Entries to record unrequited transfers are required in the balance of payments when one party provides to a nonresident a good, a service, or cash, without the recipient assuming any obligation for payment. The country of residence of the provider of a gift in the form of goods, for example, records in its balance of payments a credit entry for the value of the goods in the merchandise account and offsets this by a debit contra-entry for the same amount in unrequited transfers.

Unrequited transfers are reported in the balance of payments under two main headings—private and official. Private transfers are, in principle, only those in which both transactors belong to the private sector, while official transfers cover transactions in which one or both parties belong to the official sector. In practice, however, it may be difficult to determine the sector of the foreign transactor, so that for convenience or lack-of information, all transactions which involve the resident private sector may be classified as private.

The main types of official transfers are development assistance provided in cash or in kind, or through the voluntary cancellation of debt, budgetary contributions to international organizations, and non-development grants provided by international organizations; government pension payments; and the receipt and payment of taxes, particularly withholding taxes. Private transfers include, inter alia, migrants’ transfers, reflecting the net worth of people who migrate from one country to another; remittances of private nonprofit institutions, normally for aid or welfare purposes; gifts and inheritances; and tickets sold by, and prizes won from, nongovernment lotteries.

As shown in Table 73, the discrepancies in both private and official unrequited transfers have opposite signs and have remained relatively stable from 1978 to 1984. The credit balance on private transfers peaked in 1980 at $7 billion, while the debit balance on official transfers reached a high of $20.8 billion in that year. The decline in this discrepancy is partly arbitrary, because it reflects the lack of reported data after 1981, in the 1985 data base, for a large Middle Eastern donor, whose debits in this account were of the order of over $5 billion annually. It may be that the appropriate amounts are being included under another heading, so that the overall current account discrepancy may not be affected to this extent. It is not known to what extent the recipients of any continuing streams of transfers reported these amounts as credits.

Table 73


(In billions of U.S. dollars)

article image
Source: International Monetary Fund, Balance of Payments Statistics Yearbook, Vol. 36 (Washington, 1985).

The discrepancies on private and official transfers largely arise out of independent factors. The fact that one partly offsets the other is mainly fortuitous, except for the minor classification problem, referred to earlier, that may arise when an official sector donor makes a grant to a foreign private sector recipient. In such cases, the donor country should have no difficulty assigning such a transaction to the resident official sector, but the recipient country might be either unable to allocate the transaction to a foreign official sector or have no such classification in its balance of payments. In either case the recipient country would include the transaction with the resident private sector, tending to generate a net credit in this account.

The regional breakdowns given in Table 73 show that for official transfers, the excess of debits largely reflects an imbalance between industrial and developing countries, for reasons discussed in this chapter. Also, the abrupt drop in debits of Middle Eastern oil exporters stands out. For private transfers, as reported in the Balance of Payments Statistics Yearbook, one of the outstanding characteristics is the large net debit reported by the Middle Eastern oil producers, while other developing countries are shown with large net receipts. A substantial factor in this pattern is the remittances of temporary workers who are considered to be residents of the country of temporary residence for balance of payments purposes.


In order to try to pinpoint the sources of discrepancy in the series for official and private transfers, it was decided to obtain more data from the countries accounting for the bulk of the transactions in transfers reported to the Fund’s Bureau of Statistics for 1983. The questionnaire used for this survey is shown in Appendix VIII. It was sent to 48 countries, with 33 responding, including all major industrial donor countries. The questionnaire asked for information on transfers broken down by the official and private sectors. Within the official sector, extensive detail was sought regarding the kinds of development assistance donated and received. In addition, respondents were asked to identify their major partner countries to provide the basis for making intercountry comparisons.

A comparison of the data obtained on the special questionnaires with those contained in the Balance of Payments Statistics Yearbook for the same countries is shown in Table 74. To make the comparisons valid, it is necessary to adjust for some reclassifications made by the Fund staff to the accounts as originally reported.

Table 74


(In billions of U.S. dollars)

article image

Balance of Payments Statistics.

The adjustments are to incorporate reclassifications by the Fund staff of data as originally reported by the countries. The major reclassification involved shifting reported credits from factor services to private unrequited transfers.

Comparing the data in columns 2 and 3 of Table 74, it is apparent that the biggest adjustment, of $5,083 million, has been made to private unrequited transfer credits. This reflects the reallocation by the Fund of personal remittance receipts to transfers, rather than inclusion of them in service receipts as reported by the countries concerned. The last column indicates the extent of revisions suggested by the questionnaire replies.


The results of the questionnaire survey by type of transfer are shown in Table 75. Because the coverage of the small number of major donors is much more complete than the coverage of recipient countries, the survey coverage of total official debits was much more comprehensive than that of official credits. Nevertheless, the coverage was sufficiently extensive to be able to support some conclusions about the discrepancies in the global data for unrequited transfers. The analysis is hampered at this stage, however, because of the inability of some countries to provide data on their transfers in the detail called for in Table 75.

