Knowledge that Transforms

PpT HIS article is a first report on some aspects of a new program of collection on the number of overtime hours worked in industries in the manufacturing segment of the economy begun by the Department of Labor's Bureau of Labor Statistics in January I956. The Bureau of Labor Statistics has been collecting data on working hours a long time and currently publishes information on the length of the workweek in considerable industrial and geographical detail. The new program was initiated in response to the increasing attention being given to the changing length and structure of the workweek and to the need an additional sensitive indicator of current economic developments. As a major subject legislation and collective bargaining, the workweek has, of course, become a considerable element in affecting labor income and labor cost. As will be shown below, fully ten per cent of the average factory worker's paycheck in I956 was accounted by premium payments work outside the regular scheduled The workweek also has become an important mechanism short-run adjustments of labor supply and labor demand. Especially during the past I 5 years of comparatively high levels of economic activity and scarce manpower resources in many occupational fields, changing man-hour input by changing the number of hours rather than the number of men has offered considerable advantages. It has become a significant factor in current economic analysis. Recent studies show the workweek to be one of the lead series in cyclical changes. Finally, it has become an important item in projections and the building of models of future economic growth. The interactions of a decline in hours of work and increases in productivity, the impact of increasing leisure time on consumption patterns and levels are just two examples of the kinds of factors which have to be considered in any assessment of future trends. Two points should be made at the beginning of this report. First, this is a new collection program and the data so far cover only one year. Obviously, this finds various industries at different points in their seasonal cycles, at different stages in their long-run development, and, in many cases, with situations peculiar to various forces affecting the year I956. In general, this was a period of slight downturn both in employment of factory production workers and their over-all hours of work. All of this should be taken into account in assessing the data presented below. Second, the definitions employed by the BLS in its collection program on employment and hours are important to a clear understanding of the data. For information on the length of the workweek, the BLS asks the employer to report the number of hours worked or in the payroll period ending nearest the I5th of each month. The phrase paid is critical: A worker getting $2 an hour a 4o-hour week, but on vacation during the payroll week, is reported and counted at 4o hours with $8o pay. The BLS series, therefore, does not measure actual labor input or actual hours of work put in place. The definition or concept employed is nearer one of labor income (or cost) although it by no means includes all such items of income or cost. Excluded, example, are Christmas bonuses and other irregular payments to employees. For information on overtime hours the BLS asks the employer to report the number of hours for which premiums were because the hours were in excess of the number of hours of either the straight-time workday or workweek. Critical here is the word premiums. Overtime hours are counted only if a premium above the straight-time rate was whether it be extra hours during any working day, extra hours during any working week above those stipulated by the Fair Labor Standards Act or a collective bargaining agreement, work on what would otherwise be a holiday, etc. i. Although overtime averaged between 2 12 and 3 hours manufacturing as a whole in I956, there were significant and substantial differences among the various industries during the

T HE study of the structure of world trade and payments which has been under way for several years at the National Bureau of Economic Research aims at preparing and analyzing a record of the economic relationships between different parts of the world. The technique, following the path marked by the League of Nations in its Network of World Trade,1 shows within the framework of a relationships between reporting areas and partner areas. But we seek to make two major advances: (i) to link the record of merchandise transactions to records of other types of transactions so that associated services transactions, transfers, capital movements, gold transactions, and multilateral settlements can all be observed in relation to merchandise; and (2) to link the record of merchandise transactions to the record of real movements in particular commodities so that the analysis of international financial problems can be related to the price and quantity elements underlying the money figures. We have compiled much of the material we set out to secure. This paper reports our progress in the first major step. It presents the rationale and results of our review of country payments accounts for a single year, I95I, showing the structure of inter-area transactions and the emerging pattern of net settlements. To give some perspective on the changing structure of international relationships, I present in the final section an anlysis of the merchandise account for the four years I950-I953. Herman Karreman, in a companion paper, reports on transportation transactions of the same period. I have not attempted in this paper to develop the analysis in the other direction, which would be an elaboration of the merchandise account into its commodity components. Nor have I completed the analysis of the payments accounts for I95I. For one thing, final judgment on the method of constructing the record of international transactions for a single year rests on the materials for all types of transactions for a period of years. For another, now that we have a preliminary five-area matrix for I95I, I intend to enlarge the analysis, dipping below the aggregates to see what patterns can be discerned by a more elaborate analysis. Many features of international economic life are hidden by the particular scheme of aggregation which we followed for practical reasons. But these aspects of the I95I story must await exposition. The paper summarized here is the fifth and Karreman's the sixth public report to emerge from the Bureau's work on the subject. The first was my I954 report, On the Elaboration of a System of International Transaction Accounts. 2 The second, Observations on the Structure of World Trade and Payments, I presented before the Subcommittee on Foreign Economic Policy of the Joint Committee on the Economic Report, November IO, I 9 5 5. The estimates given here are a revision and extension of those given before the Subcommittee. Cornelius J. Dwyer summarized his progress in the study of international petroleum transactions and Robert M. Lichtenberg reported on his in the study of the role of middlemen in world trade, before the annual meeting of the American Statistical Association in December

