⁽¹⁾ A full list of references would be very long indeed. The basic methodology was established almost thirty years ago in a famous article by R. M. Solow, ‘Technical change and the aggregate production function’, Review of Economics and Statistics, vol. 39, 1957. Estimates on British data for the period 1856-1973 are given in chapter 7 of British Economic Growth, 1856-1973, by R. C. O. Matthews, C. H. Feinstein and J. C. Odling-Smee, Clarendon Press, 1982.

⁽¹⁾ One cannot conclude that investment and growth would have been higher necessarily if government saving had been higher. It is possible that the problem, at least after 1973, was not one of inadequate savings to support a higher investment rate but weak investment demand. If this were so, raising government saving-by reducing public consumption or raising tax rates and hence tightening the fiscal stance-might not have financed more investment, but might instead have reduced aggregate demand and output—and possibly investment still further. The reduction in government saving may have been less a cause of lower investment than a reflection of it, slower growth tending to reduce tax receipts and raise current transfers.

⁽²⁾ Only domestic fixed investment is covered, i.e. stockbuilding and net investment abroad, which are included in total investment in tables 1 and 2, are here excluded.

⁽¹⁾ See, for instance, M. Wall and S. Wren-Lewis, ‘The capital stock, survey information on capacity utilisation and the Kalman filter’, National Institute of Economic and Social Research, mimeo, 1985; A. D. Smith, ‘A current cost accounting measure of the stock of equipment in British manufacturing industry’, National Institute Discussion Paper, no. 115; and A. D. Smith, ‘The feasibility of fire insurance measures of capital stock’, National Institute Discussion Paper, no. 116.

⁽²⁾ A faster growth of gross than of net stock is typical when the quantity of capital increases substantially more slowly than in a preceding period, as was the case after 1973.

⁽³⁾ Defined as the gross trading profits of companies and surpluses of public enterprises (after providing for stock appreciation) and imputed rent from the ownership of dwellings as a percentage of GDP (at factor cost). This may tend to understate the share of capital as it excludes income from self-employment, an unknown part of which is a return on capital. On the other hand, to the extent to which the estimates of profits in the national accounts include an element of ‘pure rent’, the share will tend to be overstated.

⁽⁴⁾ For instance technical progress is taken as given; to the extent to which it is in fact ‘embodied’, capital accumulation will be more influential on growth than would appear from growth accounting estimates.

⁽¹⁾ Y - L = 1.5 + 0.25K - 0.25L where Y - L, L and K are, respectively, the percentage changes in potential (non-oil) labour productivity, labour input and capital input. Hence, if Y -L = 2, K = 2 + L.

⁽²⁾ In 1985, the average of retirements and capital consumption was 2½ per cent of the average of the gross and net stocks. This percentage has tended to rise during the 1980s because of the changing composition of the capital stock- in particular the declining importance of industrial buildings. By assuming a constant depreciation rate we assume that the age structure of the capital stock remains as in 1985.(3) The trend fitted to the logarithm of relative investment prices over the period 1975-85. This relatively short period was chosen in order to exclude the 1972-4 housing boom, which caused an extraordinary rise in the deflator for dwellings.

⁽⁴⁾ The aggregate stock-output ratio has no consistent long-term trend. Exponential trends we fitted were insignificantly negative (at the 5 per cent level) for the 1950s, insignificantly positive for the 1960s and for the 1970s, and significantly negative for the 1980s.

⁽¹⁾ Expenditure on benefits to the unemployed rose from less than ½ per cent of GNP in 1973/4 to about 2 per cent of it in 1985/6.

⁽²⁾ As rent and non-trading income are deducted as well as interest receipts, the concept employed here differs from ‘net debt interest’ in the national accounts.

⁽³⁾ This ratio is not without problems as an indicator of the tax burden. There is, for instance, the difficult question of the ‘inflation tax’. In so far as the erosion by inflation of its holdings of government debt relieves the public of discretionary spending power, and affects incentives to save and invest, it could be argued that this is a form of taxation of capital like any other.

⁽¹⁾ B = G - T + i.D

where B is the budget deficit, G is government expenditure (excluding debt interest) and i is the average rate of interest on public debt (D). We assume: D = k.GNP; so that ΔD GNP =k (g + p), where g is the real rate of growth in GNP and p is the rate of inflation. As B = ΔD,

⁽¹⁾ The weighted average growth rate of GDP of the four countries that are most industrialised among the LDCs and have conventionally been considered as NICs par excellence -South Korea, Hong Kong, Singapore and Taiwan—declined from 9.7 per cent in 1965-75 to 8.1 per cent in 1975-83; it is likely to continue declining, albeit from a high level.

⁽¹⁾ On account of the decline in the relative price of oil since 1980, the shift of real resources would be smaller if measured (say) at 1986 prices.

⁽¹⁾ Prepared by Malcolm Levitt and Michael Joyce.

⁽²⁾ Reform of Social Security, Background papers, Vol. 3, HMSO, Cmnd. 9519.