This chapter is intended to bring together and sum up the findings of the Working Party resulting from its research into the sources of the discrepancies in several sectors of the world current account—especially the discrepancy in accounting for income on international investments. It is mainly devoted to formulating recommendations of a general nature. More extended discussions and detailed recommendations are given in the relevant chapters.

Based on our studies we believe considerable progress can be made in reducing these discrepancies, both at the level of global aggregation and at the level of the statistics prepared in individual countries. The data reported by individual countries are the foundation of the world accounts, of course, but since gaps or inconsistencies in the national data are almost inevitable, there is also a need for the Fund to work with national compilers to improve international consistency, to press for the correction of errors, and, where necessary, to supply missing data or adjustments of the type described herein. In fact, lasting improvements can only be accomplished through close cooperation and effective interchange of views and information between the Fund and the national authorities responsible for preparation of the international accounts. Moreover, policymakers must express their interest in and support for these recommendations if they are to have the intended effect.

The recommendations listed below fall into two main groups: those that apply directly to actions to be taken by responsible officials and statisticians in individual countries, and those that require actions to be taken by the Fund, often in conjunction with national compilers or other international agencies. Since the production of the balance of payments accounts usually involves inputs from more than one official agency and should be integrated into the overall national accounts of each country, some of the recommendations refer to this broader framework as well as to the specific aspects of the balance of payments accounts.

At the broadest level of generality, the experience of the Working Party is that the producers of the international accounts can often generate better statistics, or additional details, when their attention is focused on areas that are deficient or have been neglected. Also, the Working Party considers that compilers and the Fund are able to improve the accuracy and usefulness of the balance of payments accounts by taking greater advantage of relevant data collected for other purposes, such as the international banking data. In that connection, the Fund can work with compilers to help in establishing priorities and to prepare improvements in data where the Fund staff finds deficiencies. At the same time, the Working Party believes that an important aim should be the limitation of costs and the enhancement of timeliness, and that these aims can be furthered by flexible and imaginative use of existing resources, coupled with a concentration of effort on the items of greatest significance for each country.

The Working Party is very much aware that it is easier to identify sources of error for a particular year and to suggest remedial measures than it is to carry through the reforms that will reduce discrepancies in the future. Therefore, a major theme of this report is that much of the effort will be wasted unless a sustained effort is organized to follow up and capitalize on the findings of the Working Party. Moreover, the explorations by the Working Party have certainly not exhausted the possibilities for ongoing research into existing problems with data on international transactions or into the problems that are certain to arise from continuing changes in practices on international financial markets and in the conduct of trade in goods and services. It should be noted that the absence of a world “discrepancy” in one or another of the categories of international transactions does not establish that there is not room for improvement in the data for that type of transaction.


Two observations about the statistical results of this study are of general applicability. First, the double-entry nature of the balance of payments accounts makes it immediately obvious when a discrepancy begins to appear in the accounts of a country, while conformity across countries with the principles of the Fund’s Balance of Payments Manual should bring to light any discrepancy in the global accounts for any given transactions category. This consistent statistical framework also lends itself to the detection of sources of error along the lines pursued in this report. Therefore, the Working Party believes that even though statistical difficulties are likely to increase, there will be scope for tracking down excessive imbalances in the reported data.

Second, the results of the analysis by the Working Party indicate that the large discrepancies observed in recent years are fairly widespread geographically, so that it does not appear that the unadjusted data have been the cause of serious miscalculations of the external economic situation of particular countries or regions. Nevertheless, the Working Party emphasizes that it would be wrong to expect too much precision from the corrective measures suggested herein. A considerable part of the information on stocks of assets and liabilities that countries are advised to use cannot be allocated geographically at present, and the possibility of unearthing hidden stocks or flows using available statistical resources is limited.

a. Direct Investment Income

The Working Party found that the causes of discrepancy in reported direct investment income for 1983 could be traced to two principal factors. For the reinvested earnings portion of this income, there were excess reported credits of $9.9 billion, and examination of the data showed that this was the automatic result of the fact that major creditor countries regularly collected data on their share of these earnings, while the debtors tended not to collect such data. Thus, the global discrepancy on this account could be greatly reduced either by making use of the geographic data on reinvested earnings as computed by creditors or by eliminating reinvested earnings from global summations of income. However, the appropriate response is clearly to encourage the collection of this information by all countries.

