^{note 1 in page 6} The unusually large discrepancy betweeen the expenditure and output estimates of GDP in recent quarters created a particular problem for the forecast. We have assumed as usual that in the future the two estimates will move together. This gives the ‘residual’ in table 2, which, while constant as a proportion of GDP from the second quarter of this year onwards, changes by almost 1½ per cent of GDP between 1982 and 1983 (i.e. the forecast increase in the output estimate of GDP is 1½ per cent less than the increase in the sum of the expenditure components).

^{note 1 in page 9} Among other things, including expected price inflation and deviations in real earnings growth from its historical trend.

^{note 1 in page 10} Most benefits, including the retirement pension, are to go up by 3.7 per cent, which is the increase in the retail price index over the twelve months to May. Supplementary benefit is to go up by 4.3 per cent, which is the increase in the index after the exclusion of housing costs. Standard rates of unemployment benefit are to go up by a little over 8 per cent (after restoration of the 5 per cent abatement made in 1980) and child benefits by 11 per cent.

^{note 1 in page 12} There is no recent evidence on the relative saving propensities out of different sources of income. A study of data for 1955-65 found marginal propensities to consume out of current grants, wages and salaries and other personal incomes of the order of 1.0, 0.75 and 0.5 respectively. (Surrey, M.J.C., ‘Personal income and consumers’ expenditure', in Hilton, K. and Heathfield, D.F. (eds.), The Econometric Study of the United Kingdom, Macmillan, 1970.) These estimates suggest that if the components of income shown in table 7 had changed in the same proportion as the total between the second quarter of 1981 and the first quarter of 1983 (i.e. they had all fallen by 0.2 per cent), consumers' expenditure would have been lower by only about ½ per cent by the end of the period.

^{note 1 in page 14} In response to the question asked in the July CBI Survey What are the main reasons for any expected capital expenditure authorisations over the next twelve months', 76 per cent of firms replied ‘to increase efficiency’, 55 per cent ‘for replacement’, and only 16 per cent ‘to expand capacity’.

^{note 1 in page 15} Our econometric investigations of stockbuilding by the manufacturing sector have failed so far to uncover a significant interest-rate effect. In what seems, at first glance at least, a some what surprising result, we have found that the nominal level of short-term interest rates seems to have a relatively large effect upon the level of stocks held by distributors and retailers. It may be that the interest rate variable is serving as a proxy for variations in the availability of bank loans which affects primarily the smaller firms found in retail trade and distribution.

^{note 1 in page 17} On the adjusted estimates, employment in Great Britain fell by 639,000 between September 1981 and December 1982 and self- employment rose by 125,000. Since registered unemployment rose by 342,000 and HM forces fell by 14,000, these imply a reduction of 186,000 in the working population, much smaller than the 511,000 reduction implied by the unadjusted estimates of a 839,000 fall in employment and no change in self-employment, and closer to the estimated small increase in the population of working age over the period. The remaining gap between the estimated change in the working population and in the population of working age may be partly explained by technical differences in definition and coverage, as well as by ‘discouraged’ workers withdrawing from the labour force or seeking early retirement because of poor employment opportunities.