⁽¹⁾ See pages 10-12.

⁽²⁾ National Institute Economic Review no. 47, February 1969.

⁽³⁾ R. C. O. Matthews, ‘Postwar business cycles in the United Kingdom’, in (ed.) M. Bronfrenbrenner Is the business cycle obsolete? page 99, John Wiley, 1969.

⁽¹⁾ Presumably in the upswings largely influenced by the expansionary fiscal and monetary actions which occurred in 1958/9 and again in 1963, and in the downswings by contractionary policies in 1960/1 and 1965/6.

⁽²⁾ The scale and frequency of policy interventions can be largely held responsible for the apparent collapse of this mechanism. The behaviour of durable expenditures, exports and public spending, for example, can obviously be related to devaluation and the accompanying deflationary measures.

⁽³⁾ The ‘propensity to register’ may be identified as α in the equation ΔU = α (ΔL-ΔE), where U is registered unem ployment, L is the labour force, and E is the level of employ ment.

⁽¹⁾ This is subject to a qualification relating to the estimate of the number of self-employed, as noted below.

⁽²⁾ J. R. Shepherd, ‘Productive potential and the demand for labour’, Economic Trends, August 1968.

⁽³⁾ There is one proviso attaching to this conclusion, which relates to the reliability of the employment figures. These include an estimate of the number of self-employed, and it is possible that this estimate is too low (and the fall in employ ment consequently exaggerated), as one of the side-effects of SET has been to encourage former employees to become self- employed. There has been no firm estimate of the number of self-employed since the 1966 Census of Population.

⁽⁴⁾ Equations fitted to the pre-1967 period suggest a long-run trend in productivity of about 3 per cent and a reaction coefficient of employment to output of about 0.5 per cent.

⁽⁵⁾ There is some evidence that this has happened. Average hours worked had fallen fairly steadily (ignoring cyclical fluctuations) since the mid-fifties, but the trend seems to have been reversed in the last two years. Firm evidence is only available for manufacturing industries, however (and these are of course least affected by SET), and not for the service industries.

⁽⁶⁾ The high unemployment levels of 1963 were, however, partly associated with exceptional weather conditions, so that the 1963 figures must be treated with some caution as a basis for comparison. As can be seen, the 1958-59 duration figures are much closer to those for 1968-69.

⁽⁷⁾ The ‘selectivity’ hypothesis does not logically entail an increase in the duration of unemployment, although as a matter of casual empiricism, it seems an eminently likely result.

⁽⁸⁾ On the ‘structural’ side, for example, there is only very weak and tentative evidence that the variance of individual industry growth rates around the average has significantly increased in recent years. See page 37.

⁽¹⁾ It might also be pointed out that vacancies have in any case had a long-run tendency to move up in relation to unemployment.

⁽²⁾ See, for example, the leading article in the Financial Times, 9 February 1970.

⁽³⁾ Within manufacturing, against the rise in share of engineering and chemicals, the typical ‘consumer-orientated’ industries tend to lose ground—the food, drink and tobacco, textiles, leather, clothing and footwear industries have been in this category. Much the same can be said of vehicles. The traditional ‘staple’ industries—iron and steel, shipbuilding and textiles—have also declined.

⁽¹⁾ In a statistical sense. Industries (split into 18 sub-groups) were ranked according to their growth rates before and after 1967, and the relative difference in ranking was tested by the method of rank correlation. At the 10 per cent level of significance (although not at the more demanding 5 per cent level) the Spearman rank correlation coefficient of 0.48 suggested a significant difference between the two periods.

⁽²⁾ On this, the Reddaway report on the effects of SET should provide some clarification.

⁽³⁾ The variance of individual industry growth rates, year-on-year (using a nine-fold industrial classification) has risen over the period 1958-1969; the variance has been high in the years 1967-69 compared with the average for the preceding decade. However, the increase did not appear to be statistically significant; the variance normally moves in a rather erratic manner.

⁽⁴⁾ Furthermore, those industries with above-average produc tivity in 1963, and which lost in relative share, also lost in terms of relative productivity and the same is true for the below-average productivity industries which lost in relative share.

