⁽¹⁾ Our analysis covers only the post-war period. A much longer span of history is studied in the recent book, British Economic Growth, 1856-1973, by R. C. O. Matthews, C. H. Feinstein and J. C. Odling-Smee, Clarendon Press, 1982.

⁽¹⁾ See the article on ‘Output growth and unemployment’, p.89.

⁽¹⁾ The detailed assumptions involved in the projection are set out in the appendix.

⁽²⁾ If measured at constant 1980 prices. If measured, for example, at 1986 prices, after the huge decline in the relative price of oil, the effects on the growth rate of changes in North Sea production would be very much smaller. This ‘index number problem’ is discussed in the appendix.

⁽¹⁾ The effect of compositional changes in employment on aggregate productivity growth has been explored in a number of studies for the United States. See, in particular, E. F. Denison, ‘The shift to services and the rate of productivity change’, Survey of Current Business, October 1973.

⁽¹⁾ If Y is output and L is employment, the change in output per employee between period 0 and period 1 is normally calculated, i.e. at current employment weights, as the change from

The change at beginning-of-period employment weights is the change from

and the change at end-of-period employment weights is the change from

The reweighting was carried out at the level of disaggregation by industry shown in table 5.

⁽¹⁾ That is to say, real income in the sense of Real National Disposable Income and real output in the sense of GDP. In the estimation of the former concept imports and exports are deflated by a single price index (the price of imports), whereas in the estimation of the latter imports are deflated by import prices and exports by export prices.

⁽²⁾ See, for example, Robert H. Rasche and John A. Tatom, ‘The effects of the new energy regime on economic capacity, production and prices’, Federal Reserve Bank of St. Louis Review, vol. 59, May 1977; Edward A. Hudson and Dale W. Jorgenson, ‘Energy prices and the US economy, 1972-1976’, Data Resources US Review, September 1978; and the discussion of these and other American studies in Edward F. Denison, Accounting for Slower Economic Grewth (pp.138-42), The Brookings Institution. Some preliminary results for the British economy are given in A. D. Roy and G. Wenban-Smith, ‘Trends in UK productivity: a production function approach’, National Institute Discussion Paper, no. 59. The results of the studies vary widely, some suggesting a very substantial effect, others a negligible one. The figuring given here, while extremely simple, does illustrate the likely dimensions of the effect, we believe.

⁽³⁾ The conclusion that the effect must have been small is supported by the results from a questionnaire survey of the National Institute's Industrial Panel members carried out at the end of 1981. Only two firms, out of the forty-eight which returned completed questionnaires, perceived any effect from higher energy prices on labour productivity, and both these said that the adaptation to higher energy prices involved a significant decrease in their labour requirements. See G. C. Wenban-Smith, ‘Factors influencing recent productivity growth— report on a survey of companies’, National Institute Economic Review, no. 101, August 1982.

⁽⁴⁾ Productivity in the economy as a whole will also have been reduced if the initial level of productivity in coal mining was lower than the level in industries in which miners would otherwise have worked. It is not clear that this was the case however.

⁽¹⁾ For an analysis of some of the factors involved in the case of married women, see Joshi, Layard and Owen, ‘Why are married women working in Britain?’, Journal of Labour Economics, vol. 3, January 1985, pp. S147-76.

⁽²⁾ Such straight-line trends could hardly continue indefinitely of course. Attempts to fit non-linear trends were unsuccessful.

⁽³⁾ The assumptions are broadly in line with the estimates of ‘frictional’ unemployment described below.

⁽⁴⁾ We used the geometrical rather than the arithmetical average. This measure was first proposed by J. C. R. Dow and L. A. Dicks-Mireaux, in ‘The excess demand for labour’, Oxford Economic Papers, February 1958. It assumes that the relationship between unemployment and vacancies is highly non-linear (a rectangular hyperbola in fact) on the grounds that unemployment, which cannot shrink below zero, becomes increasingly insensitive to the demand for labour as demand rises and, similarly, that vacancies become increasingly insensitive to the demand for labour as demand falls.