IT is apparent that the expansion of the United States economy in recent years is in large part traceable to debt-financed expenditures on GNP. Individuals with current incomes in excess of their purchasing needs lend out the excess to other individuals whose current incomes are below their purchasing needs. The latter are making purchases at the expense of future incomes since these debts must eventually be paid. If these deficit purchases are suddenly to cease, the economy will be left with an enormous amount of excess producing capacity. It is the object of this paper to study certain implications of the process of debt repayment for the volume of saving. Repayments are defined as transfers to sinking funds and other similar reserves maintained for bonded debt retirement by enterprises and governments, and periodic payments to retire short-term loans and mortgages held by all units. (See Table i.) An important characteristic of repayments is that in large part they are compulsory payments which must be made by the borrower, like insurance premiums. Also there is a significant lag between the time the debt is contracted and the final repayment made on it. In the case of certain consumer loans the time span is brief, but for bonds and mortgages it may exceed twenty or thirty years. Another characteristic is that repayment flows enter into gross savings in the saving and investment account, in the same way as any other individual saving flow. In the business account, repayments are included in the item, undistributed profits; in the household account, they are routed through personal saving; in the government account, through surplus in the United States accounts and savings in the UN accounts; in the external account, through foreign investment. However, the flow of repavments in the national accounts is complex. Only repayments made directly or indirectly out of income flows (more precisely, charges against GNP) are included in the national accounts. Repayments which originate during the year in question in financial and existing asset transactions are excluded. Nor are the repayments included in the national accounts fully represented or measured in the saving totals. Savings in the household, government, and external accounts are netted magnitudes, i.e., positive saving less dissaving or disinvestments. (See earlier Consumer Finance Surveys of the Michigan-Federal Reserve Board and also the National Resources Planning Committee Study Family Expenditure Survey 1935/1I936 for the concept of positive and negative saving.) Despite this, changes in the flow of repayments are fully reflected in the netted savings flows either through changes in positive saving or through changes in dissaving.' This fact is sufficient for the hypotheses presented in this paper, despite the difficulties involved in trying to visualize the flow of repayments through netted savings. Statistics on repayments are difficult to obtain. Crude approximations intended to convey a notion of their order of magnitude are shown in Table I. Repayments on long-term debt (to be designated long-term repayments) amount to about one-fourth of gross savings in the I955 national accounts. Shortand long-term repayments are almost as large as gross savings.2 These figures suggest that the rise of consumer borrowing (a small part of personal consumption expenditures in economies less developed than the United States) in the United States economy today has made repayments a magnitude of consequence, deserving closer examination.3

A SUMMATION of the amounts on the transportation line of the balances of payments submitted to the International Monetary Fund does not give a good picture of the amounts actually paid and received by the reporting countries. The explanation is, partly, that (i) many countries report the amounts paid for freight under merchandise trade instead of under transportation; (2) earnings as well as disbursements of substantial portions of the world fleet are not reported at all; and (3) many countries do not report the receipts from bunker sales or port services. But even if allowance is made for all these deficiencies in the reporting, the receipts are still much lower than the payments: the submitted figures are either too high for payments or too low for receipts, or both. My paper deals first, with elimination of the deficiencies in the reporting, and second, with methods to correct the submitted figures. To calculate the freight paid on imports by countries which do not report it separately but include it in merchandise payments, socalled freight factors were applied. These freight factors stand for the proportion of freight in the total paid by the importing country for delivery of the merchandise to its ports (the total comprising the price received by the exporting country, the freight by the carrier, the insurance premium by the underwriter, and so forth). Freight factors differ among groups of merchandise; in general, they are high for raw materials and foodstuffs, medium for semimanufactures, and low for manufactured products. During the four years studied, they varied greatly over time for the first group because of the sharp fluctuations in tramp rates but very little for manufactured products mainly transported by liners. These freight factors were applied to the amounts paid by the importing country for each of three broad groups of commodities in each of the four years of this study. This calculation is a rather crude one, particularly so since the number of selected merchandise subgroups was rather small for each country, usually not more than twenty. A more accurate but much more time-consuming method, based on imported quantities and corresponding freight rates, will be discussed later. Neither earnings nor disbursements of tankers operated by British oil companies are reported under transportation; rather they are included with all other foreign exchange receipts and payments of these companies under miscellaneous in the United Kingdom balance of payments. To assess their earnings on account of transportation, use was made of the results of a companion study aimed at the reconciliation of payments and receipts on account of international transactions in petroleum and its derivatives. Based on quantities of petroleum products traded between countries, a world freight bill for petroleum was computed by multiplying the moved quantities by the appropriate freight rates. The calculated freight was then divided between American and British oil companies according to each group's share in petroleum exported and/or imported by each country. Charter hire payments by British oil companies for foreign tankers were calculated on the basis of tanker tonnage owned (or controlled via subsidiaries) by British and American oil companies and non-oil companies. Estimates of other disbursements in foreign exchange of tankers operated by British oil companies were based on operating costs of Norwegian tankers. Earnings and disbursements of ships flying the Panamanian, Honduran, and Liberian flags of convenience, as they are often called, are not reported at all, since the three countries do not consider these fleets as part of their economy. Such earnings and disbursements were estimated and added to transportation receipts and payments of the other reporting countries. These estimates were based on the financial accounts of Norwegian ships which have virtually the same freedom in the area of opera-