Other direct investment income (dividends, interest, and branch profits) showed excess debits of $11.5 billion as reported in 1983. The Working Party was able to eliminate about half of this discrepancy by using (1) a combination of corrections to data that were incorrectly reported and (2) bilateral comparisons of reported direct investment income that helped to expose certain anomalies in reporting. A sizable negative discrepancy remained after these adjustments were made, but the Working Party believes it has shown where further corrections can be made on the basis of additional consultations between the Fund staff and country compilers.

b. Portfolio Income

The largest single discrepancy for 1983 was reported in connection with income on positions vis-avis banks, plus income derived from securities and miscellaneous assets—excess debits of $32.0 billion. By a series of adjustments the Working Party was able to virtually eliminate this discrepancy, though it does not attach a high degree of precision to its results. The largest source of error was the underreporting of income from cross-border positions with banks, which accounted for about two thirds of this discrepancy. Most of the remaining discrepancy resulted from the understatement of income derived from cross-border investments in securities. Several other sizable adjustments were made to account for missing data (including the net income of international organizations themselves), to correct or reclassify particular country submissions, and to account for interest paid by nonmembers of the Fund.

The statistical procedures used by the Working Party to effect these adjustments can be repeated as necessary by the countries concerned or by the Fund staff, with such modifications of assumptions or other factors as may be required by changes over time. However, as noted above, the characteristics of available data limit the degree of accuracy to be expected.

c. “Shipment” and “Other Transportation”

The combined discrepancy for these two interrelated accounts was an excess of debits of $35.2 billion in 1983 and has been of that order of magnitude since 1980. These accounts are rather complex because of the nature of the shipping industry, which typically involves intricate chains of ownership and operating responsibility, and also has a high proportion of its operations under the control of residents of countries or economies that do not compile the necessary statistics or report balance of payments accounts. In the limited time available for its research on this topic, the Working Party has attempted to quantify and locate the source of the shipping revenues—mostly ocean freight—that are omitted and therefore result in an excess of reported shipping debits. Similarly, the Working Party has shown that many countries understate their earnings from port expenditures (a major portion of “other transportation”) and has suggested a procedure for testing the adequacy of national data reported under this heading.

Although the Working Party can only report the results of preliminary research on these two accounts, it considers that the essentials for further progress have been delineated sufficiently to serve as a basis for substantial reduction of this discrepancy.

d. Unrequited Transfers

By using corrected data based on a special survey and by introducing the missing transactions related to the operations of a number of international institutions, the Working Party reduced the initial reported discrepancy on official unrequited transfers (minus $12.9 billion) by about half. Other corrections to the reported data were approximately offsetting. For private unrequited transfers, the Working Party found that part of the discrepancy ($6.7 billion) reflected reclassifications of workers’ remittances by the Fund staff, but this does not change the overall current account discrepancy.

Considerable gaps in the reporting of official and private transfers remain, even when the adjustments suggested by the Working Party are carried out. However, follow-up analysis by the Fund and the OECD, along the lines indicated in Chapter VIII, should succeed in effecting further improvements.

e. Financial Centers and Innovation

Cutting across the analysis of particular categories of international transactions is the question of the extent to which the operations of offshore financial centers, together with the accelerated pace of financial innovation, are undermining countries’ ability to obtain the statistics necessary for accurate accounting. The Working Party looked at the major offshore banking (MOB) centers and some other intermediary economies in the context of direct and portfolio investment incomes, insurance, and shipping transactions, and found that these centers caused difficulties in at least two ways. First, some of these centers do not provide balance of payments accounts to the Fund, so that the transactions of their residents—including banks, insurance companies, holding companies, etc. established primarily or exclusively to conduct activities unrelated to the domestic economies of these centers—are missing from global tabulations. The missing income accounts were already quite large in 1983, involving $6.0 billion of missing credits on portfolio income and minus $5.9 billion of debits on direct investment income. It happens that these adjustments were nearly offsetting in 1983, so that the net effect on the global income residual was small, but the potential for large discrepancies is quite evident. Similarly, some economies serve as the countries of residence (at least in terms of balance of payments accounting principles) for the sizable part of the world’s merchant fleet whose international earnings and expenditures are missing from the shipping accounts, and they are also very important as centers through which insurance premiums and losses, and associated capital flows, are transmitted among other countries. The lack of complete reporting on these activities leaves a gap in the statistics, the full dimensions of which are not yet known.

