This analysis and forecast for the economy is mainly concerned with the effects of devaluation and the associated measures taken. It begins, therefore, with a separate analysis of these effects. These are then superimposed on the pre-devaluation forecast for 1968. Briefly, this pre-devaluation forecast showed a rise in real gross national product of 3½ per cent between the years 1967 and 1968, and a rather higher rise than this—of nearly 4 per cent—between the fourth quarters of 1967 and 1968. It also showed a ‘basic’ deficit, on current and long-term capital account (including a ‘normal’ balancing item) of £275 million in 1967 and £250 million in 1968.

Developments prior to devaluation had not greatly altered the position as it appeared to us in August. The fuller information now available confirms that there was little change in the second quarter either in industrial production or in the value of world trade. The check to the growth of trade, which continued into the early part of the third quarter, was, however, rather more severe than we had supposed, particularly in the case of industrial countries' trade with one another.

The National Institute is currently studying ways of improving its methods of short-term economic forecasting. The two articles which follow report on some of the recent work undertaken in two parts of the field, namely exports and imports.

In the case of exports, our methods hitherto have been rather subjective. The first article is the account of a more rigorous procedure proposed for one part, though a very important one, of total exports of goods, namely exports of manufactures to industrial countries. Attempts to find procedures of analogous rigour for exports to primary producing countries have, so far, failed to throw up anything very useful. For imports, our procedure has for some time been rigorous, in the sense of making use of postulated econometric relationships. Thus the second article does not so much propose a new method as elaborate and modify existing methods in a way which may prove to be more satisfactory.

‘The forecasting of exports, because it depends on trends and policies in all the rest of the world, is likely to remain the most intractable problem in economic forecasting. Since changes in exports have repercussions elsewhere, this limits the success one can hope for in forecasting demand as a whole.’(1)

This article was prepared before the announcement of the devaluation of sterling on 18 November 1967

In August 1965, an article in this Review considered the problem of the short-term forecasting of imports into this country. It argued that some aggregate relationship was probably necessary, for forecasting total imports of goods and services, mainly because stockbuilding appeared to have an important and distinctive influence on imports. The import content of stockbuilding seemed to be much higher than the import content of other forms of expenditure. It was not possible to distinguish the separate influence of stockbuilding on different categories of imports—food, basic materials, fuel, and so on.