⁽¹⁾ It is possible to read some shift in the relationship between the growth of manufacturing output and total output in the post 1967 period. There seems to have been some downward shift in the relationship though it is not statistically significant.

⁽²⁾ Again the evidence of change in the structure of produc tivity was tested by rank correlation methods and confirmed a change in structure at a weak (10 per cent) level of significance both for major and more detailed categories of production.

⁽¹⁾ The figures are subject to the caveat expressed earlier with regard to the cyclical productivity components.

⁽²⁾ Rank correlation techniques were used to test Verdoorn's Law that high output growth industries are also industries with high productivity growth. On an 18-fold industrial classification the results support this thesis over the period 1965-69. ⁽³⁾Cmnd 4234.

⁽⁴⁾ See Chapter I.

⁽⁵⁾ See ‘The Economic Report of the President’, 1962 (Report of the Council of Economic Advisers).

⁽⁶⁾ R. A. Musgrave, ‘On Measuring Fiscal Performance’, Review of Economics and Statistics, 1964; see also R. A. Musgrave and P. B. Musgrave, ‘Fiscal Policy’ in ‘Britain's Economic Prospects’, R. E. Caves and associates, Brookings Institution, 1968.

⁽¹⁾ Two main problems may be briefly mentioned; first, it has been found that estimates of average marginal rates of tax (needed in order to work out the level of tax revenues at the normalised income level) are very imprecise; secondly, there is a difficult conceptual problem to resolve in determining the assumed pattern of expenditure at the full employment income level, to which the estimate of taxation receipts is highly sensitive.

⁽²⁾ The leverage concept can be defined as L = k (w₁ E-w₂R) where k is the multiplier, E and R are (actual) government expenditures and receipts and w₁ and w₂ are appropriate weights. In confining our attention to (w₁ E-w₂R) we are implicitly assuming that k is constant.

⁽³⁾ In view of the problems mentioned in footnote (1) above it would be interesting to know more about the basis for these figures, but the White Paper gives no such detailed information.

⁽⁴⁾ The kind of adjustment which conversion to a ‘normalised’ activity level would make to this run of figures would probably reinforce the trend as read from the figures in table 7. Assuming that 3 per cent (at constant prices) represents the rate of productive potential growth throughout the period, the figures for 1968/9-1971/2 are already comparable in this normalised sense. If we also assume as a first approximation that taxation receipts vary proportionately with GDP and that expenditures are invariant then the figures in the last line of table 7 for 1964/5 to 1967/8 should all be rather higher than as shown. (The normalised GDP levels would be lower than actual levels, taxation receipts lower in proportion and the weighted budget deficit as a proportion of the normalised GDP would be higher than the table 7 figures.)

⁽¹⁾ This sum is composed of the following measures: July 1966 (regulator) +0.7 per cent; September 1966 (SET) +0.6 per cent; March 1968 (budget and SET) +1.7 per cent; November 1968 (regulator) +0.8 per cent; April 1969 (budget and SET) +1.1 per cent; total = +4.9 per cent. The effect of changes in corporation tax is omitted.

⁽²⁾ DCE can be regarded as a crude attempt to adjust the increase in money supply for the effect of the balance of payments.

⁽¹⁾ S. J. Maisel, for example, distinguishes four widely held views about the transmission process:

1. 1.

   1. ‘Monetary policy influences interest rates, which affect spending. Interest rates may alter the desire to consume or save. They also determine the cost of borrowing, which influences the profitability of investment. Higher rates may limit the ability to borrow.

2. 2.

   2. Spending is a function of the wealth or the assets of individual units. Monetary policies have an impact on wealth. The measurement of policy movements, for this purpose, can be performed in many ways. Some consider movements of the money supply narrowly defined as a measure of proxy for such changes. Others use broader definitions including other commercial bank deposits, deposits in all financial institutions, all liquid assets, or all wealth.

3. 3.