A second difficulty caused by the operations of international financial centers is that the identities of the beneficial owners of assets or of the actual parties to transactions are often impossible to determine. This would not necessarily, in itself, cause discrepancies in the accounts if the countries ultimately involved in the transactions had adequate information on the activities of their residents, but, of course, a main feature of the intermediaries is their facilitation of concealment of assets and transactions from the countries in which the transactors reside.

In a similar vein, the surge in financial innovation in recent years has created statistical problems, largely because the transactions using the newer instruments and methods of conducting international business are difficult to capture in existing statistical structures—especially because the records are no longer so concentrated at banks and are less likely to provide the essential resident/nonresident distinction. The Working Party cannot evaluate quantitatively the impact of these institutional changes on the quality of international financial statistics but is concerned that unless adaptations are made, the measurement of international capital flows and related income streams will be severely impaired.


No. 1: National statistical agencies, with the assistance of the Fund, should implement the specific suggestions for adjustments given in the Report, adapting them when necessary to meet individual circumstances, but aiming at overall international consistency.

In the view of the Working Party, the Report lays the groundwork for methods and procedures that can be applied quite rapidly in many cases and that would yield considerable reductions in the existing discrepancies. Thus, member countries should be invited to implement the suggestions presented in the Report. Some countries may disagree with the procedures or results given in the Report, but in those cases the Working Party believes such countries should respond by notifying the Fund of the nature of their objections. If better options are available, these should also be brought to the attention of the Fund. At a minimum, however, national compilers should feel an obligation to check their data against the suggestions made in the Report, and to make adjustments or corrections where they are justified.


No. 2: The Working Party is convinced that much of the world discrepancy in the international investment accounts results from inadequacies in the methods of compiling statistics on international transactions and strongly recommends that each country conduct a reappraisal of its methodology in the light of the findings in the Report and in conformity with the Balance of Payments Manual.

There are several aspects of national compilation procedures that should be reviewed. These are discussed separately below.

a. Direct Investment Earnings

Inquiries conducted by the Working Party indicate that while many countries have developed a systematic method of accounting for the international transactions of direct investment enterprises, either as creditors or as host countries, there are many countries that do not collect adequate data. This is hardly surprising given that data on the earnings of direct investment enterprises, whether distributed or retained, are among the most difficult to collect. The disarray that exists in accounting for direct investments is manifest in the chapter on direct investment income in this report and in the description of the varying methodologies now being employed. It is the view of the Working Party that the only hope for reasonably comprehensive and accurate rendering of these accounts is the more general adoption of standard survey (questionnaire) techniques designed to obtain uniformly defined sets of data directly from the enterprises themselves. Such surveys need not be nearly as elaborate as those now employed by some countries—they should be confined to immediately relevant information, and they could concentrate on those enterprises with considerable economic weight. They should cover both capital and income accounts in an internally consistent framework.

Since uniformity in the format of such surveys is important to both the statistical agencies and the respondents, the international agencies, especially the Fund, should take the initiative by reviewing the existing questionnaire systems and attempting to establish a minimum set of uniform data. Some work along these lines has already been initiated by the OECD, and this should be coordinated with the work of the Fund.

One important means of filling some of the gaps in countries’ data on direct investment income is to use data compiled by partner countries. The Working Party strongly recommends that countries should be invited to initiate comparisons of bilateral data on direct investments with partner countries that produce detailed statistics to check on the validity of their own data, to track the complicated flows characteristic of multinational corporations, and to fill gaps where data collection systems do not exist. The Fund should be prepared to organize such exchanges of data.

(Additional specific recommendations are given in Chapter IV.)

b. Portfolio Interest and Dividends

With respect to the collection of data, or preparation of estimates, or portfolio investment income credits and debits, the Working Party finds that statistics depending entirely on exchange transaction records are insufficient and should be supplemented by estimates based on all available data on residents’ international assets and liabilities.

In carrying out its assignment, the Working Party finds that the only effective and credible way to judge the accuracy of a country’s data or estimates for interest and dividends received or paid on resident positions with nonresident banks or on portfolio holdings of stocks and bonds is to relate the income figures to the best available figures for the stock of cross-border assets and liabilities. This process also helps to ensure that the adjustments made are consistent across countries. Much of Chapter V of the Report is based on such an estimating procedure, though, as noted therein, great precision cannot be expected.

