^{(note 1 in page 32)} They would appear even weaker if North Sea output were excluded from industrial production.

^{(note 1 in page 34)} Of course, other things being equal, a still further dis cretionary reduction in the PSBR would have tended to reduce interest rates. But it is just the use, by others, of ceteris paribus to which the Economic Progress Report was objecting.

^{(note 1 in page 35)} Over the last ten years the average (absolute) difference between annual changes in M1 and sterling M3 has been 8 1/2 per cent, which is more than double the tolerance between the upper and lower bounds of the target range.

^{(note 1 in page 36)} In 1980 0.35 per cent of total UK exports (£173 million) went to Argentina, and 0.22 per cent of imports (£114 million) came from there. 92 per cent of UK exports to Argentina are manufactures and accounted for 0.43 per cent of total UK exports of manufactures (£159 million). 51 per cent of UK imports from Argentina were food (£58 million) and amounted to 1.05 per cent of total UK imports of goods. Non-mineral raw materials, excluding fuel, accounted for 24.6 per cent (£28 million) of imports from Argentina, 0.78 per cent of total UK imports in that category.

Complete data for 1981 are not available, but in the last four months of 1981 the percentages were similar to those given above. If all UK exports to Argentina ceased we would thus expect a fall of about 0.4 per cent in exports of manu factures, assuming (a) no other countries joined a boycott of British goods and (b) no other export markets absorbed the displaced exports.

On the imports side there are substitutes for everything we import from Argentina. It may cost more to buy from alternative sources; but even if alternative supplies cost 20 per cent more, it would add only about 0.05 per cent to the price index of imports of goods.

^{(note 1 in page 45)} The recent productivity performance is discussed in the February Economic Review, pp. 26-28.

^{(note 1 in page 46)} It is difficult to assess the magnitude of the effect of deficient demand on the balance of payments. Such an assessment will depend on the view taken as to the causes of the deficiency in demand: for example, the current low level of output may be in considerable part a result of a low level of exports because of a deterioration in competitiveness associ ated with the rise in the exchange rate up to the beginning of 1981; the effect of this on the balance of payments is, of course, different from that of a low level of output arising from deficient domestic demand. But it is possible to arrive at the crudest sort of estimate, taking account only of the effect of a lower level of demand on the imports side of the balance, in the following way. The elasticities in the NIESR model suggest that a 1 per cent reduction in total final demand would reduce the volume of imports by a little less than 2 per cent, which is roughly equivalent to a £1 billion improvement (at current prices) in the balance of payments. This simple arithmetic, which is intended to do no more than convey an impression of the broad orders of magnitude involved, would imply that, if real demand had grown at ‘trend’ (say 2-3 per cent per annum) rather than fallen by 5 per cent between 1979 and 1981, the greater part of the current account surplus would have been eliminated by higher imports.

^{(note 1 in page 48)} This note was prepared by Mr P. S. O'Brien.

^{(note 1 in page 49)} These are all single equation simulations; any feedback effects of exports on other price variables or world trade, or inter-relationships between the explanatory variables, are not allowed for.

In particular it should be noted that the comparisons ignore the interdependancies between the different measures of com petitiveness. It was argued earlier that the greater stability of the relative price index probably reflects manufacturers absorbing part of large movements in labour costs in order to prevent large price changes, which may cause marketing problems. Under these circumstances one might expect that if RNULC actually had stayed at the level of late 1980 then manufacturers would have eventually been forced to adjust prices so that PRPEX would have continued rising in 1981 and 1982, rather than remaining flat. Without a model of the relationship between the two series these effects cannot be considered here, but they would certainly reduce to some extent the differences between the simulations represented on chart 3.

^{(note 1 in page 50)} See S. J. Brooks, ‘Systematic econometric comparisons: exports of manufactured goods’, National Institute Economic Review, no. 97, August 1981.