Table 75


(In billions of U.S. dollars)

article image

European Economic Community.

For official transfers, the largest factors causing the discrepancy are the following:

(1) The lack of credits covering transfers of non-development support to international institutions themselves, in particular the transactions of the European Economic Community (EEC). For questionnaire respondents, which include all major industrial countries, reported net transfers to such institutions, other than for development assistance, reached $6.5 billion in 1983, about half of the total discrepancy on official transfers. These institutions receive larger amounts from their member countries than they pay out to them in the form of transfers, and this imbalance is reflected in the excess of reported debits by the countries transferring funds, on balance, to these institutions. As shown in Table 75, under “other multilateral,” the reported debit balance for reporting EEC countries (with reports received from all except Greece and Ireland) was over $3.4 billion, reflecting the imbalance between extremely large gross flows. Gross flows to non-EEC institutions under the same heading were much smaller, but because transfer receipts from these institutions, whether reported by industrial or developing countries, were very small, a net reported debit balance of almost $3.1 billion resulted.

(2) To a lesser extent a similar problem exists for development assistance channeled through international agencies, where the discrepancy is a net debit of $1.6 billion. That discrepancy arises from gross debits by reporting countries of $2.1 billion, which is partly offset by reported aid receipts from the institutions of $0.5 billion. The imbalance partly reflects the financing of the overhead costs of the institutions themselves, but the questionnaire results exaggerate the discrepancy somewhat because of the limited coverage of credits by developing countries. The actual discrepancy would be reduced if relevant data for the international agencies themselves were recorded, and this could also throw some light on the shortfall in credits reported by recipient countries.

(3) The excess of debits reported in the questionnaire for bilateral development assistance was $3.4 billion, which to some extent reflects the less complete coverage of recipient countries than of donor countries in the questionnaire survey. Since no data were available for one major non-industrial donor, Saudi Arabia, the discrepancy in this category for all countries is understated to the extent that recipient countries recorded such flows, and, of course, the gross flows are probably too low. Recipient countries were able to identify and report much higher amounts of transfers in the form of cash than did donor countries, where perhaps the information was not readily available to balance of payments compilers. The high proportion of cash to total transfer credits as reported by recipients suggests that recipient countries might have difficulty in identifying and valuing noncash development assistance, particularly technical cooperation, and that this difficulty could account for a large part of the discrepancy of $3.4 billion. Alternatively, the donors may be assigning a relatively higher value to noncash assistance than the recipient countries.

(4) The discrepancy for pension payments was a net debit of $2.6 billion. All major paying countries were covered in the survey, but the coverage of recipient countries was incomplete. If total receipts were increased by, say, $600 million to include an estimate for this incomplete coverage, this would still leave a discrepancy for all countries of over $2 billion, reflecting the difficulties of recipient countries in identifying and measuring these flows. Of course, it is also possible that pension credits may be erroneously included elsewhere in the current account.

(5) Within the total of miscellaneous transfers for the official sector, the small net credit balance of $104 million considerably understates the real discrepancy between gross credits and debits owing to offsets within components of this balance. For example, only three countries (Australia, Canada, and the Federal Republic of Germany) reported withholding tax transactions, which resulted in a net credit balance of $0.7 billion, composed of gross credits of $1.5 billion and gross debits of $0.8 billion. Discrepancies in other components of this category of transfers cannot be identified because no further details are available. To the extent they are reported, military grants should also be included here.

It should be noted that the discrepancy caused by the omission of data for international agencies, described in (1) and (2) above, will not affect the overall current account to the extent that these agencies have expenses in the countries where they are located that are incorporated as service receipts in the balances of payments of those countries. Similarly, it should be noted that the treatment of withholding taxes discussed in (5) above will have an effect on discrepancies in the accounts for transfers and the items on which withholding taxes are levied, such as investment income; but to the extent that these taxes are treated consistently throughout the accounts, there will be no effect on the overall current account asymmetry.

For private transfers, the questionnaire asked only for a disaggregation between institutional remittances and other transfers. Data on institutional remittances were requested to complement the information obtained on official sector development assistance. The numbers reported were, however, rather small, particularly for the countries where these nongovernment organizations are located. This might reflect the fact that to the extent these organizations are government-funded, the data are in the official sector, while recipient countries cannot, of course, classify aid received from these organizations on the basis of the latter’s funding.

A major problem area for private transfers, which was not probed by the questionnaire, appears to be workers’ remittances. This might to some extent be a definitional problem, relating to residence. With a few exceptions, if a person is resident in an economy for one year or more, he is deemed, according to the Balance of Payments Manual, to be a resident of that economy. If he makes payments to family members in his country of origin, these payments are unrequited transfers. However, some countries consider that their nationals working abroad remain residents of their country of origin, so that the payments received are classified as services rather than transfers.