The stumbling block for many countries is that they themselves do not produce comprehensive statistics on the cross-border assets and liabilities of their residents. Some of this information is available from reliable external sources—notably the cross-border banking data produced by the BIS and the Fund, and the statistics on the debts of developing countries produced by the OECD and the World Bank. Country compilers should make use of these data—and also data compiled by partner countries. National authorities who collect their own data on the cross-border assets and liabilities of their residents can check their results against the data compiled via the international agencies and should investigate differences. In general, it is likely that banks’ reporting on their liabilities to nonresidents will be more comprehensive than data obtained from the holders of these assets, if only because the asset holders are diffused throughout the whole population of persons and enterprises. Further, the data on changes in outstanding cross-border banking assets can be used (with some adjustments) to measure capital flows in given periods, in addition to being useful as a check on income statistics.

More generally, the Working Party believes each country would improve its understanding of its external position if it would attempt to develop comprehensive data on its international investment position and relate changes in that position to the flow data in their balances of payments.

(Additional specific recommendations are given in Chapter V.)

c. Offshore Financial Centers and Financial Innovation

With respect to the statistical problems associated with offshore financial centers, the Working Party considers that these centers should be urged to give a higher priority to producing data for those sectors of their balance of payments accounts that have a significant counterpart in the international accounts of other countries. The Fund should prepare estimates of some of the missing elements, such as gross interest received or paid or the earnings of intermediary financial enterprises, but it is unlikely that this can be done effectively without considerable cooperation from the authorities of these economies.

The Working Party is also concerned that the rapid spread of new instruments and techniques for carrying out international financial transactions will blur the lines between some of the categories now deemed significant in the Fund’s Balance of Payments Manual and, more importantly, may seriously weaken the ability to identify those cases where a resident/nonresident transaction has taken place. It will become increasingly difficult to measure cross-border assets and liabilities that no longer are identified as international on the books of financial institutions or even in the accounts of the end users or suppliers of funds. Solutions to this problem may differ among countries, depending on the compilation system in use, and, in the interest of arriving at consistent solutions, the Working Party would suggest (as noted below) that the Fund put this issue high on the agenda of a joint session with national compilers. It seems to the Working Party that the necessary information on resident/nonresident positions must come either from financial intermediaries or the ultimate transactors, or possibly some mixture of these, and that advice on this subject should be sought from financial institutions and national compilers. Since the problems associated with innovations in the field of international finance are persistent, the Working Party would also suggest that a consultative group on this subject might be organized as a resource for the Fund and national compilers when new problems arise.

(Additional specific recommendations are given in Chapter VI.)


No. 3: In order to make the best use of all the information available in a country that is relevant for construction or analysis of the international accounts, national authorities should require all their agencies to coordinate and share their relevant information, and should provide for a continuing mechanism to act as a focal point for this activity. Work on the international accounts should not be isolated from other macroeconomic statistical activities.

In most countries, responsibility for the preparation of the overall balance of payments accounts is centered in one office, which usually depends to a considerable degree on data supplied by other official agencies or the banking sector. The actual compiler is often the central bank or a central statistical office, or the responsibility may be given to a specialized agency. Sometimes these statistics are collected explicitly for inclusion in the balance of payments accounts, using definitions and concepts designed for those accounts, but more often these collateral sources of relevant data serve for administrative purposes (as with tax returns or bank supervision) or as part of a broader collection of national statistical aggregates. Based on discussions with many national compilers of the international accounts, the Working Party has a strong impression that coordination and interchange of information among all the various agencies engaged in statistical work is not always as good as it could be.

In several countries, the central bank uses exchange transactions data as a basis for current account entries, but since these are not adequate for the purpose the banks’ economists infer the greater part of their overall non-trade current account from surveys of industry conducted by the national statistical office and make the necessary adjustments by means of some relatively arbitrary allocations. This process is often carried out with a minimum of contact between the two agencies, nor is there a serious effort to apply existing surveys to yield results more useful for the balance of payments. In one country the agency responsible for preparing the balance of payments does not even have timely access to vital trade data available in another ministry.

Interchange between the main agencies charged with preparing such statistics, typically the central banks and national statistical offices, is especially important and should be as comprehensive as possible.

The Working Party was created because of serious concern that large and growing discrepancies in measures of national and world current accounts could lead to misdirected policy recommendations. This consideration points out the intimate relationship between the collection of raw data, the assembly of those data in ways meaningful for economic analysis, and the content and conduct of economic analysis itself. Consequently the Working Party must emphasize that this connection should be recognized at the policymaking levels in national governments and in international organizations.

