You are here

STATISTICAL DISCREPANCY

29 October, 2015 - 14:23

At the beginning of this article it was noted that a country's balance of payments is commonly defined as the record of transactions between its residents and foreign resiGents over a specified period. Compiling this record presents difficult problems, and errors and omissions in collecting the data are sometimes made.

Take first the matter of coverage. In spite of attempts to gather data on them, some international transactions go unreported. One category of transactions that probably is often substantially underreported is purchases and sales of short-term financial claims; such unreported movements of short-term capital are widely believed to be a major component of total errors and omissions. No attempt is made to collect complete data on certain other transactions, which are estimated by balance-of-payments statisticians. The sample observations on which these estimates are based are sometimes of doubtful reliability, and even the best sampling and estimating techniques will not prevent errors of estimation.

Or take the matter of valuation. While import documentation, for example, may state a precise value for the merchandise imported, a different amount may eventually be paid the exporter. The discrepancy can arise for a number of reasons, ranging from default by the importer to incorrect valuation of the merchandise on the import documents.

Because of such problems total recorded debits do not equal total recorded credits in the actual balance-of-payments accounts in any given year. To provide a specific illustration of how this discrepancy arises, suppose that u.s. export documentation valued an item at $500, while in fact the terms of sale called for payment of only $400 by the foreign importer, who drew down his bank balance in the United States to make the payment. On the basis of the export documentation, balance-of-payments accountants would credit merchandise exports by $500; but when they turned their attention to U.S. short-term liabilities to private foreigners, they would find that U.S. banks had reported a decrease of only $400 (assuming no other transactions). Consequently, recorded credits would mistakenly exceed recorded debits by $100. In fact, of course, the credit entry should have been in the amount of $400.

It is to accommodate such discrepancies that the residual account, "Statistical discrepancy," was created. An excess of credits in all other accounts is offset by an equivalent debit to this account, or an excess of debits in other accounts is offset by an equivalent credit to this account. The account thus serves at least two purposes; it gives the balance-of-payments analyst an indication of the net error in the balance-of-payments statistics, so that he can have some idea of the reliability of the balance-of-payments data, and it provides a means of satisfying the requirement of double-entry bookkeeping that total debits must equal total credits. Of course, there is no need for the account in our hypothetical balance-of-payments table, which displays an equality between total debits and total credits (line 27).