   3. Expenditures may be influenced through the creation or intermediation of credit. Monetary authorities, through their creation of reserves and their impact on relative interest rates influence the amount and type of credit creation, of lending, and of borrowing. Policy effects are multiplied by legal and other institutional forces. Because of usury laws, ceiling rates, limits on types of lending, and similar forces, when interest rates rise entire sets of institutions and groups of customers may be completely eliminated from the market. Their re-entry will depend on changes in long-standing laws, regulations and contracts. This availability or rationing of credit will affect spending by individual units of the economy.

4. 4.

   4. Shifts in monetary policy cause changes in attitudes and expectations. These, in turn, may influence the spending of particular units.’

‘The effects of monetary policy on expenditures in specific sectors of the economy’, Journal of Political Economy, July/August 1968.

⁽²⁾ In an American context, it has been strikingly pointed out that ‘Whether one attributes prescience or perversity to the Federal Reserve depends largely on the particular indicator of Federal Reserve policy that one selects. Observers of money market variables, such as free reserves (excess reserves less borrowings from the Federal Reserve) or the Treasury bill rate, inevitably assess Federal Reserve actions very favourably while observers of money-stock related variables, such as bank credit or the money stock itself, habitually judge the Federal Reserve quite critically.’ (P. H. Hendershott, ‘The Neutralised Money Stock’, Irwin, 1°68, page 1.)

⁽³⁾ Adjustments and definitions are further i dicated in the notes to the chart.

⁽¹⁾ The work of Hendershott (op. cit) referred to above (page 42, footnote (2)) provides an interesting example of an attempt to normalise one kind of indicator (money stock) for non-policy influences, in the American context. Some other investigators have attempted to approach the normalisation procedure by calculating a potential money supply constructed on the assumption that the banking system obtains all available credit base assets. This procedure does not seem at all readily applicable in the United Kingdom context. Another, less partial, approach would be to build financial and policy dimensions into econometric models and determine policy stance by simulation.

⁽²⁾ As we did last year : see National Institute Economic Review no. 47, February 1969, pages 40-45.

⁽³⁾ However, as is indicated later, the scale of the reflation requirement is so large, and the size of personal consumption relative to public sector consumption and the total of invest ment sufficiently large that plausible variations on the forecast for the latter items make rather little difference to the sug gested consumption stimulus. Furthermore, the public current expenditure and investment forecasts are based on the currently known plans.

⁽¹⁾ This was an objective explicitly stated in the Green Paper (page 44, paragraph 23), which postulated a (perhaps too low) growth in productive potential of 2.9 per cent per annum. (See National Institute Economic Review no. 47, February 1969, page 41.)

⁽²⁾ This more than offsets an apparently much faster rate of growth assumed for exports of goods and services. Goods and services import and export figures are not shown separately in the table because the Green Paper contained no comparably defined figures. The differences in assumption about the rate of growth of exports and imports of goods alone can, however, be readily seen.

⁽¹⁾ See pages 39-41 above. Indirect tax increases (including SET) since July 1966 have added nearly 5 per cent to the consumer price index. Devaluation may have added a further 3 per cent. Hence from these measures alone real disposable income has been deflated by nearly 8 per cent; nor do these figures take account of other deflationary measures which have been taken.

⁽²⁾ There is the issue of ‘reflation through wages explosion’. As explained above (pages 26-27) we do not believe that wage inflation has a close bearing on the volume of consumption. For the medium term we have assumed that wages and salaries should tend to revert to normal patterns (implying a 5-6 per cent annual rate of rise); but even substantially higher figures would not much affect the reflationary calculation. No special allowance, however, was made for any resultant deterioration of the external balance due to a faster rise in wages; this depends on the rate of rise of costs in other countries as well (also accelerating) and could hardly reduce the extremely large forecast surplus for 1972 to an untenably small number.

⁽³⁾ National Institute Economic Review no. 47, February 1969, page 12.

⁽¹⁾ That is including payments for aircraft but making no allowance for anticipation of removal of import deposits.

⁽²⁾ Import price forecasts this far into the future are obviously highly uncertain.