In practice, this means that the persons or offices responsible for these separate activities should not only be encouraged to communicate with each other but also should be required to integrate data requirements and analytical priorities. Where necessary to reinforce this integration, authorities might consider formal organizational ties between agencies, or offices within an agency, charged with these associated responsibilities. The Working Party has observed that the relatively narrow focus of those whose only explicit responsibility is to collect bits of data leads to failures to recognize emerging difficulties or gaps in the data, while those whose preoccupation is the preparation of analyses intended to inform policy discussions are often unaware of the fragility of the statistical underpinning of their conclusions. Moreover, the analysts are more likely to give their support to the work of the statistical offices when they are in closer touch with the day-to-day problems of those offices. One result of such support may be improved morale in the offices collecting basic statistics and more readiness to take initiatives to improve data. As well, the allocation of resources to data collection can be made on a better-informed basis.


No. 4: The agency responsible for preparation of the international accounts should be given adequate support to enable it to obtain necessary information from market participants and from other agencies.

The Working Party has noted that whereas in some countries the central bank or some other agency has full power to require reporting of any desired level of detailed data, in many others there is no legislative or regulatory mandate for such data collection, so that the compiler must depend on voluntary compliance. It is obvious that strong authority does not necessarily yield reliable data, but it is even more obvious that under many circumstances purely voluntary reporting would be futile. Of course, this is true of all statistical programs, not only those dealing with balance of payments data, but the lack of authoritative backing may be a particularly acute problem for the international sector because it is difficult to check on the accuracy and completeness of reporting, and the data themselves may not be readily available in the desired form from the records of the transactors.

On the whole, the Working Party believes that to achieve improvements in data, the statisticians need to have the support and authority of the government, not merely in terms of legal powers but especially in terms of the expressed interests of national officials. At the same time, any such authority should be used with discretion to minimize reporting burdens and avoid the accumulation of irrelevant or out-of-date information.


No. 5: National compilers should develop systematic programs for bilateral comparisons of major transaction categories with their most important partner countries, and this activity should be promoted by international agencies.

Several countries collect data on their international accounts with considerable geographic detail. This information provides an opportunity for crosschecking of several items with partner countries.

In addition to comparing the statistics for the same transaction as they appear in the accounts of partner countries, and potentially determining where corrections should be made, countries often find that a partner country has access to data that are not available at all in the other country, or may be available only at considerable cost. In such cases, the two countries may agree to share statistics rather than duplicate their efforts. An example would be the arrangement between the United States and Canada, by which the United States uses information collected by Canada on Canadian tourists in the United States, and vice versa. In addition, the two countries also make mutual adjustments to conform their trade and other current account data and also exchange data on capital flows. Another example would be the cooperation between the United States and Mexico in arriving at estimates of transactions taking place in the border area. In Europe, several countries have undertaken bilateral comparisons of their data, and such work is of considerable interest to Eurostat.

While the Working Party would not expect immediate results from such bilateral discussions in reducing discrepancies, we believe experience shows they are a useful tool for fostering improvements in basic data and their interpretation over the longer run. Furthermore, the international agencies could facilitate any such arrangements by, first, identifying those cases where bilateral or broader discussions may be most urgent because of the kinds of discrepancies that are emerging and, second, providing a mechanism for organizing such discussions. For example, the Fund will have to decide to what extent it should try to induce countries that do not have data on reinvested earnings to accept the figures produced by partner countries.


No. 6: The Working Party considers it essential that the Fund staff prepare an immediate plan of action to implement these recommendations and develop a longer-run program to guard against a recurrence of major discrepancies.

One conclusion reached by the Working Party is that the staff of the Fund, especially of its Bureau of Statistics, has a central and critical role not only in ensuring the greatest possible accuracy in national balance of payments accounts but also in providing a worldwide perspective on the international consistency of the reported accounts and of the major types of transactions that enter into the accounts. To carry out this role a number of specific changes are suggested by the Working Party in the existing Fund program for handling these statistics for enhanced contact with national compilers. However, the effect of these suggestions is likely to be dissipated if they are not brought together into an organized plan of action, with specific goals and a firm timetable.

The Working Party is aware that the implementation of this recommendation would require, at least for some time ahead, a commitment of resources beyond those customarily devoted to this set of statistics, as well as some restructuring of existing resources. It may be possible to reduce this commitment when the various adjustments suggested in the Report have been built into the regular flow of balance of payments statistics. However, the experience of the Working Party clearly indicates that without a special effort to implement its proposals and a high-level assignment of responsibility, the underlying statistical problems are likely to continue indefinitely.