To the extent possible, the Fund’s Bureau of Statistics, in its publications, has tried to ensure that the conceptually correct treatment is followed and has adjusted countries’ data accordingly. These adjustments are, however, easier to identify and to incorporate as credits than as debits, because they tend to bulk large among the credits reported by some countries. As noted above, the Fund’s adjustments add $5 billion to reported private transfer credits in 1983, helping to create a positive residual.

If adjustments were made to official transfers for the factors identified above which can be most readily addressed, the results for 1983 (and, probably, similar results for preceding years) would be as follows:

The net discrepancy of $5.5 million, after taking into account the adjustments of $6.3 billion made in Table 76, probably largely reflects underrecording of credits for development assistance and possibly for some military aid.

Table 76


(In billions of U.S. dollars)

article image

Balance of Payments Statistics

From Table 74.

Derived from the $6.5 billion imbalance reported in the questionnaire, less an estimate for undercoverage in the questionnaire survey, compared with BOPS Yearbook net credits of $0.5 billion.

Derived from the $1.6 billion imbalance reported in the questionnaire, less an estimate for undercoverage in the questionnaire survey, compared with BOPS Yearbook net credits of $0.6 billion.

This net amount is added, conjecturally, to compensate for the lack of data after 1981 in the 1985 BOPS Yearbook.

The problems concerning the recording of development assistance transfers have recently been explored by the OECD, which, as a contribution to the work of the Working Party, has produced a report entitled Official Development Assistance Grants and the Discrepancy in the Reporting of Official Unrequited Transfers, dated June 17, 1985. That report concentrated on comparing the data on official unrequited transfer debits and credits reported in the Balance of Payments Statistics (BOPS) Yearbook with the data on official development assistance (ODA) supplied to the Development Cooperation Directorate (DCD) of the OECD.

The analysis contained in the above-mentioned OECD report was impeded by a lack of detail in the data published in the BOPS Yearbook. Total grants for ODA by DAC countries in 1983 are reported as $18.1 billion, and it was presumed by the DCD that this amount would be consistent with the balance of payments debit entry. However, the maximum amount of development assistance that can be included in the gross balance of payments debit of $49.0 billion, as reported in the BOPS Yearbook adjusted for the questionnaire results, is $17.4 billion,36 and this covers all countries included in the balance of payments tables and not just DAC countries reporting ODA data to the DCD. It would therefore appear that, if the ODA total as compiled by the OECD is accepted as correct, the gross debits in the balance of payments are understated by perhaps $700 million, which would increase the reported discrepancy.

However, in using the ODA data it should be recognized that there are definitional and coverage differences between these data and the balance of payments data. For example, transfers to French overseas departments are included in ODA but not in the French balance of payments as reported to the Fund. Similarly, while DAC members include in ODA the administrative costs of their aid programs which cannot be allocated to specific parts of these programs, they do not, with the notable exception of the United States, appear to include such costs in their balance of payments statements.

To quantify these differences and to try to resolve the discrepancies, further comparisons should be made, at the finest possible level of detail, between the data obtained from the questionnaire survey and those in the DCD information system. To the extent possible, bilateral flows of transfers should be compared. Unfortunately, however, the availability of geographic data in the questionnaire survey is somewhat limited, since the balance of payments compilers did not seem to have access, in all cases, to the data base from which data are supplied to the DCD.


(1) The Fund should incorporate information for international institutions into its balance of payments data base. This should be done not just for unrequited transfers but also for all current and capital account transactions. The information should be collected directly from the institutions concerned through specially designed report forms. The information for specific institutions should, to the extent possible, be compared by the Fund with data obtained from member countries of these institutions and, in particular, with the accounts of the countries in which these institutions are located.

(2) Data collected on official transfers in the balance of payments should show development assistance separately, in order to permit comparisons with the information collected by the DCD of the OECD. If balance of payments data are available on a bilateral basis, comparisons with the DCD data will be facilitated and problem areas can be narrowed down. In the meantime the Bureau of Statistics should explore with the OECD discrepancies between the bilateral and other data obtained from the questionnaire survey and the DCD reporting system, in order to learn more about the nature of these discrepancies and what can be done to reduce them.

(3) The Fund should establish a focal point for the compilation and exchange of bilateral data on such transfers as payments of pensions, workers’ remittances, and development assistance.

(4) Only a few countries include withholding taxes in their balance of payments statements, resulting in a sizable net credit of $0.7 billion under that heading, with an offsetting discrepancy, in principle, under the heading of the income or the other factor payment involved. Consideration should be given to treating such taxes separately in global totals of official transfers, so that part of the discrepancy can be explained.

(5) The present treatment of workers’ remittances in the Balance of Payments Manual should be reviewed to determine whether it is consistent with actual experience.