A New Approach to the Measurement of Terms-of-Trade Effects

Abstract

FOR a proper evaluation of the development over time of national accounts data referring to product flows, a distinction has to be made between the developments of volume and price which together account for changes in value. In order to make this possible these data are usually presented in constant as well as in current prices. In the United Kingdom accounts, for instance, we find a special table presenting the gross domestic product at I948 prices by categories of expenditure (see, e.g., Table i i of National Income and Expenditure, I946-I953 1) alongside tables which present the same data in current prices. The data concerned are shown in the form of a special accounting statement, known as the domestic production account, which shows all product flows, duly aggregated, for a number of years. If the data were not related to each other in this way then they would only permit an analysis of the current value development of each individual item into a price and a volume component. Thanks to the accounting form in which the data are presented, we can take the analysis one step further by distinguishing changes in pattern from changes in level for both prices and volumes. The fact should be stressed that this further statistical analysis of the data, which would appear to be of value to macro-economic model-building and economic analysis in general, can only be carried out successfully in the framework of a national accounts statement. It is one of the special features of these statements that they provide a consistent picture of the structure of the economic process without leaving any loose ends. The definitional relationship between the aggregates appearing in these statements is duly reflected in them. They are always complete in the sense that they present a rounded-off picture of economic activity, though the amount of detail shown may of course differ from one statement to another. In economic analysis a certain interest attaches to each of the four components mentioned: price level, price structure, volume level, and volume structure. Changes in the general price level, for instance, play an important part in monetary theory, and there is little doubt that the first price indexes to be constructed were meant to measure the change in the purchasing power of money. Changes in the price pattern, on the other hand, are often considered as being of more importance than changes in the general price level in the analysis of supply and demand, since in supply and demand analysis particular emphasis is put on relative prices, that is, on the ratios between the price of one product or group of products and that of another. Although the distinction is perhaps less pronounced in economic theory with regard to volumes than with regard to prices, the separation of the two elements of volume change, that is to say, the change in the level and the change in the pattern of economic activity, often proves to be equally helpful in economic analysis. Once an attempt is made to apply this comparatively simple idea of distinguishing four instead of two components of change in the statistical analysis of changes in the values of national accounts aggregates, it soon becomes clear that matters are more complicated than one might have expected. This being so, it is not surprising that there is a good deal of confusion as to how to calculate the eff ct of changes in the terms of trade on the amount of resources which an economy has at its disposal for satisfying its needs. After all, the measure of this effect is in essence nothing but a measure of the effect of a change in price structure, namely that of world market prices. Different calculations of this effect show different results, and all of them would seem equally justifiable at first sight. It is actually in attempts to throw more light on this particular problem that the present ' Central Statistical Office, London, I 954.