No. 7: The Working Party recommends that the Fund take the lead in arranging for the interchange among compilers of practical advice, to develop improved techniques for collecting and evaluating the data essential for the balance of payments accounts. This could be done either by using the Fund staff or by encouraging the exchange of technical experts among countries. This is the direct way that countries with well-developed systems can be helpful to countries that are in the early stages of developing these and other macro-economic accounts.

It is taken for granted that the Fund will keep on providing national compilers with advice on questions of accounting principles, including the use of the Balance of Payments Manual and occasional visits by Fund experts in response to requests from member countries. In addition, the Fund conducts a training course, which could become a vehicle for providing advice at an operational level, with less emphasis placed on the elaboration of the precepts of the Manual. The Working Party believes the Fund could improve its interaction with national compilers on the practical problems of data collection or evaluation and could help national compilers to evaluate the relative importance of various sectors of the international accounts.

The Working Party considers that advice on the development of statistical systems is of the utmost importance, especially in newly organized statistical agencies. The Working Party believes progess in improving international accounting could be accelerated if the necessary practical advice on procedures and priorities could be provided at an early stage and followed up as necessary.


No. 8: The Working Party believes the Fund should give a high priority to activities aimed at reducing the discrepancies in the shipping and unrequited transfer accounts. In both of these cases, the corrections that might be desirable for individual countries will depend to a considerable degree on overall analyses of the global accounts by the Fund and other interested agencies.

With respect to the shipping and transportation accounts, basic data for the industry must be developed before either the appropriate global totals of credits or debits, or a reasonable allocation by country, can be determined and put into a consistent framework. This will probably require contributions from industry sources and may require a special consultation among national experts on this subject.

Data for official unrequited transfers are reported to both the OECD and the Fund, and comparisons of these sources by those agencies should help to identify problem areas. In addition, however, the Fund should continue its analyses of specific problems, such as reporting by international institutions, that were initiated in the Report and should exploit existing bilateral data to identify anomalies in the reporting of these transactions by donor or recipient countries. The Fund is also in a position to provide a focal point for sharing information on such items as pensions, workers’ remittances, and tax collections.

(Additional specific recommendations are given in Chapters VII and VIII.)


No. 9: The Working Party believes that the Fund’s Bureau of Statistics and Research Department could contribute to improvement of the collection and analysis of balance of payments data over the years ahead by instituting a few operational changes.

A number of topics related to the compilation or publication of the balance of payments accounts are concerned mainly with the statistical program of the Fund itself in this area. For convenience, they are listed below and are discussed only briefly. Some of the items listed are matters for consideration rather than firm recommendations.

(1) The Fund’s Balance of Payments Manual should be reviewed and updated and should benefit from the advice of national compilers. The objectives should be to take into account new developments in international transactions since the last edition was published in 1977, to improve the presentation of various subjects, to simplify the language and distinguish more clearly between major and minor issues, and to bring the Manual closer to operational problems of national compilers. The Fund, in collaboration with compilers, should also consider issuing a companion to the Manual that would present a simplified set of basic accounts, principles, and definitions. In addition, the Manual or a companion publication might suggest sources and techniques for obtaining the data necessary for the accounts and for checking on the consistency of data.

Meetings are scheduled for early 1987 to discuss definitional consistency between the concepts expressed in the Manual and the concepts of the System of National Accounts, recognizing that the international transactions of a country are an integral part of its macroeconomic statistical framework. However, it is essential that the broader problems outlined above also be covered in a revision of the Manual.

As a corollary to a revision of the Balance of Payments Manual, the Fund should review the balance of payments publications of the Bureau, especially the Balance of Payments Statistics Yearbook. In addition to changes in categories or presentation which may result from changes in the Manual, consideration should be given to changes in the statistical publications which would reflect developments in the economic environment and to modifications that would make these publications more useful to their audiences.

Part 2 of the Balance of Payments Statistics Yearbook, which has been central to the Working Party’s inquiry into world discrepancies in the balance of payments accounts, is of particular interest. Strengthening Part 2 and developing it into a more accessible and well-rounded publication would be a significant contribution by the Fund to information on world economic matters.

(2) As data are received from member countries, they should be viewed in a relatively broad perspective for their consistency not only within the country’s own accounts but also in terms of data available from sources external to the country. This may be more time-consuming and suggests that there might be several stages in the review of incoming reports. Also, the process for incorporating Fund estimates into the Balance of Payments Statistics Yearbook should be brought under systematic control.

