⁽¹⁾ We estimate that, when valued at 1980 prices, the volume of coal production lost during the dispute amounted to 1.1 per cent of GDP in 1984 and 0.2 per cent of GDP in 1985. The direct effect of the strike was, therefore, to subtract just over 1 per cent from the growth rate in 1984 and to add just under 1 per cent to the growth rate in 1985. Indirect effects (e.g. on the output of coal-using industries and on the expenditure of miners' households) cannot be measured accurately, but were probably comparatively small.

⁽¹⁾ Excluding special sales of assets, which were planned to increase from £2 billion to £2½ billion. (Such sales score as negative expenditure in the definition of the planning total.)

⁽¹⁾ The method of computation of these indicators is described in ‘The measurement of fiscal stance’, by Rajiv Biswas, Christopher Johns and David Savage, National Institute Economic Review, no. 113, August 1985.

⁽¹⁾ As for example in the article ‘Fiscal policy and interest rates’ by Jane Darby in the National Institute Economic Review, no. 109, August 1984.

⁽¹⁾ The estimates, particularly those for 1984 and 1985, are liable to be revised. The initial estimates of profits, always to be taken with caution, are more than usually uncertain when the residual error in the national accounts is large, since (with the possible exception of income from self-employment) profits are the component of income which is least reliably measured. The residual error of over £5½ billion in 1984 and almost £4½ billion in the first three quarters of 1985 (reflecting a higher income-based than expenditure-based measure of GDP) suggests that the estimates are more likely to be revised down than up. The balancing item in the flow of funds accounts (indicating a smaller company sector financial surplus when measured by the net acquisition of financial assets than by the excess of income over expenditure) also suggests this.

⁽¹⁾ This follows because changes in the labour force are relatively small and because unemployment constitutes a much smaller fraction of the labour force than employment.

⁽¹⁾ If a part-timer is counted as one half of a full-timer.

⁽²⁾ 57,000 households (about one in 350 in Great Britain) were interviewed in 1984, and the response rate was 83 per cent. Preliminary results from the 1985 survey will be made available in May. The Department of Employment is exploring ways of improving the data on self-employment.

⁽³⁾ The miners' strike was still in progress when our February forecast was made, but it was assumed, correctly as it turned out, that the dispute would end around the end of the. first quarter.

⁽¹⁾ Financial Statement and Budget Report 1985-86, para. 3.14. On average over the year, the effective index was almost 10 per cent higher than in January-February.

⁽²⁾ The errors in our forecasts of the level of stockbuilding were much smaller than the errors in the forecasts of its change between 1984 and 1985, the reason being that the CSO's initial estimates for 1984, showing substantial destocking, have since been heavily revised.

⁽³⁾ We had remarked in the February Review that these estimates appeared low and ‘might well be revised in time’ (p. 14), but no allowance was made for this in the forecast.

(4)The forecast is not therefore strictly comparable with the Treasury forecast accompanying the Chancellor's November Statement, which apparently (though this was not explicitly stated) allowed for a ‘fiscal adjustment’ of unknown size.

⁽¹⁾ All the results cited assume that the policy change is unantici pated, real interest rates are constant, and the exchange rate is allowed to float freely.

⁽¹⁾ Nominal 3-month rates are assumed to decline to 10-10½ per cent by the end of the year. In our last forecast in November, we assumed they would be down to 6½ per cent. The main reason for this difference is that interest rates currently are higher than we had expected, a reflection of the sharp fall in oil prices and the weakening of the pound.

⁽¹⁾ Before allowing for any expenditure out of the Contingency Reserve.

⁽¹⁾ Cmnd 9702, table 2.17. Expenditure on defence, which is classified mainly as current expenditure in the national accounts, needs to be excluded from the figures given here.

⁽²⁾ Autumn Statement 1985, para. 1.42.

(3)lt is assumed that British Gas and British Airways (with invest ments totalling an estimated £870 million, at current prices, in 1985/6) will be privatised in 1986/7 and the British Airports Authority and the National Bus Company (£219 million) in 1987/8. This follows the White Paper (Cmnd 9702, table 5.2).

⁽⁴⁾ It should be borne in mind that estimates of stockbuilding are subject to considerable error and, with such large discrepancies between the output and expenditure estimates of GDP in 1984 and 1985, the position is particularly uncertain. It is possible, for exam ple, that part of the substantial statistical discrepancy shown in table 11 represents ‘unrecorded’ stock changes.

⁽¹⁾ It is possible that the real trend in exports was steadier than shown by the DTI's seasonally-adjusted estimates. For several years now, these estimates have consistently shown exports to be relatively high at the beginning and the end of the year and low in the middle, suggesting that the seasonal adjustments might be out of date.

⁽²⁾ Statistical Appendix, table 14.

⁽¹⁾ The two estimates were derived as follows. The contributions of the relative price terms to the volume of exports in the two equations were computed by multiplying past values of relative prices by the distributed lag coefficients. The resulting series were then expressed as a percentage of their values in the first quarter of 1985. Thus according to estimate A, for example, the total lagged effect of movements in relative prices up to the end of 1985 will be to reduce the volume of exports of manufactures by about 3 per cent relative to its level at the start of 1985.

⁽¹⁾ The very high level of unemployment in recent years may have weakened the effectiveness of the seasonal adjustments. The Department of Employment does not publish the standard errors of its adjustments. However our own estimates, based on data for the last four years only, suggest that the standard errors of adjust ments for the months of December and January are of the order of 20,000 (at the 95 per cent level of confidence). On these esti mates, there has been no significant change in unemployment over the two months.