^{note (1) page 4} Published in House of Commons Debates, vol. 755, cols. 648-52.

^{note (1) page 6} House of Commons Debates, vol. 756, cols 921 ff.

^{note (2) page 6} Ibid., col. 923.

^{note (3) page 6} A revision of public expenditure plans was overdue, and some reductions in existing programmes would have been necessary in any case, short of further increases in taxation or other measures for restraining quite severely the growth in private consumption. (See National Institute Economic Review no. 39, February 1967, Chapter II, especially pages 26-27.)

^{note (4) page 6} This target was compatible with that announced by the Chancellor— ‘I think we can safely allow a rate of expansion of roughly 4 per cent a year during 1968 and 1969’ —on the assumption that the Chancellor's 4 per cent applied to growth from the last quarter of 1967 on, as this would have implied for the year 1968 a rate of output about 3 per cent above the average level for 1967.

^{note (5) page 6} The taxes concerned (see table 1) were the export rebate, corporation tax, and the selective employment tax. The effects on home demand of the modifications made to the first two of these (which accounted for £200 million in nominal terms) were probably particularly small. The total effect on consumption demand of these changes could not have amounted to much more than a reduction of 1/2-3/4 per cent.

^{note (6) page 6} See Economic Trends, December 1967, page ii. The Assessment (which was dated and released to the Press for review on 14 December) specifically noted that’… there can be little if any scope for an increase in personal consumption next year, given the need to shift resources into the balance of payments’. (Ibid.)

^{note (7) page 6} This is not to suggest that the high import figures were solely due, either then or later in the year, to the level of con sumption. Other factors were also important, and apart from the rising level of import prices, the first quarter's figures were perhaps affected to a greater degree than then realised by the rebound from the previous year's dock strikes (see below, page 18).

^{note (1) page 7} All this was more graphically put by the Chancellor him self in a speech to the House of Commons on 5 December 1967, when he noted that ‘As the export demand builds up, we have to make room for it in the economy. Some room has been made already, but almost certainly not enough, and this is intentional. We do not want to dig a hole and leave it empty. We want it to be there only when the export demand is ready to fill it.’ (House of Commons Debates, vol. 755, no. 1199.)

^{note (2) page 7} Thus, it would only require the time-lag of fiscal effects to be about one quarter longer than assumed (or for the responses to devaluation to come in so much earlier), for the whole Government case for postponing the budget to become instead a credible case for deflation in December 1967.

^{note (1) page 8} However, with the resolution of this crisis, and under the favourable response in sentiment to the budget, it became at last possible to cut Bank rate from its ‘crisis’ 8 per cent level to 7 1/2 per cent. A further 1/2 per cent reduction was not possible until 19 September, when advantage could be taken of another resurgence of confidence.

^{note (2) page 8} For example, in the National Institute Economic Review, no. 44, May 1968, page 13.

^{note (3) page 8} In the Notice issued by the Bank of England on 23 May 1968, the banks were asked ‘especially to intensify restrictions on the granting of credit associated with imports of manu factured goods for home consumption or imports for stock accumulation’. (Bank of England Quarterly Bulletin, June 1968, page 120.)

^{note (4) page 8} Details were subsequently published in a White Paper ‘The Basle Facility and the Sterling Area’, (Cmnd 3787).

^{note (5) page 8} There were very substantial press ‘leaks’ in advance of this action which may have had the effect of causing the changes to be implemented earlier than first planned.

^{note (6) page 8} This inducement had been slightly reduced by the simultaneous withdrawal of certain tax rebates which had previously been incentives to export.

^{note (1) page 10} He was reported as saying, ‘I think we can safely allow a rate of expansion of roughly 4 per cent during 1968 and 1969’, House of Commons Debates, vol. 756, col. 1788, 17 January 1968.

^{note (1) page 11} The ‘actual outcome’ incorporates estimates for the fourth quarter of 1968, for which official national accounts data are not yet available. In addition, it should be pointed out that the data already available are liable to revisions, which may be substantial. For these reasons, the figures shown should be treated with caution.

^{note (2) page 11} Attribution can only be very approximate : See National Institute Economic Review no. 46, November 1968, pages 20-21.

^{note (1) page 12} A commodity/area breakdown of exports by value is given in table 3, page 50, which includes forecasts for 1969 on the same basis.

^{note (1) page 13} The sharper than usual drop in the United Kingdom share of trade in manufactures within a year of devaluation contrasts unfavourably with the experience of other countries. (France, after the devaluation of 1957 and 1958, the gradual devaluation by Canada in 1961-62, or the converse experience of Germany following the 1961 revaluation, see chart 4.)

^{note (2) page 13} National Institute Economic Review no. 43, February 1968, page 9.

^{note (3) page 13} Excluding imports of US military aircraft.

^{note (1) page 14} See I. G. Black, J. E. Kidgell and G. F. Ray, ‘Fore casting imports : a re-examination’, National Institute Economic Review no. 42, November 1967, and the earlier article by W. A. H. Godley and J. R. Shepherd, ‘Forecasting imports’, National Institute Economic Review no. 33, August 1965.

^{note (2) page 14} The equation in question is M = −60 + 0.19 CE + 0.10 PA + 0.15 GFI + 0.25X + 0.48S, where, with all values in £ million, at 1958 prices and seasonally adjusted, M = imports; CE = consumers' expenditure; PA = public authorities' current expenditure; GFI = gross fixed invest ment; X = exports and S = stockbuilding. The equation was fitted over the period 1955-67. It was estimated by regression subject to the constraint that the ratios one to another, of the co-efficients of the variables CE, PA and GFI should be consistent with those implied by the input-output tables for 1963. For purposes of estimating calculated values for 1968, figures of stocks and exports were adjusted for the estimated effect of the hangover from the 1967 dock strikes.

^{note (3) page 14} See National Institute Economic Review no. 42, Novem ber 1967, pages 4-9. The ultimate effect was then put at over £500 million.

^{note (4) page 14} Purchases of US military aircraft excluded.

^{note (1) page 15} Purchases of US military aircraft excluded.

^{note (1) page 16} The net balance of trade in the diamonds, etc. category was in fact slightly favourable in 1968, although on the silver- platinum account, there was quite a large deficit, which may be diminished later.

^{note (2) page 16} The January 1969 trade figures extend this impression.

^{note (3) page 16} The diminishing residuals of the aggregative equation (above, chart 5) support the same suggestion.

^{note (1) page 17} Some results of an earlier enquiry into the operation of the surcharge were discussed in I. G. Black, J. E. K. Kidgell and G. F. R. Ray, ‘Forecasting imports : a re-examination’, published in the National Institute Economic Review no. 42, November 1967.

^{note (2) page 17} That is, the ratio of the percentage change in imports by volume to the percentage change in prices.

^{note (1) page 18} This would imply an overall elasticity on imports of goods of something like −0.45, compared with the originally assumed −0.7. See National Institute Economic Review no. 42, November. 1967, page 5.

^{note (2) page 18} This would imply slightly more extensive lags than were assumed in the November 1967 Review (ibid. page 5, table 1).

^{note (3) page 18} Very little stockbuilding was recorded as having taken place in 1968 (+£43 million in the first three quarters) and it is worth noting that the discrepancy between the estimated ‘compromise’ GDP and expenditure-based GDP, which might well be regarded as in part an unrecorded addition to stock was larger in these first three quarters than the recorded stockbuilding. In addition, it may be noted that the series of ‘stock-changes in mainly imported commodities' (Statistical Appendix, table 15) shows a large rise, in volume terms greater over the first three quarters of 1968 than the past annual averages for more than ten years.

^{note (4) page 18} The pattern of arrivals of imports from distant sources compared with that of arrivals from nearer sources suggests that some allowance should be made for the dock strike effects. But it probably cannot be at all a large one—perhaps only £20-£30 million. Official statements concerning the level of importing in the first quarter of the year never sought to make much of any dock strike effect.

^{note (1) page 19} This Inquiry gave a +1 per cent balance of those revising their intentions upward over those revising down since the previous (November-December 1967) Inquiry.

^{note (2) page 19} Which is, of course, partly estimated (for the fourth quarter).

^{note (3) page 19} Assuming that reliable information on existing pro grammes was available to those who prepared this forecast. The Financial Statement suggested (page 34, para. 5) that the annual rate of growth in public sector investment—including housing—would be over 2 per cent between the second half of 1967 and the first half of 1969; but it looks as though there was an actual fall in public sector investment up to the second half of 1968 of well over 1 per cent. (The new White Paper ‘Public Expenditure 1968-69 to 1970-71’ (Cmnd 3936) con firms that unanticipated falls occurred in capital spending.)

^{note (4) page 19} In table 2, the ‘actual’ stock figures are shown inclusive of the discrepancy between the GDP compromise measure and GDP as measured from the expenditure side. This has the effect of changing the estimate of stockbuilding (com pared with that shown in Economic Trends, January 1969) by -£14 million in the second half of 1967, +£74 million in the first half of 1968, and +£45 million in the second half (where the adjustment affects only the third quarter data).

^{note (5) page 19} These measures were, however, preceded by some anticipatory purchases, so that the net effect on the second half figures, if this is taken into account, would be rather small.

^{note (1) page 20} It is presumably a particular hazard of official forecasting that such a statement would either impugn the public policy commitments of the authorities (if, for example, it allowed for a discount in the realisation of announced objectives) or weaken them (if, by not including such a discount, the less palatable consequences of strict adherence to policy commit ments were made explicit). These points are particularly relevant to prices and incomes policy.

^{note (2) page 20} All these quotations are from The Financial Statement 1968-69, page 35, para. 8. They refer to a course of develop ment between the second half of 1967 and the first half of 1969, but must appear to apply a fortiori as a description of expectations for the period from the second half of 1967 to the second half of 1968.

^{note (3) page 20} Economic Trends, January 1969, pages iii-iv.

^{note (4) page 20} Economic Trends, October 1968, page iii.

^{note (5) page 20} The sum of money incomes plus a ‘credit effect’ deflated by the consumer price index. See table 3, page 37.

^{note (6) page 20} The ratio of savings to personal disposable funds was very probably about the average for the 1960s, and slightly higher than in 1966 or 1967.

^{note (7) page 20} The savings figure is calculated as a residual so that small revisions to the income and consumption estimates can make a large difference to it.

^{note (8) page 20} Cf. National Institute Economic Review no. 46, November 1968, pages 9-11.

^{note (9) page 20} The failure of credit to fall, however, was probably not a negligible factor. In the case of our own May forecast (National Institute Economic Review no. 44, May 1968, page 15, table 6) for example. the credit effect in the last half of the year was expected to be negative (-£18 million and -£7 million in the third and fourth quarters); with the positive out-turns (see page 37, table 3), the total ‘turnround’ implied amounts to over a 1/2 per cent unexpected addition to disposable funds. (The credit effect referred to is a weighted sum of changes in outstanding bank advances and hire purchase debt, added to disposable income to give disposable funds, on the basis of which the consumption forecast is prepared.)

^{note (10) page 20} In the sense of wages and salaries per person in employ ment (see page 37, table 3, footnote^(b)). Theo fficial average earnings figures show a similar rate of rise.

^{note (11) page 20} See ‘Productivity, Prices and Incomes Policy in 1968 and 1969’ (Cmnd 3590).

^{note (1) page 21} See the Employment and Productivity Gazette, January 1969, page 41, for details.

^{note (2) page 21} The consumer price index referred to is the implicit deflator of the constant price series for consumers' expenditure. It usually moves in approximate conformity with the retail price index, which enables us to estimate its value in the fourth quarter of 1968.

^{note (3) page 21} Between the first three quarters of 1967 and the first three quarters of 1968.

^{note (4) page 21} Allowing for a six months lag in the feed-through from higher unit factor costs to consumer prices, and a lesser lag in the case of import prices. Some account also needs to be taken of the change in expenditure patterns over the period, which, with a relative reduction in purchases of cars, would lead to a small increase in the price index.

^{note (5) page 21} It could be argued that the ‘prices’ side of prices and incomes policy has been independently responsible for the fall in profit margins : but the hypothesis is not very com pelling.

^{note (6) page 21} National Accounts definitions.

^{note (1) page 22} House of Commons Debates, vol. 755, cols. 648-52.

^{note (2) page 22} House of Commons Debates, vol. 743, col. 1788.

^{note (3) page 22} None of these figures includes any allowance for the normal surplus on unidentified transactions (shown in table 4 as the ‘normal’ balancing item).

^{note (4) page 22} The detailed analysis underlying these conclusions is given on pages 11-13, for exports, and pages 13-18, for imports.

^{note (5) page 22} See, for example, G. A. Renton, ‘Forecasting British exports of manufactures to industrial countries’, National Institute Economic Review no. 42, November 1967, pages 41-2.

^{note (1) page 27} Attention was drawn to this some time ago : Cf. I. G. Black and C. Gillion ‘Some characteristics of unemploy ment’, National Institute Economic Review no. 37, August 1966.

^{note (2) page 27} National Institute Economic Review no. 46, November 1968, pages 12-13.

^{note (1) page 28} All the same, it may be worth noting that the official seasonal adjustment factors for unemployment combine both constant and multiplicative elements (Ministry of Labour Gazette, September 1965, page 382). It could be suggested that in a period of sustained high levels of unemployment, greater than those prevailing in the period from which the seasonal adjustments were estimated, the constant term will give rise to distortions, and that a wholly multiplicative adjustment would be better. Experiments with a wholly multiplicative adjustment indicate that figures processed in this way show a rather smoother movement through 1967 and 1968 than the officially adjusted series.

^{note (2) page 28} There is also the fact that the employment indicators (for industrial and manufacturing employment, at least—the June 1968 total unemployment figures are not yet available) and the hours worked series are sympathetic with vacancies rather than with unemployment.

^{note (3) page 28} This assertion is grounded in some work which has been carried out in association with the Institute's regional studies, and the methods employed to describe the duration profile and account for variations in it will be fully explained in a forthcoming paper. Briefly, the duration profile at any time can be described by an equation which is linear in the logarithms of the number unemployed (N_i) and the period (P_i) for which they have been on the register (say, 2 weeks, 3 weeks, 4 weeks and so on) so that log N_i = a — b log P_i. The slope of this linear relationship (the value of the co efficient b) changes according to the level and recent change in unemployment. See also the article ‘Duration of Unemploy ment’ in the Employment and Productivity Gazette, August 1968.

^{note (4) page 28} The taxonomy of unemployment is a vexed subject. By ‘structural’ here would be meant either a situation of ‘Galbraithian technological unemployment’ (or more generally one where there is mismatching of skills demanded and offered) or one perhaps more correctly defined as ‘spatial frictional’ unemployment, where vacancies are available in one area which the unemployed in another would be equipped (in terms of skill) to fill.

^{note (5) page 28} There are variations on this theme. See, for example, R. M. Solow, ‘The Nature of Sources of Unemployment in the U.S.’, Wicksell Lecture, 1964, page 21, ff.; Eleanor G. Gilpatrick, ‘Structural Unemployment and Aggregate Demand’, (Johns Hopkins 1966); Knowles and Kalachek, ‘Higher unemployment rates 1957-60, structural transformation or inadequate demand?’, Report to the U.S. Congress Sub-committee on Economic Statistics (Washington; US Government Press, 1961).

^{note (6) page 28} This estimate was obtained by inspection of a graph relating unemployment and vacancies, in logarithmic form. The relation is linear, and can be thought of, diagramatically, as having shifted outward from the origin since the period 1963-65.

^{note (1) page 29} Again, using the vacancy-unemployment relationships of 1963-65 as the benchmark. All the figures quoted are subject to error as they are the result of visual inspection; but they seem accurate enough to indicate a wide spread in experience, which is the essential point.

^{note (2) page 29} This discounts a low 5 per cent figure for Wales, where the unemployment-vacancies relationship appears already to have shifted markedly by 1964-65. In most other cases the pattern is stable until 1966.

^{note (3) page 29} The increase in hours worked as unemployment rose over this period is notable, being the first time since October 1958-April 1959 that this has happened, and being substained over a much longer period. In this latter respect the experience is unique in the whole period for which data is available.

^{note (4) page 29} As measured by the variance of regional unemployment levels about the national average.