(3) From the vantage point of the Working Party, the country-by-country orientation of the staff of the Balance of Payments Division of the Bureau of Statistics has had certain advantages. The Working Party has at times, however, found that its studies required it to cut across country lines to deal with the totality of such important categories as investment income, shipping, transportation, or official transfers. At this stage there is no focal point in the Division for analyzing major types of transactions. In the view of the Working Party, it would be helpful to have subject matter responsibilities assigned in the Division, as a means of recognizing an emerging discrepancy earlier, and as an organized way to analyze the difficulties that may appear in particular global accounts. Therefore, the Working Party recommends that the Division organize its work along topical lines as well as along the strictly geographic lines of responsibility that are now the rule.

(4) The Working Party suggests that the Bureau of Statistics strengthen its capability to provide practical advice on data collection problems, tempering its emphasis on the intricacies of definitions and principles in the light of the realities of the difficulties faced by national compilers on the ground. This emphasis parallels the Working Party’s recommendations for some reordering of the emphasis in the present Manual. Some of the world discrepancy problem is, no doubt, a reflection of local difficulties that the Fund might be able to ameliorate by working with compilers at the most elementary level.

(5) As its work has progressed, the Working Party has encountered a number of problems on which it would not offer firm recommendations, but which the Fund is urged to take under review. Without elaboration, these topics include the following: the problem of dealing with offshore financial centers that do not report the activities of the institutions in their jurisdictions that have little or no contact with the local economy; the best way to publish accounts when an element is clearly badly reported, as in the case of reinvested earnings, or is omitted; the utility of continuing to publish the Balance of Payments Statistics Yearbook using the SDR as numeraire rather than the U.S. dollar; the validity or usefulness, in current circumstances of floating rates and heavy official dependence on credit, of the groupings into partial balances that are now published and of the category of “exceptional finance.”


No. 10: The efforts of the Bureau of Statistics to obtain improved reporting from national sources should be effectively supported at the highest levels of the international institutions.

Part of the difficulty found by the Working Party in studying the current account discrepancies is that a number of countries do not submit adequate accounts on a timely basis despite requests by the Bureau of Statistics. Moreover, the Working Party found, in connection with its questionnaire survey, that compilers often were able to improve their statistics when they intensified their efforts. When reporting is incomplete or inaccurate, the Bureau faces the dilemma that it is constrained to publish country data as submitted, with any corrections or extrapolations appearing only as an unidentified adjustment to area or global totals published in the Balance of Payments Statistics Yearbook. The Working Party commends the efforts of the Bureau to carry through these adjustments, but believes this process leaves significant gaps in the statistical record. Moreover, for some countries the data as published in the Yearbook may be unsuitable for use in the surveillance process or in bilateral discussions.

Based on its experience, the Working Party believes the efforts of the Bureau to obtain improved national balance of payments accounts would be strengthened considerably if they were directly based on a renewed determination at a high level that the Fund attached great importance to these statistics and expected members to make every effort to comply with the Bureau’s requests.


No. 11: Since much of the analysis of the Working Party was focused on discrepancies between countries and regions, the question of possible restoration of the country or regional breakdowns that were at one time requested of member countries by the Fund and the OECD was discussed.

While the first priority is to get correct global data for each balance of payments sector, and to do so promptly, there is also considerable interest in the geographic distribution of each of these sectors. Countries vary greatly in the degree of geographic detail they collect for each part of their balance of payments accounts, reflecting in some cases a lack of interest in such details and in other cases a lack of the necessary resources. The process of integration of financial markets is eroding the ability to obtain data segregated by residence, and there can be no doubt that obtaining country or area detail is an increasingly burdensome part of the collection process, not only for the compilers but also for those supplying the data. On the other hand, many users of the data are interested in country or regional breakdowns. In view of the unsettled balance of the pros and cons of collecting geographic detail, the Working Party suggests that this issue be put on the list of topics to be discussed at the next assembly of national compilers.


No. 12: The Working Party commends the Fund, the BIS, and other international agencies for producing very useful statistics on international banking positions and on the debt of developing countries. Moreover, the Working Party notes the Fund’s publication explaining some of these data37 and understands that other agencies are also preparing explanations of these data. Nevertheless, the Working Party recommends consideration of a broader handbook for the public bringing together the diverse data on these subjects and explaining their similarities and differences.

In its work on the discrepancy in the world income accounts, the Working Party has worked intensively with data on international banking developed by the BIS and the Fund, and data on international debt issued by the OECD and the World Bank. These data are the result of devoted work by the agencies concerned, which deserve much credit for the results they have achieved. The path of the Working Party in using these computations has been smoothed by ready access to the compilers themselves, so that questions of definition, coverage, or interpretation could be easily resolved. However, even the Working Party, with its ready access to the sources of the data, found that it was rather time-consuming to establish what series was best suited for its purposes or to keep up with revisions and modifications of the data. (See Appendix III.)

It would be a considerable service to users of the data if a readily available handbook could be prepared that described the nature and relationships of the several series on international banking data and international debt produced by the BIS, the Fund, the OECD, and the World Bank, and provided time series to the extent possible. These data are increasingly in demand for studies of international debt, capital flows, and investment income.


No. 13: The Working Party recommends that the OECD and the Fund confer on possible coordination, or at least preparation of regular reconciliations, of their respective reports on official transfers. National compilers should be asked for their comments on their experience in reporting these data for the two purposes and on the possibilities of coordinating their national submissions.

For the purposes of its study of the discrepancy in the world balance on official unrequited transfers, the Working Party has attempted a reconciliation of data on assistance programs collected by the OECD and data on official unrequited transfers—a broader concept than development assistance—reported to the Fund as part of national balance of payments accounts. These two sets of data are compiled for different purposes and should not be expected to be the same, but they are related and should be consistent wherever the definitions and coverage are, in principle, identical. The Working Party understands that some of the differences arise from the fact that different national agencies supply the data for the two series. Some improvement in the quality of the data on official unrequited transfers and perhaps some reduction in the reporting burden on member countries might be achieved if reporting to the OECD and to the Fund on this subject were better coordinated.

The Working Party has conducted a questionnaire survey on this topic, the results of which should facilitate the reconciliation process, as well as helping to identify errors or gaps in the reporting done for balance of payments purposes.


No. 14: The Working Party recommends that compilations be prepared at the Fund to fill reporting gaps for some nonreporting entities; it also suggests that within the Fund, regular consultations on this subject among the interested departments should be intensified. The results of this work should be made available to the public and incorporated more clearly in Part 2 of the Balance of Payments Statistics Yearbook.

In carrying out its assignment the Working Party has had occasion to develop data for certain entities for which Fund sources are inadequate or missing. These groupings include the countries of Eastern Europe (with respect to their income accounts), international organizations, certain economies that were not members of the Fund, and economies that do not report the activities of large shipping fleets operated by their residents. The Working Party has carried these estimates as far as necessary for its purposes, mainly for the year 1983. However, continued preparation of data for these missing accounts will be necessary if the gaps they leave in the world’s international accounts are to be closed. The Fund staff should prepare such data for international organizations and for certain countries whose activities are, in whole or in part, not regularly submitted to the Fund. One important class of such countries would be the offshore financial centers.


No. 15: Since the aggregation of corrected world and regional current accounts will inevitably require a high degree of judgment, and will be intimately related to analyses of world economic developments, the Working Party recommends that the Research Department of the Fund, in cooperation with the staff of the OECD, take the lead in evaluating the adjustments suggested by the Working Party in terms of how best to integrate them into revised statistical summations and into the analysis of world international transactions.

As a result of its work the Working Party has provided the basis for regional allocations of some of the gaps in the compilation of current account balances. It is for the Fund, in cooperation with the OECD, to consider how best to integrate these results into revised statistical summaries of world international transactions. Of course, the Working Party has not provided pinpoint estimates of data deficiencies, nor is it able to carry all of these adjustments to the level of particular countries. Even the assignment of estimates to broad geographic areas must be viewed as somewhat impressionistic. Because these allocations of adjustments have a direct bearing on the analytical inferences that may be drawn, it is important that anybody making such inferences keep clearly in mind the assumptions on which they depend. Fund staff may well want to attempt to refine these assumptions and procedures.

Given the nature of the result, the Working Party feels that the reconstruction of the accounts, incorporating estimates of the gaps in the country data, will be essentially a research project rather than a purely statistical operation. This consideration also reflects the view, expressed above, that there should be better links between the research and statistical functions that are concerned with balances of payments.

Finally, the Working Party has, of necessity, made little or no attempt to consider the discrepancies and asymmetries that are evident in the global capita accounts. However, it is recommended that consideration be given to a review of these discrepancies and their interactions with the global current account discrepancy.