⁽²⁾ Derek F Channon, The Service Industries, Macmillan, London, 1978; and J Whiteman, The Services Sector: a Poor Relation, NEDO, 1981.

⁽³⁾ Professor Douglas McWilliams, The IBM Lecture, ‘The Renaissance of British Management’, Kingston Business School, 25 May 1988; and the Liverpool Research Group in Macroeconomics, ‘The Underestimation of Productivity’, Quarterly Economic Bulletin, Vol 9, No 2, June 1988.

⁽⁴⁾ Current official practice for measuring real changes in service output is described in: CSO, Series and Weights Used in the Output-Based Estimate of Gross Domestic Product at Constant Factor Cost, Occasional Paper No 20,1987.

⁽⁵⁾ See, for example, GF Ray, ‘Productivity in Services’, National Institute Economic Review, February 1986; and AD Smith (Ed), Technological Trends and Employment: Commercial Service Industries, University of Sussex, Science Policy Research Unit, Gower, 1986.

⁽⁶⁾ A publication which proved helpful in identifying such unofficial sources is David Mort and Leona Siddall, Sources of Unofficial UK Statistics, Warwick Statistics Service, University of Warwick Library, Gower, 1985.

⁽⁷⁾ Strictly, for purposes of GDP measurement, these prices should be the price dimension of value added associated with each of the industry's products/services.

⁽⁸⁾ R Hirschhorn and R Geehan, ‘Measuring the Real Output of the Life Insurance Industry’, The Review of Economics and Statistics, Vol 59, 1977, p212.

⁽⁹⁾ The latter price index should be, but rarely is, one compiled with end-year weights.

⁽¹⁰⁾ Including the national income adjustment for financial services.

⁽¹¹⁾ The official method currently used for measuring real output changes in the finance leasing industry is based on two elements: the gross capital stock of leased assets held by each lessee industry in 1980; which are moved over time by the respective output indices of these lessee industries. This measure has recently replaced an indicator based on the gross capital stock of the finance leasing industry valued at 1980 prices. The modification was introduced because it was thought that, given the output measurement methodology employed for industries which use the leased assets, an element of double counting was effecting real changes in GDP. However, the above solution adopted is far from ideal below whole economy level as the double counting should really be eliminated from the output of the lessee industry not from finance leasing output.

⁽¹²⁾ For each industry the 1980 GDP and 1985 employment weights were respectively: building societies 2 and 6 per cent; hire-purchase 5 and 3 per cent; insurance 17 and 22 per cent; advertising 2 and 4 per cent; construction plant hire 1 and 4 per cent; banking 44 and 36 per cent; accountants 8 and 12 per cent; architects 4 and 5 per cent; finance leasing 10 and 2 per cent; and computer services 7 and 6 per cent.

⁽¹³⁾ Although the number of sports hours transmitted by television and radio is also included in the output of the broadcasting industry this no more constitutes duplication than does counting tons of steel produced separately from, and in addition to, the output of vehicles. In principle potential duplication is avoided when the output indicators are aggregated using value added weights.

⁽¹⁴⁾ Official output indicators do not align with the 1980 SIC ‘group’ classification for which the new measures have been produced so that it is possible only to compare the alternative and official output indices for recreational services taken overall, not for individual activities.

⁽¹⁵⁾ This capital stock measure takes as its starting point an estimate, based on current cost accounting valuations, of the gross stock of equipment in the catering trades in 1983. This figure has been carried back to 1971 and brought forward to 1986 by means of the perpetual inventory method with the assumption of a 14 year asset life for equipment. See AD Smith, ‘A current cost accounting measure of Britain's stock of equipment’, National Institute Economic Review, May 1987.

⁽¹⁶⁾ AD Smith, Technological Trends and Employment, op cit, p32 et seq.

⁽¹⁷⁾ Hotel and Catering Training Board, Innovation and Change in the Hotel and Catering Industry, Research Report, 1988, p13.

⁽¹⁸⁾ On one score only may the introduction of alternative service output measures have led to any upward bias in the adjusted GDP growth rates. To allow for the fact that sectoral contributions to GDP are measured before the deduction of interest payments to financial service industries, the official practice is to modify GDP growth rates using the output measures for the relevant financial services in conjunction with appropriate negative weights. Logically therefore any changes to GDP growth rates resulting from the introduction of new output measures for financial services should to some extent be offset by such an adjustment. In the present case this adjustment proves to be small. For the period 1978 to 1986, for example, the effect of the new measures would be to lower the GDP growth rate by no more than a twentieth of a percentage point, from the rate of 1.6 per cent per annum to about 1.55 per cent.

⁽¹⁹⁾ English Tourist Board, Registration Statistics and English Hotel Occupancy Survey.