Statistics.

Regional totals include IBS amounts only for countries reporting corresponding questionnaire amounts. IBS data are also shown for replies lacking the item.

Total debt as reported in the questionnaire can also be compared with the statistics on debt published by OECD for developing countries. (See Appendix IV.) The OECD data are taken mainly from creditor reports rather than from the debtor countries. Table 88 shows the comparative data for the questionnaire respondents that reported debt figures in their replies and, for the same countries, the OECD tabulations.39 The 22 countries shown owed more than half of the OECD total of developing country debt in 1983, even though Brazil, Nigeria, and China are omitted for lack of a comparable questionnaire amount. For the countries listed in Table 88, results indicate that in general the creditor-based amounts tend to be only slightly larger than totals compiled by national authorities. There are a few sizable differences, however, that may be causing difficulties. As Appendix IV mentions, there is no parallel comprehensive independent measure of industrial country debt beyond the bank-related amount in IBS. For cross-border assets only the IBS exists to give some information as an “other party” measure of positions for either industrial or other countries. The special questionnaire attempted to cover the whole spectrum of international investment positions, but its coverage is incomplete. This is an area of international economic relationships deserving of greater attention.

Table 88

TOTAL DEBT REPORTED IN QUESTIONNAIRE COMPARED WITH OECD DATA FOR SELECTED COUNTRIES, 1983

(In billions of U.S. dollars)

article image
Source for OECD Data: Statistics on External Indebtedness: Debt and Other External Liabilities of Developing, CMEA, and Certain Other Countries and Territories At End-December 1983 and End-December 1984 (Paris: OECD, December 1985). Includes use of Fund credit and “other” liabilities.
Table 89

DIRECT INVESTMENT INCOME, 1977–841

(In billions of U.S. dollars)

article image

Published in International Monetary Fund, Balance of Payments Statistics Yearbook, Vol. 36 (